THESE DE DOCTORAT Université Paris VI- Pierre et Marie ...hydrologie.org/THE/LE.pdf · Université...

THESE DE DOCTORAT

Présentée et soutenue publiquement pour obtention du titre de DOCTEUR des

Université Paris VI- Pierre et Marie Curie &

Académie de la Science et de la Technologie du Vietnam (Co-tutelle)

Spécialité: Biogéochimie des hydrosystèmes Ecole Doctorale : Géoscience et Ressources Naturelles

LE Thi Phuong Quynh

FONCTIONNEMENT BIOGEOCHIMIQUE DU FLEUVE ROUGE

(NORD –VIETNAM) : BILANS ET MODELISATION

Soutenue le 7 Juillet 2005

Composition du jury:

M. Venu ITTEKKOT Prof. Dr, CTME, Bremen, Allemagne Rapporteur M. Quang Cu BUI Prof. Dr, VAST, HoChiMinh, Vietnam Rapporteur M. Georges VACHAUD Prof. Dr, CNRS, Grenoble, France Rapporteur M. Ghislain DE MARSILY Prof., Univ. Paris VI, Paris, France Examinateur M. Wolfgang LUDWIG Dr, CEFREM, Perpignan, France Examinateur Mme. Josette GARNIER Dr, CNRS- Univ. Paris VI, Paris, France Directrice de thèse M. Gilles BILLEN Dr, CNRS- Univ. Paris VI, Paris, France Directeur de thèse M. Van Minh CHAU Prof. Dr, VAST, Hanoi, Vietnam Co-Directeur de thèse

Thèse préparée au sein des laboratoires Sisyphe, UMR 7619, CNRS (France) – INPC, VAST (Vietnam)

THESIS

Written and defended for obtaining the doctorate degree of

Pierre et Marie Curie University (France) &

Vietnamese Academy of Science and Technology (Vietnam) (Co-supervision)

Speciality: Biogeochemistry of hydrosystems PhD School: Geoscience and Natural Resources

LE Thi Phuong Quynh

BIOGEOCHEMICAL FUNCTIONING OF THE RED RIVER

(NORTH VIETNAM): BUDGETS AND MODELLING

Defended on July 7th 2005

Composition of the Committee:

M. Venu ITTEKKOT Prof. Dr, CTME, Bremen, Germany Reporter

M. Quang Cu BUI Prof. Dr, VAST, HoChiMinh, Vietnam Reporter

M. Georges VACHAUD Prof. Dr, CNRS, Grenoble, France Reporter

M. Ghislain DE MARSILY Prof., Univ. Paris VI, Paris, France Examinator

M. Wolfgang LUDWIG Dr, CEFREM, Perpignan, France Examinator

Mme. Josette GARNIER Dr, CNRS- Univ. Paris VI, Paris, France Advisor

M. Gilles BILLEN Dr, CNRS- Univ. Paris VI, Paris, France Advisor

M. Van Minh CHAU Prof. Dr, VAST, Hanoi, Vietnam Co- Advisor

This thesis is prepared at the laboratories

Sisyphe, UMR 7619, CNRS (France) – INPC, VAST (Vietnam)

Acknowledgements

Acknowledgements

First of all, I am extremely grateful to my advisors Dr. Josette Garnier and Dr. Gilles Billen for accepting me as their PhD student and for their enormous assistance and helpful discussions during my thesis. They have helped me in understanding concepts in a simple and intuitive way, while introducing me to new ideas. They always know how to solve the problems and always encourage me during the difficult periods. They have always offered me special cares during my stays in France so that I could feel happy and comfortable. I would express my particularly thanks to them.

I would like to thanks my Vietnamese co-advisor, Prof. Dr. Chau Van Minh who gives me the opportunity to work in the ESPOIR project and to realize the cotutelle Ph.D. thesis. He always provides me the favorable working conditions in the Institute of Natural Products Chemistry (INPC). Without his helps in experiments, samplings and administrative papers in INPC in Vietnam, the thesis would never be finished.

The PhD thesis was performed in the ESPOIR project, a French-Vietnamese program for water quality and water treatment in the period from 2000 to 2004. I would like to thank Prof. Georges Vachaud, Prof. Chau Van Minh and Prof. Nguyen The Dong to give me the chance to pursue this thesis in the framework of the ESPOIR project.

I am also indebted to Prof. Ghislain de Marsily, the ex-director of the Ecole Doctorale “Géosciences et Ressources Naturelles” for analysing my Vietnamese degree courses and accepting my inscription. I must also thank Prof. Laurent Jolivet, the present director of the Ecole Doctorale for his kindness with the administrative forms that permits my continuing during the last period of this thesis.

Furthermore, I am deeply thankful to all the members of the jury: Prof. Venu Ittekkot, Prof. Georges Vachaud, Prof. Bui Quang Cu, Prof. Ghislain De Marsily, Dr Wolfgang Ludwig, Dr. Josette Garnier, Dr Gilles Billen and Prof. Chau Van Minh, who gave many interesting and helpful comments and critics for my thesis manuscript and also for the enrichment of my scientific knowledge.

During this work, I have been granted by the French Embassy in Vietnam at Hanoi. I would like to express my thanks to the French Embassy in Vietnam and I especially thank Mr Bruno Paing, attached to the cooperation of Science and Technology for his interest in this programme and for always helping me with kindness in finding administrative solutions.

This work is a cotutelle thesis. I would also like to acknowledge the Leaders of Institute of Natural Products Chemistry, the Leaders of University of Pierre and Marie Curie, who

i

Acknowledgements

permitted me to carry out this work. Helpful financial supports was provided by the Direction of the International Cooperation of the Pierre and Marie Curie University.

I wish to extend a sincere gratitude to the director of Sisyphe laboratory, Prof. Alain Tabbagh, to give me the warm welcome in this laboratory.

I express my sincere thanks to Sylvain Théry, a very humorous, friendly and hard working person, for his huge helps, especially in the Red River data base elaboration, logical programs creation and map drawing.

Moreover, I would like to thank the kind colleagues Nguyen Van Tuan, Tran Bich Nga and Nguyen Van Tue in Meteorological and Hydrological Institute for their useful helps in Vietnamese meteo-hydrological information. I would like to thank the sympathetic colleagues in the Son Tay, Yen Bai, Hoa Binh, Vu Quang hydrological stations for their helps in water samplings.

Among all the numerous people who have contributed to valuable ideas and experiments related to this work, I would like to mention Dr Michel Meybeck and Dr Agnes Ducharne (Sisyphe), Dr Pham Van Cu (Institute of Geography in Vietnam), Dr. Pham Huu Dien (Hanoi, Pedagogic University I), Dr Nguyen Kien Cuong, Prof. Ngo Ngoc Cat and Dr Nguyen Thanh Van (VAST). I express a deep gratitude to them.

I also wish to express my gratitude to my colleagues in the Institute of Natural Products Chemistry: Luong, Thao, M. Ha; they have given so much help in the sampling campaigns and sample analyses. My thanks are also due to Nicolas Prieur, who spent two years as a CNRS Engineer making the link between French and Vietnamese team, and had a large contribution in the organisation of the sampling campaigns for Nhue-Tolich urban rivers in the framework of the ESPOIR project.

At Sisyphe, I truly thank Nadine and Valérie at the management and secretaryship, Maya responsible for the informatics. I would like to sincerely acknowledge the generous assistance provided by the following colleagues: Maïa, Séverine, Anun, Mohamed, Samia, Maïté. My thanks are sent also to all these so kind friends: Agata, Véronique, Aurélie, Julien, Harouna, Denis, Anne, Angelbert, Hans, Noémi, D. Thuy, Tam … for cheering me during the four half-year stays in France.

At last but above all, I would like to extend my sentiments to my closest relatives. I am greatly indebted to my parents and sisters for their morale and love supports as well as their confidence in my scientific orientations and decisions. I am happy to get diploma, but my parents are proud of that. I am deeply grateful to my husband who no only always understands, believes and encourages me, but also helps me. My best Vietnamese friends Binh, Trang, Long, Phong, Vu, Loi, Thuc are thanked for the wonderful days we spent together at school and/or University and for their continuous encouragements.

ii

Résumé

Résumé

Le Fleuve Rouge (au Nord Vietnam et en Chine méridionale) couvre une surface de bassin versant de 156 450 km2, avec une population de près de 30 millions d’habitants. L’axe principal du Fleuve Rouge (aussi appelé Yuan, Thao ou Hong) reçoit deux affluents principaux, le Da et le Lo, puis forme un large delta avant de se jeter dans le Golfe du Tonkin (en Mer de Chine méridionale). Les trois sous-bassins supérieurs et le delta diffèrent largement en terme de densité de population (de 101 hab.km-2 dans les bassins amont à plus de 1000 hab.km-2 dans le delta), d’usage du sol et de pratiques agricoles.

Le but général du présent travail est de développer une compréhension d’ensemble du fonctionnement biogéochimique de ce système sub-tropical de dimension régionale, et de son contrôle par les processus naturels et anthropiques. L’épine dorsale du travail a consisté dans l’implémentation du modèle RIVERSTRAHLER, développé antérieurement pour décrire le lien entre la qualité de l’eau et les activités humaines dans le bassin de la Seine et d’autres fleuves européens (Billen et al., 1994, 1997, 1999, 2005; Garnier et al., 1995, 1999, 2000, 2002), pour le cas particulier du système Fleuve Rouge.

La première étape dans cette étude a consisté dans la modélisation du régime hydrologique et du transport solide du Fleuve Rouge (Le Thi Phuong Quynh et al., subm). Les estimations antérieures de la charge solide du Fleuve Rouge variaient entre 100 et 170 106 t.an-1, c-à-d de 640 à 1060 t.km-².an-1. La forte dépendance du transport solide à l’hydrologie est responsable d’une large variabilité inter-annuelle. Sur la base de données hydrologiques relatives à la période 1997-2004, et d’un suivi journalier de la matière en suspension à l’exutoire des 3 principaux tributaires du Fleuve Rouge en 2003, un modèle simplifié a été établi pour estimer la charge solide moyenne interannuelle du Fleuve Rouge sous les conditions actuelles. La valeur obtenue est de 40 106 t.an-1, correspondant à une charge spécifique de 280 t.km-2.an-1. Elle reflète une réduction de 70% de la charge solide totale suite à la mise en eau des réservoirs de Hoa Binh et de Thac Ba réservoirs dans les années 1980s. Le modèle prévoit une réduction supplémentaire de 20% de la charge en suspension suite à la construction planifiée de deux grands réservoirs supplémentaires. Utilisant les mesures de contenu en phosphore total dans la matière en suspension réalisées dans ce travail, le flux de phosphore exporté par le Fleuve Rouge peut être estimé à 36 106 kgP an-1.

Les données de concentrations en nutriments dans le réseau hydrographique du Fleuve Rouge étant assez rares, un suivi de la concentration des formes de l’azote, du phosphore, de la silice, du carbone organique et de la chlorophylle à l’exutoire des principaux sous-bassins amont, dans l’axe principal du Fleuve dans le delta et dans quelques rivières polluées de la

iii

Résumé région d’Hanoï, a été réalisé à une fréquence mensuelle durant les années 2003 et 2004, permettant de définir le niveau général de concentration en nutriments dans les eaux de surface.

En vue d’examiner le degré de perturbation anthropique du cycle de l’azote et du phosphore à l’échelle du bassin, des bilans de ces deux éléments ont été établis pour le système sol et pour l’hydrosystème des 4 principaux sous-ensembles (Da, Lo, Thao et Delta) du bassin du Fleuve Rouge (Le Thi Phuong Quynh et al., 2005). En terme de production agricole, d’une part, de consommation de nourriture et de fourrage d’autre part, les sous-basins amont apparaissent comme des systèmes autotrophes, exportant des produits agricoles, tandis que le delta dépend d’importations de biens agricoles. Le bilan des sols agricoles révèle de fortes pertes d’azote, principalement attribuables à la dénitrification dans les rizières, et de phosphore, principalement dues à l’érosion. Le bilan du réseau hydrographique montre une importante rétention/élimination d’azote (de 62 à 77 % dans les basins amont et de 59 % dans le delta), et de phosphore, avec un taux de rétention de plus de 80 % dans le Da et le Lo, à l’aval desquels sont localisés les grands réservoirs (Hoa Binh sur le Da et Thac Ba sur le Lo). L’exportation spécifique estimée à l’exutoire du Fleuve Rouge est estimée à 855 kg.km-².an-1

d’azote total et 325 kg.km-².an-1 de phosphore total. L’azote plutôt que le phosphore semble être l’élément limitant principal de la croissance algale dans les zones côtières influencées par le Fleuve Rouge dans le Golfe du Tonkin.

Une base de données sous SIG a été assemblée à l’échelle du bassin du Fleuve Rouge, avec des couches d’informations renseignant la géomorphologie du bassin, sa lithologie, la météorologie, l’usage du sol et les pratiques agricoles, la population et les rejets d’eau usées domestiques et industrielles. Cette base de données est conforme au format requis par le logiciel SENEQUE/Riverstrahler (Ruelland et al, 2004), une version du modèle Riverstrahler encapsulée dans une interface SIG constituant un outil de modélisation générique et spatialement explicite de la qualité de l’eau à l’échelle des grands réseaux hydrographiques. La première application de ce logiciel au système Fleuve Rouge est décrite et validée sur la base des données acquises lors des suivis mensuels de qualité d’eau à l’exutoire des grands sous-bassins et sur l’axe principal du Fleuve lors des années 2003 et 2004.

Enfin, le modèle a été utilisé pour explorer l’effet, en terme de qualité de l’eau et de fonctionnement biogéochimique de divers scénarios décrivant de possibles changements futurs du bassin du Fleuve Rouge concernant son aménagement hydraulique, l’usage de ses sols et son agriculture, sa population et sa gestion des eaux usées.

Mots clés: Rivière Tropicale, Fleuve Rouge, Vietnam, modèle Riverstrahler/Sénèque, nutriments, cycle de l’azote, du phosphore, de la silice, charge solide.

iv

Résumé

Summary

The Red River (in North Vietnam and South China) covers a watershed area of 156 450 km2

with a total population near 30 million inhabitants. The main branch of the Red River (also

called Yuan, Thao or Hong River) receives two major tributaries, the Da and Lo Rivers, then

forms a large delta before discharging into the Tonkin Bay (South China Sea). The 3 upstream

sub-basins and the Delta area differ widely in population density (from 101 inhab km-2 in the

upstream basins to more than 1000 inhab km-2 in the delta), land use and agricultural

practices.

The general goal of this work is to develop a comprehensive understanding of the

biogeochemical functioning of this sub-tropical regional system, and its control by natural

and anthropogenic processes. The backbone of the work consisted in implementing the

RIVERSTRAHLER Model, previously developed for describing the link between water quality

and human activities in the watershed in the Seine river and other European river systems

(Billen et al., 1994, 1997, 1999, 2005 ; Garnier et al, 1995, 1999,2000, 2002) to the special

case of the Red River system.

The first step of the study consisted in modeling the hydrological regime and the suspended

solid transport of the Red River (Le Thi Phuong Quynh et al., subm). Previous estimates of

its suspended matter loading range from 100 to 170 106 t.yr-1, i.e. from 640 to 1060 t.km-².yr-1.

The strong dependence of suspended solid transport on hydrology results in a large year-to-

year variability. Based on available data on the hydrology over the period 1997-2004, and on

one -year survey of the daily suspended matter of the three main tributaries of the Red River

system in 2003, a simplified modeling approach is established to estimate the mean suspended

loading of the Red River under present conditions. The obtained value is 40 106 t.yr-1,

corresponding to a specific load of 280 t.km-2.yr-1. It reflects a 70% decrease of the total

suspended load since the impoundment of the Hoa Binh and Thac Ba reservoirs in the

1980’ies. The model predicts a further reduction by 20% of the suspended loading of the Red

River with the planned construction of two additional reservoirs. Using measurements of the

total phosphorus content of the suspended material in the different Red River tributaries, we

could estimate the present phosphorus delivery by the Red River as 36 106 kgP yr-1.

As data on nutrient concentration in the Red River drainage network are rather scarce, a

survey of nutrient concentration (N, P, Si, organic carbon and chlorophyll a) at the outlet of

the three main sub-basins, the main branch in the delta and some polluted rivers in the Hanoi

v

Résumé region was carried on at monthly intervals in 2003 and 2004, allowing to define the general

levels of nutrient concentrations in surface water.

In order to examine the degree of human-induced alteration of the nitrogen and phosphorus

cycles at the scale of the watershed, budgets of these elements were established for the soil

and the drainage network of the 4 main sub-basins (Da, Lo, Thao and Delta) of the Red River

(Le Thi Phuong Quynh et al., 2005). In terms of agricultural production, on the one hand, and

consumption of food and feed on the other, the upstream sub-basins are autotrophic systems,

exporting agricultural goods, while the delta is a heterotrophic system, depending on

agricultural goods imports. The budget of the agricultural soils reveals great losses of

nitrogen, mostly attributable to denitrification in rice paddy fields and of phosphorus, mostly

caused by erosion. The budget of the drainage network shows high retention/elimination of

nitrogen (from 62 to 77 % in the upstream basins and 59 % in the delta), and of phosphorus,

with retention rates as high as 80 % in the Da and Lo sub-basins which have large reservoirs

in their downstream course (Hoa Binh on the Da and Thac Ba on the Lo). The total specific

delivery estimated at the outlet of the whole Red River System is 855 kg.km-².y-1 total N and

325 kg.km-².yr-1 total P. Nitrogen rather than phosphorus seems to be the potential limiting

factor of algal growth in the plume of the Red River in Tonkin Bay.

A GIS data base has been assembled at the scale of the whole Red River basin, with layers

documenting geomorphology, lithology, meteorology, land-use and agriculture, population,

domestic and industrial wastewater release, etc. This data base follows the format required

for running the SENEQUE/Riverstrahler software (Ruelland et al, 2004), a version of the

Riverstrahler model encapsulated into a GIS interface in order to build a generic and spatially

explicit water quality modelling tool. The first application of this model to the Red River

system is described and validated with the data acquired by the monthly surveys of water

quality at the outlet of the 3 sub-basins and in the main branch of the Red River during the

years 2003 and 2004.

Finally, the model is used to explore the effect in terms of water quality and biogeochemical

functioning of a variety of scenarios describing possible future changes in the Red River

basin concerning hydrological management, land use and agricultural practices, population

increase and wastewater treatment policy. Key words: tropical river, Red River, Vietnam, Riverstrahler/Seneque model, nutrient budgets, nitrogen, phosphorus, silica cycle, suspended solids.

vi

Résumé

Tãm t¾t

L−u vùc s«ng Hång n»m trªn ®Þa phËn miÒn B¾c ViÖt Nam vµ miÒn Nam Trung Quèc víi diÖn

tÝch toµn l−u vùc kho¶ng 156 450 km2 vµ d©n sè trong toµn l−u vùc ®¹t 30 triÖu ng−êi. Nh¸nh

chÝnh cña s«ng Hång (cßn gäi lµ s«ng Nguyªn, Thao, Çi, Hång) nhËn hai nh¸nh s«ng kh¸c lµ

s«ng §µ vµ s«ng L« t¹i ViÖt tr×, vµ b¾t ®Çu t¹o vïng ®ång b»ng ch©u thæ réng lín tr−íc khi ®æ

ra vÞnh B¾c Bé (biÓn §«ng). Ba tiÓu l−u vùc th−îng nguån vµ tiÓu l−u vùc ®ång b»ng hoµn

toµn kh¸c nhau vÒ mËt ®é d©n sè (tõ 101 ng−êi/km2 t¹i vïng th−îng nguån ®Õn h¬n 1000

ng−êi/km2 t¹i vïng ®ång b»ng ch©u thæ), vÒ t×nh h×nh sö dông ®Êt vµ çc ho¹t ®éng n«ng

nghiÖp trong tiÓu l−u vùc.

Môc tiªu chung cña luËn ¸n lµ ph¸t triÓn sù hiÓu biÕt vÒ çc ho¹t ®éng sinh th¸i ®Þa hãa cña

hÖ thèng b¸n nhiÖt ®íi chÞu t¸c ®éng cña çc qu¸ tr×nh tù nhiªn vµ cña con ng−êi. M« h×nh

RIVERSTRAHLER tr−íc ®©y ®· ®−îc x©y dùng ®Ó m« t¶ mèi quan hÖ gi÷a chÊt l−îng n−íc

vµ ho¹t ®éng cña con ng−êi trong l−u vùc s«ng Seine vµ mét sè l−u vùc s«ng lín ë Ch©u ¢u

(Billen et al., 1994, 1999, 2001; Garnier et al., 1995, 1999, 2002), lÇn ®Çu tiªn ®−îc ¸p dông

cho hÖ thèng s«ng nhiÖt ®íi, s«ng Hång.

B−íc ®Çu tiªn cña luËn ¸n lµ nghiªn cøu chÕ ®é thñy v¨n vµ chuyÓn t¶i hµm l−îng phï sa

trong hÖ thèng s«ng Hång. Çc nghiªn cøu tr−íc ®©y cho r»ng mçi n¨m s«ng Hång chuyÓn t¶i

ra biÓn kho¶ng 100-170tÊn, tøc lµ vµo kho¶ng 640-1060 tÊn/km2/n¨m. Sù phô thuéc m¹nh mÏ

cña hµm l−îng phï sa vµo chÕ ®é thñy v¨n ®· t¹o ra sù kh¸c biÖt râ rÖt vÒ tæng l−îng phï sa

chuyÓn t¶i ra biÓn hµng n¨m. Dùa vµo çc sè liÖu thu thËp ®−îc vÒ chÕ ®é thñy v¨n trong giai

®o¹n 1997-2004 vµ sè liÖu hµng ngµy vÒ hµm l−îng phï sa trong n¨m 2003 t¹i ba nh¸nh chÝnh

cña s«ng Hång, mét m« h×nh ®¬n gi¶n hãa ®· ®−îc thiÕt lËp ®Ó ®¸nh gi¸ t¶i l−îng trung b×nh

vÒ hµm l−îng phï sa víi çc ®iÒu kiÖn hiÖn t¹i. KÕt qu¶ cho thÊy h»ng n¨m s«ng Hång ®æ ra

biÓn kho¶ng 40.106tÊn/n¨m, tøc lµ kho¶ng 280 tÊn/km2/n¨m. §iÒu nµy ph¶n ¸nh 70% tæng

l−îng phï sa ®· bÞ gi¶m tõ khi cã sù vËn hµnh cña hå Hßa B×nh vµ hå Th¸c Bµ vµo nh÷ng n¨m

1980s. KÕt qu¶ dù b¸o cña m« h×nh cho thÊy sÏ cã kho¶ng thªm 20% tæng l−îng phï sa sÏ bÞ

gi¶m khi cã thªm 2 hå chøa n÷a ®i vµo ho¹t ®éng (S¬n La vµ §¹i ThÞ). Sö dông çc phÐp ®¸nh

gi¸ vÒ tæng l−îng phètpho trong phï sa t¹i çc nh¸nh chÝnh kh¸c nhau cña s«ng Hång, cho

thÊy, hiÖn nay mçi n¨m, s«ng Hång chuyÓn ra biÓn kho¶ng 36 106 kgP/n¨m.

Do thiÕu çc d÷ liÖu vÒ hµm l−îng chÊt dinh d−ìng trong m¹ng l−íi s«ng Hång nªn quan tr¾c

hµm l−îng çc chÊt dinh d−ìng (N, P, Si, Cacbon h÷u c¬ vµ chlorophyll a) t¹i çc h¹ nguån

cña ba nh¸nh s«ng chÝnh vµ trªn trôc chÝnh ë vïng ®ång b»ng vµ mét sè s«ng « nhiÔm t¹i Hµ

vii

Résumé néi ®· ®−îc thùc hiÖn hµng th¸ng trong suèt hai n¨m 2003-2004, cho phÐp x¸c ®Þnh møc ®é

chung vÒ chÊt l−îng n−íc s«ng Hång.

Môc tiªu thø hai cña luËn ¸n lµ ®¸nh gi¸ møc ®é ¶nh h−ëng cña con ng−êi trong l−u vùc tíi

chu tr×nh nit¬ vµ phèpho. C©n b»ng dinh d−ìng cña hai nguyªn tè nµy ®−îc thiÕt lËp trong

bèn tiÓu l−u vùc §µ, L«, Thao vµ vïng ®ång b»ng cña hÖ thèng s«ng Hång. VÒ mÆt s¶n xuÊt

n«ng nghiÖp vµ tiªu thô l−¬ng thùc vµ thùc phÈm, çc tiÓu l−u vùc th−îng nguån ®−îc ®¸nh

gi¸ lµ çc hÖ thèng tù d−ìng, tøc lµ cã kh¶ n¨ng xuÊt khÈu hµng n«ng nghiÖp, trong khi vïng

®ång b»ng s«ng Hång l¹i ®−îc ®¸nh gi¸ lµ hÖ thèng dÞ d−ìng, phô thuéc vµo hµng n«ng

nghiÖp nhËp khÈu vµo l−u vùc. Nghiªn cøu vÒ c©n b»ng dinh d−ìng trong vïng ®Êt n«ng

nghiÖp cho thÊy nit¬ bÞ mÊt mét l−îng lín, hÇu hÕt lµ do qu¸ tr×nh khö nirat hãa trong vïng

®Êt trång lóa, trong khi l−îng phètpho mÊt chñ yÕu lµ do qu¸ tr×nh xãi mßn ®Êt. Nghiªn cøu vÒ

c©n b»ng dinh d−ìng trong hÖ thèng thñy v¨n cho thÊy qu¸ tr×nh l−u gi÷/lo¹i bá nit¬ diÔn ra

rÊt m¹nh (tõ 62-77% ë vïng th−îng nguån vµ 59% ë vïng ®ång b»ng) cßn phètpho th× ®−îc

l−u gi÷ rÊt nhiÒu trong çc hå chøa (Hßa B×nh, Th¸c Bµ) trong çc tiÓu l−u vùc s«ng §µ vµ

s«ng L«. T¶i l−îng tæng nit¬ vµ tæng phètpho chuyÓn t¶i ra biÓn cña toµn bé hÖ thèng s«ng

Hång ®−îc −íc tÝnh kho¶ng 855 kg/km²/n¨m vµ 325 kg/km²/n¨m. Nit¬ cã kh¶ n¨ng lµ yÕu tè

giíi h¹n sù ph¸t triÓn cña t¶o t¹i vÞnh B¾c Bé h¬n lµ phètpho.

HÖ d÷ liÖu GIS cña toµn bé l−u vùc s«ng víi çc líp vÒ ®Þa m¹o, ®Þa chÊt, thæ nh−ìng, khÝ

hËu, sö dông ®Êt- çc ho¹t ®éng n«ng nghiÖp, d©n sè, n−íc th¶i sinh ho¹t, c«ng nghiÖp … ®·

®−îc tËp hîp. HÖ d÷ liÖu nµy ®ßi hái d¹ng format ®Æc biÖt ®Ó cã thÓ ch¹y trong phÇn mÒm

SENEQUE/Riverstrahler (Ruelland, 2004), phiªn b¶n cña m« h×nh Riverstrahler) ®· ®−îc gãi

gän d−íi bÒ mÆt GIS ®Ó x©y dùng thµnh mét c«ng cô phÇn mÒm râ rµng thÓ hiÖn tÝnh tæng

qu¸t vµ tÝnh kh«ng gian cho phÐp x¸c ®Þnh chÊt l−îng n−íc. ¸p dông ®Çu tiªn cña m« h×nh

nµy ®èi víi s«ng Hång ®· ®−îc m« t¶ vµ ®¸nh gi¸ víi bé d÷ liÖu ®ßi hái chÊt l−îng n−íc cÇn

®−îc quan tr¾c hµng th¸ng trong suèt hai n¨m liªn tôc 2003-2004 t¹i h¹ l−u cña çc tiÓu l−u

vùc vµ trªn trôc chÝnh cña s«ng Hång.

Cuèi cïng, m« h×nh ®−îc sö dông ®Ó khai th¸c çc ¶nh h−ëng cña con ng−êi trong t−¬ng lai vÒ

çc mÆt qu¶n lý thñy v¨n, sö dông ®Êt vµ çc ho¹t ®éng n«ng nghiÖp, t¨ng d©n sè vµ çc chÝnh

s¸ch xö lý n−íc th¶i trong l−u vùc s«ng Hång ®Õn chÊt l−îng n−íc vµ çc ho¹t ®éng sinh th¸i

cña hÖ thèng s«ng Hång.

Tõ khãa: s«ng nhiÖt ®íi, s«ng Hång, ViÖt Nam, m« h×nh Riverstrahler/Seneque, c©n b»ng

dinh d−ìng, chu tr×nh nit¬, phètpho, silic, chÊt r¾n l¬ löng.

viii

Biogeochemical functioning of the Red River (North Vietnam): Budgets and Modelling

Main contents Introduction 1CHAPTER 1: Site description and major issues 91.1 Geographical presentation of the Red River basin 9 1.2 Geomorphology 11 1.3 Climate and hydrological regime 13 1.4 Hydrology 16 1.5 Social-economical context in the Red River basin and impacts 20 1.6 References 25 CHAPTER 2: General approach and methodology 292.1 Modelling the quality of the Red River hydrographic network 30 2.2 Experimental works 42 2.3 Nutrient budgets 48 2.4 References 50 CHAPTER 3: Hydrological regime and suspended load: observation and modelling 573.1 Introduction 58 3.2 General characteristics of the Red River basin 59 3.3 Hydrological regime of the Red River and its tributaries 66 3.4 Suspended solids loading of the Red River and its tributaries 74 3.5 Future scenarios of suspended solids loading 82 3.6 Conclusions 82 3.7 References 83 CHAPTER 4: Water quality 894.1 Discharge variations 89 4.1 Physical-chemical variables 90 4.3 General pattern of nutrients 94 4.4 Organic matter 102 4.5 Conclusions: water quality in the Red river 105 4.6 References 109 CHAPTER 5: Nutrient budgets (N, P) 1155.1 Introduction 116 5.2 Description of the Red River Basin 117 5.3 The budget of the soil system 120 5.4 Domestic and industrial N, P loadings 130 5.5 The budget of the hydrographical network 133 5.6 Discussions 136 5.7 References 141 CHAPTER 6: Modelling the nutrient transfers in the river system 1496.1 Introduction 1496.2 The Riverstrahler model 1506.3 Geomorphology 1516.4 Hydrology 1536.5 Role of reservoirs 1556.6 Land use and non-point sources of nutrients 1566.7 Wastewater point sources 1586.8 Validation 1606.9 References 169CHAPTER 7: Exploring future trends of nutrient transfers 1737.1 Impacts of new reservoirs constructed in the Red River basin 1737.2 Fast increasing population and impact on water quality 1767.3 Agricultural evolution and its impact on water quality 1807.4 Prospective simulation at the 50 years horizon 1817.5 References 184General conclusions and perspectives 185Contents 189Annex 193

ix

Introduction

Introduction Together with the Mekong, the Red River is the one of two largest rivers in Vietnam (Figure 1). Both play an important role in the economic, cultural and political life of Vietnamese people.

Red-ThaiBinh

Mekong

BangGiang-Kycung

Ma-Chu

Ca

ThuBon

Sesan

Ba

DongNai

Huong (parfum)

Sre pok

Red-ThaiBinh

Mekong

BangGiang-Kycung

Ma-Chu

Ca

ThuBon

Sesan

Ba

DongNai

Huong (parfum)

Sre pok

Red-ThaiBinh

Mekong

BangGiang-Kycung

Ma-Chu

Ca

ThuBon

Sesan

Ba

DongNai

Huong (parfum)

Sre pok

200km

Red-ThaiBinh

Mekong

BangGiang-Kycung

Ma-Chu

Ca

ThuBon

Sesan

Ba

DongNai

Huong (parfum)

Sre pok

Red-ThaiBinh

Mekong

BangGiang-Kycung

Ma-Chu

Ca

ThuBon

Sesan

Ba

DongNai

Huong (parfum)

Sre pok

Red-ThaiBinh

Mekong

BangGiang-Kycung

Ma-Chu

Ca

ThuBon

Sesan

Ba

DongNai

Huong (parfum)

Sre pok

200km200km

Figure 1: Main river basins in Vietnam.

1

Introduction

The Red River brings many advantages with its abundant water resources. In general, the

water sources of the Red River in Vietnam are not only significantly used for irrigation but

also for domestic demand in country-village (Nguyen Ngoc Sinh et al., 1995). The river water

is also utilized for industries in the provinces of its upstream basin, Viet tri and Thai Nguyen

being typical examples of industrial zones. In addition, the water of the Red River is largely

exploited for power generation, since about 8.58 109 KW.h are provided each year by two

dams (the Hoa Binh and the Thac Ba dams) located in the Red River system (Trinh Quang

Hoa, 1998). Furthermore, the extensive network of wide and deep waterways in the Red River

basin represents an interesting potential for providing efficient means of transport, particularly

of heavy bulk cargo. Numerous inter-linked rivers, estuaries and coastal waters in the Red

River basin can be viewed as an excellent scope for the development of inland water-borne

transport facilities (Nguyen Ngoc Sinh et al., 1995). Whereas the benefits of the Red River are

clearly identified in Vietnam, its role in China has not been clearly recognised, perhaps less

important because of its morpho-geography unfavourable to human activities (94% of hills

and mountains in Yunnan province (Chinadata, 1998)).

The Red River has been strongly influenced by human activities in Vietnam. The

environmental pollution has regularly increased in the Red River basin, especially in its delta.

In the upstream of the Red River basin, deforestation (clear cutting or other harvesting

techniques) and land use changes are considered to cause a variety of environmental impacts

such as increased flooding and dramatically increased soil erosion from denuded watershed

exposed to high intensity tropical rainfall (Vo Tri Chung, 1998). In its downstream sector, the

high intensive farming areas attached to the use of nitrogen and phosphorus fertilizers, the

increase of population, the economic industrial development and urbanization as well as the

increased transportation network have strongly affected the water quality of the Red River

system and also influenced the coastal zone ecosystem (Ministry of Science and Technology -

MOSTE-: MOSTE 1998; MOSTE, 1999; Ministry of Environment and Natural Resources -

MONRE-, 2003).

The main objective of this Ph-D thesis, realized in a cooperative research program, was to

develop a comprehensive understanding of the linkage between land use and human activities

in the watershed in order to quantify the water quality and the transfer of nutrients (N, P, Si)

in the Red River drainage network (Vietnam and China). The mathematic model that has been

utilized for the Red River to establish this linkage is the RIVERSTRAHLER model. This has

been firstly developed for the Seine River (Billen et al., 1994; Garnier et al., 1995; Billen and

Garnier, 1999; Garnier et al., 1999), and then for several large European rivers (the Danube:

Garnier et al., 2002; the Mosel: Garnier et al., 1999; the Scheldt: Billen et al., 2005; the Rhine

2

Introduction and the Loire: Garnier et al., 1997) to address the questions of organic pollution and oxygen

balance, nutrient contamination and related eutrophication, transfer and retention in the whole

basin. Moreover, this model would allow establishing the diagnostic of nutrient balance

(N:P:Si ratios), a key for controlling the eutrophication problem not only in the drainage

network but also at the coastal zone (Billen et al., 1985; Billen et al., 1997; Garnier and

Billen, 2002; Cugier et al., 2005). On the point of view of basic research, such an ecological

model has been applied to a sub-tropical river system for the first time, an approach devoted

to enlarge our knowledge on the ecological functioning of river ecosystem. Regarding the

management aspects, this study is also expected to serve as a guide for planning

environmental decisions at both regional and local scales. We implemented the

RIVERSTRAHLER model for the recent period of 8 years (from 1997 to 2004).

This work was undertaken in the framework of the ESPOIR on WATER project

aiming at identifying the water quality controls and at developing new processes for water

treatment. This three-year project (2001-2004) was supported by the activities of scientific

cooperation between different Vietnamese laboratories of the VAST (the Vietnamese

Academy of Sciences and Technology) and the French laboratories of CNRS (The French

National Centre for Scientific Research). Although this programme focused on the study on

water pollution and water treatment of urban rivers surrounding Hanoi, i.e. the Nhue-Tolich

river system located in the Red River delta, a special interest was given to the upstream

drainage network of the Red River, the Nhue river being one of diverted branched of the Red

River, upstream Hanoï (Figure 2). The Nhue receives directly the Tolich River draining Hanoï

(about 3.5 million inhabitants) therefore it is seriously polluted by the domestic and industrial

wastewater. It is important to note that Hanoi is equipped neither for domestic wastewater

collection and treatments nor for treatment systems of industrial wastewater; consequently the

Tolich River is extremely polluted and this pollution strongly impacts on water quality of the

Nhue River. Beside the Hanoï domestic and industrial pollution, the Nhue is also affected by

agricultural (irrigation in rice field and vegetation culture) and aquacultural (fish culture)

activities. The Nhue-Tolich hydrosystem is typically representative of the anthropogenic

rivers in the Red River Delta. As the Nhue River is supplied by the major branch of the Red

River through the Lien Mac dam, immediately upstream of Hanoi city (Figure 1), it was not

out of the scope of the programme to obtain a general knowledge of the quality of the Red

River, which constitutes the upstream limit condition of the Nhue River. A better regulation

of the inputs of water from the Red River to the Nhue River is indeed one of the possible

measures that can be proposed to improve the water quality of the Nhue River. Thus, although

the present study does not focuses on the small polluted urban rivers of the delta, a dialogue

3

Introduction

will now be possible between the model we developed for the Red River, and the one

developed in parallel in the framework of the ESPOIR programme on the special case of the

Nhue (Trinh Anh Duc, 2003).

Duong R.Red R.

Day R.

Ba Lat

Tra Ly

HaiphongD

ayR

.

Ninh C

o

Hanoï

Son Tay

Nhue R.

Thao R.

Da

R.

Lo R.

Boî R.

Luc Nam R.

Tonkin Bay

Hoa Binh

Duong R.Red R.

Day R.

Ba Lat

Tra Ly

HaiphongD

ayR

.

Ninh C

o

Hanoï

Son Tay

Nhue R.

Thao R.

Da

R.

Lo R.

Boî R.

Luc Nam R.

Tonkin Bay

Hoa BinhTolichR

.

Duong R.Red R.

Day R.

Ba Lat

Tra Ly

HaiphongD

ayR

.

Ninh C

o

Hanoï

Son Tay

Nhue R.

Thao R.

Da

R.

Lo R.

Boî R.

Luc Nam R.

Tonkin Bay

Hoa Binh

Duong R.Red R.

Day R.

Ba Lat

Tra Ly

HaiphongD

ayR

.

Ninh C

o

Hanoï

Son Tay

Nhue R.

Thao R.

Da

R.

Lo R.

Boî R.

Luc Nam R.

Tonkin Bay

Hoa BinhTolichR

.

Duong R.Red R.

Day R.

Ba Lat

Tra Ly

HaiphongD

ayR

.

Ninh C

o

Hanoï

Son Tay

Nhue R.

Thao R.

Da

R.

Lo R.

Boî R.

Luc Nam R.

Tonkin Bay

Hoa Binh

Duong R.Red R.

Day R.

Ba Lat

Tra Ly

HaiphongD

ayR

.

Ninh C

o

Hanoï

Son Tay

Nhue R.

Thao R.

Da

R.

Lo R.

Boî R.

Luc Nam R.

Tonkin Bay

Hoa BinhTolichR

.

Figure 2: Schematic representation of the Red River and its connections to the Nhue-Tolich

system.

This Ph-D thesis contains 7 chapters, several of them under the form of scientific papers

already published or submitted.

Chapter 1 is devoted to a general presentation of the Red River and its watershed, oriented

towards the construction of the model, the data required for the modelling approach being

physical constraints such as the geomorphology, geology and lithology and also of hydro-

meteorological nature, i.e. temperature, rainfall and hydrology.

Chapter 2 presents the general approaches and methodologies appropriate for the study of a

large regional system like the Red River basin. The general principles of the Riverstrahler

model, which has structured the whole study, are presented first. The experimental work that

was necessary to document the model regarding point and diffuse sources, as well as to

validate the modelling results, is then presented in this chapter. Indeed, whereas we have been

able to gather the data presented in chapter 1 from literature, or internet websites, water

4

Introduction quality data in the Red River basin are scarce. Sampling campaigns were therefore realized

biological and chemical analyses were performed in the Vietnamese laboratory INPC

(VAST), after several trainings and inter-comparison have been organised with the French

Sisyphe laboratory (UMR 7619, CNRS and University Paris VI). The sampling strategies and

the methods used for these campaigns are described in this chapter. Lastly, the principles of

regional nutrient budget calculations, which offer a useful way of summarizing the overall

biogeochemical functioning of a regional system as well as of testing the coherency of the

data collected, are presented in this chapter.

Chapter 3 focuses on the modelling of the hydrology of the hydrographical network and on

the transport of suspended solid in the Red River basin. Daily meteorological and discharge

data have been analysed for a period of 8 years (1997-2004) with the simplified hydrological

model used as a part of the RIVERSTRAHLER MODEL. In addition, we have analysed the

behaviour of suspended solids in the drainage network in the context of the recent and future

large dam constructions. This chapter constitutes a scientific paper submitted in the Journal of

Hydrology.

The results of water quality observation in the rivers of the Red River drainage are reported in

Chapter 4. This chapter mentions the experimental results obtained in both INPC and

Sisyphe laboratories on water quality at the outlet of the three main sub-basins and in the

main branch of the Red River system in the period from 2002 to 2004. A comparison is made

with the data obtained in parallel on the much more polluted Nhue and ToLich rivers.

The establishment of nutrient budgets in the 4 sub-basins of the Red River is reported in

Chapter 5. In this part, nutrient budgets have been calculated using many statistical sources

within the Red River basin, and our own measurements in the hydrographic network. For the

first time, nutrient budgets were established for the agricultural soils using an agronomical

point of view and nutrient transfers calculated in the drainage network. This work is the

material of a paper published in the Journal of Global Biogeochemical Cycles.

The modelling of nutrient transport in the rivers of the Red River system is reported in

Chapter 6. This chapter describes how the Riverstrahler model takes into account the various

constraints to the drainage network functioning, and how the corresponding information has

been gathered for the special case of the Red River watershed. The results of the application

of the Seneque/Riverstrahler software to the Red River system are presented to validate the

model and illustrate its capabilities. This part will be submitted as a paper to the Journal of

Biogeochemistry.

5

Introduction

Lastly, in Chapter 7, we discuss the scenarios aiming to explore future conditions that could

be found in the Red River basin taking into account socio-economical trends observed and

new plans, for a rehabilitation of impacted systems of the urbanised areas, such as the delta,

but also to avoid ecosystem damage in zones of still good ecological status. A main objective

is to demonstrate that the tools implemented during this Ph-D thesis can be utilisable in

Vietnam to test scenarios for management purposes of human impacts in the watershed.

Explorations by the model such as rapid increase in population, reservoir construction in the

upstream basin of the Da and the Lo Rivers, are all subjects that are discussed in this chapter.

This part is also intended to form the basic material of a paper to be submitted to a scientific

journal.

The Conclusions stress the usefulness of our modelling approach as a framework to gather

pertinent information on a regional territory and to test the coherency of the data available at

this regional scale. We will defend the view that this approach, tested here on the Red River

system, can be extended for improving our knowledge on other poorly documented river

systems of the world.

References

Billen G., Somville M., DeBecker E. and Servais P., 1985. A nitrogen budget of the Scheldt

hydrographic basin. Neth J. Sea Res., 19: 223-230.

Billen G., Garnier J. and Hanset P., 1994. Modelling phytoplankton development in whole drainage networks: The RIVERSTRAHLER model applied to the Seine river system. Hydrobiologia, 289: 119-137.

Billen G. and Garnier J., 1997. The Phison River plume: coastal eutrophication in response to change in land use and water management in the watershed, Aquat. Microb Ecol., 13: 3-17.

Billen G. and Garnier J., 1999. Nitrogen transfer through the Seine drainage network: a budget based on the application of the RIVERSTRAHLER Model. Hydrobiologia, 410: 139-150.

Billen G., Garnier J. and Rousseau V., 2005. Nutrient fluxes and water quality in the drainage network of the Scheldt basin over the last 50 years. Hydrobiologia (in press).

Chinadata 1998. Statistical yearbook of Yunnan, Vol. 1997, Vol. 1998, Vol. 2000 China Statistical Publishing House, (Basic Information of Yunnan, China). (http://chinadatacenter.org)

Cugier Ph., Billen G., Guillaud J.F., Garnier J. and Ménesguen A., 2005. Modelling the eutrophication of the Seine Bight (France) under historical, present and future riverine nutrient loading. J. Hydrol. 304: 381-396

6

Introduction Garnier J., Billen G. and Coste M., 1995. Seasonal succession of diatoms and chlorophyecae

in the drainage network of the River Seine: Observations and modelling. Limnology. and Oceanography, 40: 750-765.

Garnier J., Billen G. and Hannon E., 1997. Biogeochemical Nutrient Cycling in Large River Systems (Binoculars). Final Technical Report. EC Environment Programme (ref PL932037). 33 pp + Annexes.

Garnier J., Leporcq B., Sanchez N. and Philippon X., 1999. Biogeochemical budgets in three large reservoirs of the Seine basin (Marne, Seine and Aube reservoirs). Biogeochemistry, 47: 119-146.

Garnier J., Billen G. and Palfner L., 1999. Understanding the oxygen budget and related ecological processes in the river Mosel: the Riverstrahler approach. Man and Rivers System. J. G. J. M. M. Hydrobiologia. Netherland, 1999 Kluwer Academic Publishers. 410: 151-166.

Garnier J., Billen G., Hannon E., Fonbonne S., Videnina Y. and Soulie M., 2002. Modeling transfer and retention of nutrients in the drainage network of the Danube River. Estuarine, Coastal and Shelf Science, 54: 285-308.

MONRE, 2003. Report on water environment monitor in Vietnam in 2003. In “Studies on Vietnam environmental statement in 2003”. Vietnam Ministry of Environment and Natural Resources, 150pp., Hanoi.

MOSTE, 1998. Documentation on the Red River Delta (1997-1998), Ministry of Science, Technology and Environment of Vietnam, Scientific and Technical Publisher, 214pp., Hanoi.

MOSTE, 1999. Environmental statement in Vietnam in the years 1990s. Ministry of Science, Technology and Environment of Vietnam, Scientific and Technical Publisher, 219pp., Hanoi

Nguyen Ngoc Sinh, Hua Chien Thang, Nguyen Chu Hoi, Nguyen Van Tien, Lang Van Ken, Pham Van Ninh and Nguyen Vu Trong., 1995. Case study report on Red River Delta in Vietnam - Project on integrated management and conservation of near shore coastal and marine areas in East Asia region (EAS-35) United Nations Environment program. Regional coordinating for the East Seas (ESA/RCU), report, 78pp., U.N. Environ. Programme, Nairobi.

Trinh Anh Duc, 2003. Etude de la qualité des eaux d’un hydrosystème fluvial urbain autour

de Hanoi (Vietnam); suivi expérimental et modélisation. Thèse de doctorat d'Etat de

l'Université Joseph Fourrier, Grenoble 1, France and Vietnam Academy of Science and

Technology (VAST). 265 p.

Trinh Quang Hoa, 1998. Water balance for purpose of socio-economic development in the Red River delta. Proceedings of International Conference on Economic development and environmental protection of the Yuan-Red River watershed, Hanoi 4th-5th Mar.

Vo Tri Chung, 1998. Forests on the Red River basin, Vietnam. Proceedings of International Conference of Economic development and environmental protection in the Yuan-Red River watershed, Hanoi 4th-5th March.

7

Introduction

Articles in press, submitted or to be submitted in the framework of this PhD thesis:

Le, Thi Phuong Quynh, Billen, G., Garnier, J., Théry, S., Fézard, C. and Chau, Van Minh (2005). Nutrient (N, P) budgets for the Red River basin (Vietnam and China). Journal of Global Biogeochemical cycles. Vol 19, GB2022, doi 10.1029/2004GB002405.

Le Thi Phuong Quynh, , Garnier J., Billen G., Thery S. and Chau V. M., 2005. Hydrological regime and suspended matter flux of the Red River system (Vietnam): Observations and modelling. Journal of Hydrology (submitted).

Le Thi Phuong Quynh, Billen G., Garnier J., Thery S., Ruelland D. and Chau V. M., 2005. Nutrient transfers through the Red River basin (Vietnam): Observations and modelling. Biogeochemistry (in prep.).

8

Site description and major issues

CHAPITRE 1

Site Description and Major Issues

1.1 Geographical presentation of the Red River basin

The Red River basin (Figure 1.1) is located in the South-East Asia, from the latitude 20°00 to

25°30 North and from the longitude 100°00 to 107°10 East. The Red River is bordered by the

Truong Giang and the Chau Giang River basins (in China) in the North, by the Langcang

River (Mekong) basin in the West, by the Ma River basin (in Vietnam) in the South, and by

the Thai Binh River and the Tonkin Bay in the East (Nguyen Ngoc Sinh et al., 1995). In this

sub-tropical region, where chemical and mechanical erosion are among the highest of the

world (500 mm/1000 years), large rivers transport considerable amount of suspended solids

(Meybeck et al., 1989; Dupré et al., 2002). The climate is of monsoon type, with summer

dramatic inundations. The biggest floods in the Red River delta occurred in 1913, 1915, 1945

and 1971 when the serious dyke breakage happened in many places. The floods in 1971

submerged 250140 ha and affected about 2.71 million people, damaged 7 millions tons of

paddy (To Trung Nghia, 2000). In the Mekong delta, dramatic floods occurred in 2000 and

2001, affecting about 900 Vietnamese people. In the Red River delta, dikes dating back to the

early 1800s are maintained to protect the population in the delta area (To Trung Nghia, 2000).

Hoa Binh reservoir

Thac Ba reservoir

Hoa Binh reservoir

Thac Ba reservoir

Figure 1.1: The Red River and its watershed

9


The Red River (or Thao River) originates in the mountainous region of South China

(100° 00’20’’ longitude, 25°30’10’ latitude), at the foot of the Himalaya mountains (Nguyen

Huu Khai and Nguyen Van Tuan 2001) in Dali city, in the Yunnan province, between the

Langcang and Jinsha river watersheds (Figure 1.2). The altitude of the source is about 3000

m. In the Chinese part, the Red River is named the Yuan River (the Yuanjiang or Yuanjiang-

Hong), located beside some other important rivers in Southeast Asia such as Nujiang-

Salween, Nanpan, Jinsha, Lancang - Mekong, Dulong - Irrawaddy rivers. All of them are

originated from (eg. Yuanjiang and Nanpan) or go through the Yunnan province, and are

important pathways between China and Southeast and South Asia (Chinadata, 1998). In

average, the Red River has smaller discharge than other biggest rivers in South Asia (table

1.1).

Table 1.1: Characteristics of some largest rivers in South and Southeast Asia River Drainage area

km2 Water discharge

m3.s -1 References

Pearl (Zhujiang) 442585 10033 Zhang J., 1996

Yangtze (Changjiang) 1808500 24443 Zhang J., 1996

Mekong (Langcang) 803000 11000 Meybeck, 1989

Irrawaddy (Dulong) 430000 13600 Meybeck, 1989

Red River (1997-2004) 151448 3577 This study

The Red River is known as the “six-head

river” that enters into Vietnam at Lao Cai

province with its name of Thao (or Cai, or

Hong) River (Dang Anh Tuan, 2000). The

name of the Red or Hong River originates

from its reddish-brown colour water, due to

the transport of large quantities of

sediments, rich in iron dioxide. It runs

directly through Yunnan, Lao Cai, Yen Bai,

Phu Tho, Hanoi, Hung Yen and Thai Binh

provinces forming the Red River delta

before flowing into the China Sea (Gulf of

Tonkin) through four distributaries called,

Ba Lat (106° 32’10’’ longitude and

20°20’00’ latitude), Lach Gia, Tra Ly, and

Day (Dang Anh Tuan, 2000). Figure 1.2: The source of Red River in China

Dali

Yuanjiang

Red River

Mekong River

CHINA

River

Dali

Yuan River

Red River

Mekong River

Dali

Yuanjiang

Red River

Mekong River

River

Dali Dali

Yuanjiang

Red River

Mekong River

Yuanjiang

Red River

Mekong River

CHINA

RiverRiver

Dali Dali

Yuan River

Red River

Mekong River

Yuan River

Red River

Mekong River

10

Site description and major issues The Thao River receives two major tributaries: the Da (or the Black) River on the right bank

and the Lo (the Clear) River on the left bank.

The source of the Da is also located in the Yunnan province. It flows directly through

Yunnan, Lai Chau, Son La, Hoa Binh and Ha Tay provinces before reaching to the Thao

River at Ha Nong district, in Viet tri city (Figure 1.1). The Da River originates from a region

with a mean elevation of 2000m (Nguyen Huu Khai and Nguyen Van Tuan, 2001).

The Lo River also originates from in China and joins with the main branch at Viet Tri city.

The elevation of the source of the Lo River is 1100m (Nguyen Huu Khai and Nguyen Van

Tuan, 2001).

From the Viet tri confluence point to the estuary, the Thao River is named the Red (or Hong)

River.

1.2 Geomorphology

The area of the whole Red River basin takes different values depending on the authors,

because of the different ways of estimating, within the delta, the complex hydrographic

network of the Red-ThaiBinh River system, i.e., the Red River delta from the ThaiBinh river

network. In this study, the total area of the Red River catchment was first estimated to 156

451 km2. A subsequent analysis based on the treatment of the digital elevation model of the

NASA (global SRTM 3” resolution) lead to a slightly different watershed area of 142 950

km². Within the Red River watershed area, 47.9% is in Chinese (Chinadata, 1998), 51.2% in

Vietnamese (MOSTE, 1997) and 0.9% is Laotian territories.

In the Yunnan province (394000 km2, 4.1% of China), the Red River watershed occupies

about 20 % of the area of the province. It is important to note this proportion that will be used

below, to calculate figures related to the Red River basin, when we only obtained information

for the whole Yunnan.

The relief of the Red River basin that much varies from headwaters to the downstream areas

can be divided into three sections (figure 1.3).

i) In the Chinese part, mountainous landscapes dominate. Mean elevation of the

Yunnan province is at about 2000 m, but maximal elevation reaches 6740 m and the minimal

one is of 76.4 m (Chinadata, 2000). Within the total Yunnan province area, about 84% are

rugged mountains; 10% are highlands and hills; and only 6% are lowland and valleys

(Chinadata, 2000). Mountain areas are tectonically active and unstable, and this, combined

with intense rainfall, causes high erosion (Fullen et al., 1998). In Eastern Yunnan, the Red

11


River valley is surrounded by the Karst Plateau, composed of red stratum, called the Central

Yunnan Red Soil Plateau. Sandstones or mudstones of mixed colors including red, purple,

bluish gray, yellow and gray-white are widely exposed to erosion giving the red color water

of this river (Chinadata, 1998).

ii) In the Vietnamese part, about more than half of the Red River basin lies in the

mountainous region. The East-North Vietnam area is dominated by the Hoang Lien Son

Mountain with the highest pick as Phanxipan (3143m) in Sapa town, in LaoCai province.

Some other high mountains also locate in this area. In the North Vietnam, soils are mostly

(70%) grey and alluvial soils (MOSTE, 1997). Red soil occupies only 7% and rugged

mountains about 10%.

iii) The delta, the third section of the Red River basin, covers a very flat and low land,

elevation ranging from 0.4 to 12 m above sea level, with 36% lying below 2m (Dang Quang

Tinh, 2001). There are however higher areas in the delta which take the form of steep

limestone karsts, type formations which occur as isolated hills in Ninh Binh, Nam Ha, Ha

Tay, Ha Bac, Quang Ninh provinces including the famous Ha Long Bay (Nguyen Ngoc Sinh

et al., 1995).

Figure 1.3: False perspective view of the relief of the Red River basin (viewed from the

delta mouth), generated by treatment of a digital elevation model (global SRTM 3”

resolution, NASA, www:\\NASA.org)

12

Site description and major issues Considering the 3 main watersheds of the whole Red River catchment, the mean elevations

are rather similar for the Da river basin (965 m), the Lo River region (884m) and the Thao

river watershed (647m) (Nguyen Viet Pho, 1984).

The total length of the Red River course is about of 1126 km from the source to the mouth, of

which 556 km is in the Vietnamese territory (To Trung Nghia, 2000). The mean slope of the

whole Red River basin is of 29.9% (Nguyen Huu Khai and Nguyen Van Tuan, 2001).

The Da and Lo rivers respectively have its length of 1010km (560 km in Vietnam) and of 470

km (275 km in Vietnam). Note that the Red River course can be split into the Thao (about 910

km) and the Hong River (delta, about 216 km, (Nguyen Viet Pho, 1984)).

1.3 Climate and hydrological regime

The climate in the Red River basin, of sub-tropical East Asia monsoon type, is controlled by

the North East monsoon in winter and South West monsoon in summer. The climate is

characterized by two distinct seasons. The rainy season lasts from May to October and the dry

season covers the period from November to the next April.

During the study, we have gathered the meteorological data during the period from 1997 to

2004: daily rainfall, monthly temperature, monthly humidity, and monthly solar radiation,

obtained from 13 meteorological stations in the Red River basin (see Figure 1.4). The

evapotranspiration (ETP) data have been calculated by using Turc’s formula (Turc, 1961),

based on monthly temperature and sunshine duration data obtained from the respective

meteorological stations (see chapter 3).

The climate of the Red River basin is well described in the chapter 3. In the period from 1997

and 2004, the annual mean temperature, humidity, annual rainfall and ETP data in the

Vietnamese part are higher than values obtained in the China part.

The annual mean temperature varied from 14 to 27 °C in the whole Red River basin. The

monthly temperature varied from 14 to 25 °C in the upstream sub-basins and is higher in the

delta region (16 to 28°C) (IMH, 1997-2004).

As other tropical river basins, the humidity always remains in high level. In the whole Red

River basin, humidity averaged from 82 to 84% all over the year in the Vietnamese part of the

basin (IMH 1997-2004), while it was lower, about of 67÷70 %, in the Chinese part

(Chinadata, 1998; Chinadata 2000).

The rainy season cumulates 85 – 90% of the total annual rainfall in the Red River catchment.

It is also interesting to note that July and August are two months with the highest incidence of

13


typhoons in the Red River. The mean annual rainfall is 1587 mm in the whole Red River

basin.

N

Kunming

Ha Noi

Thai Binh

Nam Dinh

Ha Giang

Son La

Lai Chau

Sa Pa

Tuyen Quang

Hoa Binh

Phu Tho

Yen Bai Lao Cai

0 20 50 70 100km

meteorological station

hydrological station

Son Tay

N

Kunming

Ha Noi

Thai Binh

Nam Dinh

Ha Giang

Son La

Lai Chau

Sa Pa

Tuyen Quang

Hoa Binh

Phu Tho

Yen Bai Lao Cai

0 20 50 70 100km

meteorological station

hydrological station

Son Tay

Figure 1.4: Meteorological and hydrological stations in the Red River basin

The climate of the Red River, characterized by a monsoon sub-tropical regime, confers the

typical hydrologic regime characterized by large runoff during summer and low runoff during

winter. Figure 1.5, constructed with data borrowed from Guilcher (1965) and other sources,

compares the climatic and hydrologic behaviour of the Red River with that of Arctic,

Mediterranean and Temperate Oceanic regions of the world. Both Mediterranean and

Temperate oceanic types of rivers have their maximum discharge during winter, because

evapotranspiration is the lowest in this season. Except for arctic rivers, which are

characterized by large discharge in spring due to snow melt at that time of the year (figure

1.5), the sub-tropical rivers are the only ones characterized by highest specific discharge

during the period of occurrence of highest radiative energy and temperature.

14


Arctic

0

50

100

150

200

250

300

J F M A M J J A S O N D

rain

or

etr.

, mm

/mon

th

-20-15-10-505101520

tem

p., °

C

rainetrtemp

Mediterranean

0

50

100

150

200

250

300

J F M A M J J A S O N Dra

in. o

r et

r., m

m/m

onth

0

5

10

15

20

25

30

tem

p., °

C

Oceanic Temperate

0

50

100

150

200

250

300


rain

. or

etr.

, mm

/mon

th

0

5

10

15

20

25

30

tem

p., °

C

Monsoon tropical

0

50

100

150

200

250

300


rain

or

etr.

, mm

/mon

th

0

5

10

15

20

25

30

tem

p., °

C

Red R., Vietnam

01020304050607080


spec

. dis

ch.,

l.s-1

.km

-²Seine R.

0

5

10

15


spec

. dis

ch.,

l.s-1

.km

- ²

Ardèche R.

0

5

10

15


spec

. dis

ch.,

l.s-1

.km

- ²Kalix R.

010203040506070


spec

. dis

ch.,

l.s-1

.km

- ²

spec. disch.

Figure 1.5: Climatic regime: rainfall (rain: mm/month); evapotranspiration (etr.: mm/month) and temperature (temp.: 0C) and specific discharge (spec.

disch.:L.s-1.km-2) of some rivers located in the different climatic regimes in the world. (Guilcher, 1965)

15


1.4. Hydrology

1.4.1 Hydrology in Vietnam

1.4.1.1 Surface water in Vietnam

Vietnam has an abundant water resource with a dense river network, of which 2360 rivers

have a length of more than 10 km (Nguyen Viet Pho, 1984). Within these rivers, eight have

large basins with a catchments area of 10000 km2 or more (table 1.2). The drainage density

varies from 0.25 to 1.94 km.km-2. Along the Vietnamese coastline (3260 km), about 20 km

separate the various river mouths. With an annual rainfall average in Vietnam of 1957 mm

and an annual evaporation of 983 mm, the total runoff of Vietnam is about 880.109 m3.y-1

(SEAMCAP, 2001).

Table 1.2: Major rivers and their watersheds in Vietnam (SEAMCAP, 2001)

Watershed area, km2

Mean annual discharge

Population in Vietnam (in 1995)

River

total area area in Vietnam

total, 109m3

% of the total Vietnam river

discharge Inhabitants

(106) Pop. Dens*, inhab.km-2

Mekong 795000 72000 520.6 59.2 16.8 233

Red-ThaiBinh 169000 86660 137.0 15.6 24.2 279

DongNai 42655 36261 30.6 3.5 10.2 282

Ma 28490 17810 20.1 2.3 2.9 163

Ca 27200 17730 24.2 2.7 3.1 175

Ba 13900 13900 10.4 1.2 0.9 61

Bang Giang-KyCung 12880 11220 8.9 1.0 1.0 91

ThuBon 10496 10496 19.3 2.2 0.9 82

*: population density (Pop. Dens*) in inhabitants.km-2

Note that the Red-Thai Binh and Mekong rivers carry 74.8 % of the total surface water

resource in Vietnam, while each of the other basins represents only 1÷3 % (table 1.2).

About two thirds of the water resources originate from catchment in neighbour countries.

Vietnam is the lower country for both the Mekong and the Red Rivers, and depends on the

water resource management and decisions taken in the upstream countries. This might

amplify the highly variable seasonal and geographical distribution of water (droughts in the

dry season and flood during in the monsoons) (MONRE, 2003).

16

Site description and major issues Most dams and reservoirs in Vietnam have been constructed for multipurpose, including flood

control, irrigation, hydropower, water supply and other flow management. There are about

3600 reservoirs of various size of which less than 15% have a capacity above 1 million m3 or

a depth higher than 10 m). Some biggest reservoirs in Vietnam are presented in table 1.3.

Sedimentation from erosion within the watersheds leads to a decline in the reservoir capacity:

most reservoirs and dams were constructed since 20 - 30 years and some of them have lost up

to 70-30 % of their original capacities (MONRE, 2003).

Surface water is utilized for agricultural irrigation, aquaculture, domestic supply, livestock,

industry and service. In Vietnam, agriculture remains the largest consumer of water (about

82% of the total demand). Industry (6.5% of the total demand) and domestic use (about 2.5%

of the total demand) are however rising with population growth and economic development

(MONRE, 2003).

Table 1.3: Major reservoirs in Vietnam (MONRE, 2003).

Reservoir Catchment km2

Volume km3

Hydropower MW

*Hoa Binh 51700 9450 1920

* Thac Ba 6100 2940 108

Tri An 14600 2760 420

Dau Tieng 2700 1580 -

Thac Mo 2200 1370 150

Yaly 7455 1037 720

Phu Ninh 235 414 -

Song Hinh 772 357 66

Ke Go 223 345 -

* The reservoirs within the Red river basin

1.4.1.2 Groundwater in Vietnam

The groundwater resource in Vietnam is abundant, with a total potential exploitable reserve of

the aquifer with the whole country estimated at nearly 60 km3.y-1 (MONRE, 2003). Over

50 % of these resources are in the central part, about 40 % in the north and 10 % in the south

of Vietnam. A large amount of water is stored in unconsolidated alluvial sand and gravel

geological formations found in plains and valleys. A substantial part of these resources

(estimated at 35 km3.y-1) returns to the rivers as base flow, underground water being an

important river flow component in the dry season (MONRE, 2003). Groundwater is exploited

17


for irrigation of crash crops or for drinking water but less than 5% of the total underground

reserves is exploited for the whole country (MONRE, 2003).

1.4.2. Hydrology of the Red River

1.4.2.1 Drainage density

Within the Red River basin, the drainage density is quite high, in the range of 0.5 to 1.5

km.km-2 with about 500 streams and rivers (Le Bac Huynh, 1997).

In the upstream basin of the Red River, the Yunnan province territory is a vast land with

plentiful rivers: over 600 rivers and lakes (Chinadata, 1998).

The drainage density is much more complex in the delta areas, ranges from 0.7 to 1 km.km-2.

A dense system of irrigation channels for agricultural activities adds to the natural complexity

of the system. Trinh Quang Hoa (1998) reports that 30 main irrigation channels have been

constructed in the Red-Thai Binh river delta providing water for 735370 ha. Tran Duc Thanh

et al. (2004) mentions that the demand for irrigation water in dry season ranges from 25 to

50% of the river discharge in the Red River delta.

For this work, the hydrographic network of the Red River and its elementary watersheds,

constitute the first and basic layer of the GIS database. The details for the construction of the

hydrographic network representation are described in chapter 6. An important work has been

realized to geo-reference all the Vietnamese streams of the drainage network and to connect

them towards the direction of water flux. This network was then simplified, in order to adjust

the resolution to the one available for the Chinese part of the basin, finally producing the

simplified map of figure 1.6.

100 km100 km100 km

Figure 1.6: Drainage

network and elementary

watersheds of the Red

River basin, obtained by

treatment of the digital

elevation model of the

NASA (see chapter 6).

18

Site description and major issues 1.4.2.2. Water flows

The daily discharge data at the outlet of the 3 main branches and in the delta of the Red River

in the period from 1997 to 2004 were obtained at 6 hydrological stations from the Vietnamese

Ministry of Environment and Natural Resources (MONRE): the Hoa Binh station (in Hoa

Binh province) for the Da outlet; the Vu Quang station (in Phu Tho province) for the Lo

outlet; the Yen Bai station (in Yen Bai city) for the Thao outlet, and two stations along the

downstream course of the Hong river: Son Tay (in Ha Tay province) and Hanoi (in Hanoi

city) (Figure 1.4). In the period 1997-2003, the mean annual discharge of the main branch at

Son Tay station was of 3577 m3.s-1 (MONRE 1997-2004).

Whereas the flow of the Red river basin including the three main branches does not vary

greatly from year to year (within the period from 1997 to 2004), it largely varies seasonally.

The seasonal distribution of the water within the Red River basin depends on unevenly

distributed monsoon rainfalls. Such high variations combined with limited storage capacity

and insufficient flood control infrastructure result in devastating floods in the wet season and

damaging extreme low flows in the dry season.

According to long term hydrological data series, the annual discharge volume of the Red

River is around 130 109 m3 (a mean discharge of approximately 3600 m3.s-1 at Son Tay. This

accounts for about 15% of the total runoff for the whole Vietnam (Nguyen Ngoc Sinh et al.,

1995).

1.4.2.3. Reservoirs

The Hoa Binh and Thac Ba reservoirs are the two largest dam-reservoirs located in the Red

River basin (figure 1.1). Similarly to most reservoirs in Vietnam, they have been constructed

as multi-purpose reservoirs: for power generation, flood control, agricultural irrigation,

fishery and tourism. The Hoa Binh Reservoir, damming the Da River, is the largest reservoir

in North Vietnam (table 1.3). These two reservoirs on the Da and the Lo rivers represent a

storage capacity of nearly 7 km3, but only 6 percent of the mean annual flows of the Red

River (Vu Van Tuan, 2002). However, the influence of the Hoa Binh and Thac Ba reservoirs

on the flow and the suspended solid flux at Son Tay station (main branch of the Red River) is

not negligible. The detail about the hydrology and suspended solid transfers will be showed

below, in the chapter 3.

19


Table 1.4: Some major characteristics of the 3 main sub-basins (Da, Lo, Thao) of the Red

River system and its delta area.

Sub-basin Da Lo Thao Delta

Catchment area, km² 51285 34559 61169 9435

Average Discharge*, m3s-1 (max ; min)

1925

(11100 ; 283

973

(8340 ; 165)

743

(6210 ; 146)

3290

(20900 ; 555)

Reservoir, 109 m3 3.9-9.5 0.78-2.94 - -

Population density 101 132 150 1173

*average discharge for the period from the daily data from 1997 to 2004. Maximum and minimum values during

the same period between brackets.

1.5 Social-economical context in the Red River basin and impacts

Due to the high population density in the whole Red River (193 inhab.km-2) and mainly in the

delta, the impact by human activities is necessarily important. Contrarily to other densely

populated countries in Western Europe or North America, human influences on water quality

have not been well studied in South East Asia, including the Vietnam. In fact, until now the

major concerns to environmental problems are the damage caused by floods. In the Red River

delta, more attention has been paid for protecting population against flood during the rainy

season and water management to feed the population, than was devoted to water quality

issues.

1.5.1. General socio-economical context

Besides geomorphological and hydro-meteorological data which are major constraints to the

modelling approach, land use and fertilization, increasing population and domestic and

industrial pollution are also major constraints required to model water quality. Whereas these

constraints will be deeply analysed in chapter 6, general insights will be given here helping to

ask the appropriate questions.

1.5.1.1. Changes in land cover

Several changes in land cover of the Red River basin have been observed since the last 100

years.

Firstly, we have to mention about the deforestation and intensification of agriculture that have

largely occurred in both Vietnamese and Chinese parts.

20

Site description and major issues In the Chinese part, the forest cover of Yunnan has declined from about 60% in the 1950s to

24.2% in 1990 (UNEP, 1990). About 10% of land in this province is categorized as severely

eroded in the 1980s. Only 7% of Yunnan land area is suitable for agricultural activities

(Fullen et al., 1998). Agriculture is restricted to a few of upland plains, open valley and

terraced hillsides. The main food crops such as maize, rice, wheat and potatoes and the main

cash crop such as tobacco, tea, sugarcane are grown in this area. The intensification of

agriculture has occurred thanks to deforestation, increasing cultivation of steep erodible

slopes, over cultivation and adoption of non-sustainable farming practices (Fullen et al.,

1998).

In Vietnam, land use and cover change is the most pervasive and immediately observable

component of the change. Deforestation, intensification of agriculture and urbanization

processes have occurred at variable and often rapid rates over the last couple of decades. It

was noted that warfare and deforestation associated with post-war development 1975 have left

the whole nation with only about 10% cover of closed tropical forests with less than 1% in

pristine state (Collins et al., 1995; Lebel, 1996).

In North Vietnam, deforestation processes was severe, especially in the northern mountains

and midlands. In this area, the forest which covered 95% in 1943 decreased to 17% in 1991

(World Bank, 1996, Nguyen Ngoc Sinh et al. 1995); a slight increased to 19% in the period

from 1995 to 1999 was observed due to the governmental policies of conservation and

development of cultivated forest (Pham Ngoc Dang et al., 2001).

Accounting for the forest area in the Red River basin in the Vietnamese territory, Vo Tri

Chung (1998) reported 3.6 million ha of forest, representing 31% in 1990 (58% for the barren

land). After carrying out the plan of 5 million ha of reforestation of which about 1.2 – 1.5

million ha should be given to the Red River basin, the forest area occupies about 45% of the

whole Red River watershed.

1.5.1.2. Increase of fertilizers utilisation

Fertiliser utilisation (as chemical fertilizer) has much increased in agricultural land in

Vietnam and in China for the recent 50 years. China is an agricultural country where

anthropogenic activity affects strongly surface and groundwater quality through chemical

fertilizer use (23.5 million tons in 1991) and irrigation. Weijin et al. (1999) mentioned that

China is the largest producer of nitrogen fertilizers and largest consumer of mineral fertilizers

in the world. In Vietnam, according to the FAO database (FAO 1990-1998), use of nitrogen

fertilizers has increased by 66 folds during a period from 1961 to 2000 (from 2.2 kgN.ha-1.y-1

in 1961 to 150 kgN.ha-1.y-1 in 2000). For phosphorus fertilizers, the amount used has been 5

21


folds multiplied during the same period. Application of chemical fertilizers may dramatically

increase nutrient concentrations in soils which may subsequently be removed by leaching and

transferred to the river water (figure 1.7). Further, serious erosion and soil loss in watersheds

accelerate the removal of nutrient elements.

Figure 1.7: pollution sources (non-point sources and point sources) in the Red River basin

1.5.1.3. Increase of the population and urbanisation

Increase in population and urbanization might also considerably impact the river system.

The total population of the Red River basin is estimated at 30 million inhabitants and is

growing at an annual rate of about 2.0 %. 65% of the Red River population is Vietnamese,

34% is Chinese and 1% is from Laos. Contrasted population density within the whole Red

River basin must be mentioned: averaging 195 inhabitants.km-2 for the whole basin;

101 inhabitants.km-2 are found in average in the northern mountainous region and

1174 inhabitants.km-2 in the Red River Delta region.

In the Chinese part, in the Yunnan province, where the inhabitants are living in 8 autonomous

prefectures and 11 cities (127 counties, towns), population of the Red River was estimated of

8.8 million inhabitants (about 20.9% of the total Yunnan population), (Chinadata, 1998), of

which 34% population belong to the ethnic minorities. In this area, the annual population

growth rate is of 1.29% (Chinadata, 1998).

In the Vietnamese part, the present annual population grow at a higher rate than in the

Chinese part (about 2.3%). The inhabitants are located in about 21 provinces and cities, and

comprise 18 different ethnic groups of minority people with typical traditions of culture

(MOSTE, 1997).

Parallel with the increasing population, the urbanisation in Vietnam has occurred at a high

rate in recent years. The number of agglomerations (city and town,) has increased from 500 in

of the early 1990’s to 623 in 2000. Whereas population living in agglomeration averaged to

19% in 1990 and increased to 23.5% in 1999, it should reach up to 30-33% in 2010 (Pham

Ngoc Dang et al., 2001). Such a continued rapid urban growth would be a big problem for the

22

Site description and major issues future. However, Smith and Dixon (1997) reported that Vietnam has the lowest rate of urban

growth in the ASEAN group, except Singapore (the Association of Southeast Asian Nations

includes Vietnam, Laos, Cambodia, Thailand, Myanmar, Indonesia, Malaysia, Philippines,

Brunei, Singapore and East-Timor countries).

Domestic wastewater from cities and large agglomeration are mostly discharged directly into

rivers or lakes without treatment, leading to serious pollution of water environment in cities,

especially in Hanoi, Hai Phong, Viet tri (figure 1.8)…

Duong R.Red R.

Day R.

Ba Lat

Tra Ly

Haiphong city

Day

R.

Ninh Co

HanoïSon Tay

Nhu

eR

.

Viet Tri city

Thai Binh R.

Hai Duongprovince

Thai Nguyenprovince

Thai Binhprovince

Quang Ninhprovince

Cau R.

Ha Tay province

Vinh Phuc city

Duong R.Red R.

Day R.

Ba Lat

Tra Ly

Haiphong city

Day

R.

Ninh Co

HanoïSon Tay

Nhu

eR

.

Viet Tri city

Thai Binh R.

Hai Duongprovince

Thai Nguyenprovince

Thai Binhprovince

Quang Ninhprovince

Cau R.

Ha Tay province

Vinh Phuc city

Duong R.Red R.

Day R.

Ba Lat

Tra Ly

Haiphong city

Day

R.

Ninh Co

HanoïSon Tay

Nhu

eR

.

Viet Tri city

Thai Binh R.

Hai Duongprovince

Thai Nguyenprovince

Thai Binhprovince

Quang Ninhprovince

Cau R.

Ha Tay province

Vinh Phuc city

Figure 1.8: Cities and provinces with high population and industrial zones

1.5.1.4. Increase of industrial releases

Since the late 1980’s, Vietnam enters into the period of rapid economic growth that has been

closely associated with a re-engagement within the international and regional Pacific Asian

economies. The “Doi moi” programme introduced in 1986 has opened the economy to the

international monetary system and to a market economy, and reduces the central control

exercised by the State. Industrial activity in Vietnam has rapidly increased. According to the

(MOSTE, 2000), the number of industrial zones in Vietnam has increased from 16 in 1996 to

66 in 1999. In general, 90% of the industrial factories which were constructed before 1975

23


have no wastewater treatment systems. Since 1994, the factories have been located in these

new industrial zones where wastewater treatment systems have been constructed.

Industrial activities are mainly found in the delta of the Red River and Mekong River, and

have especially increased in some large cities. For example, industrial values of Hanoi

represent 8.2% of the whole Vietnamese industrial production in 1995 and increased to 9.4%

in 2002 (Le Qui An, 2003). With the rapid industrial increase and the absence of wastewater

purification infrastructures, the quality of surface and groundwater, at the local scale mainly,

but also at a larger scale, has strongly decreased.

In the Red River basin, there are several industrial zones which influence directly the water

quality of the Red River. In the middle of the basin, the Viet tri city is one of the most

important industrial zones in the North of Vietnam, where the food and drinks production,

paper, chemicals ... are concentrated. Almost all of the wastewaters related to these activities

are discharge directly into the Red River. Beside this zone, another industrial zone in the

mountainous region in the North Vietnam (Thai Nguyen, see figure 1.9) has less influence on

the Red River but strongly influences the Cau River (part of the Thai Binh river system).

Some other industrial zones such as Hai Duong, Hai Phong and Quang Ninh provinces in the

downstream delta of the Red River, and Hanoï, considered apart, have also their impacts on

aqua-ecological processes (see figure 1.8).

1.5.2. Impacts on water quality

1.5.2.1. Decline of surface water quality

In Vietnam, data on surface water quality is poor, and hardly exist in the upstream basin of

the Red River. However, the few existing researches have revealed that water quality of rivers

remains good in upstream rivers while downstream, domestic and industrial water releases

strongly pollute the river water especially in major cities. Urban rivers such as the To Lich,

Lu, Set, Kim Nguu Rivers in Hanoi are typical examples of open wastewater collectors, the

water quality of which being disastrous, especially in dry season. Suspended solid (SS) ranges

from 60 to 300 mg.L-1; dissolved oxygen (DO) within a range of 0.2 to 3 mgO2.L-1, biological

oxygen demand (BOD5) reaching values up to 180 mg.L-1 (MOSTE, 1998).

1.5.2.2. Increasing the water pollution in the delta and the coastal zone

The pollution brought by the Red River is a potential threat for coastal wetlands in the Red

River delta and coastal waters in the South China Sea. Any changes in human activities in the

basin will lead to a change in sediment discharge associated to nutrient loads at the coastal

zone of the Red river delta. For example, the deforestation in the upstream basin will lead to

an increase of floods in the delta, together with a sediment flow which will impact a coastal

24

Site description and major issues marine zone larger than before. This problem can influence the coral reefs in the Southeast of

Cat Ba Island (Nguyen Ngoc Sinh et al., 1995). On the other hand, impoundment of large

reservoirs has decrease the sediment supply to the delta wetlands, increasing the salt intrusion

and reducing the production of wet-rice… Thus, changes of land use and hydrological

management can have contradictory or balancing effects able to temporary hide the problems.

Further, eutrophication in estuaries and coastal zones is another serious consequence of

human alteration of nutrient cycles. The increase of nutrient delivery to the coastal zones, and

the changes in their ratios (N:P:Si, Redfield et al., 1963) can decrease the diversity among

planktonic organisms, and modify the transfer of organic matter within the food web, leading

to phytoplankton accumulation, that paradoxically becomes an oxygen consumer. Moreover,

in eutrophied water, toxic algal blooms have been shown to dominate in many estuaries and

coastal zones in the recent decades (Vitousek et al., 1997), causing economical problems,

such as reduction of tourism activities, prohibition of selling fish and shellfish, etc. Excessive

development in tourism activity leading to pollution together with overfishing would already

have seriously contributed to reduce the productivity of the Red River Coastal zone. The

exceptional site of the Ha long Bay, although protected by UNESCO, is henceforth seriously

threatened. 1.6. References

Chinadata, 1998. Statistical yearbook of Yunnan, Vol. 1997, Vol. 1998, China Statistical Publishing House, (Basic Information of Yunnan, China) (http://chinadatacenter.org).

Chinadata, 2000. Statistical yearbook of Yunnan, Vol. 1999, Vol. 2000; China Statistical Publishing House, (Basic Information of Yunnan, China) (http://chinadatacenter.org).

Collins N.M., Sayer J.A. and Whitmore T.C., 1995. The conservation atlas of tropical forests. Asia and the Pacific. World Conservation Monitoring Centre.

Dang Anh Tuan, 2000. The Red River Delta - The Cradle of the Nation (in Vietnamese), 53 pp., National University in Hanoi, Hanoi.

Dang Quang Tinh, 2001. Participatory planning and management for flood mitigation and preparedness and trends in the Red River basin, Vietnam. Workshop international on Strengthening capacity in participatory planning and management for flood mitigation and preparedness in large river basin, Bangkok (Thailand) 20th-23rd Nov.

Dupre B., Gaillardet J. and Allegre C., 2002. A l’interface entre ciel et terre: les grands fleuves d’Asie, Chapitre 2: 123-130. In: Himalaya-Tibet, le choc des continents. Ed. CNRS and Museum National d’Histoire Naturelle, Paris. ISBN 2-271-05934-8, 191pp.

25


FAO, 1990-1998. Faostat statistics database (Fertilizer data used in China, Vietnam, Laos…), Faostat database Results, copyright FAO 1990-1998, http://www.apps.fao.org/; http://www.fertilizer.fao.org/. 2002.

Fullen M.A., Mitchell D.J., Barton A.P., Hocking T.J., Liu Liguang, Wu Bo Zhi, Zheng Yi and Xia Zheng Yuan., 1998. Soil erosion and Conservation in the Headwaters of the Yangtze River, Yunnan Province, China. In M.J. Haigh, J. Krecek, S. Rajwar and M.P. Kilmartin (eds.), Headwaters: Water resources and Soil conservation. pp: 299-306.

Guilcher A., 1965. Prescis d’hydrologie marine et continentale, Masson, Paris. France, 389pp.

IMH, 1997-2004. Journal of Meteo-hydrology, Institute of Meteo-Hydrology in Vietnam, Hanoi. (Monthly Journal during the periods of from 1997 to 2004).

Jhang Jing, 1996. Nutrient elements in large Chinese estuaries. Continental Shelf Research. Vol 16 (8): 1023 -1045.

Le Bac Huynh, 1997. The especial flood in the Da River and downstream of the Red River in August 1996: the role of Hoa Binh reservoir for flood control in the delta. Journal Vietnamese of Meteo-Hydrology 4(439): 6-15.

Le Qui An, 2003. Environmental plan for the delta of the Red River and some environmental problems in this area. Vietnamese Project National KC. 08.02. In “Proceedings of the first conference on Science» of the Project KC.08: Environmental Protection and some methods for natural disasters, Ministry of Science and Technology. 209 pp. Hanoi.

Lebel L., 1996. Change Challenges for Environment and Development in Vietnam. Seminar on Environment and Development in Vietnam. Common Room, University House, and J G Crawford Building, National Centre for Development Studies. Australian National University, December 6-7. Funded by AusAID International seminar support scheme. Available at http://coombs.anu.edu.au/~vern/env_dev/papers/pap05.html.

Meybeck M., Chapman D.V. and Helmer R., 1989. Global freshwater quality: a first assessment. Cambridge, MA, World Health Organization/United Nations Environment Programme Basil Blackwell, Inc. 306 pp.

MONRE, 2003. Report on water environment monitor in Vietnam in 2003. In “Studies on Vietnam environmental statement in 2003”. Vietnam Ministry of Environment and Natural Resources. 150 pp., Hanoi

MONRE, 1997-2004. Vietnamese Ministry of Environment and Natural Resources. Report

annual on hydrological observation in Vietnam, Hanoi.

MOSTE, 1997. Vietnamese general statistics officer, Ministry of Science, Technology and Environment of Vietnam, general statistics editor, Hanoi. 550 pp.

26

Site description and major issues MOSTE, 1998. Documentation on the Red River Delta (1997-1998), Ministry of Science,

Technology and Environment of Vietnam, Scientific and Technical Publisher, Hanoi. 214 pp.

MOSTE, 2000. Environmental statement in Vietnam. Ministry of Science, Technology and Environment of Vietnam, rapport annual, Hanoi. 210 pp.

Nguyen Huu Khai and Nguyen Van Tuan, 2001. Geography and Hydrology in Vietnam. Vietnam National University publisher, Hanoi, Vietnam.194pp.

Nguyen Ngoc Sinh, Hua Chien Thang, Nguyen Chu Hoi, Nguyen Van Tien, Lang Van Ken, Pham Van Ninh and Nguyen Vu Trong, 1995. Case study report on Red River Delta in Vietnam - Project on integrated management and conservation of near shore coastal and marine areas in East Asia region (EAS-35) United Nations Environment program. Regional coordinating for the East Seas (ESA/RCU), 78pp, U.N. Environ. Programme, Nairobi.

Nguyen Viet Pho, 1984. Streams in Vietnam. Science Publishing House, Hanoi, 209pp.

Pham Ngoc Dang, Nguyen Quynh Huong, Nguyen Hoang Yen, Le Thai Bat, Vo Chi Trung, Dang Huy Huynh, Ho Thanh Hai, Pham Van Ninh, Tran Hieu Nhue, Nguyen Chu Hoi, Luu Van Dieu, Nguyen Kim Thai and Truong Manh Tien., 2001. Environmental statement in Vietnam in the year of 2000. CEETIA. Report in the NORAD project “Enforcing the environmental monitoring and assessment in the South Asia and the modified Mekong basin”. Printed by UNEP RRC.AP. 170pp.

Redfield A.C., Ketchum B.A and Richards F.A., 1963. The influence of organisms on the composition of sea-water. In: Hill MN (ed.) The Sea, Vol 2. Wiley, New York, p 26-77.

SEAMCAP, 2001. Report on the statement of environment in Vietnam in 2001. Strengthening Environment assessment and monitoring capabilities in South Asia and the greater Mekong sub region. SEAMCAP project, UNEP RRC.AP publisher, Klong Luang, Pathumthani, Thailand. 200pp.

Smith D.D and Dixon C., 1997. Sustainable urbanisation in Vietnam. Geoforum, Vol 28(1): 21-38.

Turc L., 1961. Evaluation des besoins en eau d’irrigation, évapotranspiration potentielle, Ann. Agron., 12 (I), I 3-49.

Tran Duc Thanh, Yoshiki Saito, Dinh Van Huy, Nguyen Van Lap, Ta Thi Kim Oanh and Masaaki Tateishi, 2004. Regimes of human and climate impacts on coastal changes in Vietnam. Regional Environmental Change 4: 49-62

Trinh Quang Hoa, 1998. Water balance for purpose of socio-economic development in the Red River delta. Proceedings of International Conference on Economic development and

27


environmental protection of the Yuan-Red River watershed, Hanoi 4th-5th Mar, 92pp., Hanoi.

To Trung Nghia, 2000. Flood control planning for the Red River Basin, International European- Asian Workshop: Ecosystem & Flood 2000, June 27-29. 210pp Hanoi, Vietnam.

UNEP., 1990. China Conservation Strategy, United Nations Environment Programme and China Environmental Science Press. Beijing.

Vitousek P.M., Chair, Aber J., Howarth R.W., Likens G.E., Matson P.A., Schindler D.W., Schlesinger W.H. and Tilman G.D., 1997. Human alteration of the Global nitrogen cycle: Causes and consequences. Issues in Ecology. Number 1, Published by the Ecological Society of America, Washington, USA, 17pp. Available at http://www.esa.org/science/issues/fileEnglish/issue1.pdf.

Vu Van Tuan, 2002. Program of Model construction for general gestation of Thac Ba reservoirs basin served to the development socio-economy, General Centre for Meteorology and Hydrology of Vietnam, 220pp. Hanoi.

Vo Tri Chung, 1998. Forests on the Red river Basin, Vietnam. Institute of forestry investment and planning. Proceedings of International Conference on Economic development and environmental protection of the Yuan-Red River watershed, Hanoi 4th-5th Mar. 92pp.

Weijin Yan, Chengging Yin and Shen Zhang, 1999. Nutrient budgets and biogeochemistry in an experimental agricultural watershed in Southeastern China. Biogeochemistry 45: 1-19.

World Bank, 1996. Vietnam water resources sector review. Selected working papers of the World Bank, ADB, FAO/UNP and NGO Water Resources Sectoral Group, Hanoi, 340 pp.

28

General approach and methodology

CHAPTER 2

General Approach and Methodology: Modelling, experimental work and nutrient budgets

The modelling approach which forms the backbone of this thesis, aims at establishing the link

between the biogeochemical functioning of a large river system and the constraints set by the

meteorology, the morphology of the drainage network and the human activity in the

watershed. We have made use of the RIVERSTRAHLER model (Billen et al., 1994, 1999,

2001b; Garnier et al., 1995, Garnier et al., 2001), developed since 15 years to describe the

ecological functioning and nutrient transfers in the large regional river systems of Europe,

characterized by oceanic regime (the Seine River: Billen et al., 1994; Garnier et al., 1995;

Billen and Garnier, 1999; Billen et al., 2001b; the Mosel River: Garnier et al., 1999; the

Scheldt River: Billen et al., 2005). The Danube, of continental hydrological regime has also

been studied by the Riverstrahler approach (Trifu, 2002; Garnier et al., 2002b). This approach

had never been applied to a sub-tropical Asian system where the hydrological regime strongly

differs from occidental ones.

The implementation of the model offers a suitable framework to collect and synthesize data

on any river system. Beside the geomorphological, climatological and hydrological data, the

model requires gathering data on the point and diffuse sources and on the water quality,

leading to a deep understanding of the human activity in the watershed.

For trans-boundary watersheds, as already experienced on the Danube, the collection of the

data is difficult. Therefore the task undertaken for the Red River was a priori ambitious, as

the watershed of Red River is almost equally distributed within China and Vietnam.

Beside the difficulty to collect data in two different countries, the scarcity of data available,

particularly in the case of point and diffuse sources and water quality, is another problem.

Whereas routine survey of water quality have been organized in most of developed countries

(at variable temporal and spatial frequencies, however) by Institutions such as Water

Agencies or Ministries because severe pollution problems were encountered as soon as the

1960’s, such survey are not well organized within the drainage network of Vietnam. The

water quality survey presently organized by Vietnamese authorities is that of the Vietnamese

coastal line, comprising the outlet of the Red River at the delta. In view of the 3260 km of

29


coast line in Vietnam, this represents a considerable effort, but such data are not well adapted

to the requirement of a whole catchment modelling of the Red River which is the objective of

this thesis.

In order to fill this gap, a seasonal survey was undertaken in the scope of the programme

Espoir on Water, carried out by the research teams involved in the programme. A sampling

strategy was commonly decided to gather the classical water quality data taken into account in

ecological modelling (measurements on the field: oxygen concentration, temperature, pH,

conductivity; water samples for laboratory analysis: biological oxygen demand, suspended

solids, nutrients –the forms of nitrogen, phosphorus- and the silica, total carbon, chlorophyll a

as an estimator of phytoplankton biomass). For other research perspectives, metallic or

organic micropollutants were investigated in parallel as well as biological compartments, such

as bacteria, algae and zooplankton that were also studied on eco-toxicological point of view

(ESPOIR on Water, 2003).

Whereas a modelling approach was developed on the Urban System (Trinh Anh Duc, 2003),

we here constructed in complement the model for the whole upstream Red River basin. The

modelling of the whole delta rivers was not investigated here, although it should become a

research perspective in near future.

2.1 Modeling the quality of the Red River hydrographic network

2.1.1 What is a model?

First of all, a model is a tool for a synthesis and a creation of knowledge. It helps the

researcher to progress into the understanding of any complex systems. A model also allows

testing the general relevance of the conceptual schema adopted as well as the coherence of the

data gathered. According to Nordstrom (2003) “a scientific model is a testable idea,

hypothesis, theory or combination of theories that provide new insights or new interpretations

of an existing problem”. In any model, we only include properties and relationships needed to

understand those aspects of the real system in which we are interested (Nordstrom, 2003).

The process of constructing a model normally goes through the following steps. Firstly, we

try to identify the dominant elements of the system, which represents the state variables of the

model. For example, to study the water quality and ecological functioning of a system such

as a river, chemical species both under dissolved and particulate forms, the algae controlled

by light and nutrients, the bacteria and the zooplankton species have to be included as

30


variables. A second stage is to give a simplified representation of the complexity of the

interactions between the variables by formulating mathematically the kinetics of the

corresponding processes and determining the parameters of these kinetics (Chen, 1995). The

mathematical relations, algebraic or statistical, are set up under the form of a system of

(differential) equations which can be solved analytically in favorite cases but may require the

implementation of complex computational methods. The quantification of the stocks and

fluxes of all components taken into account within the system is then the major objective,

necessary to reach the knowledge of the functioning of the system in terms of the circulation

of material between its constituents. The system representation, either under conceptual

schemas or as mathematical model, contributes to a rational and scientific view helping to

understand the observed phenomena. To summarize, a model originates from a naturalist

description of the different elements of a system and attempts to apprehend its dynamical

functioning as the result of a series of cause-effects relationships. In the same time, the model

allows identifying the weakness of both of our knowledge and of the conceptual schemas

from which the model was constructed.

An ecological model of a river system should be able i) to reproduce spatial and temporal

variations of the concentrations of the various variables of the river system, ii) to allow

establishing budgets for any sub-systems within the system and iii) to calculate the flux

exchanged between these sub-systems (Sferratore et al., submitted).

Moreover, a model of knowledge able to reproduce some aspects of the behavior of an

ecosystem can be utilized as a predictive tool to simulate the behavior of the same ecosystem

in changing environmental conditions. In this way, although being first a research tool, the

model is possibly becoming a tool for environmental management and planning.

Finally, a model is a powerful tool of communication, particularly in the issue of water and

watershed management, as it makes available under an organized form a large amount of

knowledge.

2.1.2. Some definitions in the context of modeling

Spatial and temporal discretisation: An ecosystem may have a geographically variable

extension. A river system is limited by its watershed whereas a lake is contained in a

topographic depression and the position of its bank varies with the water level. The spatial

dimension of a model (0, 1, 2, 3 dimensions) characterizes the resolution of the studied

domain. The model for water quality of a lake can be a zero dimensional model, when

considered as a well mixed reactor, one dimensional if the depth distribution of variables is

31


considered, tri-dimensional if the heterogeneities in all directions are taken into account. A

model for a fluvial ecosystem in which only longitudinal evolutions are considered is as one-

dimensional or mono-dimensional. A model which describes the transversal and horizontal

heterogeneity of concentrations of any pollutant could be called bi-dimensional model. We

can also take into account the vertical heterogeneity in a tri-dimensional model (Poulin et al.,

1998).

Biological complexity: The level of complexity of a biological model characterizes the way in

which the variables and the biological processes are treated and taken into account in the

model (Chen, 1995).

State variables and forcing variables: A model allows simulating the spatial and temporal

variations of the physical, chemical and biological variables such as the water temperature,

the concentration in dissolved oxygen, in nutrients, in phytoplankton biomass, etc... The state

variables are calculated by the model. The forcing variables (or constraints) are the ones

provided to the model under the form of file of numeric values. The forcing variables control

the dynamic of the system but are not influenced by it. A state variable can be treated as a

forcing variable when it is easy to measure (the temperature for example) and when it is not

influenced by other variables (in the case of any influence between this variable and other

variables, it must be considered as state variable) (Tauson and Akimov, 1997).

Measured parameters and adjusted parameters: There are many coefficients named

parameters which used in the expression of the relationships describing the dynamics of the

system. An adjusted parameter is the one which can settle the results given by the model close

to the observations. We can be led to modify such a parameter for each river system studied

or even for each river section. Another approach is to experimentally determine the kinetics of

the processes and to determine the associated parameters. The value can then be adjusted

within its confidence interval range (Garnier et al. 2004; Sferratore et al., submitted). The

models based on this kind of approach are called deterministic or mechanistic: the

RIVERSTRAHLER model, used here, is of these two types: i) some parameters being

adjusted for the hydrological part and ii) some other are experimentally determined for the

ecological sub-model (see below).

Verification, setting, validation and control of quality: After that conceptual schema is chosen

and that the mathematical equations are established and analytically or numerically resolved,

the model has to be verified, adjusted and evaluated.

32


The verification controls the accuracy of the analytical or numerical solutions, the validity of

numeric approximation adopted. In this step, the comparison between the results by the model

are compared with the field observations.

This step of parameter adjusting is a procedure of sensibility analyses of the obtained results

with the variations of each and/or all parameter(s).

The validation of the model occurred when it is applied to a data set different from the one

used for verification and adjustment and simulations by the model compared to the

measurements. The validation can interfere or not with the adjustment step. The verification

and the validation represent a step of quality control for the model (Poulin et al., 1998).

2.1.3. The ecological functioning of hydrographic networks: RIVERSTRAHLER Model

2.1.3.1. General principles

The approach used in this study is based on the adaptation of the RIVERSTRAHLER model

which has been developed in the framework of the PIREN-Seine program and some other

international research programs (Billen et al., 1994; Garnier et al., 1995) to relate the

ecological and biogeochemical functioning of the whole drainage network of a large river

system to the constraints set by the climate, the morphology of the river system and the

human activities in its watershed (figure 2.1). It combines a simplified hydrological model

(HYDROSTRAHLER), relating meteorological constraints to hydrology, to an ecological

model (RIVE), describing in-stream ecological processes. Beside the Seine River, this model

has been successfully applied to several European rivers with differing population densities

(Billen et al., 1994; Billen and Garnier, 1999; Billen et al., 2001b; Billen et al., 2005; Garnier

et al., 1995; Garnier et al., 1999; Garnier et al., 2002b).

The RIVERSTRAHLER model takes into account the whole drainage network according to

the concept of stream orders (Strahler, 1957): the complex network of tributaries is

represented by a regular scheme of the confluence of rivers of increasing stream orders with

mean morphological characteristics. One obvious limitation of this approach is the fact that it

only provides simulations of the mean behaviour of tributaries of given orders, instead of

describing a real river with its own local characteristics. However, in order to improve the

geographical resolution, it is possible to apply the approach separately to several sub-basins

and connect the results to a model of the main branch of the drainage network.

33


Figure 2.1: the structure of the RIVERSTRAHLER model (ref: Billen and Garnier, 2000)

Within the whole drainage network, the model considers three kinds of interconnected

objects. The description of the sub-basin is typically based on the concept of stream order

(Strahler, 1957) with an idealized description (see 2.1.3.2). The main branch is represented

with a finer geographical resolution describing the longitudinal profile every kilometer. All

characteristics found on the main branch have to be described: depth, wetted section, length of

the river stretch canalized, geometry of navigation dam (location and water level), confluence

point of tributaries, location of reservoirs when existing, etc. When connected to the drainage

network, lakes and reservoirs (hydraulic annexes) constitutes the third kind of objects that are

taken into account by two ways: i) for large reservoirs, we consider them under the individual

form and describe their morphology and water inflow and outflow ii) lakes and ponds are

represented statistically by mean characteristics by hydrological order.

The first versions of the RIVERSTRAHLER model were developed under Quick-Basic

computing language. Recently, a new version of the model (Seneque 3-Riverstrahler), has

34


been developed (Ruelland, 2004, Ruelland et al., in prep.), Riverstrahler being embedded

within the SENEQUE GIS interface allowing the user to run the model with any structure of

basins and branches, selected on line according to the geographical resolution required for the

studied question (Figure 2.2). Owing to this new software, the functionalities of the

RIVERSTRAHLER model are multiplied by those of a GIS, allowing an easy extraction of

the data required for separate runs of the Riverstrahler. The Seneque 3-Riverstrahler

developed under the Visual Basic in the Windows environment has been utilized here. This

version that comprises some software accessories is much friendlier for use. It requires

however to assemble a complete set of geo-referenced data on the different constraints under

the form of a GIS data base.

Figure 2.2: One of the working screens of the SENEQUE/Riverstrahler GIS software.

2.1.3.2. The hydrological model

The HYDROSTRAHLER model (Billen et al., 1994; Garnier et al., 2002a) allows simulating

the seasonal variations of the discharge at the outlet of each sub-basin with at a daily time

resolution. This model takes into account the rainfall, the potential evapotranspiration and the

geomorphological data which determine the flow rate. It is based on a simple representation

35


of the rainfall-discharge relationship considering the exchanges between 2 reservoirs: the soil,

contributing surface runoff, and the aquifer contributing base flow (Bultot and Dupriez,

1976). The model involves 4 parameters (soil saturation, infiltration rate, internal flow rate,

groundwater flow rate), and distinguishes between three components of the specific discharge

from the watershed: the base flow supplied by the water table, the internal (or hypodermic)

flow supplied by the soil reservoir, and the surface runoff supplied in periods of soil

saturation (Figure 2.3), (Billen et al., 1994; Garnier et al., 2004).

Infiltration

SWsoil

GWgroundwater

baseflow

PLU ETR

surf.runoff

superf.runoff

solsat

total spec discharge

Infiltration

SWsoil

GWgroundwater

baseflow

PLU ETR

surf.runoff

superf.runoff

solsat

total spec discharge

Figure 2.3: Representation of the rain-discharge relationship in the Hydrostrahler model

The discharge calculated for each order of any sub-basins and every km along the main

branch can be compared to the available observations. The simulations can be adjusted to the

data by considering an initial level of aquifer NAPo (mm) and by adjusting the 4 parameters

of the HYDROSTRAHLER model:

i) the level of soil saturation: SOLsat (mm)

36


ii) the infiltration rate: rinf (d-1)

iii) the internal flow rate: rssr (d-1)

iv) the aquifer flow rate: rgwr (d-1)

The daily variations of the soil water content (SW, mm) and of the groundwater stock (GW,

mm), as well as the specific discharge (mm.d-1) of any elementary watershed, are calculated

from rainfall (mm.d-1) and evapotranspiration (mm.d-1), as follows:

The evapotranspiration is taken equal to the potential evapotranspiration excepted when SW> 0.1 solsat, in which case evapotranspirtation is set to zero.

The total specific discharge is calculated as qspec tot = qbaseflow + qsurf.runoff

in which

The base flow supplied by the water table: qbaseflow = rgwr . GW

The infiltration from the soil water to the aquifer: infiltration = rinf. SW

The specific surface discharge is the sum of the superficial runoff and the (sub)surface runoff: qsur.runoff = rssr. SW + qsup.runoff ,

the superficial runoff, qsup.runoff , is only supplied in periods of soil saturation:

If SW > solsat then = PLU-ETP else = 0

Within a sub-basin, the total discharge (Q, m3.s-1) in order n tributaries is calculated as the

sum of the discharges of their two n-1 order tributaries, the discharges of lateral tributaries of

order 1 to n-1, and the flow from its direct watershed, i.e. the part of the watershed which

does not belong to the catchments of the tributaries. In the main branch, the discharge is

calculated from the discharge of the tributaries and that of the direct watershed (Figure 2.4).

The main merit of this approach is that at any point in the drainage network, the baseflow and

the surface runoff component of the total discharge can be distinguished, which is the key for

taking into account the diffuse sources of material from the watershed (see below).

From the value of the discharge, calculated by stream order, width (w, m) and slope (s, m.m-

1), mean depth (d, m) and flow velocity (v, m.s-1) are calculated by rearranging of the

Manning's empirical formula (Billen et al., 1994). The flow from the direct catchment area of

the river, or from its lateral tributaries of lower stream orders, 'dilutes' the water masses

flowing through the main channel. The corresponding dilution factor and its variations with

stream order and the season are very important for controlling the ecological functioning of

rivers.

37


In the main branch of the river, the calculation is similar, taking into account the contribution

to the flow of both the direct watershed and the considered sub-basins. In regulated sectors,

the values of the depth and the wetted section are taken into account.

nn-1

n-2

Qn = 2.Q(n-1) + Q(lateral tributaries) + Q(direct watershed)

Figure 2.4: Calculation of water flow in the HYDROSTRAHLER module

In the case where reservoirs are present, their role in the hydrological regime must be taken

into account. This will be discussed in chapter 3.

2.1.3.3. The biogeochemical and ecological model: RIVE

The basic assumption in the RIVERSTRAHLER model is the unity of the microscopic

processes (biological and physical-chemical) involved in the functioning of river systems, i.e.

the kinetics of the processes are the same from headwaters to downstream sectors, whatever

the object considered (sub-basins, branches or stagnant annexes). On the contrary, the

hydrological constraints control their expression and differ widely along the upstream-

downstream gradient as do the constraints due to inputs from point and diffuse sources.

Therefore, the specificity of the ecological structure and function of the different sectors of

the river continuum depend on the constraints, rather than on the nature of the processes

involved.

A same model takes into account ecological processes (RIVE: see Garnier et al., 1999 where

developments taken into account in this version are included), and hence allows describing

the main variables of water quality.

Coupled to the HYDROSTRAHLER model, the RIVE model calculates the seasonal and

spatial variations of 22 variables characterizing the water quality and ecological functioning,

including nutrients (nitrate (NO3-) and ammonium (NH4

+) dissolved phosphate (PO43-) and

particulate inorganic phosphorus – PIP- and dissolved silica - (SiO2) two taxonomic groups of

phytoplankton (diatoms and Chlorophyceae, Garnier et al., 1995), two kinds of zooplankton

(rotifers with a short generation time and microcrustaceae with a long generation time,

Garnier et al., 1999) and two compartments of bacteria (the small bacteria autochthonous and

38


the large bacteria allochtonous, Garnier et al., 1991), (Figure 2.5, Table 2.1). The description

of the phytoplankton dynamics is based on the Aquaphy module by Lancelot et al. (1991)

which distinguishes between photosynthesis -controlled by light intensity- and algal growth -

controlled by nutrient availability-. The module has been adapted to two groups of algae

(diatoms and non diatoms) and a formulation for loss processes by excretion and grazing has

been added (Garnier and Billen, 1993; Garnier et al., 1998). The degradation of organic

matter and heterotrophic bacterial dynamics are described according to the HSB module

(Billen and Servais, 1989) and split into two bacterioplankton compartments (the small

bacteria autochthonous and the large bacteria allochtonous, Garnier et al., 1992; Barillier and

Garnier, 1993) and also the nitrifying bacteria. The RIVE model also includes a calculation

of nutrient exchanges across the sediment-water interface (Venice) as a result of a given

sedimentation flux of organic material, taking into account organic matter degradation,

associated ammonium and phosphate release and oxygen consumption, nitrification and

denitrification, phosphate and ammonium adsorption onto inorganic material, mixing

processes in the interstitial and solid phases and accretion of the sedimentary column by

inorganic matter sedimentation (Billen et al., 1989 ; Sanchez, 1997; Billen et al., 1998).

Sedimented biogenic silica is re-dissolved (Garnier et al., 2004). Water column nitrification

(Brion and Billen, 1998; Brion et al. 2000) and phosphate adsorption on suspended inorganic

particles (and their subsequent sedimentation) are also taken into account in the model.

Table 2.1: Kinetic formulation of the processes taken into account in the RIVE model, and

values of the corresponding parameters (in Garnier et al., 1999) Process Kinetic expression Parameters

Phytoplankton dynamics meaning Diatoms Chloro-

phyc. Units

Photosynthesis (phot) kmax (1-exp-(α I/kmax)) PHY kmax* maximal rate of photosynth. 0.2 0.5 h-1

α initial slope of P/I curve 0.0012 0.0012 h-1/(µE.m-2 s -1) reserves synthesis srmax M(S/PHY,Ks) PHY srmax* max. rate of reserve synthesis 0.15 0.37 h-1

Ks 1/2 saturation cst 0.06 0.06 reserves catabolism kcr R kcr.* rate of R catabolism 0.2 0.2 h-1

growth (phygrwth) mufmax M(S/PHY,Ks) lf PHY mufmax max. growth rate* 0.07 0.14 h-1

nutrient limitation factor with lf = M(PO4,Kpp) or M(NO3 +NH4, Kpn) or M(Si02 , KpSi)

Kpp Kpn KpSi

1/2 sat. cst for P uptake 1/2 sat. cst for N uptake 1/2 sat. cst for Si uptake

15 70 0.42

46 70 -

µg P liter-1

µg N liter-1

mgSiO2 liter-1

respiration maint PHY +ecbs phygrwth maint* ecbs

maintenance coefficient. energetic cost of biosynthesis

0.002 0.5

0.002 0.5

h-1

- excretion (phyex) exp phot.+ exb PHY exp "income tax" excretion 0.0006 0.0006 h-1

exb "property tax" excretion 0.001 0.001 h-1

lysis (phylys) kdf + kdf (1+ vf) kdf* mortality rate 0.004 0.004 h-1

vf + parasitic lysis factor 0 / 20 0 / 20 - phyto sedimentation (vsphy/depth).PHY vsphy sinking rate .004 .0005 m/h NH4 uptake phygrwth /cn NH4/(NH4+NO3) cn algal C:N ratio 7 7 g C(g N)-1 NO3 uptake phygrwth /cn NO3/(NH4+NO3) PO4 uptake phygrwth /cp cp algal C:P ratio 40 40 g C(g P)-1 Si02 uptake phygrwth /cSi cSi algal C:Si ratio 2 - g C(g Si02)-1

temperature dependency p(T) = p(Topt).exp(-(T-Topt)² / dti²)

Topt dti

optimal temperature range of temperature

18 13

35 17

°C °C

39


Process Kinetic expression Parameters Zooplankton dynamics

Total zooplankton.

ZOO growth (zoogwth)

µzox.M(PHY-PHYo),KPHY).ZOO µzox KPHY PHYo

max. growth rate 1/2 sat cst to PHY threshold phyto conc.

0.02* 0.4 0.1

h-1

mgC/l mgC/l

ZOO grazing grmx.M((PHY-PHYo) KPHY).ZOO

grmx max grazing rate 0.035* h-1

ZOO mortality kdz.ZOO kdz mortality rate 0.001* h-1


Topt dti


22 12

°C °C

Bacterioplankton dynamics small

bac large bac

HPi production by lysis εpi . (phylys+bactlys+zoomort) εp1 εp2 εp3

HP1 fraction in lysis pducts HP2 fraction in lysis pdcts HP3 fraction in lysis pdcts

0.2 0.2 0.1

- -

enzym. HPi hydrolysis kib.HPi k1b k2b

HP1 lysis rate HP2 lysis rate

0.005 0.00025

h-1

h-1

HPi sedimentation (vsm/depth).Hip Vs Hip sinking rate 0.05 m/h Hid production by lysis δe . (phylys+bactlys+zoomort) εd1

εd2 εd3

HD1 fraction in lysis pdcts HD2 fraction in lysis pdcts HD3 fraction in lysis pdcts

0.2 0.2 0.1

- - -

enzym. HDi hydrolysis eimax. M(HDi,KHi).BAC e1max e2max KH1 KH2

max. rate of HD1 hydrolysis max. rate of HD2 hydrolysis 1/2 sat cst for HD1 hydrol. 1/2 sat cst for HD1 hydrol.

0.75 0.25 0.25 2.5

0.75 0.25 0.25 2.5

h-1

h-1

mgC/l mgC/l

direct substr. uptake bmax. M(S,Ks).BAC bmax Ks

max. S uptake rate 1/2 sat cst for S uptake

0.2 0.1

0.8 0.1

h-1

mgC/l bact. growth (bgwth) Y. bmax. M(S,Ks).BAC Y growth yield 0.25 0.25 - bact. mortality (bactlys) kdb.BAC kdb bact. lysis rate .01 0.1 h-1

bact. sedimentation (vsb/depth).BAC vsb bacteria sinking rate 0 0.01 m/h ammonification (1-Y)/Y.bgwth/cn cn bact. C:N ratio 7 gC/gN PO4 production (1-Y)/Y.bgwth/cp cp bact. C:P ratio 40 gC/gP temperature dependency p(T) = p(Topt).exp(-(T-Topt)² /

dti²) Topt dti


25 15

25 15

°C °C

nitrification and phosphorus dynamics meaning nitrifying bacteria Units

NIT growth (nitgwth) µnix.M(NH4,KNH4).M(O2,KO2). NIT

µnix* KNH4 KO2

max growth rate of NIT 1/2 sat cst for NH4 1/2 sat cst for O2

0.05 7 0.6

h-1

mgN/l mgO2/l

NH4 oxidation nitgwth/rdtnit rdtnit NIT growth yield NIT 0.1 mgC/mg NH4

NIT mortality kdnit.NIT kdnit* NIT mortality rate 0.01 h-1

PO4 adsorpt/desorpt. (planktonic phase)

Langmuir isotherm Pac KPads

SM max. adsorpt. capacity 1/2 saturation ads. cst.

0.0045 0.3

mgP/mgSM mgP/l


Topt dti


23 16

°C °C

benthos remineralisation susp. matter sedim. (vsm/depth)*MES vsm sinking rate m/h Diffusion (interstitial ph.) Fick law Di app. diffusion coefficient 2 10-5 cm²/s Mixing (solid phase) Fick law Ds mixing coefficient 2 10-6 cm²/s orgN mineralis. (maorg) kib.HPi/cn orgP mineralis. kip.HPi/cp k1p*

k2p* orgP hydrolysis rate of HP1 orgP hydrolysis rate of HP2

0.05* 0.0025*

h-1

h-1

benth. nitrification kNi*NH4 (in oxic layer) kNi 1st order nitrification cst 1 h-1

NH4 adsorpt/desorpt. 1st order equilibrium Kam 1st order adsorpt. cst for NH4 30 - PO4 adsorpt/desorpt. (in benthos)

1st order equilibrium Kpa Kpe

PO4 adsorpt. (oxic layer) PO4 adsorpt. (anoxic layer)

35 1.7

- -

SiO2 redissolution kdbSi.SIB kdbSi silica redissolution rate 0.01

h-1

temperature dependency p(T) = p(Topt).exp(-(T-Topt)² / dti²) Topt dti


25 20

°C °C

*These parameters depend on temperature according to the relation mentioned.

+ M(C,Kc) = C/(C+Kc) : hyperbolic Michaelis-Menten function .

+ vf: parasitic lysis amplification function. It is maintained at zero while algal density of each group remains lower than a threshold value of 65 µg Chl a.L-1 and temperature is below 15°C.

40


SS

Cyanobacteria

OXY

PO4PIP

NH4

NO3

DSi

BAC

Large heterotr. bact

DIA

Diatoms

growth

photos& resp.

CO2 S R

grazing

Lysis& excretion

GRAgrowth

photos& resp. S R

Flagell. Chloroph.

Small heterotr. bact

HD1,2

HP1,2

HD3

HP3

exoenz.hydrol

SM

mineralization

OXY

growth& resp.

mortality

microcrusteaceans

OXY

Rotifers, Ciliates

nitrif.

NIT

nitrif.

NO3 NH4organ.matterdegrad.

Oxi

cla

yer

Ano

xic

laye

r

HP1,2,3

denit.SO4

PO4

adsPO4

BSi

dissol. sedim.OXY

THE RIVE MODEL

growth& resp.

ZOO

Cyanobacteria

OXY

PO4PIP

NH4

NO3

DSiDSi

BAC


DIA

Diatoms

growth

photos& resp.

CO2 S R

grazing

Lysis& excretion

GRAgrowth

photos& resp. S R

Flagell. Chloroph.


HD1,2

HP1,2

HD3

HP3

exoenz.hydrol

SM

mineralization

OXY

growth& resp.

mortality

microcrusteaceans

OXY

Rotifers, Ciliates

nitrif.

NIT

nitrif.


Oxi

cla

yer

Ano

xic

laye

r

HP1,2,3

denit.SO4

PO4

adsPO4

BSiBSi

dissol. sedim.OXY

THE RIVE MODEL

growth& resp.

ZOO

SS

Cyanobacteria

OXY

PO4PIP

NH4

NO3

DSiDSi

BAC


DIA

Diatoms

growth

photos& resp.

CO2 S R

grazing

Lysis& excretion

GRAgrowth

photos& resp. S R

Flagell. Chloroph.


HD1,2

HP1,2

HD3

HP3

exoenz.hydrol

SM

mineralization

OXY

growth& resp.

mortality

microcrusteaceans

OXY

Rotifers, Ciliates

nitrif.

NIT

nitrif.


Oxi

cla

yer

Ano

xic

laye

r

HP1,2,3

denit.SO4

PO4

adsPO4

BSiBSi

dissol. sedim.OXY

THE RIVE MODEL

growth& resp.

ZOO

Cyanobacteria

OXY

PO4PIP

NH4

NO3

DSiDSi

BAC


DIA

Diatoms

growth

photos& resp.

CO2 S R

grazing

Lysis& excretion

GRAgrowth

photos& resp. S R

Flagell. Chloroph.


HD1,2

HP1,2

HD3

HP3

exoenz.hydrol

SM

mineralization

OXY

growth& resp.

mortality

microcrusteaceans

OXY

Rotifers, Ciliates

nitrif.

NIT

nitrif.


Oxi

cla

yer

Ano

xic

laye

r

HP1,2,3

denit.SO4

PO4

adsPO4

BSiBSi

dissol. sedim.OXY

THE RIVE MODEL

growth& resp.

ZOO

Figure 2.5: Processes taken into account in the RIVE module (from Garnier et al., 1999)

2.1.3.4. Point sources and non point sources

The point sources and non-point sources within the drainage basin are major constraints that

must be documented for modelling the water quality in any river system and are taken into

account in the RIVERSTRAHLER model. Starting from the level in the headwater streams,

whose water is a mixture of surface runoff and groundwater, the nutrient content evolves from

upstream to downstream of the hydrographic network both because of point discharges of

nutrients and because of the processes that transform, immobilise or eliminate them during

their downward transfer.

Diffuse sources are taken into account through mean nutrient concentrations in each of the

two components of runoff (surface- and groundwater flow) as calculated by the

HYDROSTRAHLER model. The documented variables are NO3, NH4, PO4, PIP, SiO2 and

suspended solids. Regarding nitrates in the surface water, the concentrations are calculated

from the land use in the watershed and from a coefficient of transfer through the riparian

zones (Billen and Garnier, 1999).

41


In the SENEQUE/Riverstrahler version of the model, suspended solids, organic carbon,

nitrogen and phosphorus composition of surface and groundwater flow are automatically

calculated, for each elementary watershed, from the GIS data base on land use, according to a

parameter file that should be documented for each new basin. Silicate content is similarly

calculated from the GIS layer on lithology. The details of the hypothesis used for calculating

the diffuse sources in the case of the Red River are presented in chapter 6.

Regarding the point sources, the variables taken into account in the domestic and industrial

wastewater are suspended mater, organic matter, and the various forms of nitrogen and

phosphorus. Note that SiO2- that typically originates from rock weathering is not considered

as a point source, although a recent work on the largest waste water treatment plant of the

Parisian region has allowed quantifying the amount of silica found in the raw and treated

wastewater (Garnier et al., 2002c). Organic matter in wastewater is an important constraint to

consider. A study carried out on the treated and untreated wastewater in the Paris urban area

(Servais et al., 1999; Garnier et al., submitted) made it possible to convert the variables

provided by sewage networks and treatment plants into state variables in the RIVE model;

biological oxygen demand (BOD) is for example converted into different fractions of organic

carbon. Bacteria brought by the effluents are also taken into account through a relationship

between BOD and heterotrophic bacteria.

Within a watershed, the distributions of all wastewater treatments are taken into account, as

well as the amount of treated or non treated effluents and the kind of treatment (through an

abatement percentage of the concerned variables).

However such kind of data, not necessarily available for European countries, hardly exist in

emerging countries, where wastewater treatment plants are rare, the polluted effluents being

brought directly to streams and rivers in the large cities. The hypothesis made to calculate the

point sources of wastewater in the case of the Red River basin are discussed in chapter 6.

To summarize, the RIVERSTRAHLER model is one of the few available means of modelling

nutrient cycling and ecological functioning of entire drainage networks as a function of the

distribution of natural constraints and human activities in the watershed.

2.2. Experimental work

2.2.1. Sampling campaigns

2.2.1.1. Monthly sampling in the sub-basin and the main branch of the Red River

42


Due to the lack of database on water quality of the Red River system, monthly sampling

campaigns were organized at the outlet of the three tributaries and in the main branch of the

Red River during the years of 2003 and 2004. At the beginning of investigation of water

quality in 2002, only two sampling campaigns were organized in dry season (in February) and

in rainy season (in August).

Hoa Binh reservoir

Thac Ba reservoir

Son Tay

Lien Mac

Hoa Binh reservoir

Thac Ba reservoir

Son Tay

Lien Mac

Figure 2.6: Sampling sites in the Red River system

The sampling sites chosen at the outlet of each of the three upstream sub-basins of the Red

River were those of the hydrological station of the Vietnam territory (see Figure 2.6). For the

Lo River, the samples were collected at the Vu Quang hydrological station, located in Vu

Quang city (Doan Hung district, Phu Tho province). For the Da River, the sampling site was

situated at the Pho Ngoc hydrological station, in Trung Minh city (Ky Son district, Hoa Binh

province). The sampling site of the Thao River was located at the Yen Bai hydrological

station, in Yen Bai city (Yen Bai province). In the delta of the Red River, due to the

complexity of the drainage network , we decided to limit our approach to the main branch of

the Red River at the Hanoi hydrological station, and chose three sampling stations located

between the confluence of the three main sub-basins (at Viet Tri city) and Hanoi city (see

Figure 2.6). In the main branch, from upstream to downstream, samples were taken at Son

43


Tay hydrological station, at Vien Son town (Son Tay district, Ha Tay province) where the

water quality of the Red River is a mixing of that of the three tributaries: Da, Lo and Thao. A

second sampling station is located just upstream the Lien Mac dam (Ha Tay province), which

is the source of the Nhue River (the urban river studied in the French-Vietnamese cooperation

program, ESPOIR on Water program).

Data obtained from this point were used for evaluating of the initial water quality of the Nhue,

seriously impacted by agricultural activities in the watershed and the Tolich River draining

the effluents of Hanoi, as already mentioned. The more downstream sampling station is

located at the Hanoi hydrological station (in Hanoi city), where the river is not completely

impacted by the wastewater of the agglomeration, as it is separated from the river by a huge

hydraulic works (dikes for protection against floods). Some images of sampling campaigns

are introduced in the figure 2.7a.

2.1.1.2. Sampling campaigns for non point source evaluations

Even fewer data are available for nutrient release from cultivated areas, especially in tropical

systems, including Vietnam. For this reason, some samples from agricultural channels in the

North of Vietnam were occasionally taken and analyzed to improve our knowledge on the

characteristics of the Red River basin, to compare the values with the ones scarcely found in

literature. The aim was to document as closely as possible the constraints of the model.

Samples were taken in 2002 and 2003 in regions of various agricultural activities, such as

vegetal culture (cabbage, salad greens…) and rice culture in the suburbs of Hanoi city, in Ha

Tay province (in the delta area) and in Viet Tri city (in the middle land) (figure 2.7b).

The main variables of interests are nutrients as nitrogen (nitrate, nitrite, ammonium) and

phosphorus (phosphate and phosphorus total). Organic carbon (dissolved and particulate

carbon) was also analyzed. Some other measurements such as pH, conductivity, dissolved

oxygen, water temperature were also realized.

2.1.1.3 Sampling campaigns for point source evaluation

Domestic wastewater

No wastewater treatment system for domestic wastewater exists in Vietnam. It can be noted

that wastewater from cities, towns or villages are occasionally diverted to canals, and then

brought to a lake, a small stream, a urban river but sometimes brought to the fields for

fertilizations (in the villages). Unfortunately, data on quantity and quality of domestic

wastewater reaching the surface water are still very poor. To fill this gap, we have therefore

44


collected samples from the various locations inside and around Hanoi city to estimate the

quality of the domestic wastewater in the whole basin of the Red River, and followed the

wastewater circulation in contrasted populated areas to roughly estimate a percentage of waste

really reaching the rivers.

a)

b)

Figure 2.7: a) Sampling campaigns in the upstream of the Red River; b) waste from the non

point-sources and point sources in the Red River basin

Industrial wastewater: sampling campaigns and data collection

Because of lack of a complete database of industrial wastewater as required by the modeling

approach, we have gathered the information of the representative enterprises within the Red

River basin as followings: daily production per enterprise, discharge of effluents, values of

variables such as pH, suspended solids (SS), dissolved oxygen, biological oxygen demand –

BOD-, chemical oxygen demand –COD-, nutrients (NO3, NO2, NH4, N total, PO4, P total).

45


Data were obtained by three ways in 2003: i) collection of available data; ii) elaboration of a

questionnaire; and iii) sampling followed by chemical analyses.

Reports of the Environmental State in Vietnam (reports of 1998, 1999, 2000, 2001) of the

MOSTE (Ministry of Science, Technology and Environment) and other reports issued from

research projects on wastewater in Hanoi and some Vietnamese provinces (JICA: a project of

Vietnam-Japan cooperation, 2000; projects of VAST (Vietnam Academy of Science and

Technology, 2000) were gathered to get the general database.

A questionnaire was elaborated and sent to a number of enterprises for which we got the

addresses by MOSTE (1998-2001). About 200 questionnaires were sent, with the duty to tick

an appropriate box; i) to document the size of the enterprises (range of wastewater effluent in

m3.s-1; number of workers; range of production in ton.day-1); ii) the quality of the wastewater

discharges (ranges of values for variables such as SS, BOD, N total and P total); and iii) the

ways of discharging the effluents (into the river, into a canal, into a lake or a pond, spread on

lands or stored in basin). Unexpectedly, we received about 20 answers.

In addition, we collected and analyzed samples taken from various industrial sectors in Viet

Tri and Hanoi cities. Several factories in Viet Tri as chemicals production, paper production

and textile plants were investigated. Around and inside the Hanoi city, we sampled the Duc

Giang district (West of Hanoi, representing chemicals, paper, wood, battery and electronics)

and the Dong Anh district (North of Hanoi representing electronics, battery, fertilizer, paper,

beer, milk and mechanics; samples were given by the Institute of Environmental Technology)

expected to lead to representative samples. It must be mentioned that it is not always easy to

enter the factories to collect samples so that the samples were mainly taken from wastewater

channels running outside the factories.

The results of water quality to evaluate the pollution sources are presented in the chapter 5.

2.2.2. In-situ measurements and samples analyses

2.2.2.1. In-situ measurements of physical-chemical variables and sampling

Water quality checker, model WQC-22A (TOA, Japan), was used in-situ to measure physical-

chemical variables during the sampling campaigns. This instrument consists of the indicator

main body, the sensors and the standard accessories. By the built-in-one type sensor, five

variables such as temperature (0C), pH, conductivity (mS.cm-1) (or salinity, %0), turbidity

(NTU), and dissolved oxygen (DO, mgO2.L-1) were measured. Before each sampling

46


campaign, the instrument was calibrated using the pure water (for DO, turbidity sensors tests)

and using the standard solution (for pH sensor test).

At the sampling site of the hydrological station, surface samples were collected (30 cm below

the surface) at the middle of the river bed, in front of a boat by a sampled auto-collector.

The water samples were kept at 4 °C to 10 °C before treatment, during transportation to the

laboratory.

2.2.2.2. Filtration and preservation of samples in laboratory

Back to the laboratory, all samples were treated to avoid any changes (enrichment in

particulate and colloidal fractions due to a coagulation or transformation by biological and

chemical processes, (nitrification, denitrification, organic matter degradation, oxidation…)

during storage. The filtration was realized with a Gelman Science filter (Pall) equipped with a

high pressure and a high flow rate. Samples were sequentially filtered through:

i) Whatman GF/F paper-filter (glass micro-fiber filters 0.47µm) for dissolved nutrient

analyses as nitrogen (nitrite, nitrate and ammonia), phosphorus (phosphate), for dissolved

carbon (dissolved organic carbon DOC and dissolved inorganic carbon DIC). For SS

determination on the filter, GF/F filter-papers were pre-weighted.

ii) Whatman GF/C paper-filter for chlorophyll a determination.

iii) Whatman Cellulose nitrate membrane filters for silica.

After treatment, all samples were contained in disposable sterile polyethylene flasks (except

the dissolved organic carbon –DOC- samples stored in glass bottles). The samples were stored

frozen (except the silica samples stored at 4°C in the fridge) to minimize any possible

transformation (volatilization or biodegradation) between the sampling and the analyses.

2.2.2.3. Analyses of samples

Nutrient analyses: A Drell 2010 spectrophotometer (HACH, American) was used for all

nutrient analyses carried out at the Vietnamese INPC laboratory. This is a microprocessor-

controlled, single-beam instrument for colorimetric testing with wavelength range of 400-

900nm and silicon photodiode detector. It can be used both in the laboratory and in the field.

Most of the analyses were also realized at UMR Sisyphe laboratory using a double-beam UV

and visible spectrophotometer (UVIKON 922, KONTRON Instruments). The methods for

nutrient analyses in the laboratories are described in the chapter 5: phosphate, silica and

ammonium were spectro-photometrically determined on filtered water according to Eberlein

and Katter (1984), Rodier (1984) and Slawyck and MacIsaac (1972) respectively; total

47


phosphorus was evaluated on non-filtered samples after sodium persulfate digestion and

mineralization at 110°C in an acidic phase; nitrate was determined after reduction into nitrite

according to Jones (1984).

Chlorophyll determination was done at the Sisyphe laboratory. The chlorophyll was extracted

in 10ml of 90% acetone solution. The optical density of sample was spectrophotometrically

measured (using a 5 cm cell optical path) at 750nm and 650nm, before and after acidification

according to the Lorenzen’s method (1967).

Suspended solids were determined on a pre-weighed standard glass-fiber filter (GF/F) through

which a well-mixed sample was filtered. The material retained on the filter was dried for

about 1 hour at 1030C to 1050C. Taking into account the filtered volume, the increase in

weight of the filter represented the total suspended solids per unit volume (mgSS.L-1).

Dissolved organic carbon: The Total Organic Carbon Analyzer equipment, ANATOC Series

II, (SGE, Australia) was used to determine the dissolved organic carbon –DOC- and of

dissolved inorganic carbon –DIC- (SGE International Pty Ltd, 2002) of a water sample. A

same analyse on filtered water (0.22 µm cellulose acetate membrane filter) allows to

determine the dissolved fraction -DOC and DIC-. The principle is an UV oxidation. At room

temperature and UV light and oxygen, titanium dioxide catalyzes the oxidation of organic

compounds in an aqueous medium, generating carbon dioxide and water. Measurements were

triplicate.

The results of water quality at the outlet of upstream three sub-basins and in the stations in the

main branch of the Red River system are presented and discussed in chapter 4.

Particulate organic carbon analyses were performed on suspended matter harvested on a 12

mm diameter filter GF/F (ignited at 550°C) using a DC-180 Carbon Analyser (Dohrman).

2.3 Nutrient budgets

Nutrient budgets (N, P) established at the basin or regional scale offer an insight into the

fluxes of biogenic material cycling in the various ecosystems constituting the terrestrial

regional system, or transferred into aquatic environments. The respective role of natural and

anthropogenic processes can be easily put in evidence. Regional systems differing in their

natural and anthropogenic characteristics can be compared in terms of their biogeochemical

functioning. Few such budgets have been established for Asian and West Pacific systems,

although rivers in this region of the world may supply about 30-40% of water and 60-70% of

sediment loads to the world’s ocean (Milliman and Meade, 1983). This is the first time

48


nutrient budgets are established for the Red River in Vietnam to evaluate the human impact to

natural nutrient cycling in this tropical region.

We present here the principle of the approach used for the establishment of the nutrient

budgets (N and P) for a given watershed. The details of the sources and hypothesis used to

establish the budgets for the Red River system are discussed in chapter 5.

2.3.1. Nutrients cycling in the soil system

The terrestrial soil sub-system of the considered watershed is divided into the forested area

(semi-natural) sub-system and the agricultural soil sub-system.

Both receives inputs from nitrogen atmospheric fixation, nitrogen and phosphorus in the wet

and dry atmospheric deposition, and are subject to losses through soil leaching and erosion.

Nutrient cycling in agricultural soils is described into more details, taking into account inputs

by chemical fertilizer and manure application, and by excretion by domestic animals, outputs

by export of agricultural goods, either consumed by human and animals in the watershed or

commercially exported outside the limits of the system. Commercial imports of food and feed

from outside the region should also be taken into account, which requires the complete

balance of food to be established for the system.

The principles of the soil budget are represented in figure 2.8.

dom.act.

agricultsoils

cattlefarming

fertili-zers

N2fixatm. depos.

Export Imp.

Forestedsoils

N2fix

agricultural goods

woodexp.

dom.act.

agricultsoils

cattlefarming

fertili-zers

N2fixatm. depos.

Export Imp.

Forestedsoils

N2fix

agricultural goods

woodexp.

Figure 2.8: Schematic representation of the nutrient budgets in the soil system

49


2.3.2. Nutrient budgets in the hydrosystem

The hydrosystem receives nutrients from the watershed as point and non-point sources. The

non-point sources comprise the amount of nutrient leached and eroded from forested or

agricultural land. The point sources include the domestic wastewater and industrial waste

discharges. The latter are extremely difficult to evaluate in the absence of a detailed census of

industrial water pollution. We have approached this question by estimating nutrient release

rate by ton of material produced from different industrial sectors, and using estimation of

industrial production in the system.

The outputs of nutrient from the hydrosystem represent the nutrient fluxes exported from

basin as calculated by the product of the annual discharge and nutrient concentration at the

outlet of the basin.

The difference between total inputs and total outputs from the hydrosystem allows putting in

evidence retention processes related either to elimination processes, like denitrification, or

retention processes, like sedimentation and storage in reservoir sediments.

The principles of the hydrosystem budget are represented in figure 2.9.

denit&

reton

river export

agricultural soil leachingForest soil

leaching

point sourcesdomestic & industrial activity

denit&

reton

river export

agricultural soil leachingForest soil

leaching

point sourcesdomestic & industrial activity

Figure 2.9: Nutrient budget in the hydrosystem.

2.4. References

Barillier A. and Garnier J., 1993. Influence of temperature and substrate concentration on

bacterial growth yield in Seine River Water batch cultures. Appl. Environm. Microbiol.,

59: 1678-1682.

50


Billen G. and Servais P., 1989. Modélisation des processus de dégradation bactérienne de la

matière organique en milieu aquatique. In : Micro-organismes dans les écosystèmes

océaniques. Bianchi et coll. Masson. p. 219-245.

Billen G., Dessery S., Lancelot C. and Meybeck M., 1989. Seasonal and year-to-year

variations of nitrogen diagenesis in the sediments of a recently impounded basin,

Biogeochemistry, 8 : 73-100.

Billen G., Garnier J. and Hanset Ph., 1994. Modelling phytoplankton development in whole

drainage networks: the RIVERSTRAHLER model applied to the Seine River system.

Hydrobiologia, 289:119-137.

Billen G., Garnier J. and Meybeck M., 1998. Chapitre12 : Les sels nutritifs: l'ouverture des

cycles. La Seine en son bassin ; Fonctionnement écologique d'un système fluvial

anthropisé,. Meybeck M ., De Marsily G. and Fustec E. (eds). Elsevier. Paris: pp531-561.

Billen G. and Garnier J., 1999. Nitrogen transfers through the Seine drainage network: a

budget based on the application of the Riverstrahler model. Hydrobiologia, 410: 139-150.

Billen G., Garnier J., Deligne C., and Billen C., 1999. Estimates of early industrial inputs of

nutrients to river systems: implication for coastal eutrophication. The Sciences of the Total

Environment, 243/244: 43-52.

Billen G. and Garnier J., 2000. Nitrogen transfers though the Seine drainage network: a

budget based on the application of the 'Riverstrahler' model. Hydrobiologia, printed in

Netherland 410: 139-150.

Billen G., Garnier J. and LeGuern G., 2001a. SENEQUE 1.3 notice d’utilisation. Programme

PIREN-Seine. UMR CNRS 7619 Sysiphe, Paris.

Billen G., Garnier J., Ficht A. and Cun C., 2001b. Modeling response of water quality in the

Seine river estuary to human activity in its watershed over the last 50 years. Estuaries,

24(6B): 977-993.

Billen G., Garnier J. and Rousseau V., 2005. Nutrient fluxes and water quality in the drainage

network of the Scheldt basin over the last 50 years. Hydrobiologia. In press

Brion N. and Billen G., 1998. Une réévaluation de la méthode d’incorporation de 14HCO3-

pour mesurer la nitrification autotrophe et son application pour estimer des biomasses de

bactéries nitrifiantes. Rev. Sci. Eau, 11 : 283-302.

51


Brion N., Billen G.., Guezennec L. and Ficht A., 2000. Distribution of nitrifying activity in

the Seine River (France) and its estuary. Estuaries, 23: 669-682.

Bultot F. and Dupriez G., 1976. Conceptual hydrological model for an average-sized

catchment’s area. Journal of Hydrology 39:251-292.

Chen D., 1995. Basic concepts of the Design methodology for production management

systems: a state-of-the-art. International Transactions in Operational Ressearch. Vol

2(1): 137-146

Eberlein K and Katter G., 1984. Automatic method for the determination of orthophosphate

and dissolved phosphorus in the marine environment. Fresenius Z. Anal. Chem. 326: 354-

357.

ESPOIR on Water, 2003. Reports on the international conference in the framework of

ESPOIR project on water in urban river system in Hanoi, Vietnam, Feb. 2003, Hanoi.

Vietnam.

Garnier J., Servais P. and Billen G., 1991. Bacterioplankton in the Seine River: impact of the

parisian urban effluents. Can. J. Microbiol., 38: 56-64

Garnier J., Billen G. and Servais S., 1992. Physiological characteristics and ecological role of

small and large sized bacteria in a polluted river (Seine river, France). Archiv. Hydrobiol.

Ergebn. Limnol. 37:83-94.

Garnier J. and Billen G., 1993. Ecological interactions in a shallow sand-pit lake (Créteil

Lake, France). A modelling approach. In: Nutrient dynamics and biological structure

in shallow freshwater and brackish lakes. Hydrobiologia, 275/276:97-114.

Garnier J., Billen G. and Coste M., 1995. Seasonal succession of diatoms and Chlorophyceae

in the drainage network of the river Seine: Observations and modelling. Limnol.

Oceanogr. 40: 750-765.

Garnier J., Billen G., Hanset Ph., Testard P. and Coste M., 1998. Développement algal et

eutrophisation. Chapitre 14. In: La Seine en son bassin Fonctionnement écologique d'un

système fluvial anthropisé, Meybeck M., De Marsily G. and Fustec E. (eds). Elsevier,

Paris.

Garnier J., Billen G. and Palfner L., 1999. Understanding the oxygen budget od the Mosel

drainage network with the concept of heterotrophic/autotrophic sequences: the

Riverstrahler approach. Hydrobiologia, 410: 151-166.

52


Garnier J., Dufayt O., Billen G. and Roulier M., 2001. Eutrophisation et gestion des apports

de phosphore dans le bassin de la Seine. Acte du Colloque, Lyon 2001. Scientifiques and

Décideurs : agir ensemble pour une gestion durable des systèmes fluviaux. 11 p. CD-rom

téléchargeable sur http://www.eaurmc.fr.

Garnier J. and Billen G., 2002a. The Riverstrahler modelling approach applied to a tropical

case study (The Red -Hong- River, Vietnam): nutrient transfer and impact on the Coastal.

SCOPE, Coll. Mar. Res. W., 12: 51-65.

Garnier J., Billen G., Hannon E., Fonbonne S., Videnina Y. and Soulie M., 2002b. Modelling

transfer and retention of nutrients in the drainage network of the Danube River. Estuar.

Coast. Shelf Sci., 54: 285-308.

Garnier J., d’Ayguesvives A., Billen G. and Sferratore A., 2002c. Silica dynamics in the

hydrographic network of the Seine River. Oceanis 29: 139-160.

Garnier J, Billen G. and Cugier Ph., 2004. Drainage basin use and nutrient supply by rivers to

the coastal zone. A modelling approach to the Seine River. p.60-87 In: Drainage basin

nutrient inputs and eutrophication: an integrated approach. P. Wassmann and K. Olli

(Eds). E-book in press, 309p. available at: www.ut.ee/~olli/eutr/

Garnier J., Billen G. and Cébron A., submitted. Modelling nitrogen transformations in the

lower Seine river and estuary (France): impact of wastewater release on oxygenation and

N2O emission. Hydrobiologia.

JICA, 2000. The Study on Environmental Improvement for Hanoi City in the Socialist

Republic of Vietnam, Final report, Data book., Nippon Koei Co. Ltd, Japan Internation

Cooperation Agency (JICA) and Hanoi People's Committee - Socialist Republic of

Vietnam., Hanoi, 350p.

Jones M.N., 1984. Nitrate reduction by shaking with cadmium, alternative to cadmium

columns. Water Res. 18:643-646.

Lancelot C., Veth C. and Mathot S., 1991. Modelling ice-edge phytoplankton bloom in the

Scotia-Weddel Sea sector of the Southern Ocean during spring 1988. J. Mar. Syst., 2 :

333-346.

Lorenzen C.J., 1967. Determination of chlorophyll and phaeopigments: spectrophotometric

equations. Limnol. Oceanogr., 12 : 343-346

Milliman J.D., and Mead R.H., 1983. World-wide delivery of river sediment to the oceans.

The Journal of Geology, 91(1), 1-21.

53


MOSTE., 1998-2001. Environmetal state in Vietnam. Ministry of Science, Technology and

Environment in Vietnam. Report annual, Hanoi. Vietnam.

Nordstrom D.K., 2003. Modelling low-temperature geochemical processes. In series books:

Treatise on Geochemistry, executive editors Holland H.D. and K.K. Turekian, Volume 5:

Surface and Ground Water, Weathering and Soils, Volume Editor: Drever J.I., Elsevier

Pergamon publisher, USA, pp: 37-72.

Poulin M., Even S., Billen G., Mouchel J.M., Garnier J., Levassor A. and Leviandier T., 1998.

Chapitre 6 Modèles: des processus au basin versant. Dans La Seine en son basin :

Fonctionnement écologique d’un system fluvial anthropisé. Meybeck M ., De Marsily G.

and Fustec E. (eds). Elsevier in Paris: 679-717.

Rodier J., 1984. L'analyse de l'eau. 7ème édition. Dunot (ed.), France, 1365 p.

Ruelland D., 2004. SENEQUE, logiciel SIG de modélisation prospective de la qualité de

l'eau. Revue Internationale de Géomatique, 14 (1), 97-117.

Ruelland D., Billen G., Brunstein D. and Garnier J., SENEQUE 3, in prep. A GIS interface to the RIVERSTRAHLER model of the biogeochemical functioning of river systems. To be submitted to Ecological Modelling.

Sanchez N., 1997. Le processus de dénitrification dans les sédiments du barrage-réservoir de

la Marne: étude de sa cinétique et modélisation. Thèse Univ. P and M. Curie. 140 pp.

Servais P., Garnier J., Demarteau N., Brion N. and Billen G., 1999. Supply of organic matter

and bacteria to aquatic ecosystems through waste water effluents. Water Research. 33:

3521-3531.

SGE International Pty Ltd., 2002. Guide manual ANATOC Series II TOC Analyzer.

“Installation and Operation Manual of Total Organic Carbon Analyser”, Publication

number: MN 0292E, Revision 2.0.0; 2.02.

Slavyck G. and McIsaac J.J., 1972. Comparison of two automated ammonium methods in a

region of coastal upwelling. Deep-Sea Res. 19:1-4.

Sferratore A., Billen G., Garnier J. and Théry S., Submitted. Modelling Nutrient (N, P, Si)

Budget in the Seine Watershed: Application of the Riverstrahler Model using Data from

Local to Global Scale Resolution. Global Biogeochemical Cycles.

Strahler A.H., 1957. Quantitative analysis of watershed geomorphology. Geophys. Union

Trans. 38: 913-920.

54


Tauson V.L. and Akimov V.V., 1997. Introduction to the theory of forced equilibria: general

principles, basic concepts and definitions. Geochimica et Cosmochimica Acta, 61(23):

4935-4943.

Trifu C. M., 2002. Transfert des nutrients dans le bassin du Danube et apports a la mer Noire:

modelisation et bilans. Thèse de doctorat d'Etat de l'Université Paris VI, 198pp.





VAST., 2000. Proceedings annual of Project Reports on Environments of Vietnam Academy

of Science and Technology in 2000, Hanoi. VAST publisher. 310p.

55


56

Hydrological regime and suspended load: observation and modelling

CHAPTER 3

Hydrological regime and suspended load of the Red River system (Vietnam): observation and modelling

Abstract

Previous estimates of the suspended-matter loading of the Red River in Vietnam range from

100 to 170 106 t.yr-1, i.e. from 640 to 1060 t.km-².yr-1. The strong dependence on the

hydrology of the suspended-solid transport results in a large year-to-year variability. Based on

available hydrology data from the period 1997-2004, and on a one-year daily survey of

suspended-matter in the three main tributaries of the Red River system in 2003, a simplified

model was built to estimate the mean suspended load of the Red River in present conditions.

The obtained value is 40 106 t.yr-1, corresponding to a specific load of 280 t. km-2.yr-1. It

reflects a 70% decrease of the total suspended load since the impoundment of the Hoa Binh

and Thac Ba dams in the 1980s. The model predicts a further reduction by 20% of the

suspended load in the Red River with the planned construction of two additional dams. Using

measurements of the total phosphorus content of the suspended material in the different Red

River tributaries, we estimated the present phosphorus delivery by the Red River to be 36 106

kgP yr-1.

Keywords: the Vietnamese Red River, suspended-solids, particulate phosphorus, dams

This chapter is submitted as an article in the Journal of Hydrology under the reference:

Le Thi Phuong Quynh, , Josette Garnier, Gilles Billen, Sylvain Théry, and Chau Van Minh (2005, submitted).

57


3.1. Introduction

Sediments transported by rivers to coastal waters play an important role in the global

biogeochemical cycles of many elements. Martin and Meybeck (1979) estimated that over

90% of the global riverine delivery of some major biogenic elements such phosphorus or iron

are transported with suspended matter. Ludwig et al. (1996) showed that 45% of the total

organic carbon annually discharged globally from rivers into the ocean is in particulate form.

At the global and regional scale, the factors controlling riverine sediment fluxes have been

extensively studied (Milliman and Sywitski, 1992; Walling and Fang, 2003; Meybeck et al.,

2003; Syvitski, 2003; Syvitski et al. 2003). Non human factors linked to climate, topography

and lithology of the watershed, obviously play a major role. However, human actions may

also be important, e.g. deforestation, farming, surface mining, road construction and

urbanization among others have led to a 50% increase of global sediment discharge in the last

2,000 years (Milliman and Syvitski, 1992). More recently, on the contrary, dam construction

has caused a significant decrease of sediment loads globally (Milliman, 1997; Walling and

Fang, 2003). Vörösmarty et al. (1997, 2003) estimated that 30% of the world’s sediment flux

becomes trapped in large dams.

South and South-East Asia, due to their peculiar tectonic status linked to the Himalayan

formation, contribute a much larger share than other areas in the world to the global riverine

flux of suspended-solids to the ocean (Holeman, 1968, Miliman and Meade 1983, Milliman

and Syvitski, 1992, Ludwig and al., 1996). Milliman and Sywitski (1992) estimated that one

third of the present global suspended-matter delivered by rivers originates from Southern Asia

(about 20 109 t.yr-1). More recently, Meybeck et al. (2003) stressed that globally, most

sediment is carried to the oceans from a small proportion of the South East Asia and Pacific

Islands land masses. Milliman and Meade (1983) estimated the mean specific sediment yield

of Asian and South-Eastern Asian rivers to 380 t.km-².yr-1 and 600 t.km-².yr-1, compared to a

global mean of 116 t.km-².yr-1 (Milliman and Meade 1983).

The Red River in Vietnam and China is a good example of a South-East Asian river system

strongly affected by human activities. Its overall sediment load has been ranked 15th in the

world (Milliman and Syvitski, 1992). It is doubtful however whether these general estimates,

which have for decades been cross-cited by many authors (Holman, 1968; Meybeck, 1989;

Ludwig et al, 1996; Van Maren and Hoekstra, 2004), take into account the recent man-made

changes (e.g. deforestation of river systems).

In this study, we analyze a set of measured discharges and suspended-solid concentrations,

gathered from several sources in Vietnam, in order to assess the present suspended-matter

58


loading of the Red River system and the trends of its variations over the last 40 years.

Furthermore, we use a simplified modeling approach to predict possible future trends of

particle transport by this large sub-tropical river system. Another aim of the study is to

examine the link between the suspended-matter and the total phosphorus load in the Red

River. Recently, a study of the nutrient budget of the Red River and its major tributaries (Le

Thi Phuong Quynh et al., 2005) related the human activity in the watershed to nitrogen and

phosphorus delivery to the South China Sea. It showed that phosphorus deserved a more

detailed investigation, because of its close relationship with suspended-sediment transport.

3.2. General characteristics of the Red River basin

3.2.1. Geomorphology

The Red River basin (Figure 3.1) is located in South-East Asia (from 20°00 to 25°30 North;

from 100°00 to 107°10 East) and drains an area of 156 451km², of which 50.3% in Vietnam,

48.8% in China and 0.9% in Laos. The Red River, bordered by the Truong Giang and the

Chau Giang River basins of China, to the North, the Langcang River (Mekong) basin, to the

West, the Ma River basin (in Vietnam), to the South, and the Thai Binh River and the South

China Sea (Tonkin Bay), to the East, is the second largest river in Vietnam (Nguyen Ngoc

Sinh et al., 1995). The Red River gets its name from the reddish-brown colour caused by its

high load of iron-dioxide-rich sediments.

The Red River has its source in the mountainous Yunnan province, in southern China, at an

elevation of 2,000 m (Nguyen Huu Khai and Nguyen Van Tuan, 2001). It is named Yuan

River in China, and flows into Vietnam at Lao Cai where it is named Cai, Thao or Hong

River. The Red River then runs through 7 Vietnamese provinces before flowing into the

China Sea through 4 distributaries called, Ba Lat, Lach Gia, Tra Ly and Day (Figure 3.1),

(Dang Anh Tuan, 2000). The Red River receives two major tributaries: the Da and Lo rivers.

The Da River reaches the Thao River in Ha Nong district, at Viet tri city, position:

105° 20’50’’E and 21°15’00’N. The Lo River joins the main branch of the Red River at Viet

Tri city, slightly downstream, at position 105° 26’40’’E and 21°17’50’N. Some general

characteristics of the sub-basins (morphology, lithology, land use, population) are listed in

Table 3.1.

59


a)

Duong R.Red R.

Day R.

Ba Lat

Tra Ly

Haiphong

Day

R.

Ninh Co

Hanoï

Son Tay

Nhue R.

Thao R.

Da

R.

Lo R.

Boitich R.

Luc Nam R.

Tonkin Bay

Hoa Binh

Hoa Binh Reservoir

Thac Ba Reservoir

100 km100 km

50 km50 km

b)

Figure 3.1: Map of the Red River basin, a) its three major upstream tributaries and b) its delta

area

60


Table 3.1: Some characteristics of the Red River and its main tributaries (Nguyen Viet Pho 1984; Nguyen Viet Pho et al., 2003; MOSTE, 1997; Nguyen Huu Khai and Nguyen Van

Tuan, 2001; Tran Duc Thanh et al., 2004; Dürr, 2003).

Thao (Yen Bai)

Da (Hoa Binh)

Lo (Viet Tri)

Hong Delta total

Red River (total)

Topography Basin area, km² 57 150 51 285 34 559 9435 156 451

Length main branch, km 902 1010 470 236.5 1138.5 Maximum elevation, m 6740 3143 3076 10 6740

Slope, ‰ 33.2 37 20 - 29.9

Lithology

Plutonic acid rocks, % 9.0 0.0 0.1 0 - Basic volcanic rocks, % 0.0 0.0 0.5 0 -

paleozoic sedimentary rocks, % 55.5 85.3 72.7 0 -

Mesozoic silicic rocks, % 18.0 0.0 21.5 0 - mesozoic carbonated rocks, % 16.7 14.7 5.2 0 -

alluvial deposits, % 0.9 0.0 0.0 100 -

Land use (in 1997) Rice, % 18.7 12.5 8.1 63 17

Industrial and other cultures, %

14.4 3.0 58.6 3.9 19.8

Grassland, % 7.2 3.6 3.9 2.6 5.0 Forest, % 54.2 74.4 22.4 17.8 51.6

Rocky areas, % 4.1 6.2 6.4 5.9 5.4 Urban area, % 1.4 0.3 0.6 6.8 1.2

Population (in 1997) population density, inhab/km² 150 101 132 1174 192

The mountain areas that form a large part of the upstream basin of the Red River are

tectonically very active and show high erosion rates (Fullen et al., 1998). The geologic

substratum of the upper basin is dominated by consolidated paleozoic sedimentary rocks of

complex lithology with variable contributions of mesozoic silicic or carbonated rocks.

Naturally alluvial deposits dominate the delta area. Soils in the upper basins are typically

Ultisols (by U.S. classification) or “red soil” (by Chinese soil classification), while in the

delta area grey soil and alluvial soil dominate (MOSTE, 1997).

61


3.2.2. Meteorology

The climate of the Red River basin is of sub-tropical monsoon type, characterised by the

alternation of a dry and a rainy season, the latter lasting from May to October and receiving

85 – 95% of the total yearly rainfall. Meteorological data (temperature, humidity, rainfall,

solar radiation) were obtained from 12 Vietnamese meteorological stations in the Red River

basin (IMH 1997-2004), as well as from the Kunming station in China (Chinadata, 1998-

2001) for the period 1997 to 2004. The 13 meteorological stations were distributed within the

different sub-basins on the basis of Thiessen polygons (Figure 3.2), and the integrated mean

values by sub-basin were calculated.

N

Kunming1228

Ha Noi1600

Thai Binh1577Nam Dinh

1602

Ha Giang2332

Son La 1341

Lai Chau2413

Sa Pa2578

Tuyen Quang1635

Hoa Binh1930

Phu Tho1242

Yen Bai 1738

Lao Cai1771

0 20 50 70 100km

Annual rainfall values in period of 1997- 2003

≥ 2000 mm

1600 ÷ 2000 mm

≤ 1600 mm

N

Kunming1228

Ha Noi1600


1602

Ha Giang2332

Son La 1341

Lai Chau2413

Sa Pa2578

Tuyen Quang1635

Hoa Binh1930

Phu Tho1242

Yen Bai 1738

Lao Cai1771

0 20 50 70 100km


≥ 2000 mm

1600 ÷ 2000 mm

≤ 1600 mm

N

Kunming1228

Ha Noi1600


1602

Ha Giang2332

Son La 1341

Lai Chau2413

Sa Pa2578

Tuyen Quang1635

Hoa Binh1930

Phu Tho1242

Yen Bai 1738

Lao Cai1771

0 20 50 70 100km


≥ 2000 mm

1600 ÷ 2000 mm

≤ 1600 mm

≥ 2000 mm

1600 ÷ 2000 mm

≤ 1600 mm

Figure 3.2: Rainfall distribution within the Red River basin: annual values for the period

1997-2003, at the different stations

In the period from 1997 to 2004, the 10-day mean temperature in the three upstream sub-

basins varied from 14-16°C in winter to 26-27°C in summer. In the Delta area, temperatures

were higher, varying from 17 to 30°C. Humidity averaged from 82 to 84% throughout the

year in the Vietnamese part of the basin (IMH 1997-2004), while it was lower, about 67÷70

%, in the Chinese part (Chinadata, 1998-2001).

62


The mean annual rainfall is 1,590 mm for the whole Red River basin. The rainfall in the

Chinese territory (annual average of 1,230 mm) is much lower than in the Vietnamese part

(1,810 mm). It varies greatly in space (Figure 3.2), with the highest values (2,000-2,600

mm.yr-1) in the upstream area of Lai Chau, Sapa, Ha Giang, smaller values in the upper

Chinese basin (above 1,230 mm.yr-1) and in the intermediate zone of Tuyen Quang, Yen Bai

and Lao Cai (from 1630 to 1770 mm.yr-1), and shows its lowest values in the median zone of

Son La, Phu Tho (1250 to 1350 mm.yr-1). In the Delta, the values are around 1,600 mm.yr-1

(Figure 3.2). Considering the 4 sub-basins, the mean annual values are 1,904 mm.yr-1 for the

Thao, 1,889 mm.yr-1 for the Da, 1,874 mm.yr-1 for the Lo and 1,677 mm.yr-1 in the Delta. In

the period from 1997 to 2004, the lowest rainfall amount was observed in 1999 and the

highest one in 1997 and 2001 in all three upstream sub-basins (Figure 3.3).

Thao sub-basin

Figure 3.3: Evapotranspiration (ETP) and rainfall (PLU) variations (mm d-1) in the upstream

sub-basins of the Red River from 1997 to 2003 (Thao, Da, Lo) and in the Delta sub-basin (Delta).

Evapotranspiration (ETP, mm) was calculated by Turc’s formula (Turc, 1961), based on

0

5

10

15

20

1997 1998 1999 2000 2001 2002 2003

PLU

-ETP

, mm

.d-1

Lo sub-basin

0

5

10

15

20

1997 1998 1999 2000 2001 2002 2003

PLU

-ETP

, mm

.d-1

Delta sub-basin

0

5

10

15

20

1997 1998 1999 2000 2001 2002 2003

PLU

-ETP

, mm

.d-1

ETPPLU

Da sub-basin

0

5

10

15

20

1997 1998 1999 2000 2001 2002 2003

PLU

-ETP

, mm

.d-1

63


monthly temperature (T°C) and sunshine duration (Sdur, h). These data were obtained from the

respective meteorological stations:

ETPmm/month = 0.4 T°C (Ig+50)/(T°C+15)

where

T°C is the atmospheric temperature in 0C during the considered period

Ig is the total solar radiation expressed in cal.cm-2.d-1 during the period, which can be

calculated by:

Ig = IgA (0.18+ 0.62 h/H)

where

IgA is the energy of solar radiation in the absence of atmospheric attenuation,

expressed in cal.cm-2.d-1.

h/H is the relative duration of sunshine, H is the duration of the astronomic day and h,

the duration of the sunshine period per day.

IgA and H values, which only depend on the latitude and the period of the year, are

provided by Turc (1961).

The mean annual evapotranspiration (ETP) (from 1997 to 2004) is rather homogeneously

distributed over the whole basin area, varying within quite a narrow range, from 880 to 1,150

mm.yr-1. Annual ETP values are: 1,040 mm.yr-1 in the Thao basin, 1,040 mm.yr-1 in the Da,

1,000 mm.yr-1 for the Lo and 1,080 mm.yr-1 in the Delta (Figure 3.3).

3.2.3. Population and land use

The population in the Red River basin was estimated at 30 million inhabitants in 1997, of

which 34 % in China, 65 % in Vietnam and less than 1% in Laos (Chinadata, 1998, MOSTE,

1997). The population density in the different sub-basins varies significantly from 101, 132

and 150 inhab.km-2 in the Da, Lo and Thao sub-basins respectively, to 1,174 inhab.km-2 in the

Delta area (Table 3.1).

Land use is very diverse from one upstream sub-basin to another, as well as between the

basins and the Delta area (Le Thi Phuong et al., 2005). Industrial crops (mainly coffee,

rubber, cotton, sugar, tobacco, etc.) dominate (58.1%) in the Lo basin, forests (74.4%) in the

Da basin, paddy rice fields (66.3%) in the Delta area, while the Thao basin is characterized by

a larger diversity of land use including forest (54.2), paddy rice fields (18.7) and industrial

crops (12.8).

64


The forest cover of the Chinese Yunnan province, in the upper part of the Red River basin,

declined from about 60% in the 1950s to 24.2% in 1990 (UNEP, 1990). About 10% of the

land in this province was categorized as severely eroded in the 1980s. Moreover, Fullen et al.

(1998) reported that, over the last 250-500 years, erosion rates in the Yunnan province have

increased fifteen folds as a result of poor land management, cultivation on steep slopes,

deforestation and lack of conservation. Similarly, deforestation has been intense in North

Vietnam, especially in the northern mountains and the center, where the percentage of forest

cover decreased from 95% in 1943 to 17% in 1991 (World Bank, 1996).

3.2.4. Dams and discharge regulation

Table 3.2: Some characteristics of the large dams already impounded (Hoa Binh and Thac Ba) or planned, in italics (Son La and Dai Thi) in the Red River basin (data gathered from Trinh Quang Hoa, 1998; Vu Van Tuan, 2002a, 2002b; Nguyen Huu Khai and Nguyen Van

Tuan 2001; To Trung Nghia, 2000).

Name of the dams

Characteristics Hoa Binh Thac Ba Son La Dai Thi

River (sub-basin) Da Chay (Lo) Da Gam (Lo)

Date of impoundment 1985 1972 2010-2015 2010

*Volume (min-max), 109 m3 3.9 – 9.5 0.78-2.94 9.3-25.5 0.5-3.0

*critical upstream discharge, m3.s-1 1750-1500 200-190 850-750

Surface area, km² 208 235 440 42

Length, m 210 60 - -

Mean depth, m 50 42 60 70

Water level (normal), m 115 58 215-265 115

Upstream watershed, km2 57285 6170 26000 9700

Electricity production, 106 KWh.yr-1 1920 386 2400-3600 313

• parameters used for the hydrological simulations (see text for explanation)

There are two large dams in the Red River watershed: Hoa Binh and Thac Ba (Figure 3.1,

Table 3.2). Hoa Binh, damming the Da River, is the largest one in Vietnam. It was

constructed in 1985, has a surface area of 208 km² and an effective storage capacity of 9.5

km3 (Vu Van Tuan, 2002b; Ngo Trong Thuan and Tran Bich Nga, 1998). Besides protecting

the city of Hanoi from exceptional floods such as the one in 1971, and providing water for

irrigation at low river flow, it serves to generate electric power and provides 40% of

Vietnam’s electricity (7.8 billion KWh). The Thac Ba dam, impounded in 1972 on the Chay

65


River (a tributary to the Lo River) is the second largest in Vietnam, with a surface area of 235

km² and a storage volume of 2.94 km3 (Vu Van Tuan, 2002.a). It provides 0.4 billion KWh

(Dang Quang Tinh, 2001; Vu Van Tuan, 2002.a).

Another large dam, the Son La, upstream of the Hoa Binh on the Da river, is planned to start

operating in 2010-2015. It will have a surface area of 440 km², a total storage capacity of 25.5

km3 (an effective storage of 16.2 km3), and a water level 265 m above sea level (a.s.l.). The

Dai Thi dam is already in construction on the Lo River and is planned to be operational in

2010. It will have a surface area of 42 km², a total storage of 3.0 km3 and a water level a.s.l. of

115 m (Dang Anh Tuan, 2000).

3.3. Hydrological regime of the Red River and its tributaries

3.3.1. Total and specific discharge of the sub-basins

The average annual discharge at Son Tay station (downstream from the confluence of the

three main tributaries) reported by To Trung Nghia (2000) for the period 1902-1990 was

3,740 m3.s-1, corresponding to a specific discharge of 26.1 L.s-1.km-². For the period 1997-

2004, for which we obtained the daily values from MONRE (1997-2004), we calculated a

similar value of 3,389 m3.s-1 (23 L.s-1.km-²).

The discharge measured at Son Tay station is equal to the sum of the discharges of the three

major tributaries Da, Thao and Lo, except in the dry season, when some water is diverted

through irrigation channels (Figure 3.4). The discharge at Hanoi, however, is lower by about

20% in all seasons. Downstream of the Son Tay station, and upstream from Hanoi, 4 main

distributaries divert water from the main branch: the Day River and the Nhue River on the

south-east side and the Ca Lo River and the Duong River on the north-eastern side (Figure

3.1). Since the implementation of the Day River Flood Diversion Scheme in 1937, the Day

River draws water from the Red River main branch through the Day Dam, located 35 km

upstream from Hanoi, open during the flood season. The Red River has an irregular flow and

is subject to flooding. In order to protect Hanoi city, the Day River system has been designed

to be the first and largest flood diversion route in case of emergency. The Nhue River receives

water from the Lien Mac dam with an annual discharge of 24.5 m3.s-1 and serves both to

evacuate wastewater from Hanoï city and irrigate the Delta. The Duong River on the left bank

of the Red River, initially a tributary 5 km upstream from Hanoi, is presently a distributary

that with a mean annual discharge as high as 1,060 m3.s-1 diverted from the Red River to the

Thai Binh River (MONRE, 1997-2004). The Ca Lo river mouth is now almost filled up with

sediment and no longer plays an important hydrologic role. As the complexity of the

66


hydrology of the delta system, comprising the multiple distributaries of the Red River, would

require a separate study, we mainly analysed the outputs of water and suspended-matter from

measurements at the outlets of the three main tributaries and/or at the upstream Son Tay

station on the main branch.

a)

b)

0

2000

4000

6000

8000

10000

2001 2002 2003

disc

harg

e, m

3 .s-1

ThaoDaLo

2001 2002 2003

0

5000

10000

15000

20000

2001 2002 2003

disc

harg

e, m

3 .s-1

Son TayHanoiTotal

2001 2002 2003

Figure 3.4: Discharge variations in 2001, 2002, 2003, a) at the outlets of the sub-basins Thao, Da, Lo, and b) in the main branch of the Red River (at the Son Tay, Hanoi stations).

The sum of the discharge (Total) at the outlet of three upstream sub-basins is shown in comparison (b).

Over the last 100 years, the maximum daily value at Son Tay station, 37 800 m3.s-1 was

observed in August 1971, while the minimum, 368 m3.s-1 was observed in May 1960. High

floods are always a threat to the highly populated delta area. In the recent history of Vietnam,

67


serious floods causing dykes to break were noted in 1913, 1915, 1945 and 1971 when the

water level in Hanoi reached respectively 11.35 m, 11.2 m, 11.45 m and 13.3 m (the highest

known), (To Trung Nghia, 2000). In fact, the 4 major floods within the return period of 100

years were caused by simultaneous strong floods in the Lo, Thao and Da rivers. The Da River

generally plays the major role, representing 53-57% of total discharge. Since the Hoa Binh

dam was constructed on the Da River (1985), the floods in Hanoi have been fairly well

controlled (Le Bac Huynh, 1997).

The seasonal variations of specific discharge at the outlets of the three upstream sub-basins

and the main branch during the period 1997-2004 (MONRE, 1997-2004) are shown in Figure

3.5. The Da and Lo basins have much higher specific discharges (34 and 25 L.s-1.km-2 as an

annual mean in 2003), than the Thao river (9.6 L.s-1.km-2), which has a large part of its basin

in the drier Chinese territory.

Thao River

0

50

100

150

200

250

1997 1998 1999 2000 2001 2002 2003

Spec

. dis

ch.,

L.s-1

.km

- ²

Da River

0

50

100

150

200

250

1997 1998 1999 2000 2001 2002 2003

Spec

. dis

ch.,

L.s-1

.km

-²

Lo River

0

50

100

150

200

250

1997 1998 1999 2000 2001 2002 2003

Spec

. dis

ch.,

L.s-1

.km

- ²

Hong River

0

50

100

150

200

250

1997 1998 1999 2000 2001 2002 2003

Spec

. dis

ch.,

L.s-1

.km

-²

Figure 3.5: Seasonal specific discharge (Spec. disch.: L.s-1 km-2) at the outlet of the three sub-basins (Thao, Da, Lo) and in the main branch of the Red River (at Son Tay station,

Hong) from 1997 to 2003.

3.3.2. Modelling the rain-discharge relationship

68


Infiltration= rinf . SW

SWsoil

GWgroundwater

baseflow= rgwr . GW

PLU ETRIf SW > 0.1 solsat then ETR=ETPelse ETR=0

surf.runoff= rssr . SW + sup.runoff

superf.runoffIf SW > solsat then =PLU-ETPelse =0

solsat

total spec discharge= baseflow + surf. runoff

Qout = Qin – Qfill + Qempt

Qin= qspec. RBA

QfillIf Vol<volmax and Qin>qcritf then Qfill = Qinelse Qfill = 0

volmax

VolQempt

If Vol>volmin and Qin<qcrite then Qempt = 0.5 Qinelse Qempt = 0

volmin

RBA

a)

b)

Figure 3.6: Principles of the hydrological model. a) The Hydrostrahler model (Billen et al, 1994). SW: soil water (mm); GW: groundwater (mm). ETR: real evapotranspiration (mm day-

1); ETP: potential evapotranspiration (mm day-1); Solsat: soil saturation content (above which all excess rainfall is evacuated as surface runoff), infr: infiltration rate, srr : surface runoff rate and gwrr: groundwater runoff rate. b) Representation of the hydrology of the large dams in the model. RBA: watershed area upstream from the dam. Qin, Qout: inflowing and outflowing discharge (m3 s-1); Qfill, Qempt: discharge of filling or emptying of the dam; volmax, volmin (m3) : minimum and maximum volume of the dam; qcritf, qcrite (m3 s-1): critical discharge above which the dam is allowed to be filled or below which it is allowed to be emptied.

69


In order to further explain the differences in mean specific discharge between the sub-basins

as well as their seasonal variations, we tried to relate the specific discharge to rainfall. In

view of the small number of meteorological stations where rainfall data are available, only a

simplified approach was possible. We chose to use the Hydrostrahler model, as described by

Billen et al. (1994). This simple and non distributed model of the rainfall-discharge

relationships considers two water reservoirs in the watershed (Figure 3.6a), i) a superficial (or

soil) reservoir, with short residence time, supplied by rainfall and feeding evapotranspiration,

infiltration and surface/sub-surface runoff, ii) a groundwater reservoir, with longer residence

time, fed by infiltration and at the origin of the base flow. The model involves 4 parameters:

surface runoff rate, soil saturation content (above which all excess rainfall is evacuated as

surface runoff), infiltration rate and groundwater discharge rate. A calculation procedure was

developed to optimize the values of these 4 parameters, based on the Nash criterion (Nash and

Sutcliffe, 1970) calculated with the observed (obsQ) and calculated (calcQ) values of daily

discharge:

Nash = 1 – [Σ (obsQ-calcQ)² / Σ (obsQ-meanQ)² ]

In order to avoid systematic bias related to poor knowledge of the total rainfall over the whole

basin, we adjusted the mean daily rainfall data of the 4-5 meteorological stations available for

each sub-basin by multiplying them with the factor required to equilibrate the annual balance

between observed cumulated discharge at the outlet of the basin and cumulated rainfall minus

potential evapotranspiration. The assumption behind this procedure is that the available

rainfall data provide a correct picture of the temporal distribution of precipitation but only a

poor estimate of its absolute value. The approach was applied to the series of daily rainfall

and discharge data available over the period 1997-2004 for the Thao basin, providing the

simulation of discharge over 3 years shown in Figure 3.7. The calibrated values of the

parameters are listed in Table 3.3. The value of the required rainfall correcting coefficient,

always below 1 for the Thao River, indicates that, in general, the rainfall data gathered from

the Chinese territory, spatially under-represented, overestimated the water balance in the Thao

basin. With corrected rainfall data, a discharge simulation with a Nash criterion above 0.7

could be obtained (Table 3.3). The main advantage of this procedure is that, in the total

discharge of the river, a component corresponding to (sub)-surface runoff can be

distinguished from another corresponding to base flow (Figure 3.7). This is the basis for a

suspended-load model (see below).

70


2003

0

1000

2000

3000

4000

j j f m a m j j a s o n d

disc

harg

e, m

3 .s-1


2002

0

1000

2000

3000

4000

0 30 60 90 120 150 180 210 240 270 300 330 360

disc

harg

e, m

3 .s-1

obs.sim.base flow


2004

0

1000

2000

3000

4000

j j f m a m j j a s o n d

disc

harg

e, m

3 .s-1


Thao R.

Figure 3.7: Simulations and observations of the discharge at the outlet of the Thao sub-basins from 2002 to 2004 (obs: discharge observations; sim.: discharge simulations; base flow).

71


Table 3.3: Adjusted hydrological parameters of the hydrological model for the three sub-basins of the Red River (Thao, Da, Lo). Solsat: soil saturation content (above which all excess rainfall is evacuated as surface runoff), infr: infiltration rate, srr : surface runoff rate and gwrr: groundwater runoff rate. Factor PLU.: factor used in the hydrological

model to correct rainfall data and Nash: Nash criterion based on observed and calculated values by ten-day periods, see text for explanations).

Parameters Thao Da Lo

solsat, mm 110 165 210 infr, d-1 0.0619 0.0375 0.05 srr, d-1 0.0384 0.0745 0.0675 gwrr, d-1 0.0132 0.0026 0.0010

Year factor PLU. Nash factor PLU. Nash factor PLU. Nash

1997 0.65 0.83 0.87 0.69 1.10 0.77

1998 0.68 0.91 1.07 0.75 1.20 0.65 1999 0.76 0.81 1.05 0.80 1.20 0.77 2000 0.61 0.73 1.01 0.81 1.00 0.68 2001 0.81 0.66 1.09 0.90 1.05 0.87 2002 0.73 0.79 1.00 0.86 1.10 0.93 2003 0.58 0.83 0.90 0.86 0.80 0.79 2004 0.51 0.73 0.95 0.51 - -

The model was adapted to take into account the filling and emptying of a dam, if it is present

in the watershed (case of the Da and Lo rivers, Figure 3.8). Four additional parameters are

taken into account to describe, in a simplified way, the management rules of each dam (Table

3.2): the minimum and maximum volume of the dam and two critical values of the river

discharge above which water is stored (Qcritf), or below which the dam is emptied (Qcrite) to

sustain the downstream flow, provided the volume of the dam has not yet reached its

maximum or minimum value, respectively (Figure 3.6b). The procedure first calculates the

daily specific discharge for the whole watershed area, then the absolute discharge entering the

dam (Qin), considering the watershed area upstream from the dam. It is allowed to store water

when its volume is below the maximum value and Qin is above Qcritf. It is emptied if its

volume is above the minimum value and if Qin is below Qcrite. In all other situations, the

discharge downstream of the dam is equal to the one entering it. The results of the discharge

simulation of the Da and Lo rivers, show major differences in May and June if their dams are

taken into account or ignored (Figure 3.8).

72


Da R., 2003

0

2500

5000

7500

10000

0 30 60 90 120 150 180 210 240 270 300 330 360

disc

harg

e, m

3 .s-1

obs.sim.sim. without damBase flow.


Lo R., 2003

0

1000

2000

3000

4000

5000

0 30 60 90 120 150 180 210 240 270 300 330 360

disc

harg

e, m

3 .s-1


Figure 3.8: Simulations and observations of the discharge at the outlet of the Da and Lo sub-

basins in 2003 (obs: discharge observations; sim.: discharge simulations; sim. Without res.:

simulation without the presence of any dam; base flow).

The optimized hydrological parameters obtained on the basis of the observed discharge and

rainfall data during the period 1997-2004 for the Lo river and, to a less extent for the Da

(Table 3.3) show higher infiltration rates and lower groundwater runoff rates than for the

Thao basin, indicating a more stable contribution of base flow in their hydrological regimes.

73


3.4. Suspended-matter loading of the Red River and its tributaries

The results of a detailed survey of daily suspended-material concentration in the Red River

and its three main tributaries in 2003 were made available to us by MONRE. This data base

is the only one available at this frequency for recent years. It concerns the stations Yen Bai,

Son Tay and Hanoï on the Thao River, Hoa Binh on the Da River, and Vu Quang on the Lo

River. We compared these data with earlier results published by several authors in the

Vietnamese and international literature, often providing only monthly or annual means.

3.4.1 Total and specific suspended load

Table 3.4: Sediment load (106 tons.y-1) transported by the Red River gathered from several studies. Authors Total suspended

load, 106 tons.y-1Remarks

Lisitzin, 1972; Holman, 1968; UNESCO, 1991; Milliman and Meade, 1983; Milliman and Syvitski, 1992

130 Before Hoa Binh dam impoundment

Nguyen Viet Pho, 1984, World Bank, 1996 116 Before Hoa Binh dam impoundment

Meybeck et al., 1989; 160 Before Hoa Binh dam impoundment

Nguyen Ngoc Sinh et al. 1995 140-150 Before Hoa Binh dam impoundment

Ludwig et al., 1996 166 Before Hoa Binh dam impoundment

Van Maren and Hoekstra, 2004 100 Period not stated This paper, observations in 2003 41 Measurements at Hanoi

station in 2003 This paper, calculations for the period 1997-2003

38.8 Calculation with the model for the period 1997-2003

Several authors have reported figures for the total annual loading of the Red River system at

its outlet (Table 3.4). Their estimates range from 100 to 166 106 t.yr-1, i.e. from 640 to 1,060

t.km-².yr-1. As mentioned above, many of these figures are cross-cited from one author to

another, and it is rather difficult to determine to which period they refer. Year-to-year

variations of the hydrology introduce a large variability. Thus, Nguyen Viet Pho (1984)

pointed out that, although the mean annual sediment load of the Red River in the period 1958

to 1971 was about 111 106 tons, it varied from a minimum of 56 106 t.yr-1 in 1963, a rather dry

year, to a maximum of 202 106 t.yr-1 in 1971, when a disastrous flood occurred. Nguyen Viet

Pho et al. (2003) reported that the suspended-solid load at Son Tay station decreased from 114

106 t.yr-1 in the period 1958-1985 to 73 106 t.yr-1 in the period 1986-1997, after the Hoa Binh

74


dam on the Da river has come into operation. The detailed data obtained from MONRE for

the year 2003 at Son Tay and Hanoï stations lead to much lower values, respectively 26 and

41 106 t.yr-1, i.e 178 - 274 t.km-².yr-1 (Table 3.4).

Data provided by Pham Quang Son (1998) and Tran Thanh Xuan and Pham Hong Phuong

(1998) show large differences between the sub-basins with the specific suspended load

varying between 262 - 417 t.km-².yr-1 for the Lo River, 228 – 1,193 t.km-².yr-1 for the Da

River and 551- 1,060 t.km-².yr-1 for the Thao River in the period from 1958 to 1995 (Figure

3.9). These differences are confirmed by our data from 2003, which however show lower

values due to rather dry hydrological conditions (Figure 3.9).

Figure 3.9: Distribution of the specific suspended-solid load (SS, tons.km-2.y-

1) between 4 time periods for the three sub-basins (Thao River at Yen Bai station, Da River at Hoa Binh station, Lo River at Vu Quang station) and the main branch (Hong River at Son Tay station), (Pham Quang Son, 1998; Tran Thanh Xuan and Pham Hong Phuong, 1998; Trinh Dinh Lu and Doan Chi Dung, 1998 and MONRE, 1997-2004).

Thao R.

0

500

1000

1500

1958-1985 1976-1985 1986-1995 2003

SS, t

ons.

km-2

.y-1

Da R.

0

500

1000

1500

1958-1985 1976-1985 1986-1995 2003

SS, t

ons.

km-2

.y-1

Lo R.

0

500

1000

1500

1958-1985 1976-1985 1986-1995 2003

SS, t

ons.

km-2

.y-1

Hong R.

0

500

1000

1500

1958-1985 1976-1985 1986-1995 2003

SS, t

ons.

km-2

.y-1

75


3.4.2. Seasonal and long-term variations of suspended load

It is a general characteristic of tropical river systems, and of the Red River as well, that most

of the suspended load is transported during the rainy season and at high discharge (Nguyen

Viet Pho, 1984, Pham Quang Son, 1998, Van Maren and Hoekstra, 2004). This is because

during the rainy summer months, both the discharge and the suspended-matter concentration

are high. During a flood, suspended-matter concentrations often increase from 1,000 to 5,000

mg.L-1 in the Thao River, from 500 to 2,500 mg.L-1 in the Da River and from 150 to 500

mg.L-1 in the Lo River (Nguyen Viet Pho et al., 2003). For a given sub-basin, there is a

significant linear relationship between suspended-matter concentration and specific discharge

and the different behaviors of the three sub-basins described above are clearly visible (Figure

3.10). The rather low suspended-matter concentrations in the Lo River, compared to those of

the Thao and the Da rivers before damming are striking. Note that the name of the “Lo” river

means “clear” in Vietnamese, indicating that its low suspended-matter content is an ancient

characteristic. Nguyen Viet Pho et al. (2003) reported that year to year mean suspended-

matter concentrations over the period 1961-1990 in the Lo river, were 710 mg.L-1 at Dao Duc

station (in Ha Giang province), 410 mg.L-1 at Chiem Hoa station (in Tuyen Quang province)

and 290 mg.L-1 at Vu Quang station (in Phu Tho province). These low values might be

surprising when it is remembered that the Lo watershed is the one with the smallest share of

forest and the largest share of industrial crops as compared with the other two sub-basins

(Table 3.2). The lower general slope and different geology of the Lo basin (Table 3.2)

probably explain this paradox.

The changing suspended-matter - discharge relationship observed in the Da River over the

long term (Figure 3.10), with a strong reduction after the filling of the Hoa Binh dam in 1985,

illustrates the prominent role of large dams in the trapping of suspended material (Vorösmarty

et al., 2003). Pham Quang Son (1998) estimated that in the first years of operation of the Hoa

Binh dam, about 50 106 tons.yr-1 of suspended-solids were deposited in the dam

(corresponding to more than 80 % of the total SS flux transported by the upstream part of the

river). Furthermore, Nguyen Viet Pho et al. (2003) reported an interannual mean of

suspended-solid concentrations over the period 1961-1989 decreasing from 1,600 mg.L-1 at

Lai Chau station and 1,430 mg.L-1 at the Ta Bu station (just upstream of the Hoa Binh dam) to

209 mg.L-1 at the Hoa Binh station (downstream of the HoaBinh dam).

The data concerning the Thao river basin shows higher suspended loading at Lao Cai, on the

Chinese border, than at the Yen Bai station, indicating that the upstream part of the basin is

subjected to greater erosion than the lower part (Figure 3.10). Nguyen Viet Pho et al. (2003)

reported mean a suspended-matter concentration of 2,730 mg.L-1 at Lao Cai against 1,760

76


mg.L-1 at the Yen Bai station during the same period, from 1958-1990 (Figure 3.10). No

significant long-term trends appear in the data from the Yen Bai station over the period 1956-

2003.

Da R.

0

1000

2000

3000

4000

5000

0 20 40 60 80 100 120

Susp

ende

d So

lids,

mg.

L-1

1964-19681981-19841990-19931986-19972003

Thao R.

0

1000

2000

3000

4000

5000

0 20 40 60 80 100 120

Susp

ende

d so

lids,

mg.

L-1

1956-19781956-19902003

Lo R.

0

1000

2000

3000

4000

5000

0 20 40 60 80 100 120

Spec. disch., L.s-1.km-2

Susp

ende

d so

lid, m

g.L

-1 1961-19701961-19901960-19902003

Figure 3.10: Relationships between suspended-solid concentrations (mg.L-1) as functions of the specific discharge (Spec. disch.: L.s-1.km-2) at the outlet of the three sub-basins: for the

Thao river at the Lao Cai station in 1956-1978 and at Yen Bai station in 1956-1990 and 2003; for the Da river at Hoa Binh station for all periods ; for the Lo river (or its tributaries, i.e. Lo-Gam-Chay river), at Thac Ba station in the Chay river in 1961-1970, at the Ghenh Ga station in the Lo in 1961-1990, at Chiem Hoa station in the Gam river in 1960-1990, at Vu Quang

station in the Lo river in 2003. Linear trends of the relationship are indicated.

77


3.4.3. Modelling the suspended load

On the basis of the simple hydrological model discussed above, providing a daily estimate of

the base-flow and the surface runoff, we proposed a simple model for calculating the

suspended load of each of the three sub-basins. We assumed that the former component of the

discharge is characterized by a constant and low base-line suspended-matter concentration,

while the latter comprises higher suspended concentrations, resulting from erosion processes

and depending on topography, lithology and land use in the watershed. These two values can

be calibrated by optimizing the reconstructed daily variations of the suspended-matter load

with respect to the observed ones, using the Nash criterion as explained above for the

discharge modeling.

Where there is a dam, a simple, steady state, model is taken into account, relating suspended

mater concentration at the outlet of the dam (SMout) to the concentration at the inlet (SMin)

and the hydraulic residence time (τ) in the dam:

SMout = SMin . [1 / (1 + ksed .τ)]

A reasonable value for ksed (day-1) is 0.5 day-1, representing the ratio of the particle setting

rate (about 1 m.h-1) to the mean depth of the dam (about 50 m).

Results of these calculations are compared with the observations in the Thao, Da and Lo

rivers for the year 2003 for which daily suspended-matter concentration data are available

(Figure 3.11). We also applied the model to a set of monthly suspended load data from the

Da River before the Hoa Binh dam was impounded using the same calibrated parameters

(Table 3.5). The Nash criterion, calculated on mean values by decades, ranges between 0.5-

0.76. Note that an alternative model in which the suspended concentration of the surface

runoff component was considered as a linear function of the specific surface runoff, instead of

as a constant, did not provide better results. The model calculations of annual loading for the

Thao, Da and Lo rivers in 2003, respectively 17.8 106, 6.2 106 and 8.8 106 t.yr-1 were very

close to the values calculated from measured daily discharge and suspended-matter

concentrations, respectively 20.0 106, 5.5 106 and 7.9 106 t.yr-1. This model, although

admittedly rather simplistic, is capable of estimating the suspended-solid loading of the three

sub-basins from hydrological data. The retention of suspended-matter by the dam is fairly

well evaluated during the high water period. It is overestimated during low flow (Figure

3.11), possibly because it neglects the role of non- or slowly settling particles, and possibly

that of algal biomass produced in the lakes. Nevertheless, this does not severely affect the

capability of the model to correctly assess the total annual sediment load.

78


Da R.

0

200

400

600

800

0 30 60 90 120 150 180 210 240 270 300 330 360

Susp

ende

d so

lids,

mg.

L-1


Lo R.

0

200

400

600

800

0 30 60 90 120 150 180 210 240 270 300 330 360

Susp

ende

d so

lids,

mg.

L-1


Thao R.

0

1000

2000

3000

4000

0 30 60 90 120 150 180 210 240 270 300 330 360

Susp

ende

d so

lids,

mg.

L-1

simulationobservation


sim.obs.

Figure 3.11: Seasonal simulations (sim.) and observations (obs.) of suspended-solid concentrations (mg.L-1) at the outlets of the Thao, Da and Lo Rivers for the year 2003.

Using the same model, and the calibrated parameters of Table 3.5, we calculated the

suspended load over the period 1997-2004, for which the discharge values were modeled and

validated (see above), but daily suspended-matter data are not available. The results provide a

79


mean year-to-year suspended-matter load of 22.5 106, 6.5 106 and 9.8 106 t.yr-1 respectively

for the Thao, Da and Lo sub-basins, and a total suspended-matter loading at Son Tay of 38.8

106 t.yr-1 (Table 3.6).

Table 3.5: Suspended-solid concentrations (SS, mg l-1) in the base flow and surface runoff determined for the three sub-basins (Thao, Da, Lo) of the Red River to calculate the

suspended-matter load.

SS concentrations, mg l-1 Thao Da Lo

Surface runoff concentration 2400 2100 520

Base flow concentration 160 30 30

Table 3.6: Calculations by the model of the mean annual sediment load (106 tons.y-1) transported by the three main tributaries of the Red River for the year 2003

at the various stations (st.) of the different rivers (R.).

Mean annual sediment load (106

tons.y-1) Yen Bai st.

Thao R. Hoa Binh st. Da R.

Vu Quang st. Lo R.

Son Tay st. Hong R.

with the presence of the two dams (Hoa Binh and Thac Ba)

22.5 6.5 9.8 38.8

without the presence of the two dams (Hoa Binh and Thac Ba)

22.5 86.8 12.5 121.8

with the presence of two additional dams (Son La and Dai Thi)

22.5 3.3 6.5 32.3

with climate change 27.9 8.5 11.5 47.9

The model was also used to calculate the suspended-matter load in the Da and Lo basins that

would have been observed in the absence of any dams in. The values of 87 106 and 12 106 t.yr-

1 were obtained for the Da and Lo respectively, resulting a value of 122 106 t.yr-1 for the total

loading at Son Tay (Table 3.6).

To what extent our estimates of the suspended load at the mouth of the three main tributaries

reflect the suspended-matter delivery of the Red River to the sea is difficult to assess. The

Delta area, where the slope of the rivers is very weak and the flow diverted into many natural

and man-made distributaries might be the site of a net sediment deposition. However,

sedimentation processes are observed mostly in areas along the coast or influenced by sea

80


water like Haiphong harbour, within Ba Lat, Ninh Co and Day estuaries, as well as on the

tidal flats of the neighbouring seashore (Nguyen Ngoc Sinh et al., 1995, Van Maren and

Hoekstra, 2004). In 2003, the daily suspended-matter and discharge data at the Hanoi station

showed a suspended-matter flow of 40 106 t.yr-1, close to our loading estimate for the three

main upstream tributaries for the same year. This indicates that no significant net deposition

of solid material occurs in the upper portion of the Delta, upstream from Hanoi.

3.4.4. Relationship between suspended-solid and phosphorus transport

The median phosphorus concentration of suspended-solids in world rivers was estimated by

Meybeck (1982) to 1.15 mgP.g-1. The phosphorus content of suspended-matter was measured

on samples taken monthly at the stations of Son Tay, Hoa Binh, Vu Quang and Hanoï in 2003

with the method described by Rodier (1984). The results show significantly different mean

values in the sub-basins, ranging from 0.42 to 0.85 mgP.g-1 in the upper tributaries (Table

3.7). The values found at the Hanoi station are consistently higher (0.93 mgP. g-1), probably

reflecting the adsorption on the suspended-matter of some phosphorus of human origin in the

lower course of the Red River basin (Table 3.7). As these values show no clear seasonal

variations, we calculated the mean annual particulate phosphorus loading by simply

multiplying the P contents by the respective suspended-solid loads estimated above. The

values obtained compare well with the phosphorus export previously calculated by another

method (Le Thi Phuong Quynh et al., 2005; Table 3.7).

Table 3.7: Phosphorus content (P content, mgP.g-1) of the suspended load (SS load, 106 tons.yr-1) measured at the outlet of the three main tributaries (Thao, Da, Lo) of the Red River system and in the main branch at Hanoi. Calculation of the corresponding mean annual phosphorus load (106 kgP yr-1) and comparison with the budget estimates reported by Le Thi Phuong Quynh et al. (2005).

Thao Da Lo Main branch

P content, mgP.g-1 0.43 ± 0.09 0.68 ± 0.17 0.85 ± 0.28 0.93 ± 0.14

SS load, 106 tons.yr-1 22.5 6.5 9.8 38.8

P load, 106 kgP yr-1 (this study) 9.7 4.4 8.3 36.1

P load, 106 kgP yr-1

(see Le Thi Phuong Quynh et al., 2005) 8.3 3.5 5.1 51

81


3.5. Future scenarios of suspended-matter loading

3.5.1. Effects of planned dams

As mentioned above, the construction of additional large dams is planned for the next 10

years in the Red River basin (Table 3.2). Assuming a constant rainfall regime over a period

of 8 years, the model described above can be used to calculate the effect of these new dams on

the suspended load of the Red river system (Table 3.6). The Son La dam, on the Da River

upstream from the Hoa Binh one will further decrease the suspended load by about 50%. The

Dai Thi dam, in the upper basin of the Lo River, will double the effect of the existing Thac Ba

dam on the total suspended-matter delivery by the Lo river, reducing its load by about 30 %.

As a whole, the total loading of the Red River will decrease from 40 to 32 106 t.yr-1 basing on

the hydrology of the period 1997-2004.

3.5.2. Effect of Climate change

As a result of increased greenhouse gas concentrations in the atmosphere and consecutive

global warming, the hydrological cycle is setting to be amplified. In Asia, in particular, the

Intergovernmental Panel on Climate Change (IPCC, 2001) predicts an annual mean rainfall

increase of about 3 ± 1% in the 2020s and 11 ± 3% in the 2080s, along with a 2-5°C mean

temperature increase. More local or detailed predictions are extremely uncertain because of

the large inter-model variations. Moreover, no models are presently able to predict the effect

of climate change on the frequency of paroxysmal events. In order to gain some insights into

the order of magnitude of the suspended load variations in the Red River induced by climate

change, we ran the described model for the last 8 years, with a 10% rainfall increase, and an

increase of evapotranspiration corresponding to a 3°C temperature rise (which leads to a

roughly 5% increase of ETP) (Table 3.6). With respect to the conditions of the period 1997-

2004, the model predicts an increase of about 20% in the suspended-matter loading, i.e. from

40 to 48 106 t.yr-1.

3.6. Conclusions

With the data presented in this article, a good estimate can be obtained of the total suspended-

matter load presently carried by the Red River main branch and its two major tributaries.

Our estimate is around 40 106 t.yr-1 over the period 1997-2004, corresponding to a specific

load of 280 t. km-2.yr-1. The figures show both the variability linked to year to year variations

in the hydrology and sub-basin to sub-basin differences related to their lithology and

82


morphology. The specific loads of the Thao, the Da and the Lo rivers during the same period

are respectively 394, 127 and 282 t. km-2.yr-1. Our results are lower than the range of previous

estimates (Table 3.4), which, however, in many cases are based on measurements made

several decades ago.

Many authors have discussed the increase in erosion and riverine suspended-solid transport

resulting from forest clearance in South-East Asian countries. The Philippines is a well

documented example where sediment load has increased from 1,100 t.km-2.yr-1 to 4,500 t.km-

2.y-1 during the last three decades (Dudgeon et al., 2000). On the other hand, the construction

of large dams has resulted in a decrease of suspended-solid transport by many Asian rivers.

Walling and Fang (2003) thus report the case of the Yellow River in China whose the

sediment flux declined by 50% between the 1950s and the 1980s.

Over the period investigated, dating back to about 40-50 years, the data we collected for the

Red River show no evidence of an increase in suspended loading, even at the scale of

individual sub-basins, despite the well-documented reduction of forested areas and increase of

bare land in the watershed. The suspended-solid loads observed at the Son Tay station in the

1930s by Pouyanne (1931), prior to the period of intense deforestation, were already as high

as 500 mg.l-1 by low flow and 3,500 mg.l-1 by high flow. It seems that the material eroded

from the upstream basin does not reach the downstream course of the Red River and its main

tributaries.

On the other hand, the impoundment of two large dams in the Da and the Lo watersheds has

resulted in a considerable reduction of the total suspended load carried to the sea by the Red

River. Using a simplified modeling approach, we estimated this reduction to about 70%, on

the basis of a calculation carried out with the hydrological data of the last 8 years. The

planned construction of two additional dams would further reduce the total suspended load by

20%, according to our model. This is also the order of magnitude of the expected increase in

suspended loading due to higher rainfall rates induced by climate change (Table 3.6), so that

over the long term, one effect should compensate for the other.

3.7. References

Billen, G., Garnier, J. and Hanset, P., 1994. Modelling phytoplankton development in whole

drainage networks: the RIVERSTRALER model applied to the Seine river system. In: J.

P. Descy, C. S. Reynolds. and. J. Padisak (eds.). Phytoplankton in Turbid Environments:

Rivers and Shallow lakes. Belgium, Kluwer academic Publishers, Belgium.

Hydrobiologia 289: 119-137.

83


Chinadata, 1998-2001. Statistical yearbook of Yunnan, Vol. 1997, Vol. 1998, Vol. 1999, Vol.

2000. China Statistical Publishing House, (Basic Information of Yunnan, China).

(http://chinadatacenter.org).

Dang, Anh Tuan, 2000. The Red River Delta - The Cradle of the Nation. Hanoi national

university publisher, Hanoi.

Dang, Quang Tinh, 2001. Participatory planning and management for flood mitigation and

preparedness and trends in the Red River basin, Vietnam. Workshop international on

Strengthening capacity in participatory planning and management for flood mitigation

and preparedness in large river basin, Bangkok (Thailand) 20th-23rd Nov.

Dudgeon, D., Choowaew, S. and Ho, S.C., 2000. River conservation in South-East Asia. In

Boon P.J, Davies B.R and G.E Petts (eds.). Global perspectives on river conservation:

Science, Policy and Practice. John Wiley& Sons Ltd., Chichester, England. 564 pp.

Dürr, H.H., 2003. Towards a typology of global river systems : some concepts and examples

at medium resolution. Ph-D, University Paris VI - Pierre and Marie Curie. 432 pp.

Fullen, M.A., Mitchel, D.J., Barton, A.P., Hocking, T.J., Liu, Liguang, Wu, Bo Zhi, Zheng,

Yi and Xia, Zheng Yuan., 1998. Soil erosion and Conservation in the Headwaters of the

Yangtze River, Yunnan Province, China. In M.J. Haigh, J. Krecek, S. Rajwar and M.P.

Kilmartin (eds.), Headwaters: Water resources and Soil conservation. pp: 299-306.

Holeman, J.N., 1968. Sediment yield of major rivers of the world: Water Resources Res., 4:

737-747

IMH, 1997-2004. Vietnamese Journal of Meteo-hydrology, Institute of Meteo-Hydrology in

Vietnam, Hanoi. (Journal monthly during the periods of from 1997 to 2004)

IPCC, 2001. Climate change 2001. Intergovernmental Panel on climate change. Third

Assessement Report. http://www.grida.no/climate/ipcc_tar/

Le, Bac Huynh, 1997. The especial flood in the Da River and downstream of the Red River in

August 1996: the role of Hoa Binh dam for flood control in the delta. Vietnamese Journal

of Meteo-Hydrology 4 (439): 6-15

Le, Thi Phuong Quynh, Billen, G., Garnier, J., Théry, S., Fézard, C. and Chau, Van Minh

(2005). Nutrient (N, P) budgets for the Red River basin (Vietnam and China). Journal of

Global Biogeochemical cycles.Vol 19, GB2022, doi 10.1029/2004GB002405.

Lisitzin, A.P., 1972. Sedimentation in the world ocean: Soc.Econ.Paleont.Mineral. Spec.Pub.

17, 218 pp.

84

http://chinadatacenter.org/


Ludwig, W., Probst, J.L., and Kempe, S., 1996. Predicting the oceanic input of organic carbon

by continental erosion. Global Biogeochemical Cycles, 10(1): 23-41.

Martin, J.M and Meybeck, M., 1979. Elemental mass-balance of material carried by major

world rivers. Marine Chemistry 7: 173-206.

Meybeck, M., 1982. Carbon, nitrogen and phosphorus transport by world rivers. Am.J.Sci.

282: 401-450

Meybeck, M., Chapman, D. and Helmer, R., 1989. Global freshwater quality: a first

assessment. Cambridge, MA, World Health Organization/United Nations Environment

Programme Basil Blackwell, Inc.

Meybeck, M., Laroche, L., Durr, H.H. and Syvitski, J.P.M. 2003. Global variability of daily

total suspended-solids and their fluxes in rivers. Global and Planetary Change 39: 65-93.

Milliman, J. D., and Mead, R. H., 1983. World-wide delivery of river sediment to the oceans.

The journal of Geology, 91(1): 1-21.

Milliman, J. D., and Syvitski, J.P.M., 1992. Geomorphic/tectonic control of sediment

discharge to the ocean: the importance of small mountainous rivers. The Journal of

Geology 100: 525-544.

Milliman, J.D. 1997. Blessed dams or damned dams? Nature 386: 325-327

MONRE., 1997-2004. Vietnamese Ministry of Environment and Natural Resources. Report

annual on hydrological observation in Vietnam.

MOSTE., 1997. Vietnamese general statistics officer, Ministry of Science, Technology and

Environment of Vietnam, general statistics editor, Hanoi.

Nash, J.E. and Sutcliffe, J.V., 1970. River flow forecasting through conceptual models.

Journal of Hydrology 10: 282–290.

Ngo, Trong Thuan and Tran, Bich Nga, 1998. Erosion of Da and Hong rivers caused by the

Hoa Binh dam operation. Proceedings of International Conference on Economic

development and environmental protection of the Yuan-Red River watershed, Hanoi 4th-

5th Mar.

Nguyen, Huu Khai and Nguyen, Van Tuan, 2001. Geography and Hydrology in Vietnam.

Vietnam National University publisher, Hanoi, Vietnam.194pp.

Nguyen, Ngoc Sinh, Hua, Chien Thang, Nguyen, Chu Hoi, Nguyen, Van Tien, Lang, Van

Ken, Pham, Van Ninh and Nguyen, Vu Trong, 1995. Case study report on Red River

Delta in Vietnam - Project on integrated management and conservation of near shore

85


coastal and marine areas in East Asia region (EAS-35) United Nations Environment

program. Regional coordinating for the East Seas (ESA/RCU).

Nguyen, Viet Pho, 1984. Streams in Vietnam. Science Publishing House, Hanoi, 209 pp.

Nguyen, Viet Pho, Vu, Van Tuan and Tran, Thanh Xuan, 2003. Water resources in Vietnam.

Vietnamese Institute of Meteo-hydrologie. Agicultural Editor (in Vietnamese).

Pham, Quang Son, 1998. Fundamental characteristics of the Red River bed evolution.

Proceedings of International Conference on Economic development and environmental

protection of the Yuan-Red River watershed, Hanoi 4-5 March.

Pouyanne, A.A., 1931. L’hydraulique agricole au Tonkin. Bulletin économique de

l’Indochine, Section A, 34e année, No2, 126 pp.

Rodier, J., 1984. L'analyse de l'eau. 7ème édition. Dunot (ed.), France, 1365 pp.

Syvitski, J.P.M., 2003. Supply and flux of sediment along hydrological pathways: research for

the 21st century. Global and Planetary Change 39: 1-11.

Syvitsk, J.P.M., Peckham, S.D., Hilberman, R. and Mulder, T., 2003. Predicting the terrestrial

flux of sediment to the global ocean: a planetary perspective. Sedimentary Geology 162:

5-24.

To, Trung Nghia, 2000. Flood control planning for the Red River Basin, International

European- Asian Workshop: Ecosystem & Flood 2000, Hanoi, Vietnam, June 27-29.

Tran, Duc Thanh, Yoshiki, Saito, Dinh, Van Huy, Nguyen, Van Lap, Ta, Thi Kim Oanh and

Masaaki, Tateishi, 2004. Regimes of human and climate impacts on coastal changes in

Vietnam. Regional Environmental Change 4: 49-62

Tran, Thanh Xuan and Pham, Hong Phuong, 1998. Impact of Hoa Binh dam on sediments

flux to the downstream of the Red River. Vietnamese Journal of Meteo-Hydrology 4(448):

7-12

Trinh, Dinh Lu and Doan, Chi Dung, 1998. Impact of Hoa Binh dam operation on

hydrological factors at the downstream of the Red River. Vietnamese Journal of Meteo-

Hydrology 4(448): 34-40.

Trinh, Quang Hoa, 1998. Water balance for purpose of socio-economic development in the

Red River delta. Proceedings of International Conference on Economic development

environmental protection of the Yuan-Red River watershed, Hanoi 4th-5th Mar.

Turc, L., 1961. Evaluation des besoins en eau d’irrigation, évapotranspiration potentielle.

Ann. Agron., 12 (I): I 3-49.

86


UNESCO., 1991. Water related issues and problems of the humid tropics and other warm

humid regions. International hydrological Programme. UNESCO, Paris.

UNEP, 1990. United Nations Environment Programme. China Conservation strategy, UNEP

and China Environemental science press, Beijing. China.

Van Maren, D.S. and Hoekstra, P., 2004. Seasonal variation of hydrodynamics and sediment

dynamics in a shallows subtropical estuary: the Ba Lat River, Vietnam. Estuarine, Coastal

and Shelf Science 60: 529-540.

Vörösmarty, C.J, Meybeck, M., Fekete, B. and Sharma, K., 1997. The potential impact of

neo-Castorization on sediment transport by the global network of rivers. In: Walling D.E.

Probst J.L. (Eds.), Human Impact on Erosion and Sedimentation. (Proc. Rabat

symposium, April 1997), IAHS Publication No. 245. IAHS Press, Wallingford, UK, pp

261-273.

Vörösmarty, C.J, Meybeck, M., Fekete, B., Sharma, K., Green, P. and Syvitski, J.PM., 2003.

Anthropogenic sediment retention: major global impact from registered river

impoundments. Global and Planetary Change 39: 169 -190.

Vu, Van Tuan, 2002.a. Report of Program of Model construction for general gestation of

Thac Ba dams basin served to the development socio-economy, General Centre for

Meteorology and Hydrology of Vietnam, Hanoi, 220pp.

Vu, Van Tuan, 2002.b. The Hoa Binh dam and the general gestation of its basin. Report of

General Centre for Meteorology and Hydrology of Vietnam, Hanoi, 250pp.

Walling, D.E and Fang, D., 2003. Recent trends in suspended sediment loads of the world’s

rivers. Global and Planetary Change 39: 111 -126.

World Bank, 1996. Vietnam water resources sector review. Selected working papers of the

World Bank, ADB, FAO/UNP and NGO Water Resources Sectoral Group, Hanoi, 340 pp.

Acknowledgements

This study was realized in the framework of a French-Vietnamese co-operation. Thanks are

due to Georges Vachaud, Research Director at the CNRS, for the coordination of the

programme ESPOIR (CNRS-CNSTV). Le Thi Phuong Quynh’s Ph-D thesis is supported by

the French Ambassy and by the Pierre and Marie Curie University (Paris 6).

87


88

Water quality in the Red River system

CHAPTER 4

Water quality in the Red River System

Few data are available on water quality in the Red River system, both in China and in

Vietnam, excepted those collected at the outlet of the rivers in the delta area by the

Oceanographic Institute of Nha Trang (Dr. Tac An, pers. comm.). For filling this gap of

knowledge, we decided to organize monthly sampling campaigns at the outlet of each three

sub-basins Da, Lo and Thao and in the main branch of the Red River over two annual cycles

(2003 and 2004). In addition, the water quality of the Nhue-To Lich urban system draining

the large city of Hanoi and its densely populated and industrialized surroundings were

analysed, so that it can be compared with the water quality of the contrasted upstream sectors.

The methods are described in Chapter 2. The results of the analyses will be used for both

establishing nutrients fluxes (Chapter 5) and validating the model (Chapter 6).

Beside nutrients (nitrogen, phosphorus, silica), other informative variables (conductivity,

dissolved oxygen, chlorophyll a, etc.) will help to better characterise the water quality and

nutrient status of the Red River.

4.1. Discharge variations

The discharge values in the years do not show much difference in 2003 compared to 2004

(Figure 4.1). During the recent period analysed (1997-2004), the year 2002 was the wettest

one, mainly due to the contribution of the Da R., (see chapter 3, Le Thi Phuong Quynh et al.

submitted).

0

4000

8000

12000

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

24.00

Dis

char

ge, m

3 s-1

Son TayHanoi

J F M A M J J A S O N D J F M A M J J A S O N D2003 2004

0

2000

4000

6000

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

24.00

Dis

char

ge, m

3 s-1

Thao R.Lo R.Da R.


Figure 4.1: Interannual variations of the discharge in 2003 and 2004, in the three main tributaries of the Red River (Thao, Da, Lo) and in the main branch at the stations Son Tay and

Hanoi).

89


To summarize, the hydrological regime of the Red River is of monsoon type, with higher

discharges in summer. The contribution of the Da River (on right bank) to the discharge of the

Red River is the highest, the levels being similar for the Thao River (the upstream Red River)

and the Lo River. The discharge of the main branch follows the same trend, but due to

distributaries in the delta, the discharge at the Hanoi station is lower than at the Son Tay

station, located immediately downstream from the confluence of the three main sub-basins

(Figure 4.1).

4.2. Physical-chemical variables

4.2.1. Temperature and conductivity

During the two years 2003 and 2004, the water temperature in the Red River system was in

the range from 17.2 to 30.4 °C, mean value averaging 250 C at the outlets of the three

tributaries (sub-basin) and at the three stations of the main branch (Figure 4.2). The similar

temperature was also found in the urban To Lich and Nhue rivers (Figure 4.2).

Regarding conductivity, whereas values varied in a narrow range around 20 µS m-1 in the Red

River tributaries and its main branch, much higher values (70 µS m-1) were found in the To

Lich, typically indicating the importance of the pollution. This small To Lich River (discharge

around 5 m3 s-1, from 1.5 to 15 m3 s-1 in extreme values) can be considered as a waste water

collector draining Hanoi city (Figure 4.2). The To Lich River represents a significant source

of pollution for the Nhue River (average discharge at 35 m3 s-1, from 8 to 50 m3 s-1 in extreme

values).

The pH values did not vary much in the Red River, as well as in the urban rivers (around 7.4,

extreme values from 6.8 to 8).

4.2.2. Suspended matter and dissolved oxygen

In addition to its role in the equilibrium of oxygen, i.e. aquatic life (production vs.

respiration, suspended solids (SS) and light climate (Garnier and Benest, 1991; Ryding and

Thornton, 1999; Garnier et al., 2001) are currently monitored in rivers because it is a major

carrier of inorganic and organic pollutants, as well as nutrients (Meybeck et al. 1989). Most

toxic heavy metals, organic pollutants, pathogens, and nutrients such as phosphorus and

appreciable amount of biodegradable organic material are associated to suspended material.

Measurements of suspended solids are also relevant to other environmental issues such as soil

conservation, land denudation, rocks weathering, inputs of elements to the ocean,

sedimentation rate in reservoirs, river bed erosion, etc. (Meybeck et al., 1989).

90


In order to evaluate the representation of our monthly sampling survey, we have plotted the

daily values available for the year 2003 only, with those we gathered during the study (Figure

4.3).

Figure 4.2: Seasonal variations, during the years 2003 and 2004, of water temperature and

conductivity (conduct.) in the three main tributaries of the Red River (Thao, Da, Lo) and in

the main branch at the stations Son Tay, Lien Mac and Hanoi). The urban rivers Nhue and To

Lich are shown for comparison for the years 2002 and 2003.

It has been long recognized that low frequency of sampling programs lead to a severe

underestimation of the mean annual suspended solid concentration. This is indeed the case

when we compare the mean of our monthly measurements with those found at the same

station from the daily values: 533 against 698 mg l-1 respectively for the Thao R., 35 against

74 mg l-1 for the Da R. and 92 against 170 mg l-1 for the Lo R..

0

10

20

30

40

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Tem

p, °C

Thao R.Lo R.Da R.

J F M A M J J A S O N D J F M A M J J A S O N D0

20

40

60

80

100

0 1 2 3 4 5 6 7 8 9 10 11 1213 14 1516 17 1819 20 21 2223 2

Thao R.Lo R.Da R.

Con

duct

, µS

m-1

0

10

20

30

40

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Tem

p, °C

Son TayLien MacHanoi


4J F M A M J J A S O N D J F M A M J J A S O N D

0

20

40

60

80

100

0 1 2 3 4 5 6 7 8 9 10 1112 13 1415 1617 1819 2021 2223 2


Con

duct

, µS

m-1

4J F M A M J J A S O N D J F M A M J J A S O N D

2003 2004

0

10

20

30

40

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

11.0

12.0

13.0

14.0

15.0

16.0

17.0

18.0

19.0

20.0

21.0

22.0

23.0

24.0

Tem

p, °C

Nhue R.

0

20

40

60

80

100

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

11.0

12.0

13.0

14.0

15.0

16.0

17.0

18.0

19.0

20.0

21.0

22.0

23.0

Nhue R.To Lich R.To Lich R.

Con

duct

, µS

m-1

J F M A M J J A S O N D J F M A M J J A S O N D

2002 20032J F M A M J J A S O N D J F M A M J J A S O N D

2002 2003

91


Thao R.

0

1000

2000

3000

4000

5000

0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00

11.00

12.00

SS, m

g l-1


Lo R.

0

500

1000

1500

2000

0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00

11.00

12.00

SS, m

g l-1


Da R.

0

250

500

750

1000

0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00

11.00

12.00

SS, m

g l-1


Figure 4.3: Seasonal variations, during the year 2003, of suspended solids (SS) from daily

(continuous lines) and monthly sampling (open circles) in the three main tributaries of the

Red River (Thao, Da, Lo).

Mean suspended solid concentrations appeared higher in 2004 than in 2003 (Figure 4.4), by a

factor of 5. This difference is significant with respect to the one caused by sampling

frequency. It cannot be explained by the hydrology that was comparable for the two years.

Note here again, as shown in chapter 3, the SS concentrations in the Thao River were much

higher than the ones in the Lo and Da Rivers (Figure 4.4). The SS concentrations in the main

branch typically represented a mixing of the three upstream water masses.

92


0

2

4

6

8

10

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

11.0

12.0

13.0

14.0

15.0

16.0

17.0

18.0

19.0

20.0

21.0

22.0

23.0

24.0

Oxy

gen,

mg

l-1

Nhue R.To Lich R.


2002 2003

0

100

200

300

400

500

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

11.0

12.0

13.0

14.0

15.0

16.0

17.0

18.0

19.0

20.0

21.0

22.0

23.0

24.0

SS, m

g l-1


2002 2003

0

2

4

6

8

10

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Oxy

gen,

mg

l-1

Thao R.Lo R.Da R.


0

2

4

6

8

10

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Oxy

gen,

mg

l-1



0

1000

2000

3000

4000

5000

0 1 2 3 4 5 6 7 8 9 1011 121314 1516 1718 1920 2122 2324

SS, m

g l-1


0

1000

2000

3000

4000

5000

0 1 2 3 4 5 6 7 8 9 10111213 14151617 18192021 222324

SS, m

g l-1


Figure 4.4: Seasonal variations, during the years 2003 and 2004, of dissolved oxygen and

suspended solids (SS) in the three main tributaries of the Red River (Thao, Da, Lo) and in the

main branch at the stations Son Tay, Lien Mac and Hanoi). The urban rivers Nhue and

To Lich are shown for comparison for the years 2002 and 2003.

Dissolved oxygen concentrations appeared in average lower in 2004 than in 2003 in the

tributaries and also in the main branch of the Red River (Figure 4.4), simultaneously with

lower concentration in suspended solids (SS).

From April to September, i.e. during the rainy season, oxygen concentrations averaged 5.5 mg

O2 l-1 in 2003 against 4.5 mg O2 l-1 in 2004 for the upstream tributaries, and 6.5 mg O2 l-1 in

2003 against 4.5 mg O2 l-1 in 2004 for the main branch. These differences might be explained

by the difference in SS that besides limiting photosynthesis and algal growth are known to be

a support for heterotrophic bacteria which consume oxygen.

93


In the urban river, concomitantly to much lower suspended solids than in the Red River (by

a factor of 100), oxygen concentration was lower due to water organic (domestic) pollution,

and much variable, the water becoming occasionally anoxic.

4.3. General pattern of nutrients

4.3.1. Inter-comparison of nutrient analyses by two laboratories

Before the beginning of this study, the Vietnamese laboratory (INPC, Institute of Natural

Products) did not currently measured nutrients with the standard methods used here.

Therefore a transfer of methods was realized and many samples have been analyzed in

duplicate with the same methodologies in the two laboratories involved in the study (Figure

4.5). The results are compared in x-y graphs (Figure 4.5).

N-NO3, mgN l-1

0.0

0.5

1.0

1.5

2.0

0.0 0.5 1.0 1.5 2.0INPC

Sisy

phe

N-NH4, mgN l-1

0.0

0.5

1.0

1.5

2.0

0.0 0.5 1.0 1.5 2.0INPC

Sisy

phe

DSi, mgSi l-1

0

2

4

6

8

10

0 2 4 6 8 10INPC

Sisy

phe

P-PO4, mgP l-1

0.0

0.5

1.0

1.5

0.0 0.5 1.0 1.5INPC

Sisy

phe

Figure 4.5: Relationship between the results of analyses carried out in parallel by the two

concerned laboratories (Sisyphe, Paris and INPC, Hanoi) for the stations of the Red River in

the years 2003 and 2004 and one of the Nhue for the years 2002 and 2003.

94


Although the results are in the same range, a considerable variability is observed, due to

several factors. Besides the quality of the spectrophotometer (see method, Chapter 2), the

quality of the chemical product used and the dilution water, the analyzes of the Paris

laboratory were realized on samples that were transported frozen but sometimes thawed at the

arrival (24 hours later) and frozen again until analysis. According to these results, although

interpreting seasonal variations could be speculative when of less than a factor of 2, we can

however state that the general levels of nutrient concentrations are correctly estimated.

4.3.2. Nutrient variations

Since the industrial revolution, human activities have caused strong impact on structure and

function of their environment, including the aquatic environment. In recent years, human

perturbations of agricultural, domestic and industrial origins have largely impacted on water

quality. Some influences, like deforestation, agricultural fertilizers, fossil fuel combustion and

urbanization, result in increasing contamination (N, P, heavy metals) in rivers, while others,

like reservoir construction, soil conservation, result in decreasing concentration of silica and

of suspended solid associated nutrients.

The enrichment of riverine water in nitrogen and phosphorus, together with decreasing

suspended solids and silica, often result in eutrophication of coastal marine (Conley et al.

1993; Billen and Garnier, 1997; Cugier et al., 2005), characterized by non-diatoms harmful

algal blooms.

Nitrate and ammonium

Nitrate content in water river originates mainly from leaching of agricultural lands (Billen et

al., 1998; Billen and Garnier, 1999), but in river sectors impacted by domestic wastewater, a

significant contribution originates from the nitrification of ammonia (Chestérikoff et al.,

1992; Brion et al., 2000, Garnier et al., 2001). In river catchment influenced by agricultural

activities, nitrate contamination increased in parallel with the quantity of fertilizer used.

In the Western European rivers, e.g. the Seine River upstream from Paris, nitrate

concentration has increased by a factor of about 5 from the 1950’s to 2000 (from 1.5 mgN.L-1

up to 8 mgN.L-1) while nitrogen fertilizers application increased from 13 kgN.ha-1.y-1 to 150

kgN.ha-1.y-1 during the same period of time. In Vietnam, according to the FAO database

(FAO, 1990-1998), the use of nitrogen fertilizers has increased 66 folds during the period

from 1961 to 2000 (from 2.2 kgN.ha-1.y-1 to 150 kgN.ha-1.y-1) but the concentrations in the

Red River system are still low (Figure 4.6), compared to those found in Western Europe.

95


0.00.51.01.52.02.5

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

20.00.51.01.52.02.5

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

24.00

NO

3, m

g N

l-1

Figure 4.6: Seasonal variations, during the years 2003 and 2004, of nitrate (NO3, mgN l-1)

and ammonium (NH4, mgN l-1) in the three main tributaries of the Red River (Thao, Da, Lo)

and in the main branch at the stations Son Tay, Lien Mac and Hanoi). The urban rivers Nhue

and To Lich are shown for comparison for the years 2002 and 2003.

Taken into account the variability of the analyses, it is difficult to put in evidence any

significant seasonal variations in the nitrate concentrations. The highest values which are

logically found in the rainy season in 2003 (nitrate is of diffuse origin), are found in April,

before the rainy season (Figure 4.6). Similarly, the nitrate concentrations in the three main

tributaries might not be significantly different, although the highest values are observed in the

Lo and Thao Rivers. Land use in the Lo basin is dominated by agriculture, while the Thao

basin is the more populated compared to the Da, less impacted. Nitrate concentrations

averaged 0.5 mg N-NO3 l-1 for the Lo and Thao R. and 0.18 mg N-NO3 l-1 for the Da R.

respectively. In the main branch, the nitrate concentrations have similar levels, reflecting the

mixing of the waters (0.31 N-NO3 l-1 at the Son Tay upstream station, 0.36 N-NO3 l-1 at the

downstream Hanoi station, in average). In the urban rivers, nitrate concentrations are much

0.00.5

01.5

05

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

11.0

12.0

13.0

14.0

15.0

16.0

17.0

18.0

19.0

20.0

21.0

22.0

23.0

24.0

NO

3, m

gN l-1

1.

2.2.

Nhue R.To Lich R.


2002 2003

0

5

10

15

20

25

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

11.0

12.0

13.0

14.0

15.0

16.0

17.0

18.0

19.0

20.0

21.0

22.0

23.0

24.0

NH

4, m

gN l-1

Nhue R.To Lich R


2002 2003

Thao R.Lo R.Da R.


0.00.51.01.52.02.5

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

24.00

NO

3, m

gN l-1



4.00

NH 4

, m

gN l-1

Thao RLo R.Da R.


0.00.51.01.52.02.5

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

Son TayLien MaHanoi

NH

4, m

gN l-1

2.

J F M A M J J A S O N D J F M A M J J A S O N D2003 2004 0

96


more fluctuating, with mean values much higher than in the Red River: 3 mg N-NO3 l-1 in the

Nhue and 2 mg N-NO3 in the To Lich.

Regarding ammonium, concentrations are usually low in natural waters, as it is taken up very

quickly by microorganisms including autotrophic algae, heterotrophic bacteria, and

autotrophic nitrifying bacteria. High ammonia level in water is typically a sign of domestic

wastewater pollution. The average ammonium concentrations at the outlet of the rivers Thao,

Lo and Da (Yen Bai, Vu Quang and Hoa Binh stations respectively) are very low: 0.1, 0.06

and 0.03 mgN-NH4.L-1, respectively (Figure 4.6).

The mean values increased considerably in the main branch, from upstream (Son Tay station)

to downstream (Hanoi station) i.e., from 0.1 to 0.85 mgN-NH4.L-1, and much more in the

urban rivers, the Nhue (2.7 mgN-NH4.L-1) and the To Lich river (9.5 mgN-NH4.L-1).

Contrarily to the nitrate concentrations that tend to increase during rainy seasons under

leaching from the agricultural lands, ammonium concentrations in the To Lich tended to show

a dilution (Figure 4.6).

Whereas the values found for nitrate are still far below the Vietnamese Standards (15 mgN-

NO3.L-1), it was not the case for ammonium in urban rivers, which were clearly above the

standards of 1.0 mgN-NH4.L-1).

Considering the proportion of nitrate, nitrite and ammonium in total inorganic nitrogen, it

appeared that nitrate was, in proportion, the dominant form (around 80 %) in the upstream

basins, decreasing in the main branch (from 69 to 25 %) at the benefit of ammonium. The

proportion in nitrite remained low (< 2 %).

In the Nhue and To Lich rivers, the proportion in ammonium reached up to 98 %. Note that,

downstream of the city of Paris, after the treated domestic effluents of the 6.5 million

inhabitants discharging their waste waters to the purification plant of Achères and then being

driven to the lower Seine River, ammonium and nitrate are in a 50 %-50 % proportion (5 mg

N-NO3 l-1 and 5 mgN-NH4.L-1), (Garnier et al., 2001).

97


Table 4.1: Proportion (%) of nitrate (N-NO3), nitrite (N-NO2) and ammonium (N-NH4)

compared to the total inorganic nitrogen at the different stations in the sub-basins (Thao, Lo

and Da), in the main branch (Son Tay and Hanoi), and in the urban river system (To Lich and

Nhue).

Location N-NO3, % N-NO2, % N-NH4, %

Thao (Yen Bai)

Lo (VuQuang)

76

84

79

2

1

1

22

15

20 Da (Hoa Binh)

Hong (Son Tay) 69

25

2

1

29

74 Hong (Hanoï)

To Lich river 2

Nhue river 9

0

1

98

90

Phosphate and total phosphorus

In freshwater, phosphorus is often the main factor that limits the production of plant biomass.

Phosphate can be dissolved or adsorbed to particles, remaining available by desorption

(Némery, 2003; Némery et al., 2005). Phosphorus was limiting in the range from 0.01 to 0.04

mgP-PO4 l-1, which corresponded to the value of the half-saturation constant for phosphate

uptake by algae (Garnier et al., 1995; Garnier et al., 1998; Garnier et al., 2005).

In the upstream tributaries, average concentrations of 0.03, 0.03 and 0.02 mgP-PO4 l-1 were

found in the Thao, Lo and Da Rivers respectively. Such low levels might be limiting for algal

growth at least at certain periods, depending on the seasonal variations (Figure 4.7).

However, total phosphorus concentrations were much higher, 0.29, 0.18 and 0.16 mg P l-1 in

the Thao, Lo and Da respectively (Figure 4.7), a significant proportion being probably

exchangeable. Taking into account the Redfield ratio (Redfield et al., 1963), and the amount

of TOC (see below), at least 20 to 40 % of the total phosphorus can be estimated to be under

mineral form. Differences in the total phosphorus levels between the sub-basins reflect mainly

their difference in suspended matter concentration, although the higher concentrations in the

Thao River are also due to its higher population density.

In the main branch, phosphates and total phosphorus, increased from 0.03 and 0.23 mg P l-1

to 0.11 and 0.27 mg P l-1 respectively, from upstream (Son Tay station) to downstream (Hanoi

station), due to the increase of population density along the river bank in the delta area

98


(Figure 4.7). The total phosphorus concentration in the downstream sector of the Red River is

very close to the one observed in the Amazon (0.24 mgP l-1) (Meybeck and Ragu, 1996).

0.0

0.5

1.0

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

24.00

PO4,

mg

P l-1

Thao R.Lo R.Da R.


0.0

0.5

1.0

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

24.00

PO4,

mgP

l-1



0.0

0.5

1.0

1.5

2.0

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

24.00

Tot P

, m

gP l-1 Thao R.

Lo R.Da R.


0.0

0.5

1.0

1.5

2.0

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

24.00

Tot P

, mgP

l-1



0.0

1.0

2.0

3.0

4.0

5.0

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

11.0

12.0

13.0

14.0

15.0

16.0

17.0

18.0

19.0

20.0

21.0

22.0

23.0

24.0

PO4,

mgP

l-1

Nhue R.To Lich R.


0.0

1.0

2.0

3.0

4.0

5.0

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

11.0

12.0

13.0

14.0

15.0

16.0

17.0

18.0

19.0

20.0

21.0

22.0

23.0

24.0

Tot P

, mgP

l-1

Nhue R.To Lich R.


Figure 4.7: Seasonal variations, during the yeas 2003 and 2004, of dissolved phosphates (PO4, mgP l-1) and total phosphorus (Tot P, mgP l-1) in the three main tributaries of the Red River (Thao, Da, Lo) and in the main branch at the stations Son Tay, Lien Mac and Hanoi. The urban rivers Nhue and To Lich are shown for comparison for the years 2002 and 2003.

According to Nguyen Viet Pho (1984), in the estuarine water of the Red River, the mean total

phosphorus concentrations varied from 0.21 to 0.56 mgP l-1 between the end of the dry season

and the flood period. This statement would imply that phosphorus is more issued from

diffuse than from point sources. Our results do not show any clear seasonal variation.

In urban rivers, phosphorus levels drastically increased, largely fluctuating up to 3 mg P l-1 for

phosphates (1.8 mg P l-1 in average) and 5 mg P l-1 (2.8 mg P l-1 in average) for total

phosphorus in the To Lich, sensibly diluted in the Nhue (in average, phosphates equal 0.5 mg

P l-1 and total P, 0.7 mg P l-1). The fraction of dissolved phosphate in the total phosphorus

concentration, represents less than 15 % in the upstream rivers where it comes mainly from

99


diffuse sources, but reaches 40 % in the Red River (at Hanoi), and up to 70 % in the urban

rivers where domestic pollution becomes a major source.

Phosphorus content of suspended solid (Tot P – P-PO4/ SS, in mgP gSS-1) in the 3 major

tributaries varied from 0.43 mgP gSS-1 in the Thao River, the most turbid of the tributaries, to

0.85 in the Lo river (Table 4.2). In the main branch, the values were slightly higher, from 0.7

to 1.2 mgP gSS-1, and increased to 18 mgP gSS-1 in the To Lich River (Table 4.2). A similar

upstream downstream gradient has been observed in the drainage network of the Seine River

by Némery (2003), with phosphorus content of suspended matter ranging from 1 mgP gSS-1 in

small streams draining agricultural soils, to values as high as 6 mgP gSS-1 in the Seine

downstream from Paris agglomeration.

Table 4.2: Phosphorus content (mgP gSS-1) of suspended solids (SS, mg L-1) at the different stations in the sub-basins (Thao, Lo and Da), in the main branch (Son Tay and Hanoi), and in

the urban river system (To Lich and Nhue). Average values for the two study-years.

Location TP, mgP gSS-1 SS, mg L-1

Thao (Yen Bai)

Lo (VuQuang)

0.43

0.85

0.65

1550

460

110

Da (Hoa Binh)

Hong (Son Tay) 0.7

1.2

640

600

Hong (Hanoï)

To Lich river 18.0

Nhue river 11.0

70

50

Dissolved silica and algal pigments

Dissolved silica concentrations in rivers mainly originate from rock weathering, and therefore

depend on the lithology (Meybeck, 1986). The lithological composition of the Red River

watershed is dominated by sedimentary rocks, with about half of carbonated rocks (Dürr,

2003; H. Durr and M. Meybeck, pers. comm., based on data from the UNESCO World

Geological Map). Meybeck (1986) assigned a silica concentration between 2 and 5 mgSi.L-1

to these lithological types. In addition, it was shown that, for a given rock composition, the

silica concentration in drainage water is much higher under warm and wet climate than under

colder climatic conditions (Meybeck, 1986; Garnier et al., in press). Dissolved silica

100


concentrations averaged 4 mgSi l-1 in the Lo and the Da Rivers (Figure 4.8); values were

notably higher in the Thao River (5.4 mgSi l-1), probably explaining by its lithology

characterized by a greater proportion of basic volcanic rocks and silico-clastic sedimentary

consolidated rocks. In the main branch, the DSi concentrations are typically intermediate (4.5

mgSi l-1).

0

5

10

15

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

24.00

DSi

, mg

l-1

Thao R.Lo R.Da R.


0

5

10

15

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

24.00

DSi

, mg

l-1



0

20

40

60

80

100

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 2

T C

hla

, µg

l-1

Thao RLo R.Da R.


0

20

40

60

80

100

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1516 17 18 19 2021 22 23 2

T C

hla,

µg

l-1Son TayLien MacHanoi


0

5

10

15

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

11.0

12.0

13.0

14.0

15.0

16.0

17.0

18.0

19.0

20.0

21.0

22.0

23.0

24.0

DSi

, mg

l-1

Nhue R.To Lich R.


0

20

40

60

80

100

0 1 2 3 4 5 6 7 8 9 1011 12 13 1415 16 17 1819 20 21 2223 2

T C

hla

, µg

l-1

Nhue R.To Lich R


Figure 4.8: Seasonal variations, during the years 2003 and 2004, of dissolved silica (DSi, mg Si L-1) and Chlorophyll a + Pheopigments (T Chl a, µg L-1) in the three main tributaries of the

Red River (Thao, Da, Lo) and in the main branch at the stations Son Tay, Lien Mac and Hanoi. The urban rivers Nhue and To Lich are shown for comparison for the years 2002 and

2003.

Higher dissolved silica concentrations were found in the urban rivers (7.9 and 5.9 mgSi l-1 in

the To Lich and Nhue, respectively). This would tend to show that effluent of the Hanoi city

could be a source for silica, the silica being then diluted in the Nhue (Figure 4.8). The origin

of these high silica concentrations in wastewater is not clear. It was already noted that

domestic wastewater in Europe contains significant dissolved silica concentrations related to

the use of sodium metasilicates as a corrosion inhibitor in modern washing powders (Billen et

101


al, 2001; Garnier et al., 2002). On the other hand, industrial effluents may also be a significant

source: analyzing the effluents of a number of industries we found particularly high

concentrations (more than 30 mgSi l-1) in some of them.

Except in the urban system where the large fluctuations of concentrations were observed, the

silica concentrations in the Red River tributaries and main branch, were rather stable (Figure

4.8), showing that biological consumption of silica is low. Diatoms use silica to elaborate

their frustules and in eutrophicated rivers where nitrogen and phosphorus are not limiting,

silica can be seriously depleted during algal blooms (Garnier et al., 1995; Garnier et al., 1998;

Billen et al., 2005). When the Redfield ratios in the water (Si:N, Si:P: Redfield et al., 1963)

are too low compared to the algal requirement, silica becomes limiting for the diatoms, which

are then replaced at the coastal zones mostly, by other non-siliceous algae, sometimes

producing toxins, a phenomenon known as harmful algal bloom (HAB). Such situations are

currently encountered in North Western Europe, in the Manche Channel and North Sea

(Lancelot, 1995; Cugier et al., 2005), Black Sea (Humborg et al., 1997) or in the Gulf of

Mississippi (Rabalais and Turner, 2001). There is, of course, no Vietnamese standard level

for silica concentrations, but to avoid silica depletion and harmful non-diatom blooms, it is

necessary to control the N and P inputs to the rivers.

Levels of phytoplankton biomass, as expressed by the sum of chlorophyll a and pheopigments

(T Chl a, µg L-1), were relatively low. The values were the highest in the Thao river, despite

its higher suspended solid concentration compared to the other two tributaries (Figure 4.8,

Table 4.3). As the Thao was also the richest in phosphorus, this would suggest that algal

growth in the Red River tributaries is more limited by phosphorus than by available light. A

further increase in phytoplankton biomass occurs in the main branch, where nutrients

concentrations increase too. In the To Lich, phytoplankton biomass was rather high, but this

biomass can originate from the several fish ponds, in communication along its course.

Phytoplankton production can occur despite suspended matter concentrations as high as 60

mg l-1, in the absence of nutrient limitation (cf. Garnier et al., 2001). In the Nhue river, the

variations of phytoplankton concentrations is closely parallel to those of the To Lich river

(Figure 4.8).

4.4. Organic matter

Organic carbon is found under dissolved or particulate form and is either autochthonous

(produced in situ by algal biomass production and subsequently released by lysis or

102


excretion) or allochtonous (brought to the river from soil leaching, or domestic and industrial

effluents).

Leaching of the organic layers of soils is the primary source of dissolved organic matter

(DOC) in rivers. The level of DOC resulting from this process is strongly influenced by the

regional vegetation, climate and hydrology (Sempéré et al., 2002; Lilienfein et al., 2001). In

Nordic countries, high DOC values (up to 15 mgC.L-1) were found in rivers draining forested

and peatland area, with considerable seasonal variations linked to variations of temperature

and hydrology (Bishop and Pettersson, 1996). Lobbes et al. (2000) estimated that TOC

concentrations (total = dissolved + particulate) of 12 Russian rivers which enter into the Artic

Ocean ranged from 2.8 to 12.1 mgC.L-1. Meybeck and Ragu (1996) reported the mean value

of DOC and the total organic carbon TOC of rivers in the Amazon zone was about 4.0 and 6.6

mgC.L-1 respectively. Meybeck (1988) concluded that dissolved organic carbon concentration

for the wet tropics were higher than those in dry tropical regions and also higher than those in

temperate zones and proposed a mean value of 8 mgC L-1 for dissolved organic carbon in the

wet tropical regions. In the head waters of temperate climates, DOC originating from soil

leaching is mostly refractory (Servais et al., 1998).

When brought by domestic effluent, a large part is biodegradable (> 50 %, Servais et al.,

1995), these inputs possibly leading to oxygen depletion (or even anoxia), due to the

respiration of heterotrophic bacteria (Servais and Garnier, 1993; Garnier et al., 2001; Garnier

et al., 2004). Similarly, when autochthonous primary production is high, due to ample nutrient

concentrations, the organic biomass of the organisms can represent a large stock of

biodegradable organic matter, the heterotrophic degradation of which can also lead to oxygen

depletion (Garnier et al., 1999; Garnier et al., 2001; Garnier et al., 2004). These two types of

organic pollution, can lead to reduce the oxygen level down to values inappropriate for

aquatic life, fish in particular.

The mean DOC values at Hoa Binh, Vu Quang and Yen Bai, Son Tay and Hanoi during the

years 2003 and 2004 were 2.5, 2.6, 2.6, 2.8, and 3.6 mgC L-1 respectively. These values might

seem low compared with the above figures proposed by Meybeck (1988), but probably

reflects the absence of alluvial forests in the Red River basin.

The results of POC occasional analyses are shown in table 4.3. These values are lower than

those proposed by Ittekkot and Laane (2002) for the different ranges of river suspended solid

concentration but the DOC/POC ratios giving a range from 0.9 to 12.8 (highest in the Da

River and lowest in the To Lich and Thao River) are very close to the data reported by the

same authors.

103


Phytoplankton biomass figures, which represent a biodegradable fraction of the total organic

carbon have been converted in carbon unit using a C / T Chl a ratio of 24 (Servais and

Garnier, submitted) (Table 4.3). The results indicate that phytoplankton biomass represent

only a small fraction of the particulate organic carbon, and, a fortiori, a small fraction, from 2

to 6 % of total organic carbon (Table 4.3).

0

5

10

15

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

24.00

DO

C, m

gC l-1

Thao R.Lo R.Da R.


0

5

10

15

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

11.00

12.00

13.00

14.00

15.00

16.00

17.00

18.00

19.00

20.00

21.00

22.00

23.00

24.00

DO

C, m

gC l-1



0

5

10

15

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

11.0

12.0

13.0

14.0

15.0

16.0

17.0

18.0

19.0

20.0

21.0

22.0

23.0

24.0

DO

C, m

gC l-1

Nhue R.To Lich R.


Figure 4.9: Seasonal variations, during the years 2003 and 2004, of dissolved organic carbon

(DOC, mgC L-1) in the three main tributaries of the Red River (Thao, Da, Lo) and in the main

branch at the stations Son Tay, Lien Mac and Hanoi. The urban rivers Nhue and To Lich are

shown for comparison for the years 2002 and 2003.

104


Table 4.3: Chlorophyll a + Pheo-pigments (T Chl a, µg L-1), Phytoplankton biomass (Phy, mgC L-1) and particulate and dissolved organic carbon (POC, DOC, mgC L-1) at the different stations in the sub-basins (Thao, Lo and Da), in the main branch (Son Tay and Hanoi), and in

the urban river system (To Lich and Nhue). Average values for the two study-years.

Location Tchla, µg.l-1 Phy (*), mgC l-1 POC, mgC l-1 DOC, mgC l-1 DOC/POC

Thao (Yen Bai)

Lo (VuQuang)

11.0

3.2

3.4

0.3

0.1

0.1

2.3

0.9

0.2

2.6

2.6

2.5

1.1

2.9

12.5(**)Da (Hoa Binh)

Hong (Son Tay) 5.6

6.9

0.1

0.2

1.6

2.8

3.6

2.0 Hong (Hanoï)

To Lich river 37.2

Nhue river 19.1

0.9

0.5

12.0

2.8

10.3

5.4

0.9

1.9 (*)

: Phy, represent the fraction of carbon content due to algal biomass, taking into account a ratio of C: T Chla equalling 24

(Servais and Garnier, submitted). (**) Although the low SS values that can explain the high DOC/POC ratio (Ittekko and Laane, 2002), this ratio was

calculated with a low number of POC data.

4.5. Conclusions: water quality in the Red River

4.5.1. General levels of nutrients in the Red River drainage network

Our two-year surveys of water quality in the Red River tributaries allow for the first time to

assess the general level of nutrient concentrations in this sub-tropical river system (Table 4.4

in the part 4.5.3). As discussed above, these levels are low compared to river systems in the

temperate region of the world with similar population densities. Note that the values found in

head water (see Table 4.4) are close to the ones found at the outlet of the sub-basins, showing

that the low anthropogenic impact within the upstream basin of the Red River. Only in small

urban river systems of the delta region, extreme signs of domestic pollution are present.

Another general conclusion drawn from our survey is the absence of important seasonal

variations of nutrient concentrations, excepted those directly related to suspended solid.

105


Table 4.4: Water quality for the three main tributaries of the Red River (Thao, Da, Lo) and in the main branch at the Hanoi station. The urban rivers Nhue and To Lich are shown for

comparison for the years 2002 and 2003. Values from head water are also given for 5 stations sampled in October 2004 in the surroundings of Lao Cai, close to the Chinese border. DO:

minimum value of oxygen concentration observed; NH4: ammonium; Tot N: sum of inorganic nitrogen; Tot P: total phosphorus; DSi: dissolved silica; T Chla: sum of chlorophyll a and

pheopigments.

Locations DO

mg L-1

NH4

mg.N L-1

Tot N,

mgN.L-1

Tot P,

mgP.L-1

DSi

mgSi.L-1

T Chl a,

μg.L-1

classification

Head waters 0.00 0.4 0.12 4.7 1.5 Oligotrophic

Da 4.9 0.03 0.2 0.16 4.3 3.4 Oligotrophic-Mesotrophic

Lo 5.4 0.06 0.6 0.18 4.2 3.2 Mesotrophic

Thao 5.4 0.10 0.6 0.29 5.4 11.0 Mesotrophic

Red-Hong-Hanoi

5.8 0.85 1.2 0.27 4.5 6.9 Mesotrophic

To Lich 0.9 9.5 9.7 2.80 7.9 37.2 Eutrophic

Nhue 2.9 2.7 3.0 0.70 5.9 19.1 Organically polluted

4.5.2. Behaviour of nutrients with increasing specific discharges in the Red River System

In order to analyse the general trends of variation of nutrients with respect to discharge for all tributaries, we plotted the measured concentration against specific discharge.

Nitrate shows an increase with specific discharge, supporting its predominantly diffuse origin, from soil leaching (Figure 4.10). Similarly, total phosphorus and suspended solid concentrations originate from erosion of soil material to which adsorbed phosphorus is associated. This trend also points out the higher concentrations of these elements in the superficial water rather than in ground waters.

Regarding ammonium and ortho-phosphates, their dilution with increasing specific discharge reveals their point source origin, the dilution being particularly evidenced for the downstream Hanoi station (Figure 4.10). The low phosphate values at high discharge, also results from an efficient adsorption of ortho-phosphates on the high concentrations of suspended solids. Silica concentrations, although showing a large dispersion of values at low specific discharge values, are rather stable within the range of specific discharges observed during the study (Figure 4.10).

106


0.0

0.5

1.0

1.5

2.0

2.5

0 25 50 75 100Spec. disch., L. km-2.s-1

N-N

H4,

mg

l-1

Thao R.Lo R.Da R.Hanoi

0.0

0.5

1.0

1.5

2.0

2.5

0 25 50 75 100Spec. disch., L. km-2.s-1

N-N

O3,

mg

l-1Thao R.Lo R.Da R.Hanoi

0.0

0.5

1.0

0 25 50 75 100Spec. disch., L. km-2.s-1

Tot P

, mgP

l-1


0

Figure 4.10: Relationship between the concentrations of nitrate (NO3), ammonium (NH4), total phosphorus (Tot P), phosphates (PO4), suspended solids (SS) and dissolved silica (DSi) and the specific discharge (Spec. Disch), in the three main tributaries of the Red River (Thao,

Da, Lo) and in the main branch at the station Hanoi for the years 2003 and 2004.

4.5.3. Classification of pollution level

Dodds et al. (1998) and Dodds and Welch (2000) proposed a general typology of rivers

according to their level of nutrient pollution (Table 4.5). On the other hand, Tran Hieu Nhue

et al. (1994) proposed a classification of nutrient pollution level specially adapted for tropical

climatic region like Vietnam (Table 4.6).

1

2

000

000

000

000

000

0 25 50 75 100Spec. disch., L. km-2.s-1

SS, m

g l-1

3

4

5Thao R.Lo R.Da R.Hanoi

0.0

0.5

1.0

0 25 50 75 100Spec. disch., L. km-2.s-1

P-PO

4, m

g l-1


0

5

10

15

0 25 50 75 100Spec. disch., L. km-2.s-1

DSi

, mg

Si l-1


107


Table 4.5: Classification of trophic levels of rivers according to Dodds et al., 1998; Dodds and Welch, 2000.

Trophic level Total N

mgN.L-1

Total P

mgP.L-1

Suspended Chl a,

µg.L-1

Benthic Chl a,

mg.m-2

Eutrophic > 1.5 > 0.075 > 30 > 60

Mesotrophic 0.7 – 1.5 0.025 - 0.075 10 - 30 20 - 70

Oligotrophic < 0.7 < 0.025 < 10 < 20

Table 4.6: Classification of pollution level on the basis of diverse variables of water quality

(DO: dissolved oxygen; BOD5: Biological oxygen demand -5 days-, Tran Hieu Nhue

and al., 1994).

Pollution level

DO BOD5 Organic degradation

Water statement

Nutrient contents Microbial contents

Eutrophic 0 ÷1 > 40 Anaerobic

degradation

Rich in nutrients [NH4+] > 10mg.l-1;

trace of CH4 and H2S

in sediment layer

Strong

development of

microbes

α-

Mesotrophic

1÷3 20÷40 Aerobic

degradation

Rich in nutrients,

occurrence of

algal blooms

[NH4+]: 8÷10mg.l-1

occurrence of

NO2-

Hundreds to

thousands

microbes per

liter

β-

Mesotrophic

3÷5 10÷20 Aerobic

degradation

Rich in nutrients,

frequent

occurrence of

algal blooms

Nitrate and nitrite

content: several

mg.l-1

Several

thousands

microbes per

liter

Oligotrophic > 5 < 10 Stable levels of

organic matter

no algal blooms Nitrate and nitrite

content low and

stable

occurrence of

macrophyte and

pink agar

On the basis of these two references, we tried to classify the nutrient pollution level of the

different sectors of the Red River drainage network as indicated in Table 4.4. The upstream of

the Red River may be classified as Oligotrophic- Mesotrophic (β), the main branch in the

delta area is Mesotrophic (β), while the urban rivers are clearly organically polluted (Table

4.4).

As a whole, the water of the Red River is oligotrophic before entering the urbanized region of

delta, as expected by the origin of the nutrients, essentially of diffuse type, with limited

anthropogenic impact.

108


Among the major types of degradation of surface water that have occurred in the recent times,

the Red River does not seem to be touched in its sub-basins; eutrophication seems to be

limited by nutrients more than by light, at least during the dry season, from September to June

whereas siltation, particularly from agriculture or deforestation, would have not changed

much, and/or counterbalanced by the role of the two reservoirs, on the Lo and the Da rivers

(cf. Chapter 3). However, i) future nutrient enrichment due to increasing population, in urban

areas mainly, and ii) the future impoundment of two additional reservoirs as planned at the

horizon 2010-2015 which will further reduce the suspended solid concentrations, could

together quickly lead to major disruptions in term of river eutrophication.

In the delta, aquatic ecosystems are seriously damaged in a number of classical ways (Wetzel,

2001). We have clearly observed the most common type of degradation through the

contamination by inorganic (NH4) and organic (DOC) pollutants. Other types of degradation

come from irrigation, channelization that modify the aquatic habitats, toxic material, etc.

Presently, the To Lich river has reached a domestic and industrial pollution level close to the

one mentioned at the end of the XIX century in Western Europe, e.g. for the Bièvre urban

tributary of the Seine in Paris intra muros (Billen et al., 1999) or to that of the Senne crossing

Brussels (Garnier et al., 1992). Note that to face such pollution, these two rivers were

covered! It is interesting to note that, after more than 50 years of wastewater treatment effort,

the re-opening of these rivers is presently under debate. The treatment of urban wastewater

appears therefore a priority for Hanoi.

4.6. References

Billen G. and Garnier J., 1997. The Phison River Plume: coastal eutrophication in response to

changes in land use and water management in the watershed. Aqu. Microb. Ecol., 13: 3-

17.

Billen G., Garnier J. and Meybeck M., 1998. Les sels nutritifs: l’ouverture des cycles.

Chapitre 12. In: La Seine en son bassin Fonctionnement écologique d’un système fluvial

anthropisé (Meybeck M., De Marsily G. and Fustec E. eds). Elsevier, Paris. p. 531-565.

Billen G. and Garnier J., 1999. Nitrogen transformation through the Seine drainage network: a

budget based on the application of the 'Riverstrahler' model, Hydrobiologia 410: 139-150.

Billen G., Garnier J., Deligne C., and Billen C., 1999. Estimates of early industrial inputs of

nutrients to river systems: implication for coastal eutrophication. The Sciences of the Total

Environment, 243/244: 43-52.

109


Billen, G., Garnier J., Ficht A. and Cun C., 2001. Modelling the response of water quality in

the Seine estuary to human activity in its watershed. Estuaries, 24(6): 977-993.

Billen G., Garnier J. and Rousseau V., 2005. Nutrient fluxes and water quality in the drainage

network of the Scheldt basin over the last 50 years. Hydrobiologia. 540: 47-67

Bishop, K. H and Pettersson C., 1996. Organic carbon in the boreal spring flood from

adjacent sub-catchments. Environment International, Vol 22 (5): 535-540.

Brion N., Billen G.., Guezennec L. and Ficht A., 2000. Distribution of nitrifying activity in

the Seine River (France) and its estuary. Estuaries, 23: 669-682.

Chesterikoff A., Garban B., Billen G. and Poulin M., 1992. Inorganic nitrogen dynamics in

the River Seine downstream from Paris (France), Biogeochemical mass-balances (C, N, P,

Si) in three large reservoirs of the Seine Catchment (France), Biogeochemistry 17: 147-

164.

Conley D.J., Claire L. S. and Stoermer E. F., 1993. Modification of the biogeochemical cycle

of silica with eutrophication. Marine Ecology Progress Series, 101: 179-192.

Cugier J., Billen G., Guillaud J.F., Garnier J., and Menesguen A., 2005. Modelling the

eutrophication of the Seine Bight (France) under historical, present and future riverine

nutrient loading. Journal of Hydrology. Volume 304 (1-4): 381-396.

Dodds W.K., Jones J.R., and Welch E.B., 1998. Suggested classification of stream trophic

state: Distributions of temperate stream types by chlorophyll, total nitrogen, and

phosphorus. Water Res. 32:1455-1462.

Dodds W.K. and Welch E.B., 2000. Establishing nutrient criteria in streams. J. N. Am.

Benthol. Soc. 19: 186- 196.

Dürr H.H. 2003. Towards a typology of global river systems : some concepts and examples at

medium resolution. Thèse de doctorat, Université Paris VI - Pierre et Marie Curie. 432pp.

FAO, (1990-1998), Faostat statistics database (Fertilizer data used for China, Vietnam, Laos),

Faostat database Results, copyright FAO 1990-1998, http://www.apps.fao.org/;

http://www.fertilizer.fao.org/. 2002.

Garnier J. and Benest D., 1991. Influence of light attenuation on primary production in a

shallow sand-pit lake. Verh. Internat. Verein. Limnol. 24: 1536-1542.

Garnier J., Billen G. and Billen C., 1992. Modélisation des modifications du paysage

écologique des rivières. Application à la Senne. Rapport de synthèse, Action "Research in

Brussels, Secrétariat d'Etat à la Région. 69 pp.

110


Garnier J., Billen G. and Coste M., 1995. Seasonal succession of diatoms and Chlorophyceae

in the drainage network of the river Seine: Observations and modeling, Limnol. Oceanogr.

40: 750-765.

Garnier J., Billen G., Hanset Ph., Testard P. and Coste M,. 1998. Développement algal et

eutrophisation. Chapitre 14. In: La Seine en son bassin Fonctionnement écologique d’un

système fluvial anthropisé (Meybeck M., De Marsily G. and Fustec E. eds). Elsevier,

Paris. p. 593-626.

Garnier J., Billen G. and Palfner L., 1999. Understanding the oxygen budget od the Mosel

drainage network with the concept of heterotrophic/autotrophic sequences: the

Riverstrahler approach. Hydrobiologia, 410: 151-166.

Garnier J., Servais P., Billen G., Akopian M. and Brion N., 2001. Lower Seine river and

estuary (France): carbon and oxygen budgets during low flow. Estuaries, 24: 964-976.

Garnier J., d’Ayguesvives A., Billen G., Conley D. and Sferratore A., 2002. Silica dynamics

in the hydrographic network of the Seine River. Oceanis, 28 : 487-508

Garnier J, Billen G. and Cugier Ph., 2004. Drainage basin use and nutrient supply by rivers to

the coastal zone. A modelling approach to the Seine River. p 60-87 In: Drainage

basin nutrient inputs and eutrophication: an integrated approach. P. Wassmann and

K. Olli (Eds). E-book in press, 309pp. available at: www.ut.ee/~olli/eutr/

Garnier J., Nemery J., Billen G. and Théry S., 2005. Nutrient dynamics and control of

eutrophication in the Marne River system: modelling the role of exchangeable

phosphorus. Journal of hydrology, Volume 304(1-4): 397-412.

Garnier J, Sferratore A., Meybeck M., Billen G. and Dürr H. (in press). Modelling silica

transfer processes in river catchments. Chapter in SCOPE book series.

Humborg C., Ittekot V., Cosiascu A., and Bodungen B. V., 1997. Effect of Danube River dam

on Black sea biogeochemistry and ecosystem structure. Nature, 386: 385-388.

Ittekkot V. and Laane R. W. P. M., 2002. Chapter 10: Fate of Riverine Particulate Organic

Matter, SCOPE 42 - Biogeochemistry of Major World Rivers, available at

http://www.icsu-scope.org/downloadpubs/scope42/chapter10.html

Lancelot C., 1995. The mucilage phenomenon in the continental coastal waters of the North

Sea. The Science of The Total Environment, 165: 83-102.

111

http://www.ut.ee/%7Eolli/eutr/


Le Thi Phuong Q., Garnier J., Billen G., Théry S. and Chau Van Minh (submitted).

Hydrological regime and suspended matter flux of the Red River System (Vietnam):

Observations and modelling. To be submitted to Journal of Hydrology.

Lilienfein J., Wolfgang W., Richard T., Lourival V., Do Carmo Lima S. and Wolfgang Z.,

2001. The effects of Pinus caribaea forests on the C, N, P and S status of Brazilian

savanna Oxisols, Forest Ecology and Management 147: 171- 182.

Lobbes J. M., Fiznar H.P. and Kattner G., 2000. Biogeochemical characteristics of dissolved

and particulate organic matter in Russian rivers entering Artic Ocean, Geochimica et

Cosmochimica Acta, Vol. 64: 2973-2983.

Meybeck M., 1986. Composition chimique des ruisseaux non pollués de France. Sci. Géol.,

Bull. (Strasbourg) 39: 3-77.

Meybeck M., 1988. How to establish and use world budgets of river material, in Physical and

chemical weathering in Geochemical Cycles, edited by A. Lerman and M. Meybeck,

p. 247-272, Kluwer, Norwell, Mass.

Meybeck M., Chapman D.V. and Helmer R., 1989. “Global environment monitoring: Global

freshwater quality, a first assessment”, Chapter 7: Suspended matter in rivers and

lakes , p. 93-104, published by WHO and UNEP.

Meybeck M. and Ragu A., 1996. River discharges to the oceans: An assessment of suspended

solids, major ions and nutrients. UNEP Env. Information and Assessment Report.

Nairobi, 245pp.

Némery J., 2003. Origine et devenir du phosphore dans le continuum aquatique de la Seine,

des petits basins à l'estuaire. Rôle du phosphore échangeable sur l'eutrophisation.

Thèse Univ. Paris 6, 258pp.

Némery J., Garnier J. and Morel C., 2005. Phosphorus budget in the Marne watershed

(France): urban vs. diffuse sources, dissolved vs. particulate forms. Biogeochemistry

(in press).

Nguyen Viet Pho., 1984. Streams in Vietnam. Science Publishing House, Hanoi, 209pp.

Rabalais N.N. and Turner R.E. (eds), 2001. Coastal Hypoxia. Coastal and Estuarine Studies

Studies, 52, American Geophysical Union, Washington, D.C.

Redfield A.C., Ketchum B.H. and Richards F.A., 1963. The influence of organisms on the

composition of sea-water.In M. N. Hill (ed.), The Sea, John Wiley and Sons, New

York, p. 12-37.

112


Ryding S.O. and Thornton J.A., 1999. Chapter 5: Types of aquatic pollutants, impacts on

water quality and determination of critical levels. In Assessment and control of non-

point source pollution of aquatic ecosystem”- Man and the biosphere series, Unesco

Paris and the Parthenon Publishing group, Vol. 23, p. 73-105.

Sempéré R., Panagiotopoulos C., Lafont R., Marroni B. and Van Wambeke F., 2002. Total

organic carbon dynamics in the Aegean Sea, Journal of Marine Systems, Vol 33-34:

355-364

Servais P. and Garnier J., 1993. Contribution of heterotrophic bacterial production to the

carbon budget of the River Seine (France). Microb. Ecol. 25:19-33.

Servais P., Barillier A. and Garnier J., 1995. Determination of the biodegradable fraction of

dissolved and particulate organic carbon in waters. Int. J. Limnol. 31: 75-80

Servais P., Billen G., Garnier J., Idlafkih Z., Mouchel J.M., Seidl M. and Meybeck M.., 1998.

Le carbone organique : origines et biodégradabilité. In: La Seine en son bassin :

Fonctionnement écologique d’un système fluvial anthropisé. (Meybeck M., De

Marsily G. and Fustec E. eds). Elsevier, Paris p. 483-529.

Servais P. and Garnier J., submitted. Organic carbon and bacterial heterotrophic activity in the

Seine estuary maximum turbidity zone France). Aquatic Sci.

Tran Hieu Nhue, Phan Ngoc Dang, Hua Chien Quang and Lam Duc Huynh. 1994. Polluted

rivers in Vietnam. Construction editor, Hanoi, 65pp.

Wetzel R.G., 2001. Limnology. Lake and river ecosystems. Elsevier Ed., Third Edition.

Academic Press. London. 1006pp.

113


114

Nutrient budgets (N, P)

CHAPTER 5

Nutrient budgets (N, P) for the Red River Basin

Abstract

In order to examine the degree of human-induced alteration of the nitrogen and phosphorus cycles at the scale of a tropical watershed of regional dimension, the budget of these two elements were estimated in the 4 main sub-basins (Da, Lo, Thao and Delta) of the Red River system (156 448 km², Vietnam and China). The 4 sub-basins differ widely in population density (from 101 inhab km-2 in the upstream basins to more than 1000 inhab km-2 in the delta), land use and agricultural practices. In terms of agricultural production, on the one hand, and consumption of food and feed on the other, the upstream sub-basins are autotrophic systems, exporting agricultural goods, while the delta is a heterotrophic system, depending on agricultural goods imports. The budget of the agricultural soils reveals great losses of nitrogen, mostly attributable to denitrification in rice paddy fields and of phosphorus, mostly caused by erosion. The budget of the drainage network shows high retention/elimination of nitrogen (from 62 to 77 % in the upstream basins and 59 % in the delta), and of phosphorus, with retention rates as high as 80 % in the Da and Lo sub-basins which have large reservoirs in their downstream course (Hoa Binh on the Da and Thac Ba on the Lo). The total specific delivery estimated at the outlet of the whole Red River System is 855 kg.km-².y-1 total N and 325 kg.km-².yr-1 total P. Nitrogen rather than phosphorus seems to be the potential limiting factor of algal growth in the plume of the Red River in Tonkin Bay.

Key-words: Nitrogen cycle, phosphorus cycle, Red River, autotrophy/heterotrophy of

regional systems, nutrient retention

This chapter is published as an article in the Journal Global Biogeochemical Cycles under the reference:

Le Thi Phuong Quynh, Gilles Billen, Josette Garnier, Sylvain Théry, Cédric Fézard, Chau Van Minh (2005, in

press). Received 8 November 2004; revised 20 March 2005; accepted 11 Avril 2005

115


5.1 Introduction

Today, human perturbation of the global N and P biogeochemical cycles is a matter of great

concern [Galloway et al., 1995; Howarth et al., 1996; Smil, 1999; Galloway and Cowling,

2002]. At the global scale, anthropogenic nitrogen fixation, either deliberate through

cultivation of nitrogen fixing crops and production of industrial fertilizer, or unintentional

through high temperature combustion, presently equals the natural rates. The resulting

increased nitrate contamination enhances the global denitrification rate and N2O emissions,

which contribute to the green-house effect and the destruction of the stratospheric ozone

layer. Similary, world-wide mining and processing of phosphorus minerals, mainly for

fertilizers production, reach a level of the same order of magnitude as natural weathering and

erosion processes [Weijin et al., 1999]. The riverine transfer of nitrogen and phosphorus to the

coastal waters has therefore increased considerably in many areas of the world, making

marine eutrophication a symptom of global change [Green et al., 2004].

To obtain a good understanding (and possibly control) of these global phenomena, they have

to be examined at a regional scale, where the diversity of climatic and socio-economical

constraints can be taken into account. The regional scale is that of the human perception of

the environment, at which management decisions are taken. Moreover, a description of the

cycling of nutrients within a given territory offers an insight into how humans have managed

their environment and, to some extent, how they live.

Numerous studies have been devoted to the calculation of the nitrogen or phosphorus budget

of regional systems in Europe and Northern America [Billen et al., 1985; Howarth et al.,

1996; Boyer et al.; 2002; Van Breemen et al., 2002]. There are few similar attempts in other

regions of the world, despite the early work by Robertson and Rosswall [1986] in the Niger

basin, and some recent studies of nitrogen budgets in Asian countries [Bashkin et al., 2002;

Xing et al., 2002].

Here the analysis concerns the nitrogen and phosphorus cycling in the terrestrial and aquatic

components of the Red River watershed, a tropical river system which has been profoundly

modified by human intervention for two millennia, as it was the cradle of an ancient

civilization. This region is now the place of an original economic development scheme where

rural population remains dominant [Pham Xuan Nam, 2001]: since 1986, the introduction of a

market-oriented socialist economy (“Doi Moi”) has resulted in the rapid growth of

agricultural and industrial production, but has avoided explosive urban growth and

uncontrolled rural exodus.

116

Nutrient budgets (N, P) 5.2 Description of the Red River Basin

5.2.1 Geomorphology

The Red River basin (Figure 5.1) is located in South East Asia (from 20°00 to 25°30 North;

from 100°00 to 107°10 East) and its watershed covers 156 448 km². It is bordered by the

Truong Giang and Chau Giang River basins in China to the North, the Langcang River

(Mekong) basin to the West, the Ma River basin (in Vietnam) to the South. The Red River

flows eastwards into the Tonkin Bay (South China Sea) [Nguyen Ngoc Sinh et al., 1995] and

rises in a mountainous region of South-eastern China, in theYunnan province, where its name

is Yuan River; it crosses into Vietnam near Lao Cai where it is named Cai, Thao or Hong

River. The main branch is about 1140 km long [Dang Anh Tuan, 2000], and passes through 8

Chinese and Vietnamese provinces before flowing into the China Sea through 4 defluent

branches named Day, Lach Gia, Ba Lat, and Tra Ly. The Thao River has two major

tributaries, the Da and Lo rivers, downstream of which the main branch is named Hong (Red)

River. The drainage density in the Red River basin is rather high, in the range of 0.5 to 1

km.km-2.

Figure 5.1: map of the Red River basin, its 3 upstreams sub-basins (Da, Lo and Thao) and its delta area. Circles indicate the gauging stations.

For budget calculations in this study, we divided the total basin area into 4 sub-basins

corresponding to the drainage area of the three main branches (Thao, Da and Lo rivers) and

the delta (Figure 5.1). Regular measurement of the discharge and water quality were carried

117


out at the outlet of the three upper sub-basins. Because of the difficulty to monitoring the

numerous diverging outlets of the Hong River in its delta area, a gauging station at Hanoï was

included, covering only about 20% of the delta, from which the output fluxes of the whole

delta were extrapolated (see below). Four sub-basins are therefore considered, i.e. those of the

three main tributaries and a portion of the delta.

5.2.2 Administrative divisions

50.3% of the Red River basin is located in Vietnam, 48.8% in China and 0.9% in Laos. Some

of the data used for budget calculations, including land use, fertilizer application, agricultural

production, livestock and industrial activity were taken from recent (1997) official provincial

statistics, i.e from 21 provinces in Vietnam [MOSTE, 1997] and one province in China

[Chinadata, 1998]. In this case, the data by province were reaffected to the 4 sub-basins on

the basis of the percentage of province surface area located inside each sub-basin, as shown in

Figure 5.1.

Table 5.1. Distribution of the surface area of Vietnamese and Chinese provinces (in %) within the sub-basins of the Red River system.

Sub-basins

Provinces Area, km² Lo Da Thao Total Hong

delta Bac Kan 4 796 37.57 Cao Bang 6 387 30.24 Ha Giang 7 831 88.40 11.60 Ha Nam 823 100 Ha Tay 2 148 0.28 7.40 24.75 67.57 Hai Duong 1 661 2.49 Hung Yen 895 64.71 Hoa Binh 4 612 33.25 1.53 54.65 Lao Cai 8 050 5.69 21.63 72.69 Lai Chau 17 133 77.99 0.06 Nam Dinh 1 669 73.57 Ninh Binh 1 387 89.97 Phu Tho 3 465 14.08 10.96 74.96 Son La 14 210 62.80 1.97 Thai Binh 1 509 46.14 Thai Nguyen 3 769 0.41 Thanh Hoa 11 106 0.29 TP. Ha Noi 921 3.21 63.88 Tuyen Quang 5 801 94.02 Vinh Phuc 1 371 87.72 3.86 8.08 Yen Bai 6 808 22.42 13.22 64.34 Yunnan (China) 394100 3.85 5.89 11.16

118

Nutrient budgets (N, P) 5.2.3 Meteorological and hydrological characteristics

The climate in the Red River basin is quite homogeneous across the 4 sub-basins and of sub-

tropical character. The average annual temperature is 19°C and the average annual rainfall

was 1470 mm in the whole basin in 1997 [IMH of Vietnam, 1997-2003], [Chinadata, 1998].

The rainy season lasting from May to October, represents 85 to 90% of the total annual

rainfall, and the dry season from November to April only 10 to 15%.

The mean annual discharge of the main branch (at Son Tay station, just downstream of the

outlets of the three main tributaries) is 3577 m3.s-1 [IMH of Vietnam, 1997-2003]. In the last

100 years, the highest daily discharge, 37 800 m3.s-1 was observed in August 1971, and the

lowest, 368 m3.s-1 in May 1960. Figure 5.2 shows the seasonal variations of discharge at the

outlets of the 4 sub-basins in 2003. The Da and Lo basins have higher specific discharges

(respectively an annual mean of 34 and 25 L.s-1.km-2 in 2003), while the Thao river, with a

large part of its basin in the drier Chinese territory, has by far the lowest specific discharge

(9.6 L.s-1.km-2 ).

0

2000

4000

6000


disc

harg

e, m

3 s-1

Yen Bai (Thao)

Vu Quang (Lo)

Hoa Binh (Da)Ha noi (Hong)

Figure 5.2: Discharges at the outlet of the 3 upstream sub-basins of the Red River system,

and at the Hanoï station in the delta area, in 2003

5.2.4 Land use and population

As shown in Table 5.2 [MOSTE, 1997], land use differs markedly between the 3 upstream

sub-basins and the delta of the Red River. Overall, forest occupies the largest part of the

upstream Red River sub-basins (54 %), while cultivated land represents 33% (12 % for the

rice culture, 20 % for industrial crops). The Lo sub-basin differs from the other two upstream

sub-basins by a greater acreage of industrial crops (58.1 %) than the Da (2.6 %) and Thao

119


(12.8 %) sub-basins. Forest dominates the Da sub-basin (74.4 %). Urban areas represent a

very small proportion (1 %) of the upstream Red River basin. In the delta however, cultivated

land (mainly rice fields) holds the largest share of the land use (63 %), far above forest (18%);

urbanized areas represents a much larger surface (6.8 %) than in the upstream basins.

Table 5.2: Land use in the upstream sub-basins and in the main branch of the Red River (delta) in 1997 (in % area)

Sub-basin Rice Industr. cult.

Dry cereals

Grassland Fruits Forest Rocks Urban areas

Da 12.5 2.6 0.4 3.6 0.0 74.4 6.2 0.3

Thao 18.7 12.8 0.7 7.2 0.9 54.2 4.1 1.4

Lo 8.1 58.1 0.4 3.9 0.1 22.4 6.4 0.6

Hong delta at Hanoî 66.3 7.6 0.7 2.2 0.6 14.9 1.0 6.7

Total Hong Delta 63.0 3.7 0.0 2.6 0.2 17.8 5.9 6.8

In the whole basin, the population was estimated at 30 million inhabitants in 1997, of which

34 % in China [Chinadata, 1998] and 65 % in Vietnam [MOSTE, 1997]. The proportion in

Laos is low, less than 1%. These values were obtained from 5235 villages and towns, all of

which were geo-referenced in the Red River basin with the help of a GIS (Arc Info). The

population density differs greatly among the sub-basins, from 101, 132 and 150 inhab.km-2 in

the Da, Lo and Thao sub-basins respectively, to 1173 inhab.km-2 in the delta area (Table 5.3).

Table 5.3: Population and population density (inhab.km-2) in the sub-basins (Lo, Thao and Da) and in the basin of the main branch (Delta) of the Red River in 1997.

Sub-basins Surface

km² Population

million inhab. Population density

Inhab.km-²

Da 51 285 5.19 101 Thao 61 169 9.17 150 Lo 34 559 4.56 132

Hong Delta at Hanoï 1578 2.47 1565 whole Hong Delta 9 435 11.08 1173

Total 156 448 30.00 192

5.3 The budget of the soil system In the nutrient budget for the soil subsystem of each sub-basin, we take into account the

following inputs and outputs, considering agricultural and forested areas separately: i) input

by atmospheric deposition, atmospheric nitrogen fixation, fertilizer application and excretion

by domestic animals, ii) output through harvested crops and grazing by domestic animals

120

Nutrient budgets (N, P) (Figure 5.3). The nutrient losses through leaching or erosion into surface- or groundwater

will be discussed below in connection with the hydrosystem budget (Figure 5.3).

dom.act.

agricultsoils

cattlefarming

fertili-zers

N2fix

denit&

reton

atm. depos.

river export

Export Imp.

Forestedsoils

N2fix

ind.act.

agricultural goods

woodexp.

53020

800510200

50060

370

37040660

30

13020

640220

1790180

640130

10010

10040

290130

610

201

390170

390335 740

70

416345

830250

-45

soildenit

90

dom.act.

agricultsoils

cattlefarming

fertili-zers

N2fix

denit&

reton

atm. depos.

river export

Export Imp.

Forestedsoils

N2fix

ind.act.

agricultural goods

woodexp.

53020

800510200

50060

370

37040660

30

13020

640220

1790180

640130

10010

10040

290130

610

201

390170

390335 740

70

416345

830250

-45

soildenit

90

Figure 5.3 (a): Da River sub-basin (51 285 km²)

Figure 5.3 (b): Lo River sub-basin (34 560 km²)

dom.act.

agricultsoils

cattlefarming

fertili-zers

N2fix

denit&

reton

atm. depos.

river export

Export Imp.

Forestedsoils

N2fix

ind.act.

agricultural goods

woodexp.

41025

9001040410

49060

110

11015

1507

38045

1420410

2140240

1450280

21030

13055

1610415 15

15

301

500220

640610 930

150

5537

-25

soildenit1160

dom.act.

agricultsoils

cattlefarming

fertili-zers

N2fix

denit&

reton

atm. depos.

river export

Export Imp.

Forestedsoils

N2fix

ind.act.

agricultural goods

woodexp.

41025

9001040410

49060

110

11015

1507

38045

1420410

2140240

1450280

21030

13055

1610415 15

15

301

500220

640610 930

150

5537

-25

soildenit1160

121


122

Figure 5.3(c): Thao River sub-basin (61 170km²)

Figure 5.3 (d): the Delta sub-basin (9 435 km²) Figure 5.3 (a,b,c,d): Nitrogen and phosphorus budgets in the 4 sub-basins of the Red River

system, expressed per km² of catchment area (nitrogen, in bold: kgN.km-2.year-1; phosphorus, in italics: kgP.km-2.year-1).

dom.act.

agricultsoils

cattlefarming

fertili-zers

N2fix

denit&

reton

atm. depos.

river export

Export Imp.

Forestedsoils

N2fix

ind.act.

agricultural goods

woodexp.

31020

1140510200

50070

280

27030

7010

23030

740230

1340130

650130

10010

15060

420195

2020

28020

570255

190170

370140

60120

970315

245

soildenit

1320

dom.act.

agricultsoils

cattlefarming

fertili-zers

N2fix

denit&

reton

atm. depos.

river export

Export Imp.

Forestedsoils

N2fix

ind.act.

agricultural goods

woodexp.

31020

1140510200

50070

280

27030

7010

23030

740230

1340130

650130

10010

15060

420195

2020

28020

570255

190170

370140

60120

970315

245

soildenit

1320

domesticactivity

agricultsoils

cattlefarming

fertili-zers

N2fix

denit&

retention

atm. depos.

river export

Exp.Import

Forestedsoils

N2fix

ind.act.

agricultural goods

woodexp.

64030

329082703230

50060

85

9010

10050

41050

86803280

83701210

57201200

850120

63602110

460120?

3390130

63602110

1060910

141905390

3670-410 ??

48701080

2240

soildenit

7410

98801790

domesticactivity

agricultsoils

cattlefarming

fertili-zers

N2fix

denit&

retention

atm. depos.

river export

Exp.Import

Forestedsoils

N2fix

ind.act.

agricultural goods

woodexp.

64030

329082703230

50060

85

9010

10050

41050

86803280

83701210

57201200

850120

63602110

460120?

3390130

63602110

1060910

141905390

3670-410 ??

48701080

2240

soildenit

7410

98801790

Nutrient budgets (N, P) 5.3.1 Atmospheric deposition

Due to the increase in nitrous oxide emissions by automobile engines and thermal power

production in industrialized countries, wet d dry nutrient atmospheric deposition has

become a significant term in the nitrogen budget of terrestrial systems [Galloway, 2002;

heldrick et al., 2003]. Van Drecht et al. [2003] mentioned a global mean N deposition rate of

50 kg.km-².y-1. A compilation of available data from non-industrial countries shows values

of atmospheric N deposition rates in the range of 100-800 kg.km-².y-1 (230 kg.km-².y-1 in a

rural area of south-east

[Chestnut et al., 1999], 580 kg.km al., 2001]), while global models

and 1000 kg.km-².y-1 in the Red River basin area [Lelieveld

and Dentener, 2000] and Pham Hung Viet et al. [1998] report the value of 2000 kg.km-².y-1

an

S

4

ern China [Weijin et al., 1999], 760 kg.km-².y-1 in the Middle Hills of

Nepal [Collins and Jenkins, 1996], 500 kg.km-².y-1 in a hillslope forest in Puerto-Rico

-².y-1 in Ecuador [Wilcke et

provide values of between 500

in a suburb of Haiphong. These figures often represents the sum of nitrate and ammonium

deposition rates, in which the proportion of both forms are generally similar. However, as

discussed by Howarth et al (1996), among other authors, only nitrate deposition should be

considered as a ‘new’ nitrogen input when examining large watershed nitrogen budget,

because most ammonium deposition depends on short distance transfer of locally volatilized

nitrogen, thus representing internal cycling within the watershed.

Phosphorus deposition, although much lower, is still significant. Values of between 7 and 156

kg.km-².y-1 are cited in the literature for regions with climate and economic conditions similar

to those of the Red River: 45 kg.km-².y-1 in agricultural areas in south-eastern China [Weijin et

al., 1999], 7-28 kg.km-².y-1 in a Ivory Coast rain forest [Stoorvogel et al., 1997], 60 kg.km-².y-

1 in Ecuador, 156 kg.km-².y-1 in a dry forest in Mexico [Campo et al., 2001].

Because of the lack of direct measurements in North Vietnam, the above data were used as a

basis for the chosen values of 500 kg.km-2.y-1 for nitrate-N deposition, and 60 kg.km-2.y-1 for

phosphorus deposition rates, considered as representative for the ‘new’ atmospheric

deposition of nutrients in the whole Red River basin (Table 5.4 a,b).

123


Table 5.4a: Nitrogen budgets of the sub-basins of the Red River (106 kg N.yr-1).

6 -1 Who

10 kg.y as N

Da Lo Thao

le Hong delta

Total Red R. basin

Soil system

Atmospheric deposition

forest 19 3.9 16.6 0.8 40.4

agriculture+grass 6.6 13 14 3.9 37.5

Nitrogen fixation

forest 19 3.9 17 0.8 40.4

Fertiliser application 26 36 31 78 171

Human manure application 15 12.9 26.1 - -

Cattle far

meat and dairy production

excretion 28 43 34 46 151

umption 33 50 40 54 179

Agriculture and food balanc

agric

mercial import 17 32

commercial

h 17.3 34.8

grass- and cropland 41 31 70 31 173

ming

8.1

25.7 4.6 7.3 5.7

grazing and feed cons

e

ultural production 92 74 82 79 327

com 1 1 -

export 27 14 19 6 -

uman consumption 20 60 132

Hydrosystem

Domestic wastewater release 60 77.9

astewater release 0.3 0.5 1.9 3.8

Leaching from agricu

I 93.2

tlet 38 32 22.5 117 169

5 4.3 8.7

Industrial w 1

Leaching from forest soil 17.5 5.3 4.3 0.7 27.7

ltural soil 20 22 12 10 64

nput from upstream tributaries - - - -

Riverine delivery at basin ou

124


Table 5. et he sub sins o e Red R r (106 kg P.yr-1)

106

Thao

otal ong

delta l

basin

4b. Phosphorus budg in t -ba f th ive

kg.y as P-1

Da Lo

TH Tota

Soil sub-systems

Atm

t 5 .0 .1

5 2.0 .5

Fe .3 4.1 12.2 .5 .1

Hum -

Cattle farming

meat and dairy production 0.6 0.9 0.7 1.1 3.3

grazing and feed consumption 6.6 9.8 8.0 35.7

Agricultur

vegetal production

rcial importation 0.03 0.04 1.4 1.2 -

1.1 0.85 1.3 0.32

human consum

Hydrosystem

ospheric deposition

fores 2.0 0. 2 0 4.6

agriculture+grass 1.0 1. 0 5.0

rtiliser application 10 1 30 67

an manure application 6.6 5.8 11.7 -

excretion 6.0 8.8 7.3 9.9 32

11.3

11.4

e

9.3 8.2 8.2 36.0

comme

commercial exportation -

ption 8.8 7.7 15.6 19.9 52

Domestic wastewater release

release 0 1.1 3.6

hing from forest soil 1.4 0.2 0.4 0.5 2.5

Leaching 17.0

Input from upstream t

R

2.2 1.9 3.9 19.9 27.9

Industrial wastewater .65 0.51 1.3

Leac

and erosion from agr. soil 21.0 10.0 8.6 57.0

ributaries - - - 16.9 -

iverine delivery at basin outlet 3.5 5.1 8.3 51.0 51.0

5.3.2 Atmospheric nitrogen fixation

Atmospheric nitrogen fixation can represent high inputs of reactive nitrogen in tropical

systems. A c at ates c pondi the m nd use asses in

the Red River basin cited in the literature Chestnut , 1999; 1999; , 1999;

Boyer et al., 2002, Vitousek et al., 2002, Xing et al., 2002; Basking et al, 2002] led to the

ompilation of specific fix ion r orres ng to ain la cl

[ et al Smil Weijin et al.

125


following values: 105 kg.ha-1.y-1 for nitrogen fixation by soyb an and pe nut crops, 0 kg.ha-

1.y-1 for -1 -1 th ltures gN.ha-1 -1 for forest and 15 kg.ha-1.y-1

for grassland. On the basis of the distribution of these land use classes in each sub-basin

(Table 5.2), the total nitrogen fixation was calculated (Tables 5.4 and 5.5). Due to its large

share in all basins, rice cultivation always repr s the r part he tot itrogen

fixation. Note however that the value of 50 kg.ha often for p rice f might

be ov sive emical rilization , 1992].

Tab N997; Chinadata, 1998] and total per sub-basins.

N fixation

e a 5

paddy rice, 5 kg.ha .y for o er cu , 5 k .y

esent majo of t al n-1.y-1, cited addy ields,

er-estimated in the case of inten ch fe [Roger and Ladha

le 5.5: Nitrogen fixation (106 kg .yr-1) for the largest land use classes in the sub-basins of the Red River basin [MOSTE 1

Sub-basins Rice Soybean & peanut

Other cultures Grassland Forest Total

per unit watershed

surface area

106 kg.yr-1 kg.km-2.y-1

Da 32.1 5.0 0.8 2.8 19.1 59.7 1165 Lo 14.0 4.6 10.1 2.0 3.9 34.6 1000 Thao 57.2 7.6 4.4 1.1 16.6 86.8 1420 Hong-delta 29.7 0.3 0.2 0.4 0.8 31.5 3334 Total Red River 133.0 17.5 15.5 6.3 40.3 212.6 1359

As far as phosphorus is concerned, the process of ‘new’ phosphorus input to soils, i.e.

phosphorus mobilization from bed rock weathering, cannot be easily estimated, and is

neglected in the budgets.

5

The u and

China in the last 50 years. According to Weijin et al. [1999] China is presently the largest

producer of nitrogen fertilizers and the greatest consumer of mineral fertilizers in the world.

In Vietnam

fold during the period from 1961 to 2000 (from 2.2 kg.ha-1.y-1 in 1961 to 150 kg.ha-1.y-1 in

2000). For phosphorus fertilizers, the increase was 5 fold during the same

1990’s (e.g. 1997), the average application rate was 115 kg.ha of

k ertiliz n crop in Vi m, on basis of the FAO data [F 90-

1 l fe r inp the River -basin e c ed fr hese

animals, either directly on grazed land or through spreading of

manure on cropland, must be considered as an input into the agricultural soil system, while

.3.3 Chemical fertilizers

se of chemical fertilizers in agriculture has increased significantly in Vietnam

, according to the FAO database, the use of nitrogen fertilizers has increased 66

period. For the late -1.y-1 N fertilizers and 45

g.ha .y-1 -1 of P f ers o land etna the AO 19

998]. The annua rtilize uts in Red sub s wer alculat om t

rates and the agricultural surface area in each sub-basin (Table 5.4).

5.3.4 Feed consumption, food production and excretion by domestic animals

The excretion by domestic

126

Nutrient budgets (N, P) grazing and feed consumption constitute an output. These terms of the budget (Table 5.4)

Poultry

were estimated on the basis of a livestock census in each sub-basin. Five livestock categories

were taken into account: pig, bovine, horse, sheep/goat and poultry, and the corresponding

data were taken from Vietnamese and Chinese statistics by province in 1997 (Table 5.6). Per

capita production rates of manure, as well as of meat and dairy products, compiled from the

literature (Table 5.7) were used to calculate the budget of animal farming in each sub-basin

(Table 5.4 and 5.8). The sum of excretion and food production was used to estimate total

feedstuff consumption by livestock. Note that pigs and bovines are responsible for more than

80% of the total fluxes in all sub-basins.

Table 5.6: Livestock census (in 103 capita) in the sub-basins of the Red River in 1997 [MOSTE 1997, Chinadata 1998]

Pigs Bovines Horses Sheep

Sub-basin x 103 x 103 x 103 x 103 x 103

Da 688 389 48 52 4148

Lo 1033 594 33 34 13095

Thao 980 436 35 85 9911

Hong-delta at Hanoi 433 91 2 1 3843

Total Red River 5130 1802 131 173 51367

Table 5.7: Per capita excretion and animal food (meat, eggs and/or milk) production for the main livestock categories in Vietnam.

Category Excretion Meat (and dairy) production

kg cap ita-1 y-1 kg capita-1 y-1

nitrogen

ita-1 y-1

Phosphorus

kg capita-1 y-1

Nitrogen

kg cap

Phosphorus

Pig 7.7 2.25 1.5 0.18

Bovine 50 .5 1

se 43 9.6 - -

t and sheep 5.8 1.9 0.9 0.11

y 0.3 .05 0.01

9.6 8

Hor

Goa

Poultr 0.04 0

( ferent s s includ 1979; SCS, 1992; Smil, 19 , ., 2002; Van der Hoek, 1 leken and Bakken, 1997; Thom

1977; Hedlund et al., 2003).

data compiled from dif ource ing Soltner 99; ITP2000; Boyer et al

Gilliam,999; B 1999,

as and Weijin et al.,

127


5.3.5 Nutrient export by crop harvesting and grazing

Nutrient outputs from the soil by harvested crops were determined from the figures of agricultural production in each sub-basin combined with the N and P content of harvested products (Table 5.8). The s con ce, w e, sta bers an potatoes), vegetabl a c o t a, r and co The production of for as also con ed, includin t directly gra y cattle. Be ta on grass produc are not avaikg Stevenson and Cole content of 2 % N and 0.25% P. Table 5.8 shows these estimates.

In order to estimate the fate utrient flux orted from ultural or g nd soils w prod cal population and the

capita diet of Vietnamese people

e part of the cereal production not included in the

uracies or gaps in the statistical data. For

mption,

main crop sidered are ri heat, maiz rchy roots (tud es, soybean, pe nut, fruit, sugar ane, tobacc , e coffee, rubbe tton. age w sider g tha zed bcause da tion lable, an overall yield value was used, i.e. 8000 .ha-1.y-1, proposed by [1999], with a nutrient

of n es exp agric rasslaith the crops and grass (either consumed locally or exported from the sub-basins), the crop

uction figures were compared (Table 5.8) with the food requirements of the lo feed requirement of the cattle.

To estimate the human consumption, the average per provided by the FAO (Table 5.9), was combined with the population figures of each sub-basin. To take into account the interregional differences in living standards, the overall FAO figure was corrected by a factor of 0.7 for the upland regions while the delta area was considered representative of the national mean [Liu, 2001]. The FAO figures for fish and seafood consumption were also corrected by region with data from MOSTE [1997].

The total animal feed requirements are estimated in Table 5.4. In order to meet these requirem nts, grass production, as well as residues of cereals, starchy roots and sugar cane were considered as fodder. When necessary, alocal human nutrition was allotted to the livestock diet. Finally, the feed budget was balanced by introducing an ‘other feed’ source represented by fodder that did not figure in the available statistics, e.g. grazing on rangeland or in forests, aquatic plants used as fodder, etc… (Table 5.8). This additional term also accounts for inaccinstance, official statistics do not correctly take into account the production of home gardens and backyard plots, which can make significant nutritional contributions. However, the size of this ‘other feed’ category, is reasonable, which demonstrates the reliability of the overall budget; it was therefore not included in the calculation of nutrient export from the agricultural soil.

Besides dividing the agricultural production between local human and animal consuthe study estimated commercial import and export to and from the sub-basins (Table 5.8). Agricultural products other than foodstuff, e.g. cotton or rubber, were considered to be entirely exported.

128


Table 5.8. Agricultural production and its destination (human and livestock consumption, or ex 1997. (Figures in 106 kg harvested products yr-1 unless

Da sub-basin Thao sub-basin

portation) in thestated)

Lo sub-basin

sub-basin of the Red River in

Whole delta

%N1 %P1 Prod

uctio

n,

kt/y

r

Hum

an

cons

umpt

ion

anim

al

cons

umpt

ion.

expo

rt/im

port

Prod

uctio

n,

kt/y

r

Hum

an

cons

umpt

ion

anim

al

cons

umpt

ion.

expo

rt/im

port

Prod

uctio

n,

kt/y

r

Hum

an

cons

umpt

ion

anim

al

Prod

uctio

n,

kt/y

r

cons

umpt

ion.

expo

rt/im

port

rice 1.1 0.22 605 596 9 497 1053 0 -556 761 523 2 769

Hum

an

cons

umpt

ion

anim

al

cons

umpt

ion.

expo

rt/im

port

38 2 1817 952 leaves 1.1 0.22 303 0 303 0 249 0 249 -556 381 0 3 38581 0 1maïze 1.2 0.35 285 20 265 0 460 36 424 0 291 18 2 178

1385 0 73 0

leaves 1.2 0.35 71 0 71 115 0 115 73 0 45 62 116 0

73 45 wheat 1.8 0.48 81 22 59 0 153 39 114 53 19 0 34 0 soja 2.2 0.46 54 4 50 0 90 6 84 0 36 3 3 3

66 -66 3 0

starchy roots 2.4 0.12 220 127 93 0 263 225 38 0 153 112 4 67 11 0 -8

1 0 leaves 2.4 0.22 110 0 110 132 0 132 77 0 34

329 -262 77 34

vegetables 3.7 0.06 410 254 156 203 449 0 -246 284 223 6 150 1 0 fruit 2.4 0.09 54 91 -37 85 160 -75 37 80 505

776 -626 -43 277 228

sugar cane 2.1 0.08 814 47 0 767 1267 83 592 592 659 41 3 25509 309 peanuts 1.3 0.23 12 5 0 7 17 9 0 8 20 4 1 41

144 111 0 6 0

tea,coffee,tobacco 2.9 0.15 93 2 0 91 127 3 0 124 62 2 13 16 0 25

0 60 cotton 2.2 0.43 2 0 2 2 0 0 2 1 0 0

6 0 7 0 1 0 0 0

rubber 2.9 0.43 34 0 34 18 0 0 18 24 0 120 24 0 0 12 grass' 2 0.26 1489 1489 0 592 0 592 0 1087 0 10 19887 0 198 0 other feed 2.9 0.06 -350 0 500 0 200 animal pdcts 3.4 0.3 135 91 0 44 169 160 0 9 214 80 239

650 0 134 277 0 -38

fish & sea food 3.4 0.3 54 18 0 36 54 32 0 22 16 16 54 0 0 188 0 -134 total kt/yr 4826 1277 2099 4492 2256 2725 4228 1122 28 94720 5 3969 3490 total ktN/yr 92 27 34 27/-1 82 47 54 19/-17 74 23 5 79 1 14/-1total ktP/yr 9 2 6 1/0 8 3 5 1/-1 8 2 11

85 54 6/-32 6 1/0 6 7 0/-1

1 Data compiled from several sources including Weijin et al, 1999; Martin-Prével et al., 1984; Stevenson and Cole, 1999; Pilbeam et a 1996; l., 2000, Morel, Beaton et al, 1995; Smil, 1999.

129


5.4 Domestic and industrial P, N loadings

W

of

di

th

R

th

5

W

discu

kg.c

valu

the V

cont

annu

Paro

hen concentrated to urban areas, domestic and industrial activities represent a major source

nitrogen and phosph

rect nt di rge. s was made of the data from which it is possible to estimate

e re e of nutrients l wastewater in the sub-basins of the Red

iver s wo ment in

e do ost of the industrial sectors in Vietnam.

.4.1 Domestic stew cities and villages

ith the human per capita food consum nd N an ent

abov ne c pi ran 5.4

a -1 for nitrogen and 0.5 - 0.65 kg.cap-1.y for phosphorus in Vietnam, the lower

e aract zing t ent

ietnamese mea opulatio se o The P loa g respo ng to P

ent of build nd ing agen cc panying the 4.8 kg of active detergents used

ally per capita in am

r , 2004, http://www.vietnampanorama.com

orus fe trans r from the agricultural soil system to the hydrosystem, by

ent

ont

8 -

res

the

etn

poi

leas

. It i

mestic and m

scha A

by

n ana

dom

lysi

estic and industria

rth ioning that presently, wastewater treatment is practically nonexist

wa ater in

ption data (Table 5.9), a the d P c

ssed

p

e, o an calculate a yearly per ca ta n-1

utrient loading in the ge of 3.-1.y

s ch eri he poorest population in the upland areas, while the higher rep

n an

ers a seques

d delta p

ter

n (

ts a

e ab

om

ve). din cor ndi

washing powders and personal care products in Vietnam [Vi

ama ; Vietparners, 2004, http://www.

vietpa rs.comrtne ] sho to res fo hosphorus idering an active

g cont of 2 5 ing

u the additional ing from these products is about 1.2 kg.cap-1.y-1 as P, ma g a

annual pho oru es es se

uropean standards ., 1 ai t a

d McKee et al. [2000] in a sub-tropical catchment in Australia (2.2 to 6.2 k ap .y-

nd 0 to 1 the

for A n cou ao Van S g [1995] estimated the specific per capita load of

V mese at 3.65 for P, in good agreement with the figures

oposed by Meybe P.

ashking et al. [2002] t al. [2003] used

uch er values, kg.c -1 -1 P,

spectively. Althoug rs probabl ok

to account the fact that only a part of the produced hum wastes is dis rged into s ce

ater, another part is spread on agricultu o

uld be added the figu r p . Cons a me

ent

cts,

ent 0%

P load

[Madsen et al., 2001], and a mean P content of % in cledeter

prod

total

to E

foun1

an

kin

clo

ita

sph s loading of 1.7 –1.8 kg.cap-1.y-1. Although th e valu are very

[see eg. Billen et al 999; Serv s e l., 1999], and within the range

by

N

g.c -1

in

for

litera

the

pr

B

m

re

in

w

, a

ture

ietna

.66 .8 kg.cap-1.y-1 for P), they are high compared to those found

sia ntries. C

kg.y

un-1 for N and 0.62 kg.y-1

ck et al. [1989] , i.e. 3.3 kg.capita-1.y-1 for N and 0.4 kg.capita-1.y-1 for

or

y to

urfa

for

18

K

-0.

ore

7

a,

kg

and

.ca

W

pit

ei

a

jin et

al

or

. [

N

1999]

d

a

0.

nd

09-

Sh

0.2

eld

5

rik e

low 0. -1.y-1 f an apita .y f

h this is not entirely clear from their paper, these autho

an cha

ral s ils.

130


131

Table 5.9: a) Average human diet per capita per year (kg capita-1 yr-1) in Vietnam (1997) (FAO), and N and P content (%N or %P); b) Average human diet per capita per year

expressed in nitrogen and phosphorus. a)

Products kg /capita/yr % N 1 % P 1

Rice 164 1.1 0.22

Maize 5.6 1.2 0.35

Wheat 6 1.8 0.38

Starchy roots 35 0.9 0.12

Soybean 1 2.2 0.46

Vegetables 70 1.1 0.06

Fruits 25 2.4 0.09

Sugar cane 13 2.1 0.08

Peanut 1.4 1.3 0.23

Tea and coffee 0.5 2.9 0.15

Meat 24 3.4 0.3

Dairy products 1 2.1 0.35

Fish and seafood 17 3.4 0.3

Total 364

b) kgN capita-1 yr-1 5.4

kgP capita-1 yr-1 0.65

1 Data compiled from several sources including Weijin et al., 1999; Martin-Prével et al., 1984; Stevenson and Cole, 1999; Pilbeam et al., 2000, Morel, 1996; Beaton et al, 1995; Smil, 1999, Boyer et al., 2002, Xing et al.,

2002, Vitousek et al., 2002, Bashkin et al., 2002

For the Red River delta region, where most of the population is agglomerated, and where

running water is available everywhere, we considered that all domestic waste is discharged

into the hydrosystem. However, in the upstream watersheds, where only 25% of the

population live in urban areas [Cao Van Sung, 1995], it was estimated that only 25% of the

domestic wastewater reaches surface waters and the rest is recycled in agriculture (Table 5.4).

5.4.2. Industrial activity

Several large industrial sites exist in the Red River basin, namely those of Viet Tri, Thai

Nguyen and Ha Bac (chemistry, textile and paper). Moreover, smaller cottage industries

(textile, food processing, …) are found everywhere in traditional villages and cause


significant pollution of surface waters. The contribution by these industrial activities to

nitrogen and phosphorus loading of the hydrosystem is extremely difficult to evaluate.

es, in order to estimate a general

these industrial branches in the present

Vietnamese and Chinese economical statistics by province were initially used to estimate the

industrial production in each sub-basin (expressed in tons of finished products) in the

following branches, thought to be the most significant ones in terms of aquatic N and P

pollution: cement production, wood processing, the production of paper industry, chemicals,

food and drink and textiles (Table 5.10). A large amount of data were then gathered to

characterise the wastewater discharged by specific factori

specific N and P loading value for each one of

Vietnamese conditions. Sectorial studies carried out for, or by, the Vietnamese Ministry of

Science, Technology and Environment [MOSTE, 1999, 2003; Le Xuan Tu and Huynh Phu,

1998, VAST, 2000…] or by International Cooperation Agencies [Japan International

Cooperation Agency, 2000] were examined. This compilation was augmented by enquiries,

collection and analysis of effluents in samples from about 20 factories in the Hanoï district.

Overall, 20-30 factories were adequately investigated in each industrial branch. The median

value of the N and P release rate by ton of material produced by the different industrial

branches (Table 5.11) together with a production census (Table 5.10), led to a calculation of

the overall N and P discharge from industrial activities for each sub-basin (see Table 5.4).

These estimates indicate that industrial activities generate nutrient fluxes amounting to less

than 10 % of those from domestic activity. The textile and chemical industries (fertilizers and

detergents) dominates nutrient point sources to surface water. Because it is difficult to obtain

reliable data on pollution fluxes generated by industries and handicraft activities in villages,

the estimates of the contribution to nutrient water contamination by industries might be

severely underestimated.

Table 5.10: Industrial production (in 106 kg yr-1 final product) of the most polluting sectors in the sub-basins in 1997 (sources: Vietnam General Statistic Office 1997, Chinadata 1998)

in 106 kg y-1 Da Lo Thao Hong delta total basin Cement 1002.5 704.3 1843.2 381.2 3931.1 Paper industry 21.3 19.5 75.2 9.0 125.0 Wood industry 827.8 268.4 2688.0 5.1 3 89 Chemical industry 176.7 168.1 690.4 167.5 1202.7 Textile industry 2.9 2.5 7.6 20.5 33.6 Food industry

drinks 16.4 11.2

7 .2

33.2 69.8 130.6

milled food 20.3 87.6 84.6 1168.6 1361.2 sugar 97.3 66.9 186.2 2.5 353.0

132


133

Taestim

dustrial sector Wastewater produced Specific N loading Specific P loading

m3. 10-3 kg of product kg. 10-3 kg product kg. 10-3 kg product Concrete 300 0.600

ble 5.11: Specific N & P loading for the most polluting industrial activities in Vietnam ated from a sample of investigated factories in North Vietnam (see text for details).

In

- Wood 1 0.002 0.0003 Paper industry 100 1.00 0.200 Chemical industry 100 0.700 0.050 Textile industry 200 30.00 4.000 Food industry

d 10 0.450 0.064 milled 50 3.000 0.700

sugar 15 0.300 0.045

0

0

rinks food

(sources: MOSTE 1999, 2003, projects on environments of VAST 1997-2003, projects JICA 2000; this study: chemical analysis results and questionary)

5.5 The budget of the hydrographical network

During their downs ough t quatic continuum, f

large river branches and reservoirs, nutrients watershed-based sources undergo several

geochemical processes with the result that a fraction of their load is immobilized or

eliminated before it reaches the outlet of the ba A comp ison of e estimates of the total

diffuse fluxes (from foreste ls) and int inp s (from domestic and

i

i

5.5.1 Diffuse nutrient loss from forested soils to the hydrosystem

while dissolved organic nitrogen often represents a large fraction (50-60%) of total nitrogen,

of 0.4 mg.L-1. These figures are significantly higher than those

tream transfer thr he a rom the headwaters to

from

bio

sin. ar th

agricultural and d soi po ut

ndustrial activities) in the watershed with the calculation of the N and P fluxes discharged at

ts outlet (our measurements) gives an insight into these “retention” processes.

Nitrogen concentrations in the surface water draining tropical forests are fairly well

documented [Forti and Neal, 1992; MacDowell and Asbury, 1994; Stoorvogel et al., 1997;

Roldan and Ruiz, 2001; Colins and Jenkins, 1996], and range between 0.05 and 0.5 (median

0.4) mg.L-1 for nitrate-N and between 0.01 and 0.07 (median 0.03) mg.L-1 for ammonium-N,

averaging a concentration

found for forested ecosystems at temperate latitudes [Howarth et al. 1996]. The total

phosphate-P content in headwaters of tropical forested watersheds amounts to around 0.015-

0.05 (median 0.04) mg.L-1; the strongest concentrations were associated with periods of high


runoff, and the weakest ones with low flow conditions [Forti and Neal, 1992; Stoorvogel et

al., 1997; Roldan and Ruiz, 2001; Colins and Jenkins, 1996].

These values were used to calculate the contribution to total N and P diffuse loading by

forested soils in the 4 sub-basins (see Table 5.4), taking into account the discharge measured

at the outlet in 2003, and the forested area of each sub-basin.

5.5.2 Diffuse nutrient loss from agricultural soils

Much fewer data are available for cultivated areas in tropical systems. Kao et al. [in press]

report nitrate-N concentrations of between 0.42 at low runoff and 3.5 mg.L-1 at high runoff in

streams draining vegetables cultures in mountain areas in Taiwan. Roldan and Ruiz [2001]

measured inorganic nutrient concentrations of 0.67 mg.L-1 for N and 0.55 mg.L-1 for P in

rivers draining industrial plantations in Columbia. Taking into account these ranges, and

considering ammonium and organic nitrogen releases similar to those from forested soils, a

value of 2.5 mgN.L-1 was used to estimate total dissolved nitrogen leaching from cultivated

soils in the upstream sub-basins of the Red River. However, due to the anaerobic nature of

waterlogged paddy-field soils, no nitrate nitrogen is exported from wet rice fields [Reddy and

Patrick, 1986]. This is confirmed by our measurements of water draining paddy rice fields in

the Hanoï area, where the nitrate concentrations were below 0.05 mgN.L-1, while the

ammonium-N concentrations amounted to 2 mg.L-1. Total phosphorus exportation was

considered to occur mainly in particulate form and to depend on soil erosion, particularly

under high flow conditions. The suspended sediment concentration, typical of the Red River

tributaries upstream from the large reservoirs, is greater than 5 g.L-1, with a phosphorus

content of 0.42 mg.g-1 and an nitrogen content of 1.3 mg.g-1 (see below). On this basis, we

attributed a mean P concentration of 2 mg.L-1 to headwaters draining cultivated soils in the

d

from trations and pecific d arge va

5.5.3 Nutrient output at the outlet of the sub-basins

Mon and nutrie lysis w rried out at the outlet of each upstream sub-

basin and at the Hanoï station during 2003. All sam ere st rozen osable

steri ne flasks. Ph sphate, si onium were determined spectro-

photometrically on water filtered through glass-fiber filters, according to respectively

Eberlein and K 4], r [1984] and Slawyck and MacIsaac

etermined after reduction into nitrite according to Jones [1984]. Total nitrogen and

hosphorus were determined on non-filtered water after sodium persulfate digestion and

Red River basin.

At the outlet of each sub-basin diffuse nutrient fluxes from cultivated areas were estimate

these concen the s isch lue measured in 2003 (see Table 5.4).

thly sampling nt ana ere ca

p wles ored f in disp

le polyethyle o lica and amm

atter [198 Rodie [1972]. Nitrate was

d

p

134


135

,

o TON 0 t w the

other samples. Higher nu

during the dry one (Figure 5.4). N and P concentrations were almost always higher in the

Thao River, both at the upstream station of Yen Bai and upstream of Hanoï, than in the outlet

of the Da and Lo rivers.

rus (Ptot, mgP.L-1) and b) total

ular analyses

Bay, prevent accurate estimates of the total riverine nutrient delivery. Nguyen Ngoc Sinh et

mineralization at 110°C in an acidic phase. Total organic nitrogen concentration (TON

N.L-1), only determined on 10 occasions, obeys the following relationship with suspended mg

s lids (SS, mg.L-1): = 0.4 + 0.0013. SS (r²= .91), from which i as extrapolated to

trient concentrations were observed during the rainy season than

a)

b)

0

2

3

4


Ntot

, mgN

.L-1

Figure 5.4: Measured concentration of a) total phospho

nitrogen (Ntot, mgN.L-1) at the outlet of the 3 upstream sub-basins of the Red River system, and at the Hanoï station in the delta area, in 2003. Ntot represents the sum of inorganic and

organic nitrogen.

The output of nitrogen and phosphorus at the outlet of each sub-basin (Table 5.4) was estimated from the monthly data combined with daily measurements of discharge at the same stations (Figure 5.2).

The complexity of the hydrological network in the Delta area and the lack of reg

at the outlet of each of the numerous branches of the Red River discharging into the Tonkin

al. [1995] estimated the flux of total nitrogen and phosphorus at 220-250 106 kg.y-1 and 61-

130 106 kg.y-1 respectively for the entire basin of the Red-Thai Binh river system, but this

represents a larger river basin than the one considered here. For this reason, the nutrient fluxes

at the outlet of the whole delta area, were calculated by extrapolating the flux measured at the

Hanoï gauging station taking into account the respective delta area in the watershed,

according to the following formula:

Flux at delta outlet =

(Flux at Hanoï – Σ flux upstr. tribut.) x tot.delta area / delta area at Hanoï +Σ flux upstr. tribut.

1

Yen Bai (Thao)Vu Quang (Lo)Hoa Binh (Da)Ha noi (Hong)

0,0

0,5

1,0


Ptot

, mgP

.L-1

Yen Bai (Thao)Vu Quang (Lo)Hoa Binh (Da)Ha noi (Hong)


5.6 Discussion

All the nitrogen and phosphorus flux estimates discussed above are expressed in watershed

area specific fluxes to facilitate comparisons between the sub-basins (Figure 5.3 a, b, c, d).

5.6.1 Balancing the budgets

The flux estimates (Table 5.4 and Figure 5.3) rely on a wide variety of sources of differing

ets of agricultural soils are also coherent. Their nitrogen budget regularly

shows an excess of inputs over the outputs, which might be explained, either by an

ith very low inputs. (Krupnik et al.,

reliability, as well as on several debatable hypotheses. Values deduced from official statistics

may be inaccurate because all activities, particularly agricultural and cottage industry

production are not always correctly registered. Population data, at least for Vietnam, were

available as a geographically referenced data base, so that they could be allocated fairly

accurately to each sub-basin. This was not the case with many data (e.g. agricultural and

industrial productions) only available at the province level, that were reallocated to the sub-

basins according to the fractions of the province surface area where they belong (Figure 5.1),

with the implicit assumption that these activities have homogeneous spatial distribution.

Moreover, for lack of direct measurements in the studied area, many fluxes, such as

atmospheric deposition or nitrogen fixation, were estimated from data in the literature

concerning similar regions. For these reasons, we a priori estimate in the order of 25-50 %

the confidence level of our figures, which must be taken with caution. Nevertheless, the

resulting budgets appear quite consistent. As mentioned above the food and feed budget

(Table 5.8) can be balanced provided an ‘unregistered’ feed source is taken into account.

However, the magnitude of this ‘missing feed’ (which might represent grazing on rangeland

or in non agricultural areas) nowhere exceeds 20% of the calculated livestock requirements.

The nutrient budg

overestimation of fertilizer or of other inputs, or by underestimation of loss processes. The

fact that cultivated plant uptake is lower than N inputs from fertilizer is not particularly

surprising: it is generally so excepted in countries w

2004). On the other hand, gaseous losses from soils have not been taken into account in our

agricultural budget, and probably explain the apparent surplus. Both denitrification and

ammonia volatilization and denitrification are known to be quite significant in paddy rice

soils (Bouwman et al., 2002). If the former process should not be taken into account in our

‘new’ nitrogen budget (as ammonium deposition has not been considered neither), the latter

might in itself explain the gap of the budget. Indeed, the shortfall in the balance of the

different agricultural soil budgets in the sub-basins is related to the relative acreage of paddy

136


137

cterized by a dominance of industrial crops both appear to show

ge erosion losses of phosphorus, while the phosphorus budget of the Thao basin and the

Delta area show a phosphorus accumulation in the agricultural soils.

od production of the basin, i.e. 3700 103 kg.y-1 (Table

mean N content of 0.2% in wood, this represents a nitrogen export of -

rice fields (Figure 5.5). Extrapolating the observed trend to 100% of paddy-field surface area

would provide a denitrification rate of about 100 kgN.ha-1.y-1, which is in the range of

reported values for denitrifcation in fertilized paddy fields or nitrate contaminated wetlands

(50-120 kg ha-1.y-1 as N) [Reddy and Patrick, 1986; Rolston et al., 1978]. As far as

phosphorus is concerned, its budget in agricultural soils shows either excess or default inputs

compared to the outputs according to the sub-basins. The mountainous Da River basin, and

the Lo River basin, chara

lar

2500

exce

ss N

00 20 40 60 80 100

5000

7500

10000

% paddy rice

, kgN

/km

²/yr

range of denitrification rate in paddy

rice fields

Figure 5.5. Estimated balance default of the agricultural soil budget in the sub-basins of the Red River, plotted against the percentage area occupied by paddy rice fields. The trend

extrapolates to plausible denitrification rates in fertilized paddy rice plots (i.e. 100 kgN. ha-

1.y-1 at 100%).

The budget of forested soil was not fully established because of a lack of reliable estimates of

forest primary production. Our estimates of nitrogen fixation and deposition on forested soils

lead to a total input of 1000 kg.km-².y-1 (expressed per surface of forested areas), far in excess

of the nitrogen output by forested soil leaching and erosion (312 kg.km-².y-1). Wood export

can be estimated from the total wo

5.10). Considering a

only 92 kg.km ².y-1 from the 80700 km² of forested area in the Red River basin as a whole.


As far as the budget of the hydrosystem is concerned, the nitrogen export calculated at the

asins is clearly smaller than the sum of the inputs. The corresponding

nitrogen retention within the hydrographic network represents respectively 36%, 0.5 %, 14 %

o

ely. The much higher retention in the Da and Lo sub-basins

.

When the total agricultural production in each sub-basin is plotted against the total

outlet of the sub-b

and 20% of all the inputs to the river network for the Da, Lo, Thao and Delta sub-basins.

Similarly, the phosphorus budget shows the retention of 83 %, 78 % and 46 % for the Da, L

and Thao sub-basins respectiv

commonly reported in the literature, is obviously related to the presence of large reservoirs

(Hoa Binh on the Da, 208 km² and Thac Ba on the Lo, 235 km²) in their downstream course,

which trap a great deal of suspended matter and associated phosphorus [Vorösmarty et al.,

1997; Garnier et al., 1999). The phosphorus budget of the delta, however, shows a deficit of

about 8%. Although this only represents a minor imbalance, it might reflect an

underestimation of the industrial contribution of phosphorus to the river system.

5.6.2 Biogeochemical functioning of the sub-basins

The four sub-basins in this study have quite different land use patterns. The population

densities of the three upstream river basins are similar (101-150 inhab.km-², Table 5.3), but

their agricultural activities differ greatly (Table 5.2): i) the Da river basin is mostly forested,

ii) the Lo river basin is predominantly devoted to industrial crops, mainly sugar cane, tea and

rubber and iii) the Thao river basin also has large areas of industrial crops but a greater

fraction of its surface area is devoted to rice production. The population is concentrated to the

delta, (population density greater than 1000 inhab.km-²) where rice production and livestock

farming are the most important agricultural activities.

The differing land use patterns result in a varied biogeochemical functioning of the systems

consumption by humans and cattle (both expressed in terms of kg.km-².y-1 of nitrogen), the

resulting diagram, similar to the classical P/R diagram of functional ecosystem analysis,

makes it possible to define the degree of autotrophy (P) or heterotrophy (R) of a regional

human system (Figure 5.6). An ecosystem is said autotrophic when its net primary production

(integrated over a certain time period) surpasses its respiration: it then accumulates or exports

biomass and represents a sink for nutrients and carbon dioxide. When respiration (either

supported by external inputs or by consumption of internal stocks of organic matter) is larger

than primary production, the system is said heterotrophic, and exports nutrient and CO2

(Odum, 1959). By analogy, a regional watershed can be said autotrophic (with respect to

human economy) when its agricultural production is greater than the consumption of

agricultural products by human and cattle. Whereas the three upstream Red River sub-basins

(particularly the Da basin), are slightly autotrophic, the delta system is clearly heterotrophic,

138


139

food and feed than the

ed River delta. The upstream Seine river basins [Billen et al., 2001], as well as the

Mississippi basin [Howarth et al, 1996] are examples of autotrophic systems with moderate

opulation densities, that export large amounts of agricultural products.

s

from the Da, Lo, Thao sub-basins). The estimated total delivery at the outlet of the delta is

which is in agreement with the fact that the former three basins export agricultural products,

while the latter imports them (Figure 5.3). Similar data derived from analyses of a few other

regional budgets published in the literature for other regions of the world are included in

figure 5.6 for comparison. In the Republic of Korea [Bashkin et al., 2002], with a population

density of 395 inhab.km-², the situation is rather similar to that of the upstream catchment of

the Red River system, but with a slightly greater autotrophy. The Scheldt basin [De Becker et

al, 1988] as well as those of the east coast of the United States [Boyer et al, 2002], both area

with high human population densities and intensive cattle farming, are examples of

heterotrophic systems, depending however much more on imports of

R

p

5000

10000

15000

agric

ultu

ral p

rodu

ctio

n, k

gN/k

m²/y

r

Delta

LoThao

Da

Autotrophy

HeterotrophyKorea

Scheldt

Seine

Mississippi

US East coast

00 5000 10000 15000

human and animal consumption, kgN/km²/yr

Figure 5.6: Characterisation of the degree of auto- or heterotrophy of regional human exploited systems: plot of agricultural production against total food and feed consumption by humans and domestic animals. The data from the Red River are compared with literature data

from other river systems. (See text for explanation).

5.6.3 Riverine nutrient export

The specific riverine export of nutrients from the three sub-basins is quite low (respectively

740, 930 and 370 kg.km-².y-1 for nitrogen and 70, 150 and 140 kg.km-².yr-1 for phosphoru


855 kg.km-².y-1 for nitrogen and 325 kg.km-².y-1 for phosphorus when expressed with respect

to the total Red River basin area, slightly lower than the specific fluxes cited by Nguyen Ngoc

Sinh et al., [1995] for the outlet of the Red-Thai Binh river system (1180-1480 kg.km-².y-1 for

nitrogen and 350-700 kg.km-².y-1 for phosphorus). According to our estimates, the

contribution by the delta area alone represents 4310 kg.km-².y-1 of nitrogen and 3600 kg.km-

².y-1 of phosphorus .

These results are in agreement with the view, expressed by Howarth et al., (1996) and Boyer

et al., [2002], that nitrogen export in streamflow is strongly related to total new inputs of

nitrogen to the catchment (Figure 5.7), although only 20-25 % of these new inputs of nitrogen

are exported by the river system. The Red River delta area appears to be one of the most

heavily loaded systems documented in the literature.

Figure 5.7: Riverine export of nitrogen plotted against total inputs of new (see text for

definition) fixed nitrogen to the watershed. The data from the Red River are compared with

literature data from other river systems. (See text for explanation).

The molar N/P ratio of riverine delivery at the outlet of the Red River basin is 5.8. This value

is much lower than the Redfield ratio (16) of marine phytoplanktonic algae, indicating that

nitrogen rather than phosphorus is the potentially limiting factor of algal growth in the plume

of the Red River in the Tonkin Bay. This conclusion is particularly important in view of the

recent work by Wu et al [2003] demonstrating that nitrogen also limits net phytoplankton

growth in the offshore waters of the South China Sea, where nitrogen fixation remains at very

low levels.

5000

1000

2000

3000

4000

river

ine

expo

rt, k

gN/k

m²/y

r Delta

00 5000 10000 15000 20000

total new inputs, kgN/km²/yr

Thao

Lo Da

whole Red R

140


Our measurements of silica flux at the Hanoï station show a silica delivery of 2920 kg.km-².y-1

as Si, indicating a molar Si/N of 1.7 in the nutrient fluxes carried by the river, in excess of the

requirements of marine diatom growth (Si/N generally close to 1, [Conley et al., 1993; Billen

and Garnier, 1997]). The increased human activity in the Red River watershed, particularly in

its delta, may further enrich the system in nitrogen and phosphorus along its aquatic

continuum. However, the Tonkin Bay does not, at the moment, seem threatened by harmful

marine eutrophication processes characterized by depletion of silica in relation to nitrogen and

phosphorus as well as by a proliferation of undesirable non diatom algae [Officer and Ryther,

1980; Billen and Garnier, 1997; Conley et al., 1993; Garnier and Billen, 2002; Garnier et al.,

in press].

141

.7 References

ashkin, V. N., S.U. Park, M. S. Choi and C.B.Lee (2002), Nitrogen budgets for the Republic of Korea and the Yellow Sea region. Biogeochemistry 57/58: 387-403.

Beaton, J.D., T. L. Roberts, E.D.H. Halstead and L.E. Cowell (1995), Global transfers of P in fertilizer materials and agricultural commodities. In Phosphorus in the global environment: Transfer, cycles and management. H. Tiessen eds. Chichester, England, John Wiley and Sons Ltd. Scope 54: 462.

Billen, G., M. Somville, E. DeBecker and P. Servais (1985), A nitrogen budget of the Scheldt hydrographic basin. Neth J. Sea Res. 19: 223-230

illen, G. and J. Garnier (1997), The Phison River plume: coastal eutrophication in response to change in land use and water management in the watershed, Aquat. Microb Ecol 13: 3-17.

Billen, G., J. Garnier, C. Deligne, C. Billen (1999), Estimates of early industrial inputs of nutrients to river systems : implication for coastal eutrophication. The Sciences of the

the Seine estu ): 977-993.

n cost of food production: Norwegian society

t

5

B

B

Total Environment, 243/244: 43-52.

Billen, G., J.Garnier, A. Ficht and C. Cun (2001), Modelling the response of water quality in ary to human activity in its watershed. Estuaries, 24(6

Bleken, M.A. and L. Bakken (1997), The nitroge26. Ambio 26(3): 134-142.

Boyer, E.W., C.L. Goodale, N.A. Jaworski and R.W. Howarth (2002), Anthropogenic nitrogen sources and relationships to riverine nitrogen export in he northeastern U.S.A. Biogeochemistry 57/58: 137-169.


Bouwman A.F, J.M Boumans and N.H.Batjes (2002) Emissions of N2O and NO from fertilied fields: Summary of available measurement data. Global Biogeochem. Cycles, 16: 6/1-6/12

Campo, J., M. Maass, V.J. Jaramillo, A. Martinez-Yrizar and J. Sarukhan (2001), Phosphorus cycling in a Mexican tropical dry forest ecosystem. Biogeochemistry 53: 161 - 179.

Cao Van Sung (1995), Environment and bioresources of Vietnam: present situation and

ogress series, Published November 4, 101: 179-192.

hestnut, T.J., D. J. Zarin, W.H. McDowell and M. Keller (1999), A nitrogen budget for late-successional hillslope tabonuco forest, P. Biogeochemistry 46: 85-108.

Statistical yearbook of Yunnan, Vol 1997, Vol 1998, Vol 2000 China Statistical Publishing House, (Basic Information of Yunnan, China), (http://chinadatacenter.org

solutions. Thê Gioi Publishers, Hanoi, Vietnam: 235 pp.

Collins, R and A. Jenkins (1996), The impact of agricultural land use on stream chemistry in the Middle Hills of the Himalayas, Nepal. Journal of Hydrology 185: 71-86

Conley, D.J., L. S Claire, E. F. Stoermer (1993), Modification of the biogeochemical cycle of silica with eutrophication. Marine Ecology pr

C

Chinadata, (1998),

), China.

ang Anh Tuan, (2000), The Red River Delta - The Cradle of the Nation, published in Hanoi, Vietnam, 53 pp.

e Becker, E., G. Billen and P. Servais (1988), Nitrogen budget at the scale of the Belgian hydrographical network, Water Supply 6: 327-333.

Eberlein, K and G. Katter (1984), Automatic method for the determination of orthophosphate and dissolved phosphorus in the marine environment. Fresenius Z. Anal. Chem. 326: 354-357

FA s),

D

D

O, (1990-1998), Faostat statistics database (Fertilizer data used for China, Vietnam, LaoFaostat database Results, copyright FAO 1990-1998, http://www.apps.fao.org/;

www.fertilizer.fao.org/. 2002http:// .

Forti, M.C. and C. Neal (1992), Hydrochemical cycles in tropical rain forests: an overview with emphasis on central Amazonia. Hydrology 134: 103-115.

Galloway, J. N., W. H. Schlesinger, H.L. III, A. Michaels and J.L. Schnoor (1995), Nitrogen fixation: anthropogenic enhancement-environmental response. Global Biogeochemical cycles 9: 235-252.

Galloway, J.N, E.B Cowling, S.P Seitzinger and R.H. Socolow (2002), Reactive Nitrogen: too much of a good thing? Ambio 31(2): 60-63

142


143

sfer processes in river catchments. SCOPE book series.

arty, M. Meybeck, J.M. Galloway, B.J. Peterson and E.W. Boyer -industrial and contemporary fluxes of nitrogen through rivers: a global

assessment based on typology. Biogeochemistry 68: 71-105

Hedlund, A., E. W anagement by nutrient balances and flows on peri-urban smallholder farms in southern Vietnam,

Howarth, R. Wndse, J. Freney, V. Kudeyarov, P. Murdoch

and Zhu Zhao-Liang. (1996), Regional nitrogen budgets and riverine N &P fluxes for

ITP, Institute Technique du Porc. (2000), Phosphore et Environnment, ITP editor, Paris,

JICA, (2000), t for Hanoi City in the Socialist Republic of Vietnam, Final report, Data book., Nippon Koei Co. Ltd, Japan Internation

-646.

Galloway, J.N and E.B. Cowling (2002), Reactive Nitrogen and the world: 200 years of change Ambio 31(2): 64-71

Garnier, J., Leporcq B. Sanchez N. and Philippon (1999)- Biogeochemical budgets in three large reservoirs of the Seine basin (Marne, Seine and Aube reservoirs). Biogeochemistry 47: 119-146

Garnier, J. and Billen G. (2002). The Riverstrahler modelling approach applied to a tropical case study (The Red –Hong- River, Vietnam): nutrient transfer and impact on the Coastal Zone. SCOPE, Coll. Mar. Res. W. 12: 51-65

Garnier, J, Sferratore A., Meybeck M., Billen G. and Dürr H. (in press). Modelling silica tran

Green, P. A., C.J. Vorosm(2004), Pre

itter and Bui Xuan An. (2003), Assessment of N, P and K m

European Journal of Agronomy, Volume 20, Issues 1-2: 71-87

., G. Billen, D. Swaney, A. Townsend, N. Jaworski, K. Lajtha, J.A.Downing, R. Elmgren, N. Caraco, T. Jordan, F. Bere

the drainnages to the North Atlantic Ocean: Natural and human influences. Biogeochemistry 35: 75-139.

IMH, (1997-2003), Journal of Meteo-hydrology, Insitute of Meteo-Hydrology in Vietnam, Hanoi. (Journal monthly during the periods of from 1997 to 2003)

France.

The Study on Environmental Improvemen

Cooperation Agency (JICA) and Hanoi People's Committee - Socialist Republic of Vietnam., Hanoi, 350p.

Jones, M.N. (1984). Nitrate reduction by shaking with cadmium, alternative to cadmium columns. Water Res. 18:643

Kao, S.J., F.K. Shiah, K.K. Liu and J. Owen (submitted), Export of dissolved inorganic nitrogen from a partially cultivated mountainous watershed (Lanyang-Hsi) in Taiwan. Biogeochemistry.


Krupnik et al., 2004. in Agriculture and the nitrogen cycle, edited by A.R. Mosier, J.K. Syers, and J.R. Freney, pp. 193-207, Island Press, Washington, D.C., 2004.

Le Xuan Tu and Huynh Phu, (1998), Studies on treatment of waste water from paper

Madsen, T, H.B. Boyd, D. Nylén, A.R Pedersen, G. Petersenand and F. Simonsen (2001),

rojekt. Danish

Martinet tropicales. Paris, Technique et Documentation

McKee, L.J. and B.D. Eyre (2000), Nitrogen and phosphorus budgets for the sub-tropical

Meybe lobal freshwater quality: a first

Morel, R. (1996), Les sols cultivés. Technique and documentation, 2ème édition. Paris,

MOSTtistics editor, Hanoi.

990s, Hanoi, Vietnam,

Nguyend Nguyen Vu Trong. (1995), Case study report on Red River Delta

production factories. Procedings of National Conference on Environment, Hanoi, Vietnam.

Lelieveld, J. and F. Dentener, (2000), What controls tropospheric ozones? J. Geophys. Res. 105: 3531-3551.

Liu, A.Y.C. (2001), Markets, Inequality and Poverty in Vietnam. Asian Economic Journal 15( 21): 217-235.

MacDowell, W.H. and C. E. Asbury (1994), Export of carbon, nitrogen and major ions from three tropical mountain watersheds. Limnol. Oceanogr. 39: 111-125.

Environmental and Health Assessment of Substances in Household Detergents and Cosmetic Detergent Products. Environmental Project No. 615 MiljøpEnvironmental Protection Agency.

-Prével, P., J. Gagnard and P. Gautier (1984), L'Analyse végétale dans le contrôle de l'alimentation des plantes tempérées (Lavoisier),

Richmond River catchment, Australia. Biogeochemistry 50: 207-239.

ck, M., D. Chapman and R. Helmer (1989), Gassessment. Cambridge, MA, World Health Organization/United Nations Environment Programme Basil Blackwell, Inc.

France.

E, (1997), Vietnamese general statistics officer, Ministry of Science, Technology and Environment of Vietnam, General sta

MOSTE, (1999), Wastewater quality of some industrial branches in Viet Nam in Resource and present condition of environment in Vietnam in years of 1219pp.

Ngoc Sinh, Hua Chien Thang, Nguyen Chu Hoi, Nguyen Van Tien, Lang Van Ken, Pham Van Ninh anin Vietnam - Project on integrated management and conservation of nearshore coastal

144


145

e east Seas (ESA/RCU), 78 pp.

Officer, e er 3: 83-91

yers (Eds.), Proceedings of the

ersity, Bangkok.

Pilbeam . Gaunt (2000), Nitrogen

Reddy, ed rice fields. In: S.K.

Robert

Rodier, J (1984), L'analyse de l'eau. 7ème édition. Dunot (ed), France, 1365 p.

Roger, etland rice fields: estimation and contribution to nitrogen balance. Int. Rice Res. Inst., Manila,

Roldan, G. and E. Ruiz (2001), Development of Limnology in Colombia. In Limnology in

Rolstonl Sci. Soc. Am. J. 41: 863-869.

office, Washington, DC.

-3531.

and marine areas in East Asia region (EAS-35) United Nations Environment program. Regional coordinating for th

Odum, E.P. 1959. Fundamentals of Ecology. Saunders, ed. 546 pp.

C.B. and J.H. Ryther (1980), The possible importance of silicon in marine utrophication. Mar Ecol Prog S

Pham Hung Viet, Pham Manh Hoai, Nguyen Xuan Trung and Vu Duc Nam. (1998), Chemical composition and acidity of atmospheric deposition in the northeastern part of Vietnam. In: Rhode, Kuylenstierna, Boonjawat, AFourth Meeting of the IGAC/DEBITS: Composition and Acidity of Asian Precipitation (CAAP), 9/12 November, Chulalongkorn Univ

Pham Xuan Nam (2001). Rural development in Vietnam, the search for ustainable development. Social Scineces publishing House, Hanoï.

, C. J., B.P. Tripathi, D.P. Sherchan, P.J. Gregory and Jbalances for households in the mid-hills of Nepal. Agriculture, Ecosystem and Environnment 79: 61-72.

K. R and W. H. Patrick (1986), Denitrification losses from floodDe Datta and W.H. Patrick (editors), Nitrogen Economy of Flooded Rice Soils. Martinus Nijhoff, Dordrecht, 15-39p

son, G.P and T. Rosswall (1986), Nitrogen in West Africa: The regional cycle. Eco. Monogr. 56: 43-72

P.A. and J.K. Ladha (1992), Biological nitrogen fixation in w

Philippines. Plant and Soil 141(1-2): 41-55.

Developing Countries, edited by R. G. Gopal.B and Wetzel. Volume 3: 69-119.

, D.E., D.L. Hoffman and D.T. Toy (1978), Field measurements of denitrification, I. Flux of N2 and N2O. Soi

SCS, (1992), Soil Conservation service, Agricultural waste management field handbook, chapter 4. US Gouvernment priting

Servais, P., J. Garnier, N. Demarteau, N. Brion and G. Billen (1999), Supply of organic matter and bacteria to aquatic ecosystems through waste water effluents. Water Research. 33: 3521


Sheldrick, W., J.K. Syers and J. Lingard (2003), Soil nutrient audits for China to estimate nutrient balances and output/input relationships. Agricultural, Ecosystems and

Slavycoastal upwelling. Deep-Sea Res. 19: 1-4

les 13(2): 647-662.

Stevenson, F.J. and M.A. Cole (1999), Cycles of soil, Carbon, Nitrogen, Phosphorus, Sulfur,

Stoorvo nt budgets of a

Thoma

re did all the nitrogen go? Fate of nitrogen inputs to large watersheds in

Van d t budgets from

Van D noop, A.H.W. Beusen and C. R. Meinardi (2003)

ater and surface water, Global Biogechem. Cycles, 17(4) 26.1-26.20, doi:

VAST. ents of Vietnam Academy

Vitouse, R. Marino, L. Martinelli, E. B. Rastetter and J. I. Sprent (2002), Towards an

Vörösm

Environment 94: 341-354.

k, G. and J.J. McIsaac (1972), Comparison of two automated ammonium methods in a region of c

Smil, V. (1999), Nitrogen in crop production: an account of global flows. Global Biogeochem. Cyc

Soltner, D. (1979), Alimentation des animaux domestiques, 13éme édition, France.

Micronutrients. John Wiley and Sons publisher, Canada, volume 1, 427 p.

gel, J.J, B.H. Janssen and N. Van Breemen (1997), The nutriewatershed and its forest ecosystem in Taï National Park in Côte d'Ivoire, Africa. Biogeochemistry 37: 159-172.

s, G.W. and J.W. Gilliam (1977), Agro-ecosystems in the USA. Agroecosystem 4: 182-239.

Van Breemen, N., E.W. Boyer, C.L. Goodale, N.A. Jaworski, K. Paustian, S.P. Seitzinger, K. Lajtha, B. Mayer, D. Vandam, R.W. Howarth, K.J. Nadelhoffer, M. Eve and G. Billen (2002), Whethe northeastern U.S.A. Biogeochemistry 57/58: 267-293.

er Hoek, K.W. and A.F. Bouwman (1999), Upscaling of nutrienagroecological niche to global scale, Book: Nutrient disequilibri in agroecosystems Smaling EMA, O. Ocenema and LO Fresco eds., CAB International publisher, 57-73p.

recht, G., A.F Bouwman, J.M. KGlobal modelling of the fate of nitrogen from point and non-point sources in soils, groundw10.1029/2003GB002060,2003.

, 2000. Proceedings annual of Project Reports on Environmof Science and Technology in 2000, Hanoi. VAST publisher. 310p.

k, P.M., K. Cassman, C. Cleveland, T. Crews, C.B. Field, N.B. Grimm, R.W. Howarthecological understanding of biological nitrogen fixation. Biogeochemistry 57/58: 1-45.

arty C.J., M. Meybeck, B. Fekene and K. Sharma (1997). The potential impact of neo-Castorization on sediment transport by the global network of rivers. In: Human impact

146


on erosion and sedimentation (Proceedings of the Rabat Symposium, April 1997), IAHS Publ, n° 245.

147

d biogeochemistry in an

Wilcke 001), Change in water quality during the

Wu Jin

Ackno

due to

program hi Phuong Quynh’s Ph-D thesis is supported by

the French Ambassy and by the Pierre and Marie Curie University (Paris 6).

Weijin, Yan, C. Yin and Shen Zhang. (1999), Nutrient budgets anexperimental agricultural watershed in Southeastern China. Biogeochemistry 45: 1-19.

, W., S. Yasin, C. Valarezo and W. Zech (2passage through a tropical montane rain forest in Ecuador. Biogeochemistry 55: 45-72.

gfeng, Shi-Wei Chung, Liang-Saw Wen, Kon-Kee Liu, Yuh-ling Lee Chen, Houng-Yung Chen and David M. Karl. (2003), Dissolved inorganic phosphorus, dissolved iron and Trichodesmium in the oligotrophic South China Sea. Global Biogeochem. Cycles, 17(1), 1008, doi 10.1029/2002GB001924

Xing, G. X. and L. Zhu (2002), Regional nitrogen budgets for China and its major watershed. Biogeochemistry 57/58: 405-427.

wledgements

This study was realized in the framework of a French-Vietnamese co-operation. Thanks are

Georges Vachaud, Research Director at the CNRS, for the coordination of the

me ESPOIR (CNRS-CNSTV). Le T


148

Modelling nutrient transfer in the river system

149

CHAPTER 6

Modelling nutrient transfer in the river system: implementation of the Seneque/Riverstrahler software

6.1. Introduction

The biogeochemical functioning of a river system is characterized by the fluxes of transfer,

transformation and retention of biogenic elements during their downwards travel from the

terrestrial watershed to the sea through the whole drainage network.

Both the importance of the inputs of these elements and the intensity of the processes they

undergo within the system depend on the complex interplay of climatic, geomorphologic and

anthropogenic factors. Understanding the role of all these factors, and assessing the

relationship between human activity in the watershed and ‘water quality’ of the river system,

given the variability of natural factors, is a prerequisite for a rational management of water

resources. The Riverstrahler model (Billen et al., 1994; Garnier et al., 1995; Billen and

Garnier, 1999) has been established for this purpose. Recently, the model has been

encapsulated into a GIS interface in order to build a generic and spatially explicit software

(Seneque/Riverstrahler) which can be implemented to any watershed, provided a suitable data

base is assembled under a specific format (Ruelland, 2004; Ruelland et al. in prep). This

approach, which was first developed for the well documented Seine river system, has proven

particularly fruitful for addressing different water management issues in temperate regions,

including the questions of oxygen deficits in regulated rivers (Garnier et al., 1999), of nitrate

contamination from diffuse sources (Billen and Garnier, 1989), of excessive algal

development either in the river drainage or in the adjacent coastal zones (Garnier et al., 2005;

Cugier et al., 2005).

Although it is conceived as a generic tool, implementing this model for the case of less

documented river systems where no Water Agencies or similar organisms are established,

might be much more difficult, by lack of access to the required data bases.

This chapter is prepared as an article to be submitted under the title: Modelling the biogeochemical

functioning of the Red River system: implementation of the Seneque/Riverstrahler software.


150

In this paper we describe how the Seneque/Riverstrahler has been successfully applied to the

case of a tropical river: the Red River, in Vietnam and China. We will describe the minimum

information required by the software to take into account the various constraints to the river

drainage functioning, and how the corresponding information has been gathered for the

special case of the Red River. Confrontation of the model results with observations on water

quality gathered at the outlet of the major tributaries of the Red River system in the scope of a

survey programme at monthly intervals throughout two years, will allow validating the

modelling approach. Some general conclusions on the biogeochemical functioning of the Red

River system and its nutrient export will be presented.

6.2. The Riverstrahler model

Riverstrahler describes the drainage network of any river system as a combination of basins,

idealized as a regular scheme of confluence of tributaries of increasing stream order, each

characterized by mean morphologic properties, connected to branches, represented more

realistically, with a higher spatial resolution. The advantage of this representation of the

drainage network is that it allows, with reasonable calculation time, to take into account both

the processes occurring in small first orders, headwater streams and those occurring in large

tributaries. The water flows in the hydrographical network are calculated from the specific

discharges generated within the watershed of the different sub-basins and branches

considered. These are calculated from rainfall and potential evapotranspiration by a simple

rainfall-discharge model with two compartments (Hydrostrahler module). The discharge in

any stream order river or branch is the sum of two components, one corresponding to surface,

or sub-root (hypodermic) runoff, the other to groundwater, base flow.

The essence of the model is to couple these water flows routed through the defined structure

of basins and branches, with a model describing biological, microbiological and

physicochemical processes occurring within the water masses. The module representing the

kinetics of the processes is known as the Rive model. The state variables comprise nutrients,

oxygen, suspended matter, dissolved and particulate non living organic carbon, as well as

algal, bacterial and zooplanktonic biomasses. Most processes important in the transformation,

elimination and/or immobilization of nutrients during their transfer within the network of

rivers and streams are explicitly calculated by the Rive, including algal primary production,

aerobic and anaerobic organic matter degradation by planktonic as well as benthic bacteria


151

with coupled oxidant consumption and nutrient remineralization, nitrification and

denitrification, phosphate reversible adsorption onto suspended matter and subsequent

sedimentation, etc. A detailed description of the Rive Model and of the physiological

parameters used can be found in Garnier et al. (2002a).

Besides morphological and climatic constraints, the Riverstrahler takes into account diffuse

and point sources of nutrient from land based anthropogenic sources. Diffuse sources of

nutrients are taken into account by assigning a constant concentration for all nutrients to

surface and groundwater flow respectively. Point sources, typically wastewater discharges,

must be specified by stream-order for the basins, and at their exact location for the branches.

The newly developed Seneque interface allows to derive from a general GIS data base

covering the watershed, all the input files required for running the Riverstrahler model, and

this for any portion of the drainage area, represented as a particular structure of basins and

branches defined by the user according to the spatial resolution required for his application

(Ruelland, 2004 ; Ruelland et al., in prep). Assembling a suitable GIS database is thus the

key for running this generic software.

6.3. Geomorphology

The whole GIS data base is structured according to the representation of the drainage

network, as a system of connected directional arcs, with each confluence marking the

beginning of a new arc. Each arc is described by data such as Strahler stream order, length,

width, slope... To each arc corresponds an elementary watershed, representing the area

directly drained to this arc, which thus constitutes the elementary spatial grid of the model.

The best way to obtain the drainage network and the elementary watershed coverage is to

derive them from a Digital Elevation Model (DEM) (Tarboton et al., 1991). For the Red

River, we made use of the digital elevation model SRTM at 3” resolution made available at a

global scale by NASA (www:\\NASA.org), (Figure 6.1a). The Standard ArcInfo Workstation

software has been used to generate the watersheds and the network coverage. First, SRTM

data sets have been converted into Arcinfo Grid, and then sinks have been filled. A minimum

watershed size of 25000 cells (approximatively 200 km²) was imposed in the process, limiting

the upstream drainage network to stream orders 2 or 3. The obtained drainage network was

compared to available topographic maps and the few required corrections were made

manually.


152

100 km100 km

Figure 6.1: Relief of the Red River basin deduced from the STRM Digital elevation model

(www:\\NASA.org) and the structure of the drainage network and elementary watershed

derived from it, using the procedure described by Tarboton et al., 1991.


153

6.4. Hydrology

The Hydrostrahler model included in the Seneque/Riverstrahler software calculates the

seasonal variations of specific base flow and superficial runoff by periods of ten days for each

sub-basin. This result is used for reconstructing the discharge anywhere in the drainage

network, but also to calculate the diffuse sources of nutrient from the watershed. The

Hydrostrahler model requires daily rainfall and potential evapotranspiration data at a number

of stations in the watershed. We could obtain daily rainfall data for 13 meteorological stations

in the Vietnamese part of the basin (IMH, 1997-2004) and at the station Kunming in China,

for the period from 1997 to 2003. Evapotranspiration was calculated using Turc’s formula

(Turc, 1961), based on monthly temperature (T°C) and sunshine duration (Sdur, h) data

obtained from the respective meteorological stations:

ETPmm/month = 0.4 T°C (Ig+50)/(T°C+15)

where

T°C is the atmospheric temperature in °C in the period considering

Ig is the total solar radiation expressed in cal.cm-2.d-1 in the period considered, which

can be calculated by the relation:

Ig = IgA (0.18+ 0.62 h/H)

in which

IgA is the energy of solar radiation in the absence of atmospheric attenuation,

expressed in cal.cm-2.d-1.

h/H is the relative duration of sunshine, H being the duration of the astronomic

day and h, the duration of sunshine period per day.

IgA and H values, which only depend on the latitude and the period of the year,

are provided by Turc (1961).

The Hydrostrahler model involves four empirical hydrological parameters (soil saturation

level (solsat), superficial runoff rate (srr), infiltration rate (infr), groundwater runoff rate

(gwrr), the value of which is calculated by the software for any sub-basin from its lithological

characteristics, provided suitable parameters are defined for each lithological class of the

watershed.

Independently of the software, we developed an automated procedure (Le Thi Phuong Quynh

et al., subm.) allowing to calibrate the values of these parameters for the three upstream sub-


154

basins of the Red River, on the basis of observed discharge values at their outlets,

communicated by the Ministry of Natural Resources and Environment (MONRE, 2004),

(Table 6.1).

Table 6.1: Hydrological parameters derived by calibration of the Hydrostrahler model on observed discharge values at the outlet of the three main sub-basins of the Red River for the

period 1997-2003 (Le Thi Phuong Quynh, subm).

Parameter Thao Da Lo solsat, mm 110 165 210 infr, d-1 0.062 0.038 0.05 srr, d-1 0.038 0.075 0.068 gwrr, d-1 0.013 0.0026 0.001

A detailed lithological map of the Red River basin is not available. We used the information

provided by the global geological/lithological database of Dürr (2003) (Figure 6.2).

Combining the distribution of the lithology in the three sub-basins, with the calibrated values

of Table 6.1, we assigned a value of the hydrological parameters to each lithological class in

order to reproduce as well as possible the combined values for the three sub-basins (Table

6.2).

Figure 6.2: Lithological map of the Red River basin derived from the global

geological/lithological map of Dürr (2003).


155

Table 6.2: Hydrological parameters for each lithological class

Parameter plutonic

acid basic

volcanic paleozoic

sedimentarymesozoic

silicic mesozoic

carbonated alluvial deposits

solsat, mm 150 100 180 30 250 400 infr, d-1 0.050 0.050 0.060 0.020 0.040 0.050 srr, d-1 0.040 0.070 0.050 0.080 0.040 0.020 gwrr, d-1 0.001 0.020 0.001 0.025 0.001 0.001

6.5. Role of dams

Two major dams are in operation in the upstream sub-basins of the Red River. Two others are planned to be impounded in the next decade (Table 6.3). Le Thi Phuong, et al., (subm) have proposed to represent the hydraulic behavior of these dams by simple management rules, based on the value of their maximum and minimum volume, as well as two critical values of the input discharge above which the dam is allowed to fill or below which it is emptied. Based on these rules, an algorithm determining their period of filling and emptying in function of upstream river discharge was constructed and corresponding parameters determined (Table 6.3). Based on these parameters and on the pre-calculated water quality of the inflowing water, a software associated to the Seneque/Riverstrahler model and very similar to the Barman model described by Garnier et al. (2000), calculates the hydrological and biogeochemical functioning of the dam, using exactly the same kinetic formulation of the ecological processes. This model provides the files required for allowing Seneque to fully take into account the role of the dams: a file providing daily values of inflowing and outflowing discharge, and a file providing the quality of dam water.

Table 6.3: Some characteristics of the large dams impounded (Hoa Binh and Thac Ba) or planned (Son La and Dai Thi) in the Red River basin (from Le Thi Phuong Quynh et al, submitted)

Name of dam Hoa Binh Thac Ba Son La Dai Thi

River (sub-basin) Da Chay (Lo) Da Gam (Lo) Date of impoundment 1985 1972 2010-2015 2010 *Volume (min-max), 109 m3 3.9 – 9.5 0.78-2.94 9.3-25.5 0.5-3.0 *Critical upstream discharge, m3/s 1750-1500 200-190 850-750 ? Surface area, km² 208 235 440 42 Mean depth, m 50 42 60 70 Upstream watershed, km2 57285 6170 26000 9700


156

6.6. Land use and non-point sources of nutrients

To calculate diffuse sources of nutrients in each sub-basin, the Riverstrahler model assigns a

yearly constant mean composition to surface- and base flow runoff respectively, according to

land use of the watershed.

a.

b.

Figure 6.3: a. Land use map for the Vietnamese part of the Red River basin, based on the data from MONRE (2004); b. Land use map for the whole Red River basin based on the

Global 1° land cover map of DeFries et al., 1998 and Hansen, et al., 2000.


157

A GIS land use coverage of the Vietnamese territory was made available from the Ministry of

Science, Technology and Environment (MOSTE, 1997). We considered the following 6

classes as the most relevant for our purpose: forest, grassland, paddy rice fields, other (dry)

cultures including industrial cultures, rocks and bareland, urban areas (Figure 6.3a). Similar

information was not available for the Chinese part of the basin. However, a global GIS

covering (DeFries et al., 1998; Hansen, et al., 2000) at a 1° resolution provide information on

the spatial repartition of forest (with a lot of details on the kind of forested formations),

grassland, cropland, bare ground, and urban area (Figure 3b). The legends of these two data

bases differ in the fact that the latter does not distinguish paddy rice fields from other

croplands. We arbitrarily assigned a constant proportion of 33% of rice fields in total cropland

to all elementary watersheds in China, a figure obtained from the general statistics of land use

of the Yunnan province as a whole (Chinadata, 2000).

The Seneque associates the spatial distribution of these land use classes with a parameter file

providing the corresponding concentrations of all variables in superficial and base flow issued

from these classes. The water composition (organic matter, nitrate, ammonium, total

phosphorus) assigned to each of the land use classes considered in the data base should be

calculated from data of empirical surveys of surface water composition of small streams

draining homogeneous basins with given land use. For the Red River basin, we relied on an

extensive survey of literature (see Le Thi Phuong Quynh et al., 2005, chapter 5) and on our

own unpublished measurements (see chapter 4), (Table 6.4).

As far as suspended matter is concerned, the rather low concentration observed in the Lo sub-

basin compared with the two other basins is rather paradoxical in view of the fact that this

basin has the greatest proportion of industrial crops in its watershed (Le Thi Phuong Quynh et

al., 2005). This difference, which probably results from a different geomorphological and/or

geological context in the Lo basin, led us to define two different values for the suspended

matter concentration associated with industrial crops in the Lo and the other sub-basins (Table

6.4). The total inorganic phosphate concentration is calculated from the suspended matter

concentrations using the measured total phosphorus content of suspended matter at the outlet

of the sub-basins, namely 0.43 mgP.g-1 for the Thao and Da Rivers, and 0.85 mgP.g-1 for the

Lo River (Le Thi Phuong Quynh et al., 2005).


158

Table 6.4: Composition of surface runoff according to land use in the Red River basin, as taken into account by the Seneque/Riverstrahler model.

Land use class Susp. matter.

mg/l NO3

-

mgN/l NH4

+

mgN/l Ptot

mgP/l Rocks 5000 0.4 0.015 3 Forest 1000 0.4 0.015 0.6 Grassland 3000 1.4 0.015 0.43 Industrial (dry) cultures 8000 (Lo 150) 2.8 0.015 4.8 (Lo 0.13) Paddy rice field 3000 0.02 2 1.8 Urban areas 8000 2.8 5 4.8

Regarding dissolved silica concentration, the concentration is similarly related to the

distribution of lithological classes, obtained from the global lithological world map of Dürr

(2003) (Figure 6.2), using the corresponding mean SiO2 proposed by Meybeck (1986, 1987),

taking into account the mean temperature of the Red River basin (Table 6.5).

Table 6.5. Dissolved silica concentration associated with each lithological class (according to Meybeck, 1986, 1987.

lithological class Dissolved silica , mgSi.L-1

plutonic acid 4.9 basic volcanic 7.7 paleozoic sedimentary 4.4 mesozoic silicic 5.5 mesozoic carbonated 3.2 alluvial deposits 3.8

6.7. Wastewater point sources

The population in the whole basin was estimated to 30.02 million inhabitants for the year

1997, of which 65 % in Vietnam (MOSTE 1997), 34 % in China (Statistical Yearbook of

China data, 1998) and 1% in Laos. For Vietnam, the population of 5235 villages and towns

could be geo-referenced in the Red River basin GIS. For China, only data on the urban

population and rural population of the Yunnan province as a whole were available. In fact, the

Red River basin in Yunnan drains none of the major cities of the province (Kunming,

Dali,…). We therefore uniformly distributed the rural population density of Yunnan (81

inhab.km-2) within the Chinese part of the upstream Red River watershed. The population


159

density of the whole basin thus varies from 81-150 inhab.km-2 in the upstream watershed to

over 1000 inhab.km-2 in the delta area (Figure 6.4).

From the analysis of the domestic consumption budget of food and washing powders in

Vietnam, Le Thi Phuong Quynh et al. (2005) estimated the human per capita release of

nutrients to 0.010 – 0.015 kgN.cap.d-1 and 0.0046 – 0.005 kgP.cap.d-1, the lower values

corresponding to the poor rural region of the upstream basin, while the highest hold for the

urban area in the delta. Moreover, in small rural villages (<10 000 inhab.) in the upstream

part of the basin, we estimated that only 25% of domestic wastewater reaches the surface

waters, the remaining part being recycled in agriculture, while in large villages, and in the

delta region, where the population is mostly agglomerated and running water is present

everywhere, we considered that all domestic wastewater is discharged to the hydrosystem.

On this basis, a database of all domestic inputs of wastewater was established for the

Seneque/Riverstrahler model. A first census of industrial wastewater discharges has been

carried on (Le Thi Phuong Quynh et al., 2005) but it remains very partial and has not been

included in the data base.

1 -50

50-100

100-200

200-500

500-1000

1000-2000

2000-5000

5000-10000

1 -50

50-100

100-200

200-500

500-1000

1000-2000

2000-5000

5000-10000

Figure 6.4: Distribution of the population density in the elementary watersheds of the Red

River basin, as used for calculating the point sources of wastewater in the

Seneque/Riverstrahler model.


160

6.8. Validation and flux calculation

The Seneque/Riverstrahler software can be used to calculate the spatial and seasonal

variations of water quality at the scale of the whole drainage network. The interface allows

the user to immediately visualize the results under three formats:

- seasonal variations of discharge or concentrations of any variable at one station (either at the

outlet of the tributary of a specified stream order for a sub-basin, or at a specified kilometric

position for a river branch);

- longitudinal variations of discharge or concentrations along a river branch at a specified

time;

- cartographic representation of the variables (with an adjustable color code) over all basins

and branches of the simulations at a specified time period.

For the two former representations, the possibility exists to automatically compare the

calculation results with measured data when these are stored in the database. This comparison

allows the validation of the modelling procedure on recent well documented situations.

A two-year survey of water quality has been carried out at monthly intervals during the year

2003 and 2004 at the stations Yen Bai (Thao River), Vu Quang (Lo River) and Hoa Binh (Da

River), Son Tay and Hanoi (Hong river) (see chapter 4). In order to adapt the resolution of

the model to the requirement of the validation with respect to these sampling stations, as well

as in order to take into account the presence of the two dams in the upper drainage network, a

suitable spatial representation of the river system in the model should be defined. The chosen

representation is shown in Figure 6.5. It involves 7 basins, 5 branches and the two presently

operating dams. It treats the upper half of the whole Red River basins as “Strahler-idealized”

basins, while the lower half courses of all three major tributaries are treated as river branches

with a kilometric resolution.

As an example, the data obtained during the field campaigns in 2003 at the 4 stations cited

above are compared with the results of the model (Figure 6.6). On the other hand, we

compare the observations and the simulations at Yen Bai, on the Thao River, for both years

2003 and 2004 (Figure 6.7). At this stage, the agreement between observed and calculated

concentrations, although far from being perfect, is in general not bad: the model is able to

reproduce the observed general levels of nutrient concentration, which is not a priori obvious.


161

Figure 6.5: Spatial representation of the Red River drainage network by the Seneque/Riverstrahler software for the validation of the results.

In fact, very few calibration procedures have been applied in the construction of the model,

excepted concerning the hydrological sub-model (see chapter 3). The agreement between the

model and the observations thus represents the exact measure of the correctness of our

representation of the system functioning and/or of our knowledge of the constraints

controlling it. The discrepancies between simulations and observations thus deserve close

examination, because they inform us on the weak aspects of our approach.

Concerning our knowledge of the constraints several weak points can be mentioned which could be responsible for a part of the discrepancies between simulation results and observation. The first lies in the meteorological data, especially in the Chinese part, which covers a haft area of the Red River basin where we have only one station (Kunming) with mean monthly meteorological data in the period from 1997 to 2001. We applied the mean monthly values of this series to the period 2003 and 2004. This leads to errors both on hydrology and water quality simulations. The same limitation of our knowledge of the conditions in the Chinese part of the Red River basin concerns the distribution of the population and industrial activity, that we know only as an overall figure for the whole Yunnan province, of which the Red River basin makes 20.9%. Even for the Vietnamese part of the basin, on the other hand, our knowledge of industrial wastewater releases is quite insufficient and the figures we used are probably largely underestimated. We hope that a


162

better knowledge of all these constraints and their distribution at the scale of the whole Red River basin would improve the simulation by the model.

As far as suspended matter is concerned, the level in the Lo River is strongly overestimated by the model, despite the lower suspended solids delivery considered for industrial cultures in the Lo basin with respect to the other sub-basins. On the contrary, the model underestimates the suspended matter of the Thao River (Fig. 6.6 and 6.7). Clearly, a general assumption concerning the relation of suspended matter yield with land use is not able to reproduce the differences in suspended loading among the sub-basins. The process of erosion, which generates the suspended solid load of the river network, should be described in a much more refined way, even at the regional scale at which we are working. This problem is also apparent on the simulations of total phosphorus, a large part of which is linked to suspended matter: the model also overestimates total phosphorus in the Lo basin. However, as far as dissolved phosphate concentrations are concerned, taking into account the variability of their measurements at the low levels occurring in most of the stations, the model predictions are rather satisfactory, which indicates that the kinetic formulation of the adsorption-desorption equilibrium used in the Rive model (Garnier et al., 2005) is valid for the Red River.

Chlorophyll a concentrations predicted by the model, although they fluctuate may be too much, reproduce some important trends revealed by the measurements: as observed, the Thao river is the only one among the three large tributaries to develop a significant planktonic algal biomass. The phytoplankton development occurs in spring, by low discharge, and to a lower extend in autumn, after the flood.

The model does not capture all the observed variability of nitrate concentrations. Although the general level is correctly reproduced for most stations, the model underestimates nitrate concentration in the Thao River. On the other hand, the very low ammonium levels are satisfactorily reproduced by the model. These low levels, in spite of significant inputs of ammonium by point sources of waste water and diffuse sources from paddy rice soils, could be the result of an active in-stream nitrifying activity. Setting the nitrifying activity to zero in the model, results in much higher calculated ammonium concentration at Hanoi station.

Finally, the agreement between the model simulations and the observations of dissolved silica concentration is so good. An important conclusion from the model simulation is that there is no significant retention process of silica along the river system. No period of silica depletion occurs, even during the limited planktonic blooms in the Thao River. The large dams on the Da and Lo Rivers are not responsible for any significant silica retention.


0

10

20

J F M AM J J A S O N D

Ch

la, µ

g.L

-1

0

1

2


N-N

H4,

mg

.L-1

0

2

4


N-N

O3,

mg

.L-1

0.0

0.6

1.2


Pto

t, m

g.L

-1

0.0

0.1

0.2


P-P

O4,

mg

.L-1

0

10

20


SiO

2, m

g.L

-1

Da R.

0

4000

8000


Dis

char

ge,

m3 .

s-1

0

2000

4000

J F MAM J J A SO NDS

S, m

g.L

-1

0

10

20

J F M A M J J A S O N DC

hla

, µg

.L-1

0

1

2


N-N

H4,

mg

.L-1

0

2

4


N-N

O3,

mg

.L-1

0.0

0.6

1.2


Pto

t, m

g.L

-1

0.0

0.1

0.2


P-P

O4,

mg

.L-1

0

10

20


SiO

2, m

g.L

-1

Lo R.

0

4000

8000


Dis

char

ge,

m.s

-13

0

2000

4000

J F MAM J J A SO ND

SS

, mg

.L-1

0

10

20


Ch

la, µ

g.L

-1

0

1

2


N-N

H4,

mg

.L-1

0

2

4

J F M A M J J A S O N DN

-NO

3, m

g.L

-10.0

0.6

1.2


Pto

t, m

g.L

-1

0.0

0.1

0.2


P-P

O4,

mg

.L-1

0

10

20


SiO

2, m

g.L

-1

Thao R.

0

4000

8000


Dis

char

ge,

m3 .s

-1

0

2000

4000

J F MA M J J A S O ND

SS

, mg

.L-1

0

10

20


Ch

la, µ

g.L

-1

0

1

2

J F M A M J J A S O N DN

-NH

4, m

g.L

-10

2

4


N-N

O3,

mg

.L-1

0.0

0.6

1.2


Pto

t, m

g.L

-1

0.0

0.1

0.2


P-P

O4,

mg

.L-1

0

10

20


SiO

2, m

g.L

-1

Hong R.

0

4000

8000


Dis

har

ge,

3 .

s

0

2000

4000

J F MAM J J A S O ND

SS

, mg

.L-1

-1m

c

Figure 6.6: Seasonal variations in 2003 of observed (open circles) and calculated (solid curve) variables (from left to right): discharge, suspended solids (SS), chlorophyll a (Chla: µg.L-1), nutrient concentrations (nitrate and ammonium: mg N.L-1; total phosphorus and phosphates mg P.L-1;

dissolved silica: mgSiO2.L-1). From top to bottom are figured the four sampled stations: Hoa Binh (Da R.), Vu Quang (Lo R.), Yen Bai (Thao R.) and Hanoi (in the delta, Red R.)

163


164


Figure 6.7: Station Yen Bai (upper Thao R.) in 2003 (left) and in 2004 (right): seasonal variations of observed (open circles) and calculated (solid curve) variables. From top to bottom are represented: discharge, suspended solids (SS), chlorophyll a (Chla: µg.L-1), nutrient concentrations (nitrate and ammonium: mgN.L-1; total phosphorus and phosphates mgP.L-1; dissolved silica: mgSiO2.L-1).

0

10

20


Chl

a, µ

g.L-1

0

1

2


N-N

H4,

mg.

L-1

0

2

4


N-N

O3,

mg.

L-1

2003

0

2000

4000


Dis

char

ge, m

3 .s-1

2004

0

10

20


Chl

a, µ

g.L-1

0

1

2


N-N

H4,

mg.

L-1

0

2

4


N-N

O3,

mg.

L-1

2004

0

2000

4000


Dis

char

ge, m

3 .s-1

0

2000

4000


SS, m

g.L-1

0

2000

4000


SS, m

g.L-1

0.0

0.6

1.2


Ptot

, mg.

L-1

0.0

0.6

1.2


Ptot

, mg.

L-1

0.0

0.1

0.2


4, m

g.L-1

P-PO

0.0

0.1

0.2


P-PO

4, m

g.L-1

0

10

20


SiO

2, m

g.L-1

0

10

20


SiO

2, m

g.L-1

165

Modelling nutrient transfer in the river system Figure 6.8 illustrates how the Seneque/Riverstrahler software can take into account the effect

of a large dam on the concentration of particulate material in river water. Two calculated

longitudinal profiles of suspended solid in the Da river are shown, respectively in March (dry

season) and July (wet season), two periods for which the level in suspended solids strongly

contrasts (250 mg.L-1 against 1150 mg.L-1 respectively). Downstream from the dam, a huge

abatement of suspended solids concentrations is shown, up to 85 % during high water flows

but hardly lower in low waters. This pattern is confirmed by the seasonal variations of

suspended matter and total phosphorus at two stations upstream and downstream the dam

(Figure 6.8); the comparison with the available observed data shows that the model is able to

correctly predict the effect of the presence of a dam in the drainage network.

Two examples of a cartographic representation of the calculated results are presented here

(Figure 6.9). Figure 6.9.a illustrates the spring algal development in the drainage network as

calculated by the model. The stronger algal growth in the Thao River is well illustrated.

Clearly, this algal development is already initiated in the upstream sector of the river. Figure

6.9 b shows the distribution of total phosphorus and the role of the two dams (Hoa Binh on

the Da River and Thac Ba on the Lo river) in the abatement of total phosphorus concentration.

The model can also be used to estimate the total flux of nutrient delivery at the outlet of the

different sub-basins (Table 6.7). As far as suspended matter is concerned, the agreement with

our previous estimate (see chapter 3) is acceptable for the Da, but the models, as already

discussed, severely overestimates the suspended solid load of the Lo while it underestimates

that of the Thao River. Regarding nitrogen and phosphorus delivery, the model estimates

differ by less than a factor 2 from those resulting from our budgeting approach (Le Thi

Phuong Quynh et al., 2005), (see chapter 5).

The model provides also an estimate of the silica delivery, which can be used to calculate the

molar Si/N, Si/P and N/P ratios at the outlet of the systems.

166


a)

0

500

1000

1500

0 50 100 150 200 250km

Susp

ende

d so

lid, m

g.L-1 Mar sim.

Jul. sim.

0.0

0.4

0.8

1.2


Ptot

, mg.

L-1

0.0

0.4

0.8

1.2


Ptot

, mg.

L-1

upstream the dam

0

500

1000

1500


Susp

ende

d so

lid, m

g.L-1

downstream the dam

0

500

1000

1500


Susp

ende

d so

lid, m

g.L-1

sim.

Obs.

b)

Figure 6.8: a) Longitudinal variations of suspended solid concentrations along the Da river branch in March and in July in 2003; b) Seasonal variations, in 2003, of the simulations (sim.) and observations (obs.) of suspended solid (mgSS.L-1) and total phosphorus (Ptot, mgP.L-1) concentrations obtained at a station upstream (left) and downstream (right) the Hoa Binh dam on the Da River. (The observed suspended solid concentration upstream from the dam are mean monthly values reported by Nguyen Viet Pho et al. (2003) for the Lai Chau station in the period 1961-1989).

167


<0.2 mgP/l

Total Phosphorus

<2 µgChla/l>2 µgChla/l>4 µgChla/l

Chlorophyll a >6 µgChla/l

Early April 2003

>0.2 mgP/l>0.4 mgP/l>0.6 mgP/l

<0.2 mgP/l

Total Phosphorus

<2 µgChla/l>2 µgChla/l>4 µgChla/l

Chlorophyll a >6 µgChla/l

Early April 2003

>0.2 mgP/l>0.4 mgP/l>0.6 mgP/l

Figure 6.9: Cartographic representation of the geographical distribution of chlorophyll a and total phosphorus concentration in the drainage network of the Red River system at the beginning of April 2003, as provided by the Seneque/Riverstrahler model. The higher (although limited) algal development, as well as the higher concentration of total phosphorus in the Thao river than in the other tributaries is quite apparent, as well as the role of the dams (open circle) in reducing the particulate phosphorus concentration.

Table 6.7: Nutrient flux (N: nitrogen; P: phosphorus; Si: silica) calculated by the model (Sim) at the outlet of the Da, Lo, Thao and Hong Rivers at the stati

ons Hoa Binh, Vu Quang, Yen Bai and Hanoi respectively, in 2003.

Total N Total P Dissolved Si Suspended solid

103 tons N.y -1

103 tons P.y -1

103 tons Si.y -1

106 tons SS.y -1

Sim. Ref.* Sim. Ref.* Sim Sim. Ref.*

Da 61 38 9.3 3.5 274 7.4 5.5

Lo 55 32 24.0** 5.1 128 35.0** 7.9

Thao 15 22.5 7.9 8.3 57 8.4 20

Hong Hanoi 146 - 48.2 - 475 58.5 40

Ref.*: data provided in Le Thi Phuong Quynh et al., 2005; Silica fluxes not calculated in Le Thi Phuong Quynh et al., 2005

** note the overestimations by the model (cf. text) of total phosphorus and suspended solids fluxes for the Lo River.

168

Modelling nutrient transfer in the river system The molar N/P ratios at the outlet of the three sub-basins Thao, Lo, Da and in the main branch

of the Red River at Hanoi are respectively 1.9, 2.3, 6.6 and 3.0. They are much lower than

the Redfield ratio of 16 characterizing the requirement of algal growth (Redfield et al., 1963).

The molar Si/N ratios at the same stations, 3.7, 2.3, 4.5 and 3.2 respectively, are much higher

than 1 (Conley et al., 1993; Billen and Garnier, 1997), indicating that silica is largely in

excess over the requirements of diatoms. We already arrived to this conclusion of great

importance (Garnier and Billen, 2002; Le Thi Phuong Quynh et al., 2005), as silica limitation

is often at the origin of harmful algal blooms at the coastal zone (see Cugier et al., 2005 and

included references).

Finally, it must be stressed that the model, in its present development stage is not able to

focus on the small urban rivers in the Delta area where water environment is the most

seriously polluted, as mentioned in the chapter 4. In fact, the Seneque/Riverstrahler model

does not here consider distributaries; it is only here able to model one main branch of the Red

River in the Delta area, which in fact limits seriously its ability to represent the real situation

of water pollution in this, very populated area. One of our aims in the next future, is to be able

to adapt the Seneque/Riverstrahler model to pursue the modelling work down to the coastal

zone through the complex drainage network of the Red River delta.

6.9. References

Billen, G., Garnier, J. and Hanset, P. (1994). Modelling phytoplankton development in whole drainage networks: The RIVERSTRAHLER model applied to the Seine river system. Hydrobiologia, 289: 119-137.

Billen, G. and Garnier, J. (1997). The Phison River plume: coastal eutrophication in response to change in land use and water management in the watershed, Aquat. Microb Ecol 13: 3-17.

Billen, G. and Garnier, J. (1999). Nitrogen transfer through the Seine drainage network: a budget based on the application of the RIVERSTRAHLER Model. Hydrobiologia, 410: 139-150.

Chinadata, (1998). Statistical yearbook of Yunnan, Vol. 1997, Vol. 1998, China Statistical Publishing House, (Basic Information of Yunnan, China) (http://chinadatacenter.org).

Chinadata, (2000). Statistical yearbook of Yunnan, Vol. 1999, Vol. 2000; China Statistical Publishing House, (Basic Information of Yunnan, China) (http://chinadatacenter.org).

Conley, D.J., Claire, L. S and Stoermer, E. F. (1993). Modification of the biogeochemical cycle of silica with eutrophication. Marine Ecology progress series, Published November 4, 101: 179-192.

169

Modelling nutrient transfer in the river system Cugier, J., Billen, G., Guillaud, J.F., Garnier, J., and Memesguen, A. (2005). Modelling the

eutrophication of the Seine Bight (France) under historical, present and future riverine nutrient loading. Journal of Hydrology. Volume 304, issues 1-4: 381-396.

DeFries, R., Hansen, M., Townshend, J.R.G. and Sohlberg, R. (1998). Global land cover classifications at 8 km spatial resolution: The use of training data derived from Landsat imagery in decision tree classifiers, International Journal of Remote Sensing; 19(16): 3141-3168.

Dürr, H.H. (2003). Towards a typology of global river systems: some concepts and examples at medium resolution. Thèse de doctorat, Université Paris VI - Pierre et Marie Curie. 732pp.

Garnier, J., Billen, G. and Coste, M. (1995). Seasonal succession of diatoms and chlorophyecae in the drainage network of the River Seine : Observations and modelling. Limnology. and Oceanography, 40: 750-765.

Garnier, J., Billen, G. and Palfner, L. (1999). Understanding the oxygen budget and related ecological processes in the river Mosel: the RIVERSTRAHLER approach. Hydrobiologia 410: 151-166.

Garnier, J., Billen, G., Sanchez, N. and Leporcq, B. (2000). "Ecological functioning of the Marne dam (upper Seine basin, France)." Regulated rivers: Research & Management, Regul. Rivers: Res. Mgmt. 16: 51-71.

Garnier, J., Billen, G., Hannon, E., Fonbonne, S., Videnina, Y. and Soulie, M. (2002a). Modeling transfer and retention of nutrients in the drainage network of the Danube River. Estuarine, Coastal and. Shelf Science, 54: 285-308.

Garnier, J. and Billen, G. (2002b). The Riverstrahler modelling approach applied to a tropical case study (The Red –Hong- River, Vietnam): nutrient transfer and impact on the Coastal Zone. SCOPE, Coll. Mar. Res. W., 12: 51-65

Garnier, J., Nemery, J., Billen, G., and Thery, S. (2005). Nutrient dynamics and control of eutrophication in the Marne River system: modelling the role of exchangeable phosphorus. Journal of Hydrology, Volume 304, issues 1-4: 397-412.

Greenlee, D.D. (1987). Raster and Vector Processing for Scanned Linework, Photogrammetric Engineering and Remote Sensing. Vol. 53(10): 1383-1387.

Hansen, M., DeFries, R., Townshend, J. R. G. and Sohlberg, R. (2000). Global land cover classification at 1km resolution using a decision tree classifier, International Journal of Remote Sensing. 21: 1331-1365.

IMH, (1997-2004). Journal of Meteo-hydrology, Institute of Meteo-Hydrology in Vietnam, Hanoi. (Monthly Journal during the periods of from 1997 to 2004).

Jenson, S.K. and Domingue, J.O. (1988). Extracting Topographic Structure from Digital Elevation Data for Geographic Information System Analysis, Photogrammetric Engineering and Remote Sensing. Vol. 4(11): 1593-1600.

170

Modelling nutrient transfer in the river system Le, Thi Phuong Quynh, Billen, G., Garnier, J., Théry, S., Fézard, C. and Chau, Van Minh

(2005). Nutrient (N, P) budgets for the Red River basin (Vietnam and China). Journal of Global Biogeochemical cycles.Vol 19, GB2022, doi 10.1029/2004GB002405.

Le, Thi Phuong Quynh, Billen, G., Garnier, J., Théry, S. and Chau, Van Minh (submitted). Hydrological regime and suspended matter flux of the Red River System (Vietnam): Observations and modelling. Journal of Hydrology.

Meybeck, M. (1986). Composition chimique naturelle des ruisseaux non pollués en France -Sci. Geol.Bull. , 39, 3-77.

Meybeck, M. (1987). Global chemical weathering of superficial rocks estimated from river dissolved loads. American Journal of Science, 287: 401-428.

MONRE, (1997-2004). Vietnamese Ministry of Environment and Natural Resources. Report

annual on hydrological observation in Vietnam, Hanoi.

MOSTE, (1997). Vietnamese general statistics officer, Ministry of Science, Technology and Environment of Vietnam, general statistics editor, Hanoi. 550 pp.

Nguyen, Viet Pho, Vu, Van Tuan and Tran, Thanh Xuan (2003). Water resources in Vietnam.

Institute of Meteo-Hydrology. Agricultural Editor. Hanoi

Redfield, A. C., Ketchum, B.H. and Richards, F.A. (1963). The influence of organisms on the

composition of sea-water.In M. N. Hill (ed.), The Sea, John Wiley and Sons, New

York, p. 12-37.

Ruelland, D. (2004). SENEQUE, a GIS software to evaluate water quality. Hermès-Lavoisier Ed., Revue Internationale de Géomatique, vol. 14(1): 97-117.

Ruelland, D., Billen, G., Brunstein, D. and Garnier, J. (in prep). SENEQUE 3A GIS interface to the RIVERSTRAHLER model of the biogeochemical functioning of river systems. To be submitted to Ecological Modelling

Tarboton, D.G., Bras, R.L. and Rodriguez-Iturbe, I. (1991). On the Extraction of Channel

Networks from Digital Elevation Data, Hydrological Processes. Vol. 5 : 81-100.

Turc, L. (1961). Evaluation des besoins en eau d’irrigation, évapotranspiration potentielle,

Ann. Agron., 12 (I), I 3-49

171


172

Exploring future trends of nutrient transfers

CHAPTER 7


The material presented above, as well as the different modelling tools developed in the scope

of this work, now offer the possibility of exploring a number of prospective scenarios

concerning the future biogeochemical functioning of the Red river system. We have

restricted ourselves to two three important aspects: (1) the impoundment of new large dams

on the drainage network of the Red River; (2) the increase of the population and its degree of

urbanization; (3) the changes in land use and the intensification of agricultural practices. The

Seneque/Riverstrahler model implemented on the Red River basin will allow us to predict the

results of possible future changes in these three aspects on the overall water quality and

nutrient delivery of the river system.

7.1. Impacts of new dams constructed in the Red River basin

As mentioned in chapter 3, the construction of two large dams, in addition to the already

existing Hoa Binh and Thac Ba dams, is planned for the next decade. The Son La dam, with a

volume of 9.3-25.5 109 m3 will be constructed on the upper course of the Da river, upstream

from the Hoa Binh reservoir. The Dai Thi dam, with a volume of 0.5-3 109 m3 will be

constructed on the Gam river, a tributary of the Lo river. We already calculated, with the

simplified approach described in chapter 3 that these dams would reduce by about 20% the

solid load of the Red River. As a consequence, the phosphorus loading should also be

significantly reduced.

We have run the Seneque/Riverstrahler model for two scenarios differing from the standard

validated scenario of the year 2003 in that (i) no dams at all are considered (scenario called

“1970”), or (ii) the four large dams are considered operating (scenario called”2050”).

Excepted for this aspect, all the other constraints (hydrology, land use, point sources of waste

water) were taken identical with those of the reference “2003” scenario. Figure 7.1 shows the

results of these simulations for suspended solid and total phosphorus concentrations at the

173

Exploring future trends of nutrient transfers outlet of the Da and Lo rivers and in the main branch at Hanoï station. The effect of the Son

La dam is less apparent than that of the Dai Thi dam (compare sc 2003 and sc 2050), because

the Hoa Binh dam already reduced severely the suspended solid load of the Da river at its

outlet (compare sc 1970 and 2003).

The annual flux of suspended solid and total phosphorus delivery were calculated and given

in Table 7.1. The results show a clear decrease of both suspended solid and total phosphorus

fluxes at the outlet of the Da and Lo rivers, as well, as in the main branch, at Hanoï. In

particular the impoundment of the Dai Thi dam on the Lo river will result in 50% reduction of

the suspended solids flux. These conclusions are in agreement with those of other authors

(Nguyen Huu Khai and Nguyen Van Tuan, 2001; Pham Quang Son, 1998; Nguyen Viet Pho,

2003).

Table 7.1: Simulated fluxes of suspended solid and total phosphorus delivery at the outlet of the Da and Lo rivers and at Hanoi station, calculated for the conditions of the year 2003, without any dam (“1970”), with the two presently existing dams (“2003”) and with two

additional dams (“2050”).

Suspended solid

Total phosphorus

106 ton SS.y -1 103 ton P.y -1

1970 2003 2050 1970 2003 2050 Da 58.1 7.4 6.7 39.6 9.3 8.8

Lo 41.5 35 21.5 27.9 24 11.6

Hong Hanoi 113.8 58.5 44.5 81.6 48.2 39.7

174


0.0

0.4

0.8

1.2


TotP

, mg.

L-1

Da R.

0

500

1000

1500

2000


SS, m

g.L-1

in 1970in 2003in 2050

0.0

0.4

0.8

1.2


TotP

, mg.

L-1Lo R.

0

500

1000

1500

2000


SS, m

g.L-1

0.0

0.4

0.8

1.2


TotP

, mg.

L-1

Hong R.

0

500

1000

1500

2000


SS, m

g.L-1

Figure 7.1: simulation results of suspended solid (SS, mg.L-1) and total phosphorus (Ptot,

mgP.L-1) at the stations Hoa Binh (in the Da River), Vu Quang (in the Lo River) and Hanoi (in the Hong River) in the scenario ‘1970’ (no dam at all), in 2003 (presence of Hoa Binh and Thac Ba dams) and in the scenario ‘2050’ (presence of two more new dams) in the Red River

system.

175

Exploring future trends of nutrient transfers 7.2 Fast increasing population and impact on water quality Several studies have been carried out concerning the factors controlling the long term demographic evolution of Vietnam. Hoang Xuyen (2000), on the basis of a detailed analysis of the present demographic structure of the Vietnamese population and of the evolution of birth, mortality and migration rates since the last 50 years; they conclude that the process of demographic transition, characterized by a reduction of the mortality rate, followed by a reduction of the birth rate, has been initiated in Vietnam, particularly in the North of the country, since the mid 1950’ies, and entered its final stage after the end of the war in 1975. This means that the population of the country should stabilize within about one generation. The same author evaluates the population of Vietnam in 2020 to 100 millions inhabitants (between 98 and 103 millions inhabitants). According to FAO statistics (FAO, 2004), the total population in Vietnam increased from 27.4 106 in 1950 to 83.6 106 in 2005, and will reach 117.7 in 2050 (Figure 7.2). The Vietnamese population should thus increase by a factor 1.4 by 2050.

0

20

40

60

80

100

120

140

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2050

106 in

habi

tant

s

Urban Rural

Figure 7.2: Evolution of the total population in Vietnam (FAO database, 2004), as well as the

urban and rural components.

The rural population always occupies the largest proportion in the whole country (about 80% in 1990s) (Figure 7.2), and this represents a typical characteristics of the social organisation in Vietnam. The analysis of the population data base by villages in the Vietnamese Red River

176


basin shows indeed that 80% of the population lives in agglomerations of less than 10 000 inhabitants (Figure 7.3).

Urban population, however, has increased at a higher rate in the last decades than rural

population. The FAO figures for the whole country show an increase of rural population at a

rate of 1.1%.yr-1 over the last 5 years, while urban population has raised at the rate of 2.2%.yr-

1 in the same period (Fig. 7.2). Based on these figures and on the total population increase

forecasted in 2050 by FAO, we estimate that urban population in Vietnam will represent

about 40% and the fraction of rural population should be 60% by the year 2050.

Figure 7.3: Cumulated population in the Vietnamese part of the Red River basin as a function

of agglomeration size in 1999. 80% of the population live in villages of less than 10 000 inhabitants.

0

2

4

6

8

10

12

0 50000 100000 150000 200000

size of agglomerations, nb inhab

cum

ulat

ed p

opul

., m

illio

n in

hab

Red River basin in Vietnam

Rural population

Analysing the population figures for the provinces of the Red River Delta area and of the

Vietnamese North Mountains regions (roughly, the upstream Vietnamese basin of the Red

River), we arrived at rather similar results concerning the rate of increase of urban versus

rural population (Table 7.2).

Obviously, these rates of increase cannot be extrapolated over the next 50 years, as we know that the total population will stabilize. Instead we made the hypothesis that the ratio between

177

Exploring future trends of nutrient transfers the urban and the rural population growth rate will remain constant in the two regions considered.

Table 7.2: a) Population census in North Vietnam (106) inhabitants in the period from 1997 to 2002; b) annual rate of population increase (%.yr-1) in North Vietnam in the period 1997-2002. Data from MOSTE 1998-2003, ‘Delta’ gathers the data from the provinces Hanoi, Hai Phong, Ha Tay, Hai Duong, Hung Yen, Nam Dinh, Bac Ninh, Thai Binh, Quang Ninh, Ninh Binh The Mountainous region comprises the provinces of Ha Giang, Cao Bang, Lao Cai, Bac Can, Lang Son, Tuyen Quang, Yen Bai, Thai Nguyen, Phu Tho, Vinh Phuc, Bac Giang, Lai Chau, Son La and Hoa Binh.

a) Region 1997 1998 1999 2000 2001 2002

rural urban rural urban rural urban rural urban rural urban rural urbanRed River Delta 12.39 3.43 12.47 3.58 12.53 3.69 12.59 3.79 12.67 3.92 12.75 4.06 Mountainous region 9.47 1.43 9.57 1.47 9.69 1.49 9.76 1.55 9.84 1.60 9.94 1.57

b)

Regions Rural Urban Delta 0.5 3.0 Mountains 0.8 2.4

Assuming in addition that the total population of both region will be multiplied by a factor of 1.4 by 2050 (similarly with the factor of increase of the total Vietnamese population), we were able to calculate, for the two regions considered, the factor of increase of rural and urban population (urbf and rurf). The reasoning is as follows:

For each region (Mountains and Delta respectively):

totP2050 = 1.4 totP2003

hence

urbP2003. urbf + rurP2003. rurf = 1.4 (urbP2003 + rurP2003)

if urbf and rurf are respectively the increase factors of rural and urban population by 2050.

Considering that urbf/rurf = ur (where ur is the observed ratio between the urban and rural population growth rate in the region considered, see table 7.2), the following relationships can be established:

urbf = 1.4 [totP2003 / (urbP2003 + rurP2003/ur) ]

178


rurf = urbf / ur

The application of the above relationships to the two regions of the Red River Basin yields the figures of table 7.3.

Table 7.3: factor of increase of urban (Urbf) and rural (Rurf) population of the Red River basin by 2050

Regions Urbf Rurf Delta 3.9 0.6 Mountains 3.2 1.1

These figures allowed us to construct the spatial distribution of the population of the Vietnamese Red River basin by 2050. For the Chinese part of the basin, that we considered mostly rural, we applied the same figure as for the mountainous Vietnamese part of the basin.

According to these calculations, the total population in the Red River basin considered in the Seneque data base will increase from 16 106 inhabitants to 23 106 inhabitants by 2050. Figure 7.4 compares the spatial distribution of the present and the future population.

2003 2050

> 100 000 inhab.

< 10 000 inhab.

179

Exploring future trends of nutrient transfers Figure 7.4: Distribution of the individual agglomerations of the Red River basin in the 2003 and 2050 scenarios. An ellipse indicate the zone of major changes in the upper delta.

Essentially, the 2050 scenario tend to reinforce the major trends of the spatial distribution, already apparent in the present situation, namely an accentuated concentration of the population in the delta area, with a few centres of population agglomeration in the upper Lo and Da river sub-basins and all along the Thao River (figure 7.4).

7.3 Agricultural evolution and its impact on water quality The increase of the population requires a parallel evolution of agricultural production. The recent trends of land use evolution in North Vietnam are the stabilization of forest cover and the increase in urban area. The latter is probably mostly increasing at the expense of agricultural land, particularly paddy rice fields, the total surface of which is decreasing in the recent years. This evolution is made possible owing to the spectacular and continuing increase in the productivity of Vietnamese agriculture observed since the last decades in terms of yield per unit cultivated surface. This increase is largely due to the use of increasing amounts of chemical fertilizers (Figure 7. 5).

in Vietnam262

140

100

200

300

1950 1960 1970 1980 1990 2000 2010

kg fe

rtilis

er.h

a -1

Figure 7.5: Evolution of fertilizer application (N and P) in Vietnam (FAO database, 2002)

Accordingly, we established a “2050” land use GIS file, by increasing the urban area, for each

elementary watershed unit of the ‘2003’ land use file, at the expense of paddy rice fields or

other agricultural surface if necessary.

180


In order to account for the intensification of agricultural practices, we also assumed that the

nitrate concentration resulting from soil leaching of ‘dry’ agricultural soils (all crops excepted

paddy rice) would have reached the levels typically observed in West European countries (10

mgN/l), and we considered a 25% increase of the export of total phosphorus from agricultural

soils.

7.4. Prospective simulation at the 50 year horizon

The Seneque/Riverstrahler model has been run for an hypothetical “2050” scenario

characterized by the hydrological conditions of the year 2003, the presence of 4 large dams, as

discussed in §7.1, a 40% increase of population distributed between rural and urban centres as

discussed in §7.2, land use and agricultural practices as discussed in §7.3. The results show

the trends of the changes to be expected from this “business as usual” scenario of the future

evolution of the human activity in the Red River basin, compared to the present “2003”

situation (Figure 7.5). Note that we have not considered any difference in wastewater

treatment practices in 2050 compared to 2003, i.e. the same hypothesis concerning the release

of wastewater in urban (no treatment and total release to surface waters) and rural areas (75%

recycling in agriculture) has been made.

The results show a very important increase in nitrate and ammonium contamination of the

Red River, while the level of phosphorus contamination remains nearly the same.

Apparently, the retention of phosphorus by the two additional dams counterbalance to a large

extend the increased release of phosphorus by agricultural soils and human population.

The calculated fluxes of nutrient delivery by the Red River and its main tributaries show the

same trends (Table 7.4). Nitrogen fluxes will be considerably increased at the outlet of the

Thao, Da, Lo rivers and in the main branch of the Hong River, while phosphorus flux at the

outlet of Da and Lo rivers tends to decrease. Only at the outlet of the Thao basin, the most

populated and free of dams basin, phosphorus flux increased in 2050 with respect to 2003.

The resulting total phosphorus flux at Hanoi station is nearly unchanged. Silica fluxes are also

predicted to remain essentially unchanged in response to the 2050 scenario.

The resulting nutrient ratios obviously reflect these trends (Table 7.5). A clear increase of the

N/P ratios is predicted for the 2050 scenario with respect to the 2003 situation, along with a

clear decrease of the Si/N ratios at the outlet of the three rivers Thao, Lo and Da and in the

main branch Hong River.

181


0.0

0.5

1.0


N-N

H4,

mg.

L-1

0.0

0.4

0.8

1.2

1.6


Ptot

, mg.

L-1

0.0

0.1

0.2

0.3


P-PO

4, m

g.L-1

Hong R.

0

2

4


N-N

O3,

mg.

L-1

0.0

0.5

1.0


N-N

H4,

mg.

L-1Thao R.

0

2

4


N-N

O3,

mg.

L-1 in 2050in 2003

0.0

0.4

0.8

1.2

1.6


Ptot

, mg.

L-1

0.0

0.1

0.2

0.3


P-PO

4, m

g.L-1

Figure 7.5: Simulation results obtained at Yen Bai station in the Thao River and at Hanoi

station for the ‘2050’ and the ‘reference 2003’ scenarios.

182


Table 7.4: Calculated nutrient fluxes (N, P and Si) at the stations Yen Bai (Thao river), Vu Quang (Lo River), Hoa Binh (Da River) and Hanoi (Hong River) in the year 2003 and in the

‘2050’ scenario with the same hydrological conditions

Sub-basin Total N Total P Dissolved Si

103 tons N.y-1 103 tons P.y -1 103 tons Si.y-1

in 2003 in 2050 in 2003 in 2050 in 2003 in 2050

Da 61 70 9.3 9 274 279

Lo 55 104 24 18 128 128

Thao 15 26 7.9 9.5 57 57

Hong Hanoi 146 234 48 49 475 479

Table 7.5: Molar nutrient ratio of fluxes delivered at the outlet of the sub-basins and the main branch of the Red River as calculated by the model for the reference year 2003 and in the

‘2050’ scenario with the same hydrological conditions

N/P Si/N in 2003 in 2050 in 2003 in 2050 Thao 1.9 2.8 3.7 2.2 Lo 2.3 5.8 2.3 1.2 Da 6.6 7.8 4.5 4.0 Hong Hanoi 3.0 4.8 3.3 2.1

The Seneque/Riverstrahler model implemented on the Red River basin enables the diagnostic

of the N:P:Si nutrient balance, which is the key for the control of freshwater and coastal

marine eutrophication problems. The implementation of such a water quality model at the

regional scale can offer an excellent framework for initiating and developing the dialogue,

both between scientists of different disciplines but also between scientists, decision-makers

and the public. It will also enable to very clearly point out the gaps subsisting in our

understanding of the system, thus indicating the need for further research.

Because of the role of water in the history of human development, the scale of the catchment

of large rivers often represents a pertinent scale regarding major environmental land planning

issues. On the other hand, downscaling should be possible with water quality models, in order

to use them operationally for local management of water quality. A coupling of this model to

other models of similar conceptual approach at more local scale, as the one describing the

functioning of polluted urban rivers in the delta (Trinh Anh Duc, 2003), could be currently

used as a support for Water Policy in the future .

183

Exploring future trends of nutrient transfers 7.5. References

FAO database, 2005. FAO Statistical Databases, available at http://faostat.fao.org/faostat.

Hoang Xuyen, 2000. La Transition démographique. In Population et Développement au

Vietnam. Patrick Gubry, ed. Karthala/Ceped Paris. p 61-82.

MOSTE, 1998 – 2003. Vietnamese general statistics officer, Annual rapport of Ministry of Science, Technology and Environment of Vietnam, general statistics editor, Hanoi.

Nguyen, Huu Khai and Nguyen, Van Tuan, 2001. Geography and Hydrology in Vietnam. Vietnam National University publisher, Hanoi, Vietnam.194pp.

Nguyen, Viet Pho, Vu, Van Tuan and Tran, Thanh Xuan, 2003. Water resources in Vietnam. Vietnamese Institute of Meteo-hydrologie. Agicultural Editor (in Vietnamese).

Pham, Quang Son, 1998. Fundamental characteristics of the Red River bed evolution. Proceedings of International Conference on Economic development and environmental protection of the Yuan-Red River watershed, Hanoi 4-5 March.





184

General Conclusions and Perspectives


The Red River system (a tropical river with its watershed of 156 448 km², mainly laid in

Vietnam and China) has been considerably influenced by human activities. While previous

studies gave mainly emphasis on the delta area of the Red River, addressing issues like flood

disasters, irrigation systems, etc., or on the upstream sectors, addressing the questions of

forest management, erosion, reservoir construction, etc., our work, for the first time, addresses

other important aspects at the scale of the whole basin.

1. The hydrological regime, characterized by irregular flow including floods in July and

August, has been studied in the period from 1997 to 2004. The Hydrostrahler model, based on

a simple description of the rain-discharge relationship, has been utilized for modelling the

discharge at the outlet of the three sub-basins Da, Lo, Thao of the Red River. The results

show that discharge simulation with a Nash criterion around 0.7 can be obtained using only

the limited number of meteorological stations available in the basin. Closely related with

hydrological regime, the suspended load carried by the Red River has also been investigated.

The results reveal that the Hoa Binh and the Thac Ba reservoirs have a major influence on

suspended solid concentrations in the Red River system. The suspended load decreased from

100-170 106 t.yr-1 to 40 106 t.yr-1 since the impoundment of the two reservoirs. With the

planned construction of two additional reservoirs (Son La and Dai Thi) in 2010, the

suspended load of the Red River in future will be further reduced by 20%.

2. A two-year field survey, at monthly frequency, has been organised in order to collect

data on water quality at the outlet of the major tributaries and in the main branch of the Red

River, as well as in some polluted rivers of the Hanoi region. This allowed defining the

general level of nutrient (N, P, Si) concentrations in the drainage network of this large river

system. It also allowed demonstrating the low level of algal growth in the major rivers as

well as in the large reservoirs. To our knowledge, water quality data did not exist up to now in

the upstream basin of the Red River, and could be used as reference in the future.

3. The degree of human-induced alteration of the nitrogen and phosphorus cycles at the

scale of a sub-tropical watershed was investigated by budgeting N, P, within the 4 main sub-

basins (Da, Lo, Thao and Delta) of the Red River system, differing in population density (by a

185


factor of 10), land use and agricultural practices. In terms of agricultural production, on the

one hand and consumption of food and feed on the other, the upstream sub-basins are

autotrophic systems (they produced more than they consumed), while the delta is, at the

opposite, a heterotrophic one. Great losses of nitrogen are attributable to denitrification in rice

paddy fields, those of phosphorus being caused by erosion. In stream elimination of nitrogen

and retention of phosphorus are the highest in the Da and Lo sub-basins which have large

reservoirs in their downstream course. The total specific delivery estimated at the outlet of

the whole Red River System is 855 kgN.km-².yr-1 and 325 kgP.km-².yr-1. Nitrogen rather than

phosphorus seems to be the potential limiting factor of algal growth in the plume of the Red

River, and further in the South Chinese Sea (Tonkin Bay).

4. For assessing the link between human activity in the watershed and water quality of

the river system, the Seneque/Riverstrahler model was successfully applied for the first time

to a tropical river system. A GIS data base has been assembled at the scale of the whole basin,

with layers documenting geomorphology, lithology, meteorology, land-use and agriculture,

population, industrial wastewater release, etc. for estimating the role of natural and

anthropogenic factors in the watershed on the water quality and biogeochemical functioning

of the whole river system. The results could be validated, by a rather good agreement

between the modeling results and the observed data of water quality, at the outlet of the three

sub-basins and in the main branch of the Red River during the years 2003 and 2004. Beside

the modeling tool, the GIS data base has allowed to check the coherence of existing data and

the synthesize them. This data base can be permanently enriched with new data, allowing a

better forcing of the model and/or a stronger validation.

5. The model has allowed to explore a variety of scenarios describing potential changes

in the watershed (climate, hydrology, land use and agricultural practices, population increase,

wastewater treatment policy), in terms of river water quality and overall export of nutrients.

As an example, with the hypothesis of a climate change in the 2080s (increase of 10% of

rainfall data and 3°C increase of temperature), the model predicts an increase of about 20% of

the suspended matter loading of the Red River (from 40 to 48 106 t.yr-1) with respect to the

conditions of the period 1997-2004. Some other changes in the Red River basin in 2050 such

as increasing population, planning of new large reservoirs and changes in land use in the next

50 years were tested to obtain the future nutrient levels of the Red River. These results are

186


expected to serve as a guide for planning environmental decisions at both regional and local

scales.

Beside all interesting results mentioned above, this work has also its restrictions. Regarding

to the results obtained, it is possible to note that the water quality of the Red River system is

not seriously polluted, especially in the upstream of the Red river basin. The most polluted

rivers are the ones mainly located in the Delta area, especially in the Hanoi city area where

rivers are considered as waste water collectors. This work has not focused on modeling these

small rivers, although we analyzed their quality level, but was complementary to a similar

modelling approach on one of the urban river (The Nhue River). Therefore, several

perspectives are fully open for the nearby future:

1. We can remind here, that this work was undertaken in the framework of the ESPOIR

project aiming at identifying the water quality controls and at developing new processes for

water treatment. As already mentioned in the Introduction, although this programme focused

on the study on water pollution and water treatment of urban rivers surrounding Hanoi, i.e. the

Nhue-Tolich river system located in the Red River delta, a special interest was given to the

upstream drainage network of the Red River, the Nhue river being one of diverted branched

of the Red River, upstream Hanoï. The Nhue receives directly the Tolich River draining

Hanoi (about 3.5 million inhabitants), being therefore seriously polluted by the domestic and

industrial wastewater from Hanoi city and also by agricultural activities (irrigation in rice

field and vegetation culture) and aquaculture (fish and crustaceans production). As the Nhue

River is supplied by the major branch of the Red River through the Lien Mac dam,

immediately upstream of Hanoi city, it could now become possible to explore how the

management of the discharge of water from the Red River to the Nhue River could be used to

improve the water quality of the Nhue River. A dialog between the Red River model

developed in this work, with the Nhue/Tolich Rivers model developed by Duc’s thesis (2003,

see reference in chapter 7) could be useful in this context.

2. In the framework of this thesis, the complex drainage network characterizing the Red

River delta area, with a lot of distributaries and irrigation channels, connected with the Thai

Binh River system, has not been examined in details. In the next step, we wish to be able to

adapt the model in terms of hydrology and water quality to be able to pursue the modelling

work down to the coastal zone. An application has just been submitted for further

187


cooperation between France and Vietnam to investigate a major branch of the delta, also

largely polluted, the Day River.

3. The Seneque/Riverstrahler model, which has now shown its capability to represent

water quality of a tropical river system, although further refinements are always necessary,

should well be utilized for addressing some other important environmental issues in other

region of Vietnam. The management of water quality in the lagoon of the Huong (Parfum)

River in Hue city is a possible example.

188


Table of contents

Introduction 1 CHAPTER 1: Site description and major issues 91.1 Geographical presentation of the Red River basin 91.2 Geomorphology 111.3 Climate and hydrological regime 131.4 Hydrology 161.4.1 Hydrology in Vietnam 16

1.4.1.1Surface water 161.4.1.2 Ground water 17

1.4.2 Hydrology of the Red River 181.4.1.1 Drainage density 181.4.1.2 Water flow 191.4.1.3 Reservoirs 19

1.5 Social-economical context in the Red River basin and impacts 201.5.1 General social-economical context 20

1.5.1.1 Change in land covers 20 1.5.1.2 Increase of fertilizers utilization 21 1.5.1.3 Increase of population and of urbanisation process 22 1.5.1.4 Increase of industrial releases 23

1.5.2 Impacts on the water quality 24 1.5.2.1 Decline of surface water quality 24 1.5.2.2 Increasing the water pollution in the delta and the coastal zone 24

1.6 References 25 CHAPTER 2: General approach and methodology 292.1 Modelling the quality of the Red River hydrographic network 30

2.1.1 What is a model? 302.1.2 Some model definitions in the context of modelling 312.1.3 The ecological functioning of hydrographic networks: RIVERSTRAHLER

model 33

2.1.3.1 General Principles 332.1.3.2 Hydrological model 352.1.3.3 Biogeochemical and ecological model: RIVE 382.1.3.4 Point sources and non-point sources 41

2.2 Experimental works 422.2.1 Sampling campaigns 42

2.2.1.1 Monthly sampling in the sub-basin and in the main branch of the Red 42

189


River 2.2.1.2 Sampling campaigns for non point source evaluation 442.2.1.3 Sampling campaigns for point source evaluation 44

2.2.2. In-situ measurements and sample analyses 462.2.2.1 Measurements of physical-chemical variables 462.2.2.2 Filtration and preservation of samples in the laboratory 472.2.2.3 Analyses of samples 47

2.3 Nutrient budgets 482.3.1 Nutrient cycling in the soils system 492.3.2 Nutrient cycling in the hydrosystem 502.4 References 50 CHAPTER 3: Hydrological regime and suspended load:

observation and modelling 57

3.1 Introduction 583.2 General characteristics of the Red River basin 593.2.1 Geomorphology 593.2.2 Meteorology 623.2.3 Population and land use 643.2.4 Dams and discharge regulation 653.3 Hydrological regime of the Red River and its affluents 663.3.1 Total and specific discharge of the different sub-basins 663.3.2 Modelling the rain-discharge relationship 683.4 Suspended matter loading of the Red River and its tributaries 743.4.1. Total and specific suspended load 743.4.2. Seasonal and long term variations of suspended load 763.4.3. Modelling the suspended load 783.4.4. Relationship between suspended solid and phosphorus transport 813.5 Future scenarios of suspended matter loading 823.5.1. Effects of planned dams 823.5.2. Effects of climate change 823.6 Conclusions 823.7 References 83 CHAPTER 4: Water quality of the Red River 894.1 Discharge variations 894.2 Physical-chemical variables 904.2.1 Temperature and conductivity 904.2.2 Suspended matter and dissolved oxygen 904.3 General pattern of nutrients 944.3.1 Inter-comparison of nutrient analyses by two laboratories 94

190


6.6 Land use and non-point sources of nutrients 1566.7 Wastewater point sources 1586.8 Validation 1606.9 References 169 CHAPTER 7: Exploring future trends of nutrient transfers 1737.1 Impacts of new dams constructed in the Red River basin 1737.2 Fast increasing population and impact on water quality 1767.3 Agricultural evolution and its impact on water quality 1807.4 Prospective simulation at the 50 years horizon 1817.5 References 184 General conclusions and perspectives 185 Contents 189Annex 193

192

Annex

Annex

Table A1: Water quality and discharge observations of the Thao River at Yen Bai station

Year Date NO3-N PO4-P DSi SS Discharge

mg.L-1 mg.L-1 mgSi.L-1 mg.L-1 m3.s-1

2003 20/01/03 0.27 0.000 5.44 109 3782003 16/02/03 0.79 0.004 5.47 34 3642003 15/03/03 0.26 0.003 4.81 55 2042003 15/04/03 0.04 0.005 4.88 39 1712003 15/05/03 0.66 0.004 6.00 257 1882003 15/06/03 0.94 0.003 4.61 1620 13802003 15/07/03 0.36 0.007 3.64 2177 8262003 15/08/03 0.63 0.009 5.25 1363 9862003 15/09/03 0.42 0.018 5.25 315 10902003 16/10/03 0.49 0.014 6.43 283 5862003 15/11/03 0.20 0.027 7.25 73 2842003 15/12/03 0.54 0.009 7.20 71 2912004 16/01/04 0.36 0.104 4.86 194 2642004 15/02/04 0.25 0.078 5.23 528 1762004 15/03//04 0.54 0.000 5.45 - 1482004 14/04/04 0.54 0.026 4.30 1640 2602004 10/05/04 0.42 0.055 4.40 2870 2762004 23/06/04 0.47 0.023 5.13 3990 11302004 20/07/04 0.60 0.011 5.72 6000 16202004 18/08/04 0.64 0.011 6.17 1650 19202004 18/09/04 0.64 0.013 6.17 10000 9992004 20/10/04 0.50 0.007 6.28 385 5022004 17/11/04 0.36 0.141 7.28 1110 3682004 05/12//04 0.64 0.137 6.03 756 275

DSi: dissolved silica concentration

SS: suspended solid concentration

193

Annex

Table A2: Water quality and discharge observations of the Da River at Hoa Binh station

Year Date NO3-N PO4-P DSi SS Discharge

mg.L-1 mg.L-1 mgSi.L-1 mg.L-1 m3.s-1

2003 20/01/03 0.18 0.001 2.69 2 8992003 16/02/03 0.16 0.004 2.51 1 9292003 15/03/03 0.25 0.001 3.51 1 7662003 15/04/03 0.18 0.004 3.67 1 10302003 15/05/03 0.02 0.002 4.64 3 13102003 15/06/03 0.01 0.009 5.02 38 16302003 15/07/03 0.01 0.012 2.64 75 26602003 15/08/03 0.01 0.008 4.30 56 26102003 15/09/03 0.01 0.011 5.56 91 21602003 16/10/03 0.05 0.015 5.53 9 12702003 15/11/03 0.04 0.018 5.47 4 6992003 15/12/03 0.43 0.001 5.68 3 7572004 16/01/04 0.19 0.003 5.28 18 7232004 15/02/04 0.06 0.016 4.16 8 6672004 15/03//04 0.11 0.017 3.89 68 6442004 14/04/04 0.13 0.003 4.28 46 8982004 10/05/04 0.19 0.020 3.70 75 13302004 23/06/04 0.46 0.001 3.84 280 24002004 20/07/04 0.27 0.003 3.56 450 41502004 18/08/04 0.28 0.003 3.80 605 20102004 18/09/04 0.32 0.005 3.46 463 21102004 20/10/04 0.20 0.004 3.75 60 15902004 17/11/04 0.33 0.073 3.42 43 9452004 05/12//04 0.44 0.073 2.90 210 824



194

Annex

Table A3: Water quality and discharge observation s of the Lo River at Vu Quang station

Year Date NO3-N PO4-P DSi SS Dischargemg.L-1 mg.L-1 mgSi.L-1 mg.L-1 m3.s-1

2003 20/01/03 0.44 0.000 3.30 15 3682003 16/02/03 0.14 0.007 4.25 6 3792003 15/03/03 0.10 0.002 4.30 7 4422003 15/04/03 0.09 0.005 4.69 5 4082003 15/05/03 0.09 0.002 3.64 242 11902003 15/06/03 1.20 0.008 4.20 107 10802003 15/07/03 1.52 0.004 4.20 72 14002003 15/08/03 0.66 0.001 3.52 487 20002003 15/09/03 0.64 0.008 4.03 76 11002003 16/10/03 0.67 0.006 3.65 152 8732003 15/11/03 0.19 0.016 4.76 16 2972003 15/12/03 0.11 0.002 4.88 6 2822004 16/01/04 0.20 0.006 3.91 66 2572004 15/02/04 0.61 0.026 3.71 42 1652004 15/03//04 0.48 0.078 4.31 40 2172004 14/04/04 1.00 0.016 3.75 80 2412004 10/05/04 0.66 0.055 4.27 4270 3912004 23/06/04 0.64 0.003 4.23 337 13902004 20/07/04 0.14 0.003 2.82 2215 13202004 18/08/04 0.29 0.010 3.28 2215 14202004 18/09/04 0.41 0.006 3.62 177 7542004 20/10/04 0.51 0.006 4.29 30 3562004 17/11/04 0.61 0.150 4.60 90 2922004 05/12//04 0.21 0.228 4.13 190 283



195

Annex

Table A4: Water quality and discharge observations of the Hong River at Son Tay station

Year Date NO3-N PO4-P DSi SS Dischargemg.L-1 mg.L-1 mgSi.L-1 mg.L-1 m3.s-1

2003 20/01/03 0.17 0.003 2.42 93 12202003 16/02/03 0.08 0.004 2.44 33 14002003 15/03/03 0.12 0.010 2.71 60 11402003 15/04/03 0.07 0.018 4.39 35 12702003 15/05/03 0.05 0.003 4.68 52 15202003 15/06/03 0.38 0.002 5.20 307 36002003 15/07/03 0.44 0.007 5.22 142 67502003 15/08/03 0.23 0.007 5.35 145 58002003 15/09/03 0.52 0.020 5.18 278 50002003 16/10/03 0.53 0.015 5.99 141 29202003 15/11/03 0.19 0.030 6.20 41 11202003 15/12/03 0.07 0.003 4.27 31 10802004 16/01/04 0.21 0.068 4.15 204 11202004 15/02/04 0.20 0.026 3.21 251 9942004 15/03//04 0.17 0.001 4.16 210 9872004 14/04/04 0.28 0.019 4.01 520 11102004 10/05/04 0.29 0.039 4.78 1705 20202004 23/06/04 0.58 0.002 4.81 1925 43702004 20/07/04 0.29 0.002 5.10 2660 66702004 18/08/04 0.41 0.001 4.22 1855 53302004 18/09/04 0.29 0.001 4.72 2340 43902004 20/10/04 0.68 0.021 6.03 900 25202004 17/11/04 0.47 0.175 2.60 690 17802004 05/12//04 0.47 0.156 1.96 675 1790



196

THESE DE DOCTORAT Université Paris VI- Pierre et Marie ...hydrologie.org/THE/LE.pdf · Université...

Documents

Transcript of THESE DE DOCTORAT Université Paris VI- Pierre et Marie ...hydrologie.org/THE/LE.pdf · Université...