Page 1

The Influence of Climate Change and Climatic Variability on the Hydrologie

Regime and Water Resources (Proceedings of the Vancouver Symposium,

August 1987). IAHSPubl. no. 168, 1987.

Sécheresse, désertification et ressources en

eau de surface — Application aux petits

bassins du Burkina Faso

Jean Albergel

Hydrologue UR B12

ORSTOM, Laboratoire d'Hydrologie

Miniparc Bât.2, Rue de la Croix Verte

34100 Montpellier, France

RESUME L'étude des conséquences de la sécheresse sur les

écoulements des petits bassins versants d'Afrique soudano-

sahélienne nécessite la constitution de chroniques de

ruissellement.

A cet effet, un modèle pluie-lame

ruisselée a été mis au point et appliqué sur deux bassins

de quelques dizaines de km . Ce modèle est basé sur la

cartographie des états de surface qui ont évolué sous

l'effet conjugué de la modification climatique et de

l'extension des surfaces cultivées. Les séries ainsi

obtenues pour la période humide 1950-1968 et la période

sèche 1969-1984 ont pu être comparées. Il en ressort que

les conditions de ruissellement favorisées par une

dégradation des bassins dans la seconde période compensent

globalement le déficit pluviométrique. Ponctuellement un

événement pluviométrique fort qui a la même probabilité

d'occurrence avant et après 1969 engendre une crue plus

forte dans la période actuelle.

Mots clefs:

Sahel, Bassin Versant Experimental, sécheresse, états de

surface, ruissellement.

ABSTRACT A chronological succession of runoff events was

built in order to study the effect of the drought on the

runoff of small watersheds in sahelian Africa. To fulfill

that aim, a specific rainfall-runoff model was derived.

This model is based on the mapping of superficial soil

features and on the related rainfall-runoff relationships

derived from rainfall simulation experiments. It was

applied to two small catchments (22 and 54 sq.km) in

Burkina Faso. Superficial soil features have evolved

along with natural climatic changes and also with changes

in land use. The runoff time series obtained for the

humid period 1950-1968 and for the dry period 1969-1984

were compared.

In the dry period, more favorable

conditions for runoff make up for the rainfall deficit, so

that globally the runoff distribution remains unchanged.

A strong rainfall which has the same probability to occur

in the two periods, produces a larger flood at the present

time.

355

356 J. Albergel

Key Words

Sahel, experimental catchment, drought, superficial soil

features, runoff.

Introduction

Les très forts déficits pluviométriques des années 1983 et 1984 sur

l'ensemble de la zone sud sahélienne de l'Afrique ont révélé qu'après

les sécheresses de 1972-1973 la région soudano-sahélienne n'a pas

retrouvé la pluviosité qu'elle connaissait dans les années anté-

rieures. De nombreuses études concordent pour dire que la période

1969-1984 se caractérise par un affaiblissement des totaux pluviomét-

riques jamais encore observés, tant par son intensité, sa persistence

et son extension géographique (Nicholson 1984, Albergel et al.1984,

Snidjers,1986). Les saisons des pluies 1985, 1986 "normales" pour

une majorité des stations ont été déficitaires dans de vastes regions

et ne permettent pas d'affirmer le retour à une période plus

clémente.

Les conséquences sur les grands systèmes hydrologiques ont été

impressionnantes et relatées tant par des scientifiques que par les

médias :

- Effondrement des débits des grands fleuves comme le Senegal

(Olivry,1983), le Niger (Billon,1985) ou l'ensemble des grands

cours d'eau tropicaux (Sircoulon,1986).

- Le bouleversement des systèmes lacustres; lac Tchad, delta

intérieur du Niger (Sircoulon,1985).

- La baisse généralisée de tous les aquifères importants (Leusink

et Tyano,1985).

L'analyse des régimes hydrologiques de cours d'eau moins impor-

tants (bassins versants de 1000 à 5000 km ) a démontré que la

sécheresse climatique a eu des répercussions bien moins importantes

sur le fonctionnement de ces systèmes: pas de modifications

significatives dans les chroniques de modules annuels ou de modules

journaliers maximum, (Pouyaud,1985).

Pour l'ensemble des petits bassins, taille inférieure à 1000 km ,

le manque de données sur les débits ne permet pas de faire le même

genre d'étude.

Rappelons cependant que des crues ponctuelles

causaient des dégâts dans une ville, comme Gorom-Gorom le 29 et 30

septembre 1984 ou emportaient des ouvrages sur des bassins versants

de surface inférieure à 100 km , barrage de Zamse au sud de

Ouagadougou à un moment où l'attention était plutôt polarisée sur des

problèmes de manque d'eau. Après avoir rappelé les principales

caractéristiques pluviométriques de la période 1969-1984, cet article

tente d'analyser les répercussions de cette sécheresse sur le fon-

ctionnement des bassins versants de petites dimensions: de quelques

km2 à 200 km2.

Principales caractéristiques des pluies durant la période 1969-1984

Snidjers (1986 op.cit) démontre la non stationarité de la série des

totaux pluviométriques sur l'ensemble des stations du Burkina-Faso.

Sécheresse, désertification et ressources en eau de surface 357

Carbonnel et Hubert (1985) confirment que la probabilité la plus

forte de "rupture" dans ces séries se situe en 1969 ou 1970. Les

moyennes interannuelles de la seconde période sont globalement

inférieures de 20% à celles de la période précédente (1920-1968).

L'analyse des hauteurs pluviométriques journalières (Albergel,

1986) a conduit aux conclusions suivantes:

- La somme des pluies journalières supérieures à 40 mm sur l'année

est significativement plus faible dans la période sèche. Les

distributions statistiques de ces pluies ont des coefficients

d'asymétrie et d'aplatissement plus forts sur les années 1969-1984

que sur les observations antérieures.

- Les pluies les plus fortes ont une égale probabilité

d'apparition dans les deux périodes; en particulier, le calcul de la

précipitation journalière de reccurrence décennale donne un résultat

équivalent sur les deux séries (résultat vérifié sur 25 stations

retenues pour la qualité et la longueur des observations).

Sur les petits bassins versants où l'essentiel de l'écoulement est

dû au ruissellement quasi immédiat des plus fortes pluies, nous

étudierons comment se sont traduites les modifications du régime des

pluies. Dans ce but, nous reconstituerons des chroniques de ruissel-

lement sur deux bassins versants expérimentaux sur lesquels nous

disposons de quelques mesures dans les deux périodes:

- Le bassin versant de Kazanga (54 km ) en zone soudanienne (11

40°N isohyète 900 mm) a été suivi en 1961-1963 et en 1983.

- Le bassin versant de Kognere dans la région de transition soudan-

sahel (12°22N isohyète 700 mm) a été suivi en 1960-1962 et 1984.

Constitution d'une chronique de ruissellement

sur un petit bassin versant

L'état actuel des connaissances sur l'influence des différentes

composantes de l'environnement sur le ruissellement met en évidence

qu'en zone soudano-sahélienne l'hydrodynamique superficielle est

contrôlée essentiellement par le couvert vegetal et les organisations

pédologiques de surfaces (Albergel _et_ al.,1985). Ce résultat a

permis la construction d'un modèle simple de constitution d'une

chronique de lames ruisselées à partir de la cartographie des états

de surfaces (Albergel et_ aL_. ,1985). La méthode cartographique

développée a cette occasion distingue deux niveaux d'organisation

(Valentin,1985):^

- La surface élémentaire, caractéristique d'un état de surface et

considérée comme homogène quant a son comportement hydrodynamique

sous pluie.

L'unité cartographique qui correspond soit a une seule surface

élémentaire soit a l'association de plusieurs (généralement interdé-

pendantes au sein de "systèmes de surface") et dont les limites

peuvent être tracées à partir des relevés de terrain et de photograp-

hies aériennes (figure 1). Sur chaque surface élémentaire, des

mesures sous pluies simulées, sur parcelles de 1 m2, permettent de

tester un comportement hydrodynamique et de déterminer une fonction

de production dépendante de la pluie et de l'état d'humectation du

sol. La fonction de production à l'échelle du bassin est obtenue par

la composition des différentes fonctions de production au prorata de

358 J. Albergel

BASSIN VERSANT DE KAZANGA

esquisse des organisations superficielles

à t/50 000e

d'apris

VALENTIN

o

fSÏEl

E2E3 -

EZ3 '

Figure 1 1. Surface éléments grossiers 2. Surface sans éléments

grossiers 3. Association surface hydromorphe-surface à

recouvrement sableux 4. Association surface vertique-

surface a recouvrement sableux 5. Surface hydromorphe

alluviale claire 6. Surface hydromorphes de bas-fond.

la surface qu'elles représentent et pondérée par un facteur de calage.

La comparaison de lames ruisselees calculées et observées pour des

mêmes événements pluvieux montre une dispersion plus importante pour

les valeurs observées (figure 2). Sur le bassin de Kazanga, pris en

exemple, le modèle expliquerait 60% de la variance de l'échantillon

"lames ruisselees fonction de la pluie journalière". Ce résultat

peut être considéré comme satisfaisant vue l'hypothèse de base

implicite: les pluies journalières ont une forme et une répartition

en intensité égales à celles du protocole de pluies simulées qui ne

sont que l'image de pluies statistiquement moyennes pour la région.

Le tableau n° 1 qui compare les paramètres statistiques de la série

de lames ruisselees calcules sur la période d'observation d'un poste

pluviometrique voisin, Manga (34 ans) et celles observées pendant la

période de fonctionnement du bassin (4 années) permet de valider ce

modèle.

Les observations de pluies journalières sont multipliées par un

coefficient d'abattement calculé par régression entre les pluies

moyennes du bassin et les pluies du poste pendant la période commune

de fonctionnement (c = 0.85).

500-

450-

q00-

R 350-

M

I

a 300-,

u

b

5 2b0-

f

E 200-

1 150-

H

M

100-

50-

0-

+ &t >

•*

#

i / 6

, ,

>>

*

1

•

*^ ,

1 0 0

2 0 0

3 0 0

iJOO

5 0 0

6 0 0

7 0 0

P L U I E JOURfJHL 1ERE - . IMH

800

900

1O00

Figure 2 Lames ruisselees observées + (4 ans) et reconstituées

* (37 ans) sur le bassin versant de Kazanga.

Comparaison des lames ruisselles reconstituées

pendant les périodes sèches et humides

Une cartographie des états de surfaces a été réalisée pour les deux

périodes à l'aide d'un relevé de terrain et des photos aériennes des

missions IGN (Institut Géographique National (Paris)) 1956 et IGHV

(Institut Géographique de Haute Volta (Ouagadougou)) 1980 sur les

bassins pris en exemple. Le tableau n°2 résume les différences

d'occupation des sols entre ces deux périodes.

Tableau 1 Comparaison des séries: lames ruisselées calculées et

observées sur le bassin de Kazanga

* Seules les lames ruisselées a 0,7 mm ont été prises en compte

Variable

Lames

ruisselées

reconstituées

sur 34 ans

Lames

ruisselées

observées sur

4 saisons

Nombre

d'observat.

453

54

Moyenne

x 0.1 mm

56,1

53,3

Ecart type

62,2

51,1

Valeur max

mm

50,1

29,7

CV

110,9

95,9

Tableau 2 Occupation des sols, comparaison des états en 1956 et 1980

Photos

KAZANGA 1956

KAZANGA 1980

KOGNERE 1956

KOGNERE1980

Champs %

16,2

36,2

16,1

37,4

Jachère %

51,3

33,8

10,7

5,6

Savane

Arborée %

32,5

30,0

73,2

57,0

360 J. Albergel

Si sur les deux bassins on assiste à un recul important des

jachères au profit des champs cultivés, sur le bassin de Kognere

l'extension des cultures s'est également faite au détriment de la

savane arborée. L'examen comparatif des photos aériennes permet

également de mettre en évidence, a Kognere, une modification

importante des unités cartographiques par une extension des zones

très érodées (figure 3) qui sont multipliées par 20 entre les deux

missions aériennes.

Figure 3 Evolution de la superficie des zones très érodées.

Sur le bassin de Kazanga les contours des unités cartographiques

restent sensiblement les mêmes, seule leurs sous divisions en zone

cultivée et en jachère est modifiée.

Dans un premier temps, pour appréhender l'effet respectif des

modifications des états de surfaces et du changement climatique nous

reconstituerons sur le bassin de Kognere une chronique de lames

ruisselees a partir des deux cartes et de l'ensemble de la série des

pluies journalières du poste voisin Boulsa.

Effets compares de la modification du régime des pluies

et des caractères physiographiques des bassins

Sur la figure n 4 sont reportées les lames ruisselees reconsti-

tuées sur l'ensemble des observations pluviometriques pour les deux

cartes d'états de surface de Kognere. On observe une homogénéité de

la relation pluie-lame ruisselee journalière sur chaque reconstitu-

tion et une difference notable entre les deux reconstitutions, la

carte de 1980 donnant des ruissellements plus forts. L'analyse des

ruissellements sur chaque reconstitution (Tableau 3) met en évidence

une diminution sensible du nombre de ruissellements moyens par an et

de la hauteur de la lame ruisselee moyenne annuelle pour la période

postérieure a 1969 dans les deux reconstitutions.

Afin de reconstituer une chronique complète des lames ruisselees

la plus probable, on se propose d'utiliser pour chacun des deux

bassins la carte des états de surface dérivant de la première mission

aérienne pour les observations pluviometriques précédent 1969 et

celle dérivant de la seconde pour la période postérieure.

Les hauteurs de lames ruisselees reconstituées ont été comparées à

Sécheresse, désertification et ressources en eau de surface 361

Eirœi'r'SffiTî^î'sîsiiâîTi»

Figure 4

Tableau 3 Lames ruisselles reconstitutées sur le bassin de Kognere

Photos

Reconstitution

d'après la carte

1956

Reconstitution

d'après la carte

1980

Date

Avant 1969

Après 1969

Avant 1969

Après 1969

Nbre moyen de

lames*

ruisseléesparan

21,3

15,1

23,4

17,2

Hauteur

ruissselée*

annuelle (mm)

174,7

116,2

241,3

166,7

Valeur max de

lalameruisselée

(mm)

64,0

37,2

84,1

51,9

"Seules les lames ruisselées ï 0,7 mm ont été prises en compte

celles observées pendant les deux périodes de fonctionnement des

bassins: les coefficients de correlation varient entre 0.75 et 0.90

Sur les figures 5 et 6 sont reportées les lames ruisselées

reconstituées pour les deux périodes; on remarque:

- pour le bassin versant de Kazanga les deux nuages de points se

confondent assez bien avec une légère tendance à un plus fort

ruissellement pour les événements de période sèche.

- pour le bassin versant de Kognere, l'ensemble des pluies

supérieures à 25 mm engendrent des lames ruisselées plus fortes en

1969-1983.

Sur les figures 7 et 8 les hauteurs de lames ruisselées rangées en

classe de 0.5 mm sont reportées en fonction des fréquences expérimen-

tales. On remarquera le peu de différence entre les distributions

statistiques des échantillons de période sèche et de période humide.

362 J. Albergel

B 350-

E* i

n 300-

s jso-

rv

MOO

500

500

700

PLUIE JOUPNnL ÎEP.F -. ]MM

Figure 5 Lames ruisselées reconstituées sur les bassins

versants de Kazanga + période antérieure a 1969

* période postérieure à 1969.

80C-J

i

1

600-

S 400^

200

300

100

500

600

700

PLUIE JOURNALIERE •.IMH

Figure 6 Lames ruisselées reconstituées sur les bassins

versant de Kognere + période antérieure à 1969

* période postérieure à 1969.

Sécheresse, désertification et ressources en eau de surface 363

Lames ruisselées reconstituées sut le bassin deKAZANG*

Seuil des

classes

mm

n

+. Period* 1949-1968

• Période 1966-1963

+

+

•S.

v*V

1

r~

1

1 ! 1 r~^—1 !

!

7

- — • ^ ^ -

10

20 30 40 50 60 70 80

90

Frequences expérimentales

Echelle gaussienne

Figure 7

Lames ruisselôes reconstitues sur le bassin de KOGNEAE

+ Période !949- 1968

• Period» 1969-1983

10

20 30 40 50 60 70 80

90

Frequences expérimentales

Echelle gaussienne

Figure 8

364 J. Albergel

Discussion

L'affaiblissement du régime pluviométrique pendant la période 1969-

1983 semble être largement compensé par la modification des états de

surfaces dans le fonctionnement des petits bassins versants. Ces

modifications qui ont amené surtout sur les bassins au nord de

l'isohyête 800 mm, des conditions de ruissellement plus favorables,

sont dues à l'action conjuguée de l'homme et des nouvelles conditions

climatiques La diminution du couvert herbacé et l'extension des zones

cultivées favorisent les tassements de la surface du sol et le

développement de pellicules imperméables ainsi que l'extension de

régions très érodées. Le bassin de Kognere présente actuellement les

caractères des paysages habituels de régions plus septentrionales,

vaste étendue de sol nus formant des glacis lisses ou caillouteux.

Les espèces de graminées pérennes disparaissent enfaveur des

annuelles, dans la végétation arborée on remarque que les rares

recrus

sont toutes des épineux qui prennent la place des

combrétacées.

La comparaison des distributions statistiques des lames ruisselées

sur les deux périodes, tout comme les crues exceptionnelles apparues

ici et là doivent mettre en garde contre une révision à la baisse des

normes de sécurité pour les ouvrages en raison de la période sèche

que nous vivons. L'apparente contradiction de fonctionnement entre

ces systèmes hydrologiques et les plus

grands rappelle

l'hétérogénéité spatiale des phénomènes hydrologiques. En effet, si

une forte pluie a une égale probabilité de se produire localement

dans la période actuelle, elle survient temporellement dans une

chronique moins pluvieuse et dans des conditions d'evaporation plus

fortes. L'alimentation des nappes alluvialcô ainsi que celle des

reserves de surface reste donc défavorisée mais ces phénomènes ne

sont pas repercutes dans l'écoulement des plus petits systèmes

hydrologiques.

Références

Albergel, J., (1986) Evolution de la pluviométrie en Afrique Soudano

-Sahelienne. Exemple du Burkina Faso, In: Colloque international

sur la revision des normes hydrologiques suite aux incidences de

la sécheresse, Cieh Ouagadougo, 17.p.

Albergel, J. (Valentin, C. (1986) "Sahelisation" d'un petit bassin

versant: Boulsa Kognere au centre nord du Burkina Faso; In:

Colloque Nordeste Sahel, Institut des Hautes etudes d'Amérique

Latine, Paris 1986.

Albergel, J. Ribstein, P., Valentin, C. (1985) Quels facteurs

explicatifs de l'infiltration? Analyse sur 48 parcelles au

Burkina Faso. Journées Hydrologiques de Montpellier, Coll. et sem.

Orstom, 26-48.

Albergel, J. Casenave, A., Valentin C. (1985). Modélisation du

ruissellement en zone soudano-sahelienne; simulation de pluies et

cartographies des états de surfaces. Journées Hydrologiques de

Montpellier, coll. et sem. Orstom, 75-84.

Albergel, J. Carbonnel, J.P., Grouzis, M. (1984). Pejoration

climatique au Burkina Faso. Incidences sur les ressources en eau

Sécheresse, désertification et ressources en eau de surface 365

et sur les ressources en eau et sur les productions végétales.

Cah. Orstom, ser. Hydrol. Vol XXIno. J_,3-19.

Billon, B. (1985) Le Niger a Niamey. Décrue et etiage 1985 Cah.

Orstom, ser. Hydrol. Vol. XXIno. 4_,3-22.

Carbonnel, J.P. Hubert, P; (1985) Sur la sécheresse au Sahel

d'Afrique de l'Ouest. Une rupture climatique dans les series

pluviometriques du Burkina Faso (ex Haute Volta); C.R. Acad. Se.

série Hydrologie Vol VII, tome 301 n 13, 941-944.

Leusink, A. Tyano, B. (1985) Observations du niveau de la nappe des

eaux souterraines et sa composition chimique et isotopique de

socle cristallin au Burkina Faso. Bull.de liaison du Cieh no. 62.

Nicholson, S.G. (1984) Rainfall fluctuations in Africa 1901 to 7973

in: Colloque OMM sur le Xeme anniversaire de l'expérience Etga

Dakar, décembre-Î984, lOlPlOlT

Olivry, J.C. (1983) Le point en 1982 sur la sécheresse en Senegambie

et aux iles du Cap Vert. Examen de quelques series de longue durée

(debits et precipitations) Cah. Orstom, ser. Hydrol. Vol XX, no.1,

47-69.

'

Sircoulon, J. (1986) Bilan hydropluviometrique de la sécheresse

1968-84 au Sahel et comparaison avec les sécheresses des années

1910 a 1916 et 1940 a 1949 in: Colloque Nordeste Sahel. Institut

des Hautes etudes d'Amérique Latine, Paris - 16 au 18 janvier

1986.

Sircoulon, J.

(1985) La sécheresse en Afrique de l'Ouest.

Comparaison des années 1982-1984 avec les années 1972-1973. Cah.

Orstom, ser. Hydrol. Vol no. 4_, 75-86.

Snidjers, T.A.B.

(1983) Interstation correlations and non

stationarity of Burkina Faso rainfall. Journal of climage and

applied meteorology, Vol.25 ,524-531.

Valentin, CT (1985) Différencier les milieux selon leur aptitude au

ruissellement: une cartographie adaptée aux besoins hydrologiques.

Journées hydrologiques de Montpellier, coll. et sem. Orstom, 50-

_ _

The Influence of Climate Change and Climatic Variability on the Hydrologie

Regime and Water Resources (Proceedings of the Vancouver Symposium,

August 1987). IAHSPubl. no. 168, 1987.

The glacial and hydrological regime under

climatic influence in the Urumqi River,

northwest China

Yao Tandong

Lanzhou Institute of Glaciology and

Geocryology, Academia Sinica,

Lanzhou, China

ABSTRACT

The glacial and hydrologie fluctuations and

their distinct features under climate influence in the

Urumqi River are analysed.

The glacial and hydrologie fluctuations in the Urumqi

River have the same pattern in which high discharge and

positive mass balance periods are mainly identical with

cold periods, low discharge and negative mass balance

periods identical with warm periods. A particular

temperature-wetness pattern, which could be explained by

prevailing general atmospheric circulation patterns is

responsible for the pattern. Distinct features between

glacier and discharge under climatic influence are

distinguished. Glacial terminal fluctuates as a slow

process with low frequency while discharge fluctuates as a

rather fast process with high frequency in the basin,

which results from the time lag difference between glacier

and discharge responding to climatic fluctuation.

Variations glaciaires et hydrologiques sous

l'influence climatique dans le bassin de la

rivière d'Urumqi, dans le Nord-ouest de la

Chine

RESUME Les variations glaciaires et hydrologiques ainsi

que leurs caractéristiques distinctes on été analysées

sous l'influence du climat dans le bassin de la rivière

d'Urûmqui.

Les variations glaciaires et hydrologiques dans ce

bassin ont la même tendance: les périodes de haut débit

et de bilan de masse positif sont principalement

identiques à la période froide, alors que les périodes de

bas débit et de bilan de masse négatif sont identiques à

la période chaude. Ceci qui dépend principalement du

modèle particulier température humidité, qui peut être

expliqué par la circulation atmosphérique régnant dans le

Nord. Sous l'influence du climat, les variations de la

glacière et du débit dans cette région ont des différences

évidentes: la glacière terminale fluctue avec un dévelop-

pement lent et une fréquence basse, tandis que le débit

fluctue avec un développement rapide et une fréquence

haute, ce qui resuite de la difference de retard de temps

367

368 Yao Tandong

entre la glacière et le débit à l'égard de la fluctuation

climatique.

Introduction

Fluctuations of glacial terminal and discharge are generally related

to climatic fluctuations. The mass budget process of glacier and

discharge is dependent on heat balance and precipitation features

which are related to climate. Some relationships between glacier,

discharge and climate are discussed in the present paper by taking

the Urumqi River as an example.

The Urumqi River originates from the eastern part of the Tianshan

Mountain in the southern part of Urumqi. The river basin has an area

of 924 km2 above Yingxiong Bridge Hydrologie Station which is the

main station measuring the outlet discharge, and an area of 46 km2 of

glaciers at the head area. The météorologie record started in 1940

at the Urumqi Station in the lower reaches of the basin, and in 1958

at the Da Xigou Station in the upper reaches. The hydrologie record

started in 1950 but it can be extended to 1940 by interpolation. The

observations and records of glaciers in the basin started in 1958.

The Urumqi River basin is a comprehensive one in earth science

studies in northwestern China, which is suitable for a comprehensive

study of the relationship between glacier, discharge and climate.

The fluctuations of glacier, discharge and their relationship to

climate and general atmospheric circulation in the Urumqi river

The study of fluctuations in glacier and discharge in the basin was

limited to the period since the Little Ice Age during which more data

are available. The Little Ice Age started in the late sixteenth

century and ended in the early twentieth century in the basin

according to Chen (personal communication). In Glacier No. 1, the

change in glacial area, glacial thickness, glacial length and volume,

and in glacial equilibrium line fluctuation were estimated based on

detailed glacial maps and radar sounding data (Table 1). Glacier No.

1 has decreased by 21% in length, 33% in area and 38% in volume from

the maximum of the Little Ice Age to present. A similar feature was

found in the fluctutations of all the glaciers in the basin. It has

been estimated that there was a decrease of 41% in glacial volume and

30% in glacial area since the Little Ice Age in the whole basin.

Table 1 Changes in Glacier No.l since the Little Ice Age

Length

Area

Thickness

Volume

Equilibrium line

(km)

(km")

(m)

(10*m')

(m)

Present

2.33

1.84

49

9000

4050

the Little Ice

2 9 8

2 ^3g

6Q

1 4 œ 0

38gQ

Age maximum

Glacial and hydrologie regime under climatic influence 369

There are three possibilities which could cause glacial advance in

the Little Ice Age in the basin: (a) drop in temperature; (b)

increase in precipitation; (c) combination of (a) and (b). There was

certainly a drop in temperature on the whole hemisphere in the Little

Ice Age, which must have influenced the basin. The temperature drop

in the Northern Hemisphere in the Little Ice Age was about 1.0°C in

annual average, 0.5°C in summer according to Flohn (1981), Lamb

(1977) and Schuurman (1981). It was estimated by Chen (personal

communication) that the temperature drop on Glacier No. 1 was about

0.6 C in the Little Ice Age. The annual temperature drop was no more

than 1°C on Glacier No.l if the smaller amplitude in climatic change

in the mountain area in the basin and the results in other regions

are considered.

According to the calculation of the relationship between

precipitation, temperature and equilibrium line in Glacier No. 1 in

the basin, a rise of 80 m in equilibrium line could be caused either

by an increase of 1°C in temperature or a decrease of 100 mm in

precipitation. The equilibrium line in the Little Ice Age (3890 m)

was about 160 m lower than that at present (4050 meter). It

estimated by using the above relationship that the precipitation

Glacier No. 1 during the maximum of the Little Ice Age is about

larger than that at present. Provided that the temperature

was

on

100

is

1.5WC lower during the maximum of the Little Ice Age than at present,

an increase of 50 mm in precipitation was estimated. Because the

estimate of temperature drop in the Little Ice Age was based on proxy

data from other areas, the above values are not the exact estimation

of the precipitation in the Little Ice Age. But it at least revealed

a trend: not only was there a temperature drop but also a

precipitation increase, characterized by cold and wet conditions,

during the Little Ice Age in the basin. To support the above

proposed relationship, three sets of tree ring data are analysed from

the basin and nearby regions (Table 2).

Table 2 The relationship between temperature and precipitation

revealed by the tree rings in the basin and nearby regions

Place

UrttiKri

River

Altai

Hami

Average

Total year

in a tree

(years)

384

158

252

265

Lew

temperature

with high

precipitation

(years)

152

32

58

81

High

temperature-

with low

precipitation

(years)

142

68

118

109

High

temperature

with high

precipitation

(years)

30

32

52

38

Low

temjjerature

with low

precipitation

(years)

60

26

24

37

Percentage of

high V' with

low P" and low

T with high P

(%)

77

63

70

70

Percentage of

high T with

high P and low

T with low P

(%)

23

37

30

30

370 Yao Tandong

Two points should be kept in mind when estimating the discharge in

the basin during the Little Ice Age. The first is that the

temperature then was lower than at present; the second point is that

it is wetter then than at present, based on the above analysis .

Because it is warmer and drier at present than in the Little Ice Age,

1976 (a year with the lowest temperature and highest precipitation

since meteorological data were recorded) was selected to estimate the

discharge in the Little Ice Age. In 1976, the temperature was 0.6°C

lower and the precipitation 46.2 mm higher than the secular average

at Da Xigou Station in the upper reaches of the basin; the tempera-

ture was 0.2°C lower and the precipitation 19.4 mm higher at Urumqi

Station in lower reaches of the basin. The discharge at Ying Xiong

Bridge Station that year was 280 670 000 m3 which was 40 000 000 m3

higher than the secular average. The discharge increased by 17% in

the Little Ice Age than at present (or decreased by 14% at present

than in the Little Ice Age) if taking the value of the discharge in

1976 as an estimation of the discharge in the Little Ice Age.

During the period with observations and records in meteorology,

glaciology and hydrology, the glacial mass balance, glacial equilib-

rium line and discharge in the basin basically kept the same trend

and experienced cycles. However, the glacial loss and discharge

decrease trend is obvious during the same period. A comparison of

two high discharge periods between the 1970's and the 1950's

indicates that the total discharge in the 1970's was 148 million m3

smaller than that in the 1950's, demonstrating a decreasing trend in

discharge. In Glacier No. 1, the average mass loss is 155 thousand m3

in water equivalence from 1959 to 1980 and the equilibrium line rise

10-12 m in the same period. The trends of the discharge and glacier

decrease are related to climatic change in the basin. The average

ten year temperature from the I960's to the 1970's has increased by

0.3°C in Urumqi, by 0.2°C in Xiao Quzi and by 0.1 C in Da Xigou. The

average ten year precipitation at Da Xigou Station which is close to

glacier No. 1 has decreased by 7.2 mm from the 1960's to the 1970's.

During the period in which recorded data are available, the relation-

ship between the fluctuations of glaciers and discharge and climatic

change is identical with that in the Little Ice Age. Although there

are different temperature-wetness patterns (warm-dry, warm-wet, cold-

dry, cold-wet), the dominant patterns were warm-dry patterns and

cold-wet patterns. The analyses of the moving curve and anomalies in

spring and summer temperature and spring and summer precipitation

indicate that the period of high temperature-low precipitation (warm-

dry pattern) and low temperature-high precipitation (cold-wet pattern)

is 76% of the whole period analysed in the upper reaches and 63% of

the whole period analysed in the lower reaches. The period of high

temperature-high precipitation (warm-wet pattern) and low temperature-

low precipitation (cold-dry pattern) is 24% and 37% respectively in

the upper reaches and in the lower reaches.

It seems that the fluctuation features of glacier and discharge

are dependent on the temperature-wetness feature in the basin.

Nevertheless, the temperature-wetness pattern in the basin could be

explained by changes of the general atmospheric circulation. The

general atmospheric circulation in the Eurasia continent was

classified into three types by Jiersi (1974). According to studies,

precipitation would decrease and temperature rise in most regions of

Glacial and hydrologie regime under climatic influence 371

middle latitude when W type prevails, there would be more opportuni-

ties for high temperature and low precipitation in most regions of

middle latitude when C type prevails, there would be low temperature

and high precipitation in most regions of the Eurasia continent when

E type prevails. In the Uriimqi River, the general trend of fluctua-

tions of climate and discharge is basically identical with the

general atmospheric circulation in the Eurasia continent (Table 3).

Table 3 Relationship between climate, discharge and general

atmospheric circulation during different periods in the

Uriimqi River

Temperature anomaly

Discharge fluctuation

n

. .

Circulation

Prec ipitat ion v.

,.

v.

., ,

Period

-,

Xiao ,, „ . Da

ïing

Northern

v. ..

type

anomaly

,, .

UrUmqi

v.

v.

?

,.v. ..

Xtniian

J

Quzi

ALgou Xiongqiao or Xinjiang

1941-1950

C

-7.3

1951-1960

E

+12.9

-0.2

-0.4

1961-1965

C

-4.5

+0.3 +0.7 +0.1

Low

discharge

High

High

High

discharge discharge

discharge

Low

Low

Low

discharge

discharge

discharge

1966-1972

E

+6.0

-0.4 -0.1 -0.2

,. L ° W

" ^

" ^

discharge discharge

discharge

In the basin, E type generally corresponds to positive precipita-

tion anomaly, negative temperature anomaly and a high discharge

period; W and C type generally correspond to negative precipitation

anomally, positive temperature anomaly and a low discharge period.

Distinct response between glacier and discharge

under climatic influence in the Uriimqi River

According to what is discussed above, the glacial volume has

decreased by 40% and the discharge has decreased by 14% from the

Little Ice Age to present in the Uriimqi River basin. It means that

the decrease rate of glacier is larger than that of discharge under

the same background of climatic warming, which results from the

distinctions in energy balance between glacier and discharge.

The input and output process in glacial system is expressed by the

mass balance equation

B = P g -M-Eg+LH : D + ; A

(1)

(where Pg

stands for annual precipitation on glacial the surface, M

for glacial surface melting, Eg for glacial surface evaporation, L

for glacial condensation, D for snow drifting and A for avalanche).

372 Yao Tandong

The input and output process in the discharge system is expressed

by the water balance equation

P, = R + E, + AW + F

d

d —

R

E, + AW

d -

F

(2)

(where R stands for discharge, P-, for annual precipitation, EJ for

evaporation, AW for underground water storage and F for seepage;. To

reveal the essential features of the input and output process of

mass, a simplification to equation (1) and (2) is necessary.

According to heat balance curves by Budyko (1974), the evapora-

tion process in the basin should belong to the continental climate

type in middle latitudes. The feature of the process at this latitude

is that evaporation is maximum in summer, dramatically changeable in

spring and autumn, and negative in winter. Summer evaporation could

therefore approximately substitute for annual evaporation. The rain

season is during the summer in the Urumqi River basin, the intense

evaporation season is also an intense condensation season in the

basin. From observations on some glaciers in Central Asia (Lvovich,

1975), evaporation and condensation are almost equal in summer.

Therefore, E and L in equation (1) could be omitted. Avalanche

influence in the studied area could also be omitted. As a secular

process, the amplitude of annual change of snow drifting is not so

great and could be taken as a constant

Equation (1) then could be simplified as

(D

W_ for example).

B •g M + W„

(3)

There are only two variables P~ and M in the equation.

As a secular process, F and in equation (2) are also relatively

stable and could approximately be taken as a constant (F + W = Wc,

for example). Equation (2) could then be simplified as:

R pd " Ed + Wc

(4)

There are also two variables, P-, and E, . in equation (4).

Equations (3) and (4) demonstrate that the input of glacier and

discharge are of the same form, but the output of these two systems

are essentially distinct: by melting in the glacial system and by

evaporation in the discharge system.

The mass output of both glacier and discharge is, in physical

essence, the result of the input of heat. The main components of heat

input are radiation balance (R), latent (L) and sensible heat (S).

The heat balance equation in the glacial system could be expressed as:

%m \ + h +

(5)

The heat balance equation in the discharge system could be expressed

as :

% = Rd + Ld + Sd

(6)

From equation (i) and (4) the heat (Q) absorbed in the glacial

system is mainly consumed in glacial surface melting, therefore

Glacial and hydrologie regime under climatic influence 373

Qg * M(Q)

(7)

The heat (Q) absorbed in the discharge system is mainly consumed in

evaporation, therefore

Qd

- E(Q)

(8)

Using latent heat of melting (80 cal.g ) and latent heat of

evaporation (597 cal.g ), equations (7) and (8) could be expressed

as :

M = Qg/80

(9)

E = Qd/597

(10)

Substituting equations (9) and (10) for the second item on the right

side of equations (3) and (4) respectively, then

B = P - Qg/80 + Wc

(11)

R = P - Qd/597 + Wc

(12)

Equations (11) and (12) indicate that the response of the input

and output process of mass to climate (actually to heat balance) is

much more sensitive in glaciers than in discharge.

Besides,

evaporation in the discharge system only consumes part of the heat

absorbed, the denominator in equation (12) is relatively enlarged.

So, the response of water balance to climate is more sluggish than

that expressed in equation (12).

The heat balance in the glacial system is, however, smaller than

that in the discharge system, and the mentioned distinction has been

weakened in some degree. But it is still evident according to the

results from Glacier No. 1 and nearby area. Taking the average

secular melting value (214 cm) at 3870 m a.s.l. on Glacier No. 1 and

evaporation value (53 cm) at Da Xigou Station near Glacier No. 1,

corresponding melting latent heat of 17 000 000 cal.cm- a-

and

evaporation latent heat of 32 000 000 cal.cm a

were obtained.

This is to say that although the heat consumed in discharge

evaporation is two times that in glacial melting, mass loss in

discharge is only 1/4 of that from the glacier. It can be deduced

from the above discussion that the fluctuation amplitude of the

glacier is larger than that of discharge in the past under the

influence of climatic change in the Uriimqi River basin.

There are obvious distinctions between the glacier and discharge

in the basin under modern climatic influence.

One of the

distinctions is that a glacial terminal responds to climatic change

as a slow process with low frequency while discharge responds to

climatic change as a fast process with high frequency. Taking

Glacier No. 1 as an example, the glacial terminal maintains its

retreating trend since observations in the late 1950's. But the

discharge in the basin experienced high and low discharge cycles

during the same period, which is basically identical with the

climatic fluctuation in short period and therefore rather sensitive

to climatic change. Because temperature in the basin shows a warming

374 Yao Tandong

trend in this century, it seems that the retreating trend of the

glacial terminal is identical with the warming trend in climate and

the glacial terminal is a rather stable indicator for the secular

climatic trend. The reason for the distinctions could be explained by

time lag differences between glacier and discharge in their response

to climate.

The time lag of the glacial terminal ranges from several years

(small mountain glaciers) to decades (larger mountain glaciers) (Nye,

1958, 1965) or even thousands of years (Budd and Smith, 1979). The

time lag in the discharge process is much shorter, ranging from

several hours to several days or 10's of days. This demonstrates

that the discharge system could reflect climatic influence

immediately in the lower reaches of a river, but the glacial terminal

could require a rather longer period to reflect it. Many factors

affect glacial time lag. The main factors are mass balance, which

mainly depends on precipitation and temperature, glacial size and

slope, glacial velocity and glacial temperature, if surging glaciers

are neglected.

There is no pattern to follow governing the

importance of these factors. It is, therefore, difficult to find a

model including all these factors and suitable for all types of

glaciers. Glacial length is, however, one of the most important

factors responsible for glacial time lag and is easy to obtain. A

model could be established by using glacial length to estimate

approximately glacial time lag in different sizes. Based on the

statistical analysis of 35 "normal" (ie. non-surging) glaciers in the

Northern Hemisphere, a simple model was established as

T - 13L°-375

(where T stands for the glacial time lag, and L for glacial length).

The relationship factor for the model has a significance of 95%.

Because only glacial length L was introduced in the model, the time

lag estimations are approximate values. It was estimated from the

model that the time lag for Glacier No. 1 may be 18+5 years.

The time lag of the discharge in the basin is different in

different seasons, but depends mainly on the character of high water

season of a year. The high water season in the Urùmqi River could be

classified as spring and summer. Summer high water season is

characterized by precipitation discharge with a short time lag. It

was calculated from the recorded data at an experimental discharge

station in the Urumqi River, that the peak flood is 7-8 hours after

peak precipitation from Hou Xia to Ying Xiongqiao. According to

this, the time lag from the head to the lower reaches of the river is

about one day. Spring high water season is characterized by snow

melting discharge and is rather complicated. Generally, the amount

of spring high water reflects the amount of precipitation during the

last winter or even previous autumn plus the heat absorbed in the

mountain area in spring. The time lag could therefore be several

months. Although this is the result of interruption by other factors,

it is still a feature in the discharge forming process. In addition,

time lag of discharge also changes with climate at different

altitudes. Usually, summer precipitation in the middle and lower

parts of the mountain is in the form of rainfall and could be

reflected immediately in the lower reaches, while the summer

Glacial and hydrologie regime under climatic influence 375

precipitation in the high mountain area is mainly snow and needs a

longer time to be reflected in the lower reaches. In conclusion, the

time lag of discharge in the basin ranges from several hours to

several months from the lower reaches in summer to the upper reaches

in spring.

If 10 years represents the order of magnitude of the time lag of

the glacial terminals of most glaciers in the basin and 1-2 days

represents the time lag of discharge from the head to the lower

reaches of the basin, the time lag of the former is several orders of

magnitude longer than that of the latter. Even taking several months

as the time lag of the discharge in the basin, the time lag of the

glacial terminal is still one order of magnitude longer than that of

discharge. This partly explains why the glacial terminal responds to

climate in low frequency while discharge responds to climate in high

frequency.

The future trend of the glacier and discharge in the Uriimqi River

Based on the discussion above, the glacial and discharge fluctuations

in the basin are controlled by climatic change. The forecasting of

the glacial and discharge fluctuation is only possible if the trend

of the climatic change is forecast. But climatic change is still a

subject which is beyond the forecasting ability of man. Most climatic

forecasts only propose possibilities. This is also the case in the

Uriimqi River. Based on tree ring analysis and forecasting of the

general atmospheric circulation by Jiersi (1974) and the relationship

between the climatic change in the basin and general atmospheric

circulation, a possibility of the future climatic trend in the basin

was estimated. It was forecast by Jiersi (1974) that W+E circulation

would prevail from 1979 to 1986, and W+C prevail from 1986 to 1996.

A possible climatic trend accompanied with the circulation is that

the precipitation from 1979 to 1986 would fluctuate near its secular

average and that the precipitation from 1986 or so would decrease and

last to the late 1990's. The decrease in precipitation may have

started in 1985. Two conclusions could be made based on the tree

ring data: (a) the present is a dryer period than the Little Ice

Age; (b) this trend would continue until the 2020's-2030's. A wet

period of several years may appear between the late 1990's and the

early 2000's. The temperature rise in the next 10-15 years is

probably 0.1-0.6°C if the temperature in the basin keeps the rising

trend at present.

In recent years, the C02 greenhouse effect has been much discussed

by Manabe & Stouffer (1980) and Manabe and Wetherald (1981). It was

estimated by them that the average global temperature would increase

by 2-3°C because of the increase of C02 and other gases in the

atmosphere in the next 50 years. A temperature increase of 1.5-4.5 C

in the next 50 years was announced at the Villach Conference in 1985.

There was not much discussion about the potential temperature

increase in the next 10-15 years. It may be still small before the

twenty-first century and an estimation of 0.5°C might be suitable

because there is a time lag from C0„ content increase in the

atmosphere to actual temperature rise. The time lag would be even

longer if the influence of oceans is considered.

376 Yao Tandong

A rise of 0.5-1.0°C in temperature would be possible in the next

15 years if taking 0.5°C as the temperature increase caused by CO2

and provided that the natural climate keeps its present warm trend.

Assuming climatic warming in the next 15 years occurs with a range

from 0.5 to 1.0°C, the forecast for glaciers and discharge around

the year 2000 could be made (Table 4).

Table 4 Estimation of changes in glacier and discharge under the

condition of temperature rise of 0.5-1.0°C in the Urumqi

River around 2000

Temperature

Precipitation Altitude of

Glacial

Discharge Discharge State in Glacial

Period

increase

decrease

glacial equi ii- mass l)alaiice of the

Glacier

glacial

tJûjinirig

(a C)

(mm)

briun 1 inu

(mm)

River

No. 1

terminal

(m)

(m)

(106 m3 ) (106 ni3 )

1985-2000

0.5-1.0

10-30

4070-4100

-100

300 22.1-23.7 1.5-2.1 retreating

2 - 8

Conclusions

The following conclusions were made from the above discussion:

(a) The glacial and discharge fluctuations in the Urumqi River

basin are related to climatic change from the Little Ice Age to

present, which were demonstrated to be glacial mass balance and

discharge decrease when the climate becomes warmer; glacial mass

balance and discharge increase when the climate becomes colder.

(b) The glacier and discharge fluctuations in the basin are

related to general atmospheric circulation through climate. It is

advantageous to glaciers and discharge in the basin when an E type of

circulation prevails, but it is not advantageous to them when W+C

types of circulation prevail.

(c) Under the influence of a warming climatic trend since the

Little Ice Age, glacier and discharge in the basin have both

decreased, 41% in glacial volume and 14% in annual discharge. The

rate of glacial decrease is larger than that of discharge decrease,

which is the result of the distinction in energy balances between

glacier and discharge.

(d) Glacial and discharge fluctuations possess different climatic

significance because of their distinctions in time lag and other

processes. Glacial terminal fluctuation (a slow process with low

frequency) could be taken as a relatively stable indicator for

secular climatic change, while discharge fluctuation (a rather fast

process with high frequency) could be taken as a relatively sensitive

indicator for climatic fluctuation in a short period.

References

Budd, W.F. & I.N. (1979) The growth and retreat of ice sheets in

response to robital radiation changes. Sea Level, Ice and Climatic

Glacial and hydrologie regime under climatic influence 377

Change (Proc. Chanberra Symposium, December 1979) 369-409.

Budyko, M.I. (1974) Climate and Life, Academia Press, New York and

London. QC 801 155 #18 C.l.

Flohn, H. (1981) Scenaries of cold and warm period of the past.

Climatic Variations and Variability: Facts and Theories. D. Reidel

Publishing Company.

Jiersi, A.A. (1972) The Secular Oscilation in General Atmospheric

Circulation and the Secular Forecasting in Hydrology and Meteoro-

logy (translated into Chinese from Russian), Academia Press.

Lamb, H.H. (1977) Climate: Present, Past and Future, London.

Lvovich, M.I. (1975) A method of studying the water balance and

estimating the water resources of glacial mountain area. Snow and

Ice Symposium, IAHS-AISH Publ. No. 104.

Manabe, S. & R. S. Stouffer (1980) Sensitivity of a global climate

model to an increase of C02 concentration in the atmosphere. J.

Geophy. R 85 (5529) C 10.

Manabe, S. & Wetherald, (1981) On the distribution of climate change

resulting from an increase in C02 content of the atmosphere. J.

Atmos. Sci. 37(3) pp. 99-118.

Manabe, S. & Wetherald, R.T. & Stouffer (1981) Summer dryness due to

an increase of atmospheric C02 concentration. Climatic Change 3

347-386.

Nye, J.F. (1958) A theory of wave formation on glaciers, IASA 47.

Nye, J.F. (1965) The frequency response of glacier, J. Glaciol. 5(4)

567.

Schuurman, C.J. (1981) Climate of the last 1000 years. Climate

Variations and Variability:

Facts and theories, D. Reidel

Publishing Company.

The Influence of Climate Change and Climatic Variability on the Hydrologie

Regime and Water Resources (Proceedings of the Vancouver Symposium,

August 1987). IAHSPubl. no. 168, 1987.

A primary study of the relationship between

glacial mass balance and climate in the Qilian

Mountain taking "July First" Glacier as an

example

Xie Zichu, Liu Chaohai

Lanzhou Institute of Glaciology and Geocryology

Academia Sinica, Lanzhou, China

ABSTRACT

The average summer (from June to August) air

temperature is 1° C lower and the equilibrium line 88m

lower in the 18 years after 1968 than the 11 years before

1968 in the Qilian Mountain. It was indicated from

calculation that an increase (decrease) of 1 C in air

temperature would result in a rise (or drop) of 80 m of

the equilibrium line. If taking the^base station (3700

m.a.s.l.) temperatures above zero C and below zero

C

at the glacial terminus as a temperature index for the

beginning and the end of the glacial melting period

respectively, the melting period is 14 days shorter in

average in the 18 years after 1968 than the 11 years

before 1968 and the maximum ablation altitude has dropped

by 130m in average on "July First" Glacier in the middle

of the Qilian Mountain. It was found from the mass balance

reconstructed according to air temperature and precipita-

tion from 1957 to 1985 on "July First" Glacier (with an

area of 3.0 km2) that it was characterised by a negative

mass balance state in the 11 years before 1968 with a

cumulative negative mass balance value of 232.9xl0tfm3

and

characterized by a positive mass balance state in the 18

years after 1968 with a cumulative positive mass balance

value of 773.4x10'+ m3. In the recent 30 years, net mass

balance has increased by 540.5xl0Lfm3 and average glacial

thickness increased by 1.8m. Relatively larger positive

mass balance appeared in 1967/68, 1975/76, 1982/83, which

showed periodical cycles of 7-8 years. Fluctuation of 1 m

of equilibrium line corresponds to a mass balance change

of O.ôSxlO^in3. Glaciers in the Qilian Mountain are typical

glaciers accumulating in summer and which possess the

following features: cs>cw, as>aw; bs >0, bw >0 (accumula-

tion area); b <0, b^ <0 (ablation area). On such kind of

glaciers, which are typical continental glaciers, mass

balance processes are mild and level of mass balances are

low. The shortening of the ablation period and drop in

summer temperature since the middle I960's and the

increase in precipitation since the 1970's have resulted

in positive mass balance and a decrease of glacial retreat.

379

380 xie

Zichu & Liu Chaohai

Etude sur le rapport entre le bilan de masse glaciaire

et le climat dans la montagne de Qilan

RESUME Dans la montagne de Qilan, la température moyenne

en été (de juin à août) est de 1°C plus basse pendant les

18 années après 1968 et la ligne d équilibre est de 88 m

plus basse que les 11 années avant 1968. Le calcul montre

que la ligne d équilibre devrait monter ( ou baisser) de

80 m, lorsque la température moyenne en été monte (ou

baisse) de 1°C. Si P'on prend à la station de base les

température moyennes journalières au-dessus et au-dessous

du zéro °C au bout des glaciers pour indices respectifs de

température au début et à la fin de la période d ablation

des glaciers, la période d ablation est en moyenne de 14

jours plus courte pendant les 18 années après 1968 que les

11 années avant 1968, et l'altitude d ablation maximale

baisse en moyenne de 130 m dans la glacière "au 1er

juillet" au milieu de la montagne de Qilan»

Le bilan de masse établi d après la température et les

précipitations de 1957 à 1984 dans la glacière au 1er

juillet (avec une surface de 3,00 km2) montre qu on a

principalement dun état de bilan de masse négatif pendant

les 11 années avant 1968 avec une valeur cumulative du

bilan de masse négatif de 232,9xl0'*m^ et état de bilan de

masse positif pendant les 18 années après 1968 avec une

valeur cumulative du bilan de masse positif de 773,4

x 10 m . Pendant ces 30 dernières années le bilan de masse

net a augmenté de 540,5xl0^m^ et 1 épaisseur moyenne des

glacières a augmenté de 1.8 m. Le bilan de masse positif

relativement large est apparu en 1967/68, 1975/76,

1982/83, avec un cycle périodique des 7-8 années» La

fluctuation de 1 m de a ligne d équilibre correspond à un

changement du bilan de masse de 0,68x10 m . Les glacières

dans la montagne de Qilian sont des glacières typiques,

accumulées en été avec les caractéristiques suivantes:

cs>cw, as>aw; bs >0, b^ >0 (la surface d accumulation);

bg <0, b^ <0 (la surface d'ablation). Dans les glacières

de ce genre, qui sont des glacières continentales typiques,

le processus de la balance de masse est lent, et le niveau

de la balance de masse est bas.

Depuis le milieu des années 60, la période d ablation

est devenue plus courte et la température en été plus

basse. Pour ailleurs, les précipitations ont augmenté

depuis les années 70, ce qui aboutit à bilan de masse

positif et a la diminution de la retraite des glacières.

Introduction

A systematic study of glacial mass balance was carried out by Lanzhou

Institute of Glaciology and Geocryology of Academia Sinica from 1975

to 1979 on 4 glaciers in the Qilian Mountain (Wang Zhongxiang, 1985).

To monitor recent glacial fluctuation, a glacial mass balance study

was made on "July First" Glacier from 1984 to 1985. Based on these

Glacial mass balance and climate 381

data, the relationship between glacial mass balance and climate was

analysed.

Methods and results

Glacial melting was recorded using poles and snow stratigraphy was

observed using snow pits. Net balance (b) at every point is the

algebraic total of ice (b^), snow (bs) and superimposed ice (bsi):

b = b i + b s + bs i

(1)

Snow pits were used to observe glacial accumulation. Snow pits

were dug in late August or early September. When the dirty layer of

snow from the previous year was found, the depth of snow layer (m)

and its density (d) were measured and the annual accumulation was

calculated:

Zdm

(2)

Glacial mass balance was calculated using the contour method. Net

balance (B), net ablation and net accumulation were obtained by

multiplying the area (Sci, Sai) between two intervals with average

accumulative depth and melting depth (c- , a^):

B

= S ciSci + E ai + Sai

(3)

Based on the above observation and calculation, it was estimated

that the annual mass balance in "July First" Glacier was +67.3x10 m

in 1983/84 and -g.lxlOV3 in 1984/85. To compare, the mass balance

results from 1974/75 to 1976/77 are listed in Table 1.

Table 1 The results of glacial mass balance of "July First'

Glacier in the Qilian Mountain

Balance

1974/75

1975/76

1976/77

1983/84

1984/85

Altitude

of zero-

equilib-

rium line

4650

4550

4620

4600

4710

Net

Area

(km1)

2.14

2.63

2.48

2.41

1.78

accumu

Depth

(mm)

367

502

532

364

288

In t i on

Amount

(xl0"m')

78.6

132.0

132.0

87.7

51.3

Area

(km1)

0.90

0.41

0.56

0.57

1.20

Net ablation

Depth

(mm)

758

373

454

358

503

Amount

(xl0"m')

68.2

15.3

25.4

20.4

60.4

Net

Amount

(xlCrn')

+10.4

+116.7

+106.6

+67.3

-9.1

balance

Balance

(mm)

+ 35

+384

+350

+226

-31

382 Xie Zichu & Liu Chaohai

The relationship between annual variation

in glacial mass balance and climate

Glacial mass balance is the direct response of climatic fluctuation

and an important indication to link air temperature(ablation

indicator) and precipitation (accumulation indicator)» It therefore

can be reconstructed and forecast by temperature and precipitation.

The temperature on the glacial surface is an important parameter to

calculate ice and snow melting and glacial dischargeo Recorded data

in the glacial area in the Qilian Mountain is discontinuous and the

temperature on the glacial surface was calculated» Taking "July

First" Glacier as an example, the summer temperature at the head of

East Niumaojuanzi Valley, which is 5.6 km away from the glacier, in

1976 and in 1984 was recorded and interpolated to understand the

temperature and its variation in the mountains» It was found from the

météorologie record on "July First" Glacier that there is a relation-

ship between the 10 days average temperature (Y) at the station and

the average temperature (X) at 650 mb in Jiuquan:

Y = 0.046 + 1.016X

(4)

The relationship coefficient for equation (4) is 0.978 and

standard variance is 0.3» The equation has a high significance, and

could be taken to interpolate and to extend recorded data for a short

period» During the recent 30 years, the summer (from June to August)

average temperature in the middle of Qilian Mountain is similar to

that in the Hexi Corridor. 1968 is a key point. Average temperature

from 1968 to 1985 is 1.0°C lower than that from 1957 to 1967 (Figure

1). A similar temperature dropping feature was found in the high

mountain area of the eastern and western parts of the Qilian Mountain

(Ding Liangfu, 1985).

u

L,

8.0

o

2

7.0

o

5

S- 6.0

o

-

5.0

<

1956

60

64

68

72

76

80

84 86

Year

Figure 1 Curves of annual fluctuation and three year moving average

summer temperature on the "July First" Glacier.

The zero-equilibrium line is a sensitive indicator of glacier

response to climatic change, which is strictly governed by radiation

and precipitation. Glacial accumulation is equal to glacial melting

at glacial zero-equilibrium line, where glacial melting is mainly

dependant on the summer temperature» An approximate altitude of the

zero-equilibrium line could be determined by the average summer

temperature. In years with equal amounts of solid precipitation,

Glacial mass balance and climate 383

surplus heat would make the zero-equilibrium line rise as well as

melting the solid precipitation in a year with high temperature, the

zero- equilibrium line would drop in a year with a low temperature,

This process reflects a dynamic equilibrium process. Therefore, the

average temperature at the zero-equilibrium line not only reflects

glacial melting, but also reflects glacial accumulation. It is a

comprehensive indicator for glacial melting and accumulation. The

freezing level over Jiuquan was selected to calculate the zero-

equilibrium line on"July-First" Glacier. The analysis shows a good

relationship:

H =

2968.93 exp(0.0988h)

(5)

Where H stand for the glacial zero-equilibrium line (m), and h for

the altitude of freezing level over Jiuquan (xl03m). The relationship

coefficient is 0.987 and standard variance 13 for equation (5). It

could be used to reconstruct the zero-equilibrium line with upper air

data.

The trend of the zero-equilibrium line on "July First" Glacier is

basically identical with average summer temperature showing a drop of

88 m in zero-equilibrium line during the 18 years after 1968 (Figure

2). Average summer temperature fluctuation corresponds to altitude

fluctuation in the zero-equilibrium line, an increase (or decrease)

of 1 °C in temperature corresponds to a rise (or drop) of 80 m in

zero-equilibrium line which is smaller than the temperature dependen-

ce on some glaciers and is smaller than that on Glacier No.l in the

Uriimqi River (an increase of 1°C corresponds to a rise of 125m

Kuhn,M. 1981). The altitude of the zero-equilibrium line on "July

First" Glacier fluctuates between 4550m and 4770m with a maximum

amplitude of 220m.

4800

I

» 4700

-o

2

<

4600

1956

60

64

68

72

76

80

84 86

Year

Figure 2 Curves of annual fluctuation and three year moving average

of zero - equilibrium line on the "July First" Glacier.

The main source of mass for the glacier is precipitation. The

precipitation in the Qilian Mountain is concentrated mainly in

summer. The summer precipitation is 70-80% of annual precipitation.

Because glaciers are at high altitudes, precipitation is mainly in

the form of snowfall. The amount of precipitation is therefore

384 Xie Zichu & Liu Chaohai

responsible for glacial accumulation. The precipitation in the

mountain is composed of dynamic precipitation and convective

precipitation which increase with altitude. The combination of these

two types of precipitation makes the mountains' precipitation stable

and the amplitude of annual change small. The variability factor is

generally smaller than 0.2. The stability of mountain precipitation

makes the amplitude of annual variation of total glacial accumulation

small. But net accumulation, which possesses secular water cyclic

function to glacier is mainly controlled by seasonal snow melting in

accumulative area, i.e., mainly controlled by temperature during

ablation period. Glacial coefficient (or accumulation area ratio of

glacier) which is governed by the altitude of glacial zero-

equilibrium line reflects comprehensively the influence of

precipitation and temperature on net glacial accumulation. A linear

relationship between average glacial accumulation (c) and glacial

coefficient (f) was found:

c = 162.12 + 58.75f

(6)

Glacial melting is a function of temperature. Krenke, A.N. (1971)

and Hoinkes, H. (1971) calculated the glacial melting and glacial

mass balance in Central Asia in the Soviet Union and in the Alps in

Switzerland respectively. Glacial melting in the Qilian Mountain

occurs mainly in the summer. It is therefore reasonable using average

summer temperature (Ts)

to calculate average glacial melting

intensity (A). Their relationship can be expressed as :

 = 425.96 +146.28 Ts

(7)

The mass balance in "July First" Glacier during the past 29 years

(1957-1985) was calculated according to the above relationship

between mass balance components and the climatic factors.The trend of

mass balance, average summer temperature in the mountains and the

zero-equilibrium line altitude is basically identical, i.e., it is

characterized by negative mass balance conditions with a cumulative

negative value of 232.9x10 m3

in the 11 years before 1968, while it

is characterized by positive mass balance conditions with a

cumulative positive value of 773.4 x lO^m3

(Figure 3) in 18 years

after 1968. The glacier has thickened by 1.8m during the same period.

Relatively larger positive mass balance appeared in 1967/68 (77.2x10

m3

) , 1975/76(116.7xl04m3), and 1982/83 (87.1xl01+m3 ) in "July First

"Glacier, with a cycle of 7-8 years. The drop of glacial zero-

equilibrium line (H) corresponds to an increase in mass balance (B)

with a good linear relationship:

B = 3 186.25 - 0.68 H

(8)

A rise (or drop) of lm in zero equilibrium line corresponds to an

increase (or decrease) of O.ôSxlO^m in mass balance.

Glacial mass balance and climatic 385

120-

£

o

I

i

84 86

Year

Figure 3 Curves of annual variation and three moving average of

mass balance from 1957 to 1985 on "July First" Glacier.

The relationship between variation in glacial

mass balance and climate within a year

The seasonal variation in glacial area in Qilian Mountain can be

classified into cold and warm seasons according to whether the

monthly mean temperature is below or above zero. The cold season

(from September to May of the next year) corresponds to the arid

season, in which the ablation and accumulation are very small.The

warm season (from June to August) corresponds to the wet season,

which is not only the period of glacial ablation, but also the period

of main glacial accumulation. Such character of summer nourishment

makes the process of mass balance mild and the mass balance level

comparatively low. Furthermore, it also has influence on ice

temperature, ice formation and glacial motive velocity as well as

other physical properties. These features make the continental

properties of the glaciers more prominent. So the Qilian Mountain is

the most typical area of extra-continental glaciers in the world.

Taking the "July First" Glacier as an example, during the cold season

in 1974/75 and 1976/77, there was stronger ablation on the tongue,

i.e. , the winter mass balance became negative. Observation during the

cold season in 1984/85 also showed a negative winter mass balance.

Even the abnormal phenomenon that the ablation during warm season is

smaller than that during the cold season was observed (Table2). Thus

it c ame to light that for a typical glacier of warm season nourishme-

nt the accumulation, ablation and mass balance have the following

correlations (Xie Zichu, 1980):

c

s>cw>

b ,>0,

b »<0,

a s

>a w ;

bw

>0 (in accumulation area);

b^ <0 (in ablation area).

386 Xie Zichu & Liu Chaohai

Table 2 The mass balance on the ablation ? of "July First" Glacier

Interval

(Year, Month)

1975,9

1976,6

1976,9

1977,6

1984,6

1984,9

1985,6

- 1976,5

- 8

- 1977,5

- 8

- 8

- 1985,5

- 8

Altitude of

equilibrium

(m)

4560

4510

4550

4590

4600

4540

4710

zero-

line

Ablation

area

(km1)

0.42

0.32

0.41

0.55

0.57

0.38

1.19

Amount of

balance

(xlOV)

-9.22

-7.65

-6.05

-22.03

-20.40

-5.53

-56.17

Depth

(mm)

-219

-239

-148

-400

-358

-147

-470

these characteristics of the mass balance process are just opposite

from that of the cold season nourishment glaciers,

The precipitation during the cold season is no more than 20-30% of

annual precipitation. The exposed glacial ice melting under solar

radiation causes the negative value of mass balance in cold season.

Besides, since glacial mass balance is observed on fixed dates, the

ablation doesn't completely coincide with the standard summer months.

This is also one of the causes for the negative value of the winter

mass balance in the tongue. Thus, the air temperature at various

altitudes from late May to early or middle September are especially

analysed to calculate the date of the beginning and end of ablation.

According to the analysis of the observed data, the following

relationship is revealed between the mean daily temperature at

certain altitudes of "July First" Glacier and the corresponding daily

mean temperature at the base station (3700 m.a.s.l.):

TQ = 0„83tB - 0.66(HG- Hg)/100 - 0.3

(9)

Where Tg is the mean daily air temperature at any altitude on the

glacier ( C); tB

is the mean daily air temperature at the base

station ( C); HQ the altitude of a given point on the glacier (m) ;HR

is the altitude of the base station (m). It is very convenient to

calculate the corresponding daily mean air temperature from equation

(9) at the base station while the temperature of a given altitude of

glacier is equal to 0 C. According to calculation, a daily mean air

temperature of 0 C at the glacial terminus corresponds to a daily

mean air temperature of 4.4 C at the base station, a temperature of

0 C at 4500 m corresponds to a temperature of 6 ° C at the base

station; a temperature of 0 C at the glacier summit (5158 m) i.e.,

the whole glacier is under ablation state, corresponds to a tempera-

ture of 11.2°C at the base station. The corresponding daily mean

temperature at the base station can be used as a temperature index

indicating the beginning and end of the ablation period while daily

mean air temperature at the end of the glacial tongue is above or

below 0 ° C.

Glacial mass balance and climatic 387

Summary and analysis

The results from calculation show that the ablation period in

different years are distincte The ablation period usually begins in

early June, and ends in early September» The persistence of the

ablation period mainly depends on the temperature conditioning in

spring and autumn. The decrease of air temperature in Spring and

Autumn results in the shortening of the ablation period, and vice

versa. For instance, because of the higher temperature in the spring

and autumn of 1963, the ablation period of the glacier was elongated

to 121 days, the lower temperature in Spring and autumn in 1971 lead

to the shortening of the ablation period to 71 dayso With increasing

altitude, the ablation period becomes shorter and shorter, up to the

snow line(4650 m), the mean ablation period is 42 days* In some

years, the maximum ablation altitude might get to 5150 m, which

causes the lift of infiltration zone and leads to the absence of cold

infiltration-recrystallization zone» Besides, during the ablation

period, the weather process with strong temperature drops would also

lead to great decreases of air temperature on the glacier, even

making the ablation altitude lower than the glacial terminus and the

meltwater would disappear temporarily»

Taking 1968 as a key point, the beginning of the ablation period

on the "July First" Glacier during the 18 years after 1968 was 8 days

later than that in the 11 years before 1968, but the end time was 6

days earlier i.e. the ablation period was shortened by about 14 days.

The mean maximum melting altitude has dropped by 130 m or so. The

shortening of the ablation period and the decrease of its temperature

made the runoff of the rivers, which are mainly nourished by glaciers,

decrease remarkably, meanwhile the retreat of glaciers has slowed

noticeably. Glacier No. 4 in Shuiguan River of the east part of the

Qilian mountain retreated 8.9 m on average from 1967 to 1984, but 16m

from 1956 to 1976 each year. The "July First" Glacier in the middle

part of the Qilian mountain has retreated for 1 m each year from 1975

to 1985, and has nearly got to a stable state. The tongue of Glacier

No. 12 in Laohugou in west part of the Qilian Mountain becomes

thicker, and the glacial area has begun to extend.

References

Wang Zhongxiang, Xie Ziuchi, Wu Guanche (1985) Mass balance of

glacier in Qilian Mountains, Memoirs of Lanzhou Institute of

Glacialogy and Cryopedology Chinese Academy of Sciences, No. 5.

Ding Liangfu, Kang Xingcheng (1985) The relationship between climatic

variances and glacier changes, Memoirs of Lanzhou Institute of

Glaciology and Cryopedology Chinese Academy Sciences, No.%.

Kuhn, M. (1981) Climate and glaciers, Sea level, ice and Climatic

Changes, IAHS publ. No. 131.

Krenke, A.N. (1971) Climatic conditions of present day glaciation in

Soviet Central Asia, Snow and Ice Symposium, IAHS publ. No. 104.

Hoinkes, H. and Steinacker, R. (1971) Hydrometeorological

implications of the mass balance Hinterisferner, 1952-53 to 1968-

69, Snow and Ice Symposium, IAHS publ. No. 104.

Xie Zichu (1980) Mass balance of glaciers and its relationship with

388 Xie Zichu & Liu Chaohai

characteristics of glaciers, Journal of Glaciology and

Cryopedology, Vol.2, No«4.

The Influence of Climate Change and Climatic Variability on the Hydrologie

Regime and Water Resources (Proceedings of the Vancouver Symposium,

August 1987). IAHSPubl. no. 168, 1987.

Global climatic changes and regional

hydrology: impacts and responses

Peter H. Gleick

Energy and Resources Group

University of California

Berkeley , California, USA

ABSTRACT

As the atmospheric concentration of carbon

dioxide and other trace gases increases, changes in global

and regional climatic conditions will lead to a wide range

of hydrologie impacts, including changes in the timing and

magnitude of runoff and soil moisture. These hydrologie

changes, in turn, will result in diverse economic, social,

and political consequences.

The nature of the regional hydrologie effects depends

on changes in the climatic conditions and the water-

resource characteristics of the region. The research

conducted to date has identified a wide range of potential

problems—as well as some possible advantages—that might

result from plausible changes in climate estimated by

state-of-the art general circulation models.

These hydrologie changes fall into a series of distinct

categories, including: changes in the timing of water

availability; changes in the magnitude of water availabi-

lity; changes in the hydrologie variability; and effects

on water quality. Similarly, diverse societal responses to

the hydrologie changes are available, including adaptation,

mitigation, and prevention. Each of these responses

depends on the quality of the information available on

future impacts and on the perceived importance of the

effects. This paper discusses the extent and character of

hydrologie changes that could result from global climatic

changes, together with the options available for hydrolo-

gists and water planners.

Introduction

Growing attention is being paid to climatic changes that may result

from increasing atmospheric concentration of carbon dioxide and other

trace gases. While the direct effects of changes in climatic

conditions can be severe—as can be seen by the effects of existing

climatic variability—we must also pay attention to the wide range

of indirect effects, such as changes in agricultural productivity,

changes in sea-level, and changes in water resources. This latter

category is one of the most important and yet least well-understood

consequence of future changes in climate. Hydrologie impacts may

include major alterations in the timing and magnitude of surface

runoff and soil-moisture availability, and changes in the quality of

freshwater resources. Associated with these effects will be a wide

389

390 P.H. Gleick

range of economic, environmental, and societal impacts» This paper

discusses the likely extent and character of important hydrologie

changes that could result from global climatic changes, together with

the options available to hydrologists and planners for dealing with

the most severe impacts»

The limited research conducted to date has identified a wide range

of potential problems—as well as possible advantages—that might

result from plausible changes in climate» These hydrologie changes

fall into distinct categories, including: changes in the timing and

magnitude of water availability; changes in the frequency and

severity of severe events, and effects on water quality» Similarly,

diverse societal responses to the hydrologie changes are possible,

including adaptation, mitigation, and prevention» Each of these

responses depends on the quality of the information available about

future impacts and on the perceived importance of the effects»

Future climatic changes

Despite the fact that hydrologists need accurate information on

climatic means and variability in order to develop appropriate water-

resource designs and rules of operation, details of future climatic

conditions cannot yet be predicted with any high degree of confidence»

The principal reasons for this inability to clearly identify future

climatic changes are the complexities of the ocean-atmosphere-land

interactions, the difficulties of developing satisfactory computer

models to reproduce these interactions, and uncertainties about our

actions that affect climatic conditions»

The problem is that, at present, while there are many ways in

which climate may be affected by human actions, we are unable to see

clearly either the direction of future climatic changes or nature of

their societal impacts» Because we are unable to "do the experiment"

directly, we must attempt to model climate and climatic changes—an

imprecise alternative because of the complexity of the global climate

system» Much of the effort of trying to understand the atmospheric

system has focused on the development of large-scale computer models

of the many intricate and intertwined phenomena that make up the

climate» The most complex of these models - general circulation

models (GCMs) - are detailed, time-dependent, three-dimensional

numerical simulations that include atmospheric motions, heat

exchanges and important land-ocean-ice interactions (see, Manabe

1969a, 1969b; Schlesinger and Gates 1980; Manabe and Stouffer 1980;

Wetherald and Manabe 1981; Ramanathan 1981; Manabe et al» 1981;

Hansen et al» 1983, 1984; Washington and Meehl 1983, 1984)»

GCMs permit us to begin to evaluate some of the implications for

global climatic patterns of increasing concentrations of radiatively-

active atmospheric gases» While many uncertainties remain, a

consensus is now beginning to form about the direction and magnitude

of certain major impacts, such as increases in global-average

temperatures and changes in the intensity and distribution of the

global hydrologie cycle»

Unfortunately, state-of-the-art general circulation models are

large and expensive to operate» Furthermore, while GCMs are invaluab-

le for identifying some climatic sensitivities and changes in global

Effects of climatic change on regional hydrology 391

climatic characteristics, they have two limitations that reduce their

value to researchers interested in more detailed assessments of water

resources: (1) they are unable to provide much detail on regional or

local impacts, and (2) they are unable to provide much detail on

small-scale surface hydrology. Until our ability to model climate

improves, we must use other methods to either enhance the information

available from GCMs or provide insights now unavailable from them.

Plausible future hydrologie changes

The attention focused on large-scale GCMs in recent years results in

large part from their relative sophistication compared to other

models. Yet this attention has also highlighted the need for new

methods of hydrologie assessment. Recently there have been some

serious efforts to evaluate the regional hydrologie implications of

climatic changes (Schwarz 1977; Stockton and Boggess 1979, Nemec and

Schaake 1982; Revelle and Waggoner 1983; Flaschka 1984; U.S. Environ-

mental Protection Agency 1984, Cohen 1986, Gleick 1985, 1986a,b,

1987c). These works have provided the first evidence that relatively

small changes in regional precipitation and évapotranspiration

patterns might result in significant changes in regional water

availability.

Methods for evaluating the hydrologie impacts of climatic changes

include using historical data to evaluate the effects of past

fluctuations in precipitation and temperature on runoff and soil

moisture; determining the sensitivity of runoff and soil moisture to

hypothetical changes in the magnitude and timing of precipitation and

temperature; and incorporating regionally disaggregated changes in

temperature and precipitation predicted by GCMs into more accurate

regional hydrologie models. While none of these methods - individual-

ly - can provide much reliable information on future changes, each

can provide insights into specific hydrologie vulnerabilities to

climatic change.

Future hydrologie changes: what can we expect?

Changes in climate may cause changes in a variety of hydrologie

variables, including the timing, location, duration, and extent of

precipitation, runoff, soil moisture, and extreme events. These

impacts can be categorized in a variety of different ways. One useful

method, shown in Table 1, is to separate the impacts by the spatial

and temporal scales involved, with additional separation for the

different statistical moments of interest and the distinction between

political and geophysical boundaries. In the following sections, the

most plausible and worrisome changes in water availability are

described. These changes are not the only hydrologie effects that

will occur, and not all of these will occur at any one place or at

any one time. Nevertheless, we should pay particular attention to

these impacts because they are more likely to occur, they are harder

to mitigate, and they may be more disruptive than other climatic

effects.

392 P.H. Gleick

Table 1 Hydrologie effects of climatic change

Hydrologie Variable of Interest

Useful Precipitation

Surface Runoff

Available Soil Moisture

Groundwater

Temperature

Monsoonality (Onset, Ending, Intensity, Location)

Storm Events

Temporal Scale of Interest

Long-Term (greater than annual)

Annual

Seasonal (two to six months)

Monthly

Daily

Spatial Scale of Interest (Political)

Global

10® km2

Continental

10^ km2

Country/Region 10° km2

Local

10J - 10s km2

Spatial Scale of Interest (Hydrologie)

Global

10® km2

Continental

10^ km2

Regional

10^ - 10° km2

Watershed

102 - 105 km2

Statistical Scale of Interest

Mean

Variance

Persistence

Skev

Higher Moments

Hydrologie Impact of Interest

Quantity

Quality

Peak Events (High and Low)

Source. Gleick (1987a)

a) Precipitation

Despite the fact that all GCMs predict an intensification of the

overall hydrologie cycle, particularly increases in global average

annual precipitation rates, this information is only marginally

useful* As the global average temperature increases, we expect an

increase in the rate of évapotranspiration and precipitations GCMs

now suggest that the annual-average increase in global precipitation

may be on the order of seven to fourteen percento Far more interest-

ing and potentially disruptive are the changes in regional precipita-

tion patterns, which are much harder to models At present, there is

little consensus about specific regional changeso

Two specific vulnerablities need attention: (1) changes in average

precipitation rates in regions with rainfed agriculture; and (2)

changes in the frequency of extreme precipitation events in areas

vulnerable to flooding and storms s In the first case, an increase in

precipitation in agricultural regions dependent on rainfall could

Effects of climatic change on regional hydrology 393

have a beneficial effect, while a decrease would have the opposite

effect. Similarly, floods and storms are already responsible for

enormous human sufferings Such events could be exacerbated by an

increase in the variability of regional precipitation»

There are a number of hydrologie effects that may be driven

primarily by temperature changes, not precipitation changes» This

permits the identification of certain impacts that are somewhat

independent of precipitation rates» Among these impacts are changes

in soil-moisture availability and changes in the timing of surface

runoff» Although both of these variables depend heavily on site-

specific

characteristics such as

soil-moisture

capacity,

precipitation rates, vegetation characteristics, topography, and soil

depth and type, some generalizations can be made»

b) Soil-moisture availability

Soil-moisture behavior in general circulation models is very simple,

and efforts to improve the representation of moisture in the soil

column are now underway (Dickinson 1986; Rind 1987)» For the last

several years, there has been a growing interest in soil-moisture

changes because of the possibility that some significant—and

potentially adverse—effects on soil-moisture availability may result

from increasing concentrations of carbon dioxide» In particular, some

general circulation model results suggest that soil moisture in mid-

continental regions in mid-latitudes may decrease during summer

months, which is often the critical period for crop productivity

(Manabe et al» 1981; Mitchell 1983; Manabe and Wetherald 1986; Rind

1987)» Although there are disputes over the magnitude (and sometimes

the direction) of these soil-moisture changes, the present active

research in this area may help to resolve the uncertainties»

Recently, some detailed hydrologie models have supported the

possibility of decreased summer soil-moisture availability in some

regions (Gleick 1986a, 1987c)» In particular, despite increases in

annual and seasonal precipitation, increases in temperatures can lead

both directly and indirectly to decreases in soil-moisture availabi-

lity during summer months» In regions with winter snowfall and

spring snowmelt (in the United States, such regions include large

parts of California, the Rocky Mountains, the Pacific Northwest, and

the Northeastern and Northcentral U»S»), increases in temperatures

may lead to decreases in the ratio of snow to rain in winter months,

increases in the speed of snowmelt in spring months, and an earlier

onset of drying in early summer (Gleick 1987c)» Figure 1 shows the

decreases in summer soil moisture in a major California watershed

that result from eight scenarios generated by three state-of-the-art

GCMs. The GCMs each predict quite different precipitation, yet the

regional model results using these scenarios all show decreased

summer soil-moisture availability» Similar results were identified

for other regions by Mather and Feddema (1986)» This robust result is

one example of the type of hydrologie impact that should be more

carefully studied on a regional basis»

394 P . H . G l e i c k

CHANGE IN SUMMER SOIL MOISTURE (JJA)

GCM S c e n a r i o s

CD

en

c

V

60

O

c

CD

O

CD

a.

UJ

=>

\—

SU

o

40

20

-20-

O

in

UJ

3

10

-40

-60-

-80

0

i -31

^

-20

%

-14

NCAR

GFDL

GISS

T ONLY

T,P(r)

T,P(a)

Figure 1 Change in summer (June, July and August) soil moisture

predicted by a water-balance model of a major California

watershed using precipitation and temperature data from

three general circulation models: the National Center for

Atmospheric Research (NCAR), the Geophysical Fluid

Dynamics Laboratory (GFDL), and the Goddard Institute for

Space Studies (GISS). Note that all three models show

decreases in soil moisture. The eight scenarios are:

Temperature only, Temperature and relative precipitation,

and Temperature and absolute precipitation. See Gleick

(1987a) for details of the model, the scenarios, and the

uncertainties.

c) Runoff

Surface runoff shows a sensitivity to increases in temperature

similar to the one described above for soil moisture. In certain

regions, seasonal runoff appears to be vulnerable to changes in the

timing of surface flows, even when the overall annual runoff does not

change significantly.

Figure 2 plots historical average-monthly runoff in a major

California watershed together with the runoff predicted by a water-

balance model using temperature and precipitation changes predicted

by a state-of-the-art GCMo In this case, while the average-annual

runoff volumes do not change significantly between the two cases, the

monthly pattern has changed» This can be seen by the large increase

in winter runoff and the decrease in summer runoff. The physical

mechanisms at work here are similar to the ones described above

Effects of climatic change on regional hydrology 395

MODEL VS. ACTUAL SURFACE RUNOFF

GCM TEMPERATURE AND PRECIPITATION CHANGES

ACTUAL

MODEL

5000

4000

CO

u

<u

<u

3 3000

o

0 2000

o

1000

~i

1

1

— I

1

1

1

1

r

r

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Figure 2 Average-monthly runoff: actual and model-predicted using

temperature and precipitation changes developed by the

Geophysical Fluid Dynamics Laboratory GCM» The annual

runoff volumes for both of these runs are the same; the

seasonal pattern has changed» See the text for details»

driving the soil-moisture changes—less total winter snowpack, more

winter runoff, faster snowmelt in the spring, and smaller spring

runoff » Figures 3 to 5 show the details of average-monthly changes in

runoff using the temperature and precipitation changes from three

GCMs to drive a regional water-balance model of the Sacramento Basin

in Northern California - perhaps the most important watershed in

California (Gleick 1987b)« In all of these cases, summer runoff

decreased and winter runoff increased, while average-annual runoff

was only slightly changed (Gleick 1987c). These runoff changes can

increase the frequency of flood events by shifting more runoff to

peak runoff months, even if overall average runoff doesn't change.

Similarly, regions dependent on minimum summer flows may be adversely

affected»

The vulnerability of water resources to climatic conditions

The availability of freshwater for agricultural, industrial, residen-

tial, and commercial use is sensitive to existing climatic variabili-

396 P . H . G l e i c k

Climate-Induced Change In Runoff

NCAR TEMPERATURE AND ABSOLUTE PRECIPITATION

— 100 J

1

1

1

1

1

1

1

r

1

i

r

1 —

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Figure 3 Percent change in monthly runoff between the NCAR

temperature and absolute precipitaion run and the long-

term average runoff. Note the increase in winter runoff

and the decrease in summer runoff.

ty. The sensitivity varies with supply and demand, water quality, and

the specific needs of the users» As the climate begins to change, the

most severe pressures rn available water resources are likely to come

in regions where the existing water resources are already constrained

during certain times. This section discusses existing vulnerabilities

that might be either exacerbated or mitigated by climatic change.

Regions with natural deficits: Arid and semi-arid lands are, by

definition, regions with natural water deficits: the potential

évapotranspiration exceeds natural water inputs during part or all of

the year* At the same time, these lands are often thought to hold the

greatest potential for future agricultural development assuming that

water can be made available for irrigation, and that the soil quality

is high enough (Rosenberg 1981; Gleick 1987a)» Improvements in the

hydrologie conditions of these regions would require increases in the

average water availability. Since évapotranspiration is likely to

increase following a doubling of atmospheric carbon dioxide, such an

increase in average availability must come through precipitation or

water transfers into the basin. At the same time, if the variability

of water resources availability were to increase, the vulnerability

of these regions to climate could remain high. For a climatic change

to be most advantageous to arid and semi-arid regions, there would

have to be an increase in mean water availability and a decrease in

the variability. It is important to note here, however, that while

Effects of climatic change on regional hydrology 397

Climate—Induced Change In Runoff

GFDL TEMPERATURE AND ABSOLUTE PRECIPITATION

o

-100 J

1

1

1

1

1

1

i

1

1

1

1

1

—

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Figure 4 Percent change in monthly runoff between the GFDL

temperature and absolute precipitation run and the long-

term average runoff. Note the increase in winter runoff

and the decrease in summer runoff.

such changes might be beneficial to agricultural productivity or

other human uses, they can lead to dramatic shifts in the natural

character of the existing ecosystems (Gleick. 1987a).

Regions with high societal demands: In many regions of the world,

the demand for water approaches the available supply during certain

periods. In these regions, efforts are often already underway to

modify either the available supply or demand. Changes in climate that

exacerbate these demands or reduce the overall supplies will have

negative consequences for the region, while overall increases in

water availability could ease some problems. As with the first

example, the most advantageous climatic change would be increased

mean availability and decreased variability. An increase in variabi-

lity would increase the frequency of severe events and may not result

in net benefits to the region.

Flood-prone regions: Areas prone to flooding, such as low-lying

floodplains, would benefit from a decrease in the variability of

precipitation and runoff and suffer f rom an increase in both the mean

and the variability of water availability. In these regions, measures

are often taken to reduce the vulnerability of society to floods,

such as the development of flood-control reservoirs. While such

facilities are often valuable, they are also expensive to design and

build. As a result, in order to properly design new flood-control

facilities, information on the nature of future hydrology is necessa-

398 P . H . G l e i c k

Climate—Induced Change In Runoff

GISS TEMPERATURE AND ABSOLUTE PRECIPITATION

100 T

•

•

-100 J

1

1

1

1

1

1

1

1

1

1

1

1

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Figure 5 Percent change in monthly runoff between the GISS

temperature and absolute precipitation run and the long-

term average runoff. Note the increase in winter runoff

and the decrease in summer runoff.

ry if a given facility is to be able to cope with future climatic

conditions» A change such as the one plotted in Figure 2, which

changes the seasonal patterns while not changing the annual average,

would increase the risks of flooding unless changes in the flood-

control system can be made»

Regions dependent on reliable seasonal supply: By far the most-

often heard hydrologie truism is that the supply of water to any

region is not uniformly distributed in space or time» Many regions

are dependent on water supplies that arrive during particular seasons,

such as the monsoons on the Asian sub-continent and winter precipita-

tion in Mediterranean-style climates » In these regions, slight

changes in the timing or magnitude of seasonality will have important

consequenceso Figure 2, which plots a possible change in the timing

of the availability of surface runoff in a major agricultural basin,

shows how the seasonality of runoff may be affected by predicted

climatic changes» Unless extensive reservoir systems permit the

storage and later distribution of seasonal precipitation and runoff,

a change in the seasonality of water availability could stress a

region» Unfortunately, as Nemec and Schaake (1982) pointed out,

climatic changes could cause problems for existing reservoir systems»

Regions sensitive to lake levels:

Major lake systems are

sensitive to inflows and outflows, which in turn depend on both

natural supply (from precipitation and runoff) and natural and

Effects of climatic change on regional hydrology 399

artificial demand (from évapotranspiration and withdrawals). Work by

Snyder and Langbein (1962) and Street-Perrott et al. (1986), among

others shows this sensitivity in the context of changing climate.

Although some work has been done to estimate the effects of future

climatic changes on lake levels (see, Cohen 1986), no clear trends

have yet been identified. Shipping, municipal and industrial water

supply, recreation and natural ecosystems will be affected by both

positive and negative changes in lake levels.

Regions with decreasing water quality:

Deteriorating water

quality due to industrial development, agricultural wastewater, and

population growth will be affected by climatic changes that alter the

availability of freshwater resources. Critical areas include rivers

used for waste disposal that may experience decreases in minimum

flows, groundwater supplies that are sensitive to pollutant inflows,

and the design methods for the adequate disposal of toxic materials.

In some of these cases, societal actions should be taken now to

reduce the vulnerability of water quality to climatic changes. The

challenge is to design such actions to be flexible enough to handle a

wider range of climatic conditions than are now normally anticipated.

Perhaps the most well-known example of this is the salinity

problems of the Colorado River near the U.S.-Mexican border. As more

water is used for agriculture, the salinity of the river increases to

the point where it becomes detrimental to further use. This problem

will increase in severity in the absence of a climatic change and

thus already requires mitigating actions. These mitigating actions,

however, should anticipate minimum and maximum flows lower than those

historically recorded because of the possibility that minimum and

maximum flows could be altered by climatic changes. This added

flexibility can be achieved at a lower cost now than after the

actions have been designed, facilities built, and operating schemes

implemented.

Regions dependent on hydroelectricity: Finally, hydroelectricity

plays a major role in many regions of the world. The reliability of

this energy depends on the reliability of water resources—part-

icularly the timing and magnitude of flow rates. As the climate

changes, one or another of these variables is likely to change, with

the risk that alternative (and more often expensive) methods of

electricity generation will be required to make up shortfalls, or

that potential hydroelectricity will be lost because of incorrectly-

sized and operated facilities (McGuirk 1982). At best, this suggests

the need for a flexible hydroelectric system-operating style; at

worst, existing facilities will have to be redesigned and new

facilities evaluated.

Discussion and conclusions

It is extremely unlikely that all the hydrologie changes induced by

changing climatic conditions will be beneficial. When the water-

resource needs of different regions are studied, we see that

different changes in the mean and variability of water resources are

required in different regions and on different time scales. The

probability is extremely low that the diverse changes appropriate for

all regions will occur in precisely the proper location and at

400 P.H. Gleick

precisely the proper time—such as increases in means in arid regions,

decreases in means in regions subject to flooding, and the appropria-

te changes in both annual and seasonal variability. Given this

problem, attention must be focused on the vulnerability of hydrologie

systems to changes in climate, so that policies to mitigate the worst

effects can be implemented should negative impacts materialize.

Despite the uncertainties that surround the nature and timing of

future climatic changes and their subsequent impacts, the research

discussed here raises some concerns about regional water availability.

In particular, certain types of impacts, such as decreases in summer

soil moisture and runoff and increases in winter runoff are robust

and consistent across widely-varying scenarios. This consistency

suggests strongly that hydrologie vulnerabilities will make the

impacts of climatic changes on water resources an issue of major

concern in many regions of the world.

Some of the results described here support recent suggestions that

summer soil-moisture reductions may occur in many regions of the

world. The principal physical mechanisms involved—the decrease in

snow as a proportion of total winter precipitation, an earlier and

faster disappearance of winter snowpack due to higher average

temperatures, and a more severe évapotranspiration demand during the

warmer summer months—are both physically plausible and hydrological-

ly consistent. While other, countervailing hydrometeorologic features

may well exist—such as cloud cover/evapotranspiration feedbacks—the

consistency of the soil moisture and runoff results described here

must be considered a first warning of possible important changes in

regional water availability.

Adverse hydrologie changes may, if they materialize, have serious

implications for many aspects of water resources, including agricul-

tural water supply, flood and drought probabilities, groundwater use

and recharge rates, the price and quality of water, and reservoir

design and operation—to mention only a few. Yet information on these

changes, by itself, is unlikely to lead to major policy responses.

Only by looking at the specific characteristics of water-resource

problems—and their vulnerability to—the types of changes in runoff

and soil moisture identified above-can details of future societal

impacts be evaluated.

References

Cohen, S.J. (1986) "Impacts of C02-Induced Climatic Change on Water

Resources in the Great Lakes Basin". Climate Change 8. pp.135-154.

Dickinson, R.E.(1986) "Modeling Hydrology in a Global Climate Model:

Applications to Impact of Amazon Deforestation." Presented at the

American Geophysical Union Fall Meeting, San Francisco, California.

December 8.

Flaschka, I.M.(1984) "Climatic Change and Water Supply in the Great

Basin". Master's thesis. Department of Hydrology and Water

Resources, University of Arizona.

Gleick, P.H. (1985) "Regional Hydrologie Impacts of Global Climatic

Changes", in E.E. Whitehead, CF. Hutchinson, B.N. Timmermann, and

R.G. Varady, (editors). Proceedings of an International Research

and Development Conference. Arid Lands: Today and Tomorrow. Office

Effects of climatic change on regional hydrology 401

of Arid Lands Studies, University of Arizona, Tuscon, October 20-

25, 1985.

Gleick, P.H. (1986a) "Large-Scale Climatic Changes and Changes in

Regional Summer Soil-Moisture Availability". Presented at the

American Geophysical Union Fall Meeting, San Francisco,

California. December 8.

Gleick, P.H. (1986b)"Methods for Evaluating the Regional Hydrologie

Impacts of Global Climatic Changes". Journal of Hydrology 88, pp.

97-116.

Gleick, P.H. (1987a) "Global Climatic Change, Water Resources, and

Food Security. Sponsored by the Indian National Academy of Science,

the American Association for the Advancement of Science, and the

International Rice Institute. New Delhi, India. February 6-9,

1987. 37pp.

Gleick, P.H. (1987b) "the Development and Testing of a Water-Balance

Model for Climate Impact Assessment : Modeling the Sacremento

Basin". In press Water Resources Research.

Gleick, P.H.(1987c) "Regional Hydrological Consequences of Increases

in Atmospheric CO and Other Trace Gases". In press Climatic

Change.

Hansen, J.E., Rind, D., Russell, G., Stone, P., Fung, I., Ruedy, R.,

and Lerner, J. (1984) "Climatic Sensitivity: Analysis of Feedback

Mechanisms". In J.E. Hansen and T.Takahashi, eds. Climate

Processes and Climate Sensitivity, American Geophysical Union

Monograph 29. Maurice Ewing Vol. 5. American Geophysical Union,

Washington, D.C.

Hansen, J., Russell, G., Rind, D., Stone, P., Lacis, A., Lebedeff,

S., Rueddy, R., and Travis, L.(1983) "Efficient Three-Dimensional

Global Models for Climate Studies: Models I and II".Monthly

Weather Review. Vol.Ill, No. 4 pp. 609-662. April.

Manabe, S. (1969a) "Climate and the Ocean Circulation. I. The

Atmospheric Circulation and the Hydrology of the Earth's Surface".

Monthly Weather Review. Vol. 97, No.11 pp. 739-774. November.

Manabe, S. (1969b) "Climate and the Ocean Circulation. II. The

Atmospheric Circulation and the Effect of Heat Transfer by Ocean

Currents".Monthly Weather Review. Vol.97, No. 11. pp.775-805. Nov.

Manabe, S. and Stouffer, R.J.(1980) "Sensitivity of a Global Climate

Model to an Increase of C02 Concentration in the Atmosphere". J.

Geo. Res.. Vol 85, No. C10, pp. 5529-5554. October.

Manabe, S. Wetherald, R.T.Q986) "Reduction in Summer Soil Wetness

Induced by an Increase in Atmospheric Carbon Dioxide". Science.

Vol. 232, pp.626-628.

Manabe, S., Wetherald, R.T., and Stouffer, R.J.(1981) "Summer Dryness

Due to an Increase of Atmospheric C02 Concentration". Climatic

Change 3^ pp. 347-386.

Mather, J.R. and J. Feddema. (1986) "Hydrologie Consequences of

Increases in Trace Gases and C02 in the Atmosphere", in J. Titus

(editor). Effects of Changes in Stratosphere Ozone and Global

Climate. Volume 3: Climate Change. United States Environmental

Protection Agency, pp. 251-271.

McGuirk, J.P.(1982) "A Century of Precipitation Variability Along the

Pacific Coast of North America and Its Impact". Climatic Change 4.

pp. 41-56.

Mitchell, J.M., Jr. (1983) "An Empirical Modelling Assessment of

402 P.H. Gleick

Volcanic and Carbon Dioxide Effects on Global Scale Temperature"»

American Meteorological Society, Second Conference on Climate

Variations , New Orleans, Louisiana» (January 10-14o)

Nemec, J» and Schaake, Jo (1982) "Sensitivity of Water Resource

Systems to Climate Variation"» Hydrological Sciences 27. No8

3,

pp» 327-343»

Ramanathan, V*(1981)"The Role of Ocean-Atmosphere Interactions in the

C02 Climate Problem/' J»Atmos»Sci Vol» 38, pp» 918-930»

Revelle, R.R. and Waggoner, P.E»(1983) "Effects of a Carbon Dioxide-

Induced Climatic Change on Water Supplies in The Western United

States". In Changing Climate » National Academy of Sciences »

National Academy Press, Washington D.C.

Rind, D„ (1987) The Doubled C02 Climate and Future Water Availability

in the United States* Goddard Space Flight Center Institute for

Space Studieso Submitted to Je Geophysical Research.,

Rosenberg, N„J0(1981) "Technologies and Strategies in Weatherproofing

Crop Production"s In L»E. Slater and S.K» Levin (editors)0

Climate's Impact on Food Supplies ; Strategies and Technologies for

Climate-Defensive Food Production» AAAs Selected Symposium 62»

(Westview Press, Boulder, Colorado) » pp., 157-180e

Schlesinger, MeEs, and Gates, W»L. (1980) "The January and July

Performance of the OSU Two-level Atmospheric General Circulation

Modèle" Ja Atmoso Scie, Vole 37, pp» 1914-1943o Septembers

Schwarz, H»E.(1977) "Climatic Change and Water Supply: How Sensitive

is the Northeast?"» in Climate, Climatic Change, and Water Supplye

National Academy of Sciences» Washington, D.C.

Snyder, C.T» and Langbein, WoB»(1962) "The Pleistocene Lake in Spring

Valley Nevada"»JeGeophys»Res» Vole 67, Noe6» pp» 2385-2394» Junee

Stockton, C»We and Boggess, W»Re (1979) "Geohydrological Implications

of Climate Change on Water Resource Development"» U»S» Army

Coastal Engineering Research Center, Fort Belvoir, Virginia» May8

Street-Perrott, F.A., M»A»J» Guzkowska» IeMe Mason, and C»G» Rapley

(1986) "Response of Lake Levels to Climatic Change- Past, Present

and Futures" in J.Titus (editor)» Effects of Changes in Stratos-

pheric Ozone and Global Climatee Volume 3: Climate Change » United

States Environmental Protection Agency» ppe211-216»

U»S. Environmental Protection Agency (1984) "Potential Climatic

Impacts of Increasing Atmospheric C02 with Emphasis on Water

Availability and Hydrology in the United States"» Strategic

Studies Staff, Office of Policy Analysis, Office of Policy,

Planning and Evaluation» April»

Washington, W»M» and Meehl, G»A» (1983) "General Circulation Model

Experiments on the Climatic Effects Due to a Doubling and Quadrup-

ling of Carbon Dioxide Concentration"» J» Geophys» Res» Vol»88,

No» CII, pp» 6600-6610» August»

Washington, W»M» and Meehl, G»A»(1984) "Seasonal Cycle Experiment on

the Climate Sensitivity Due to a Doubling of C02 With an Atmosphe-

ric General Circulation Model Coupled to a Simple Mixed-Layer

Ocean Model"»J» Geophys» Res» Vol» 89, No» D6, pp» 9475-9503» Oct.

Wetherald, R»T» and Manabe, S»(1981) "Influence of Seasonal Variation

Upon the Sensitivity of a Model Climate"» JeGeophys» Res» Vol. 86,

No» C2, pp» 1194-1204. February 20.

The Influence of Climate Change and Climatic Variability on the Hydrologie

Regime and Water Resources (Proceedings of the Vancouver Symposium,

August 1987). lAHSPubl. no. 168, 1987.

The impacts of G02-induced climate change

on hydro-electric generation potential in the

James Bay Territory of Quebec

Bhawan Singh

Professor agrège

Université de Montreal, Canada

Introduction

CO2 - induced climate change can significantly affect the economy of

water - resource - based industries. Other studies in Canada, for the

Great Lakes-St. Lawrence basin (Cohen, 1987, 1986; Howe et al; 1986,

Sanderson et al, 1985; Bruce, 1984) have shown that changes in lake

and river levels, resulting from changes in Net Basin Supply (NBS)

can have far-reaching economic consequences on hydro-electric power

production, lake shipping and shoreline erosion.

In this paper, we intend to focus on the impact of CC>2 - induced

climatic change on the hydro-electricity generating capacity of the

three basins, within the James Bay Territory, that are presently

exploited for hydro-electric generation, namely the La Grande, the

Caniapiscau and the Opinaca-Eastmain basins (Figure la and Figure lb).

The hydro-generating capacity of the three basins are to be de-

rived from actual (normals : 1951-1980) and projected (2 x CO2) changes

in NBS using the GFDL (scenario A) and GISS (scenario B) scenarios

data provided by Environment Canada (Hengeveld and Street, 1986).

These data sets provide normals (1 x CO2) and projected (2 x CO2 )

temperature and precipitation conditions for several grid points

within the James Bay Territory (Figure la and Figure lb).

These projected changes in NBS will then be used to gauge the

potential economic costs or benefits of CO2 - induced climate change,

based on both climate change scenarios (GFDL and GISS).

Methods, data sources and derivations

For the three drainage basins selected, namely, the La Grande, the

Caniapiscau and the Opinaca-Eastmain, the NBS was calculated for both

the normals(1951-80) and the projected data for the two scenarios

(GFDL) and (GISS) in question.

The general form of the formula for calculating NBS for drainage

basins consisting of both land and open lake and reservoir surfaces

is written as:

NBS = Land

x Land + (Piake-^1ak ^ x

Lake ~ Consumption

Discharge

Area

Area

(1)

NBS is expressed in m^s-!, ^lake represents the total precipitation

falling on the lake or reservoir surface and E^^g represents the

total evaporation from the lakes or reservoirs.

403

404 B. Singh

"~r~~

72° >A6/7

1A7

(H3023-)

\ ^ H 3 0 5 S ) N s ^

\cANIAPISCAU

)

1 .

'

«Afi/11

1A6-II/7-H»

'.

A7/11«N

^~*

yK

A.

ÏA6-10/10.

^ „ ^'< D 3 1 2 1 > /

^

/

V*-*OPfNACA -

BASJMAiN

e

} /

—-^, _ *

- ^ " ^

A6-1 1/10-1 1y £ •

V

>Aio/ii

IA11

G F D L grid point

I

Extrapolated grid point

(R1037) Soil unit

etc...

ITT

Figure la Drainage basins and grid points used in study

( Scenario A ).

aJ

76»

72°

68°

=i>^--/

,

-

^(R1037)

B B l 2

(H3023)

\

a B 1 3

/

V Î H s o s s ) ^ .

.CANIAPISCAU^

(Rl016)\

A"

3124^

/

)

/

"B13/13-18- \

LA

GRANDE

.^- '

\_B12/13-18 o

j(Rl037>

\

y t f ^ . , .

,

.

L_ B12-17

„

\ /

*--^

/

\_.__&.OP ( NACA-

.Bia-JBjL/

^ ~ '

lEASfMAIN

>

v-

iBl3/18

HB17

Extrapolated grid point

(R1037) Soil unit

aBl8

1^

64°

Bl4E

km

300

IB19"

ISt. Lawrence Rj-

Figure lb Drainage basins and grid points used in study

( Scenario B ),

Impact of climate change on hydro-electric generation 405

In the analyses, land discharge was calculated as the surface

runoff or water surplus (mm) using the Thornthwaite monthly water

budget method as written by Environment Canada (Johnstone and Louie,

1984). The Thornthwaite water budget formula, stated simply is:

R0

= P - AE ± ASt

(2)

R0

is the monthly surface runoff (mm), P is the monthly precipitation

(mm), AE is the monthly évapotranspiration depending on air tempera-

ture and soil moisture and, ASt is the monthly change in soil mois-

ture. Based on our knowledge of the average soil depth of the region

(Singh and Taillefer, 1984) and on values cited elsewhere for similar

regions (Cohen, 1986, 1987; Howe et al, 1986), a value of Water

Holding Capacity (WHC) equal to 100 mm was used.

Because of the remoteness of the area and the very sparse popula-

tion, water consumption is negligible and, because very little water

is consumed in hydro-electric power generation, this small total

amount of water consumed is neglected in the analysis.

Since the hydrology of the region has been modified recently by

the creation of dams for hydro-electric power generation, and artifi-

cial reservoirs have expanded natural lake surfaces, it was necessary

to slightly modify equation (1) to reflect these changes. Total basin

area, land area and reservoir area for each of the drainage basins

are supplied in Table 1. In addition, Table 1 provides mean hydrolo-

gie characteristics of each basin based on the period 1949-1980.

These conditions essentially represent the pre-damming period since

the dams were closed in November 1978 for La Grande basin, July 1980

for Eastman-Opinaca basin and, October 1981 for Caniapiscau basin

(M.T. Tran-Van SEBJ, personal communication).

Table 1 Physiographic and hydrologie parameters of drainage basins

( Source: SEBJ 1986 )

Drainage Basin

La'Grande

Caniapiscau

Eastmain-Opinaca

Total

Basin Area

(Km2)

98,031.5

36,881.6

40,274.5

Natural

Lakes Area

(Km2)

19,606.3

7,745.1

7,652.2

Reservoirs

Area

(Km2 )

6,021.8

4,273.5

1,041.2

Land Area

without

reservoirs

(Km2 )

78,425.2

29,136.5

36,622.3

Land Area

with

reservoirs

(Km2)

72,403.4

24,863.5

31,581.1

Mean Dis- 1

charge and

(Std. Dev)

m3 /S

1762 (255)

788 (122)

851 (104)

Mean dls-2

charge/

Prec. Ratio

(X)

77

77

84

1 Mean discharge and (Standard Deviation) for period 1949-80:

Measured data: La Grande; 1960-79, Caniapiscau: 1962-79, Eastmain-Opinaca: 1960-80.

Other years data is estimated (SEBJ).

2 Mean discharge Precipitation Ratio for period 1949-80.

The hydrologie data set in Table 1 (1949-1980) therefore corres-

ponds to the normals (1951-1980) data and was thus used for

calibrating the normal discharge data.

406 B. Singh

For the normals (1951-1980) data set, NBS was then calculated

using equation (l)o However for the projected scenarios data, NBS was

calculated both with equation (1) which ignores the reservoirs and

the following equation (3) which takes the surface area of the arti-

ficially created reservoirs into account.

NBS =

Land x Land + ( Plake - Elafc

) x Lake

Discharge AreaR

AreaRX

(3)

Land Areag represents the reduced land area resulting from the

creation of reservoirs, PiakeR a n d ElakeR represent the total of

monthly lake plus reservoir precipitation and the monthly evaporation

respectively, and Lake AreaR represents the sum total of lake and

reservoir areas.

In equations (1) and (3), since land discharge and P^a^e- Elakeare

in mm/year and land area and lake area are in sq. km each half of the

equations, neglecting consumption, was multiplied by the conversion

factor of 31.71xl0-6 so that the final units of NBS are in m3 s-l .

Monthly values of discharge, precipitation and evaporation were sum-

med to derive total annual values.

As mentioned previously total annual land discharge is calculated

using the Thornthwaite water budget method. Lake and lake plus reser-

voir precipitation was assumed to be equal to the land precipitation,

unlike other other studies (Cohen, 1986,1986b) since the lake and

reservoir surfaces are relatively small.

Because of the small basin areas and the coarse spatial resolution

of the grid points for both scenarios (GFDL and GISS), it was

necessary to perform a series of extrapolations between grid points

so as to derive at least two reference points for each basin. These

extrapolations were done linearly along the X (latitude) and Y

(longitude) coordinates and diagonally across these X-Y vectors.

For instance, in the case of the La Grande basin, the two derived

grid points for scenarios A are A6/ioand A 6/11 and for scenario B,

B12 and

B12/13-18 (Figure 1 and Figure 2). Grid point A6 /1 Q

is

derived by linearly extrapolating latitudinally between GFDL grid

points Agand Aio(Figure 1). On the other hand, grid point Ag/n is

derived by extrapolating diagonally between GFDL grid points A g and

A]^. For scenario B, GISS grid point B]^ is already within the basin

(Figure 2). However, grid point B^2/13-18is derived by first extrapo-

lating latitudinally between GISS grid points B ^3 and B^g to derive

B13/18 and then by extrapolating diagonally between grid points

B13/18and B12 .

Interpolated grid points for the other two basins, as shown in

Figure 1 and Figure 2, namely Caniapiscau and Eastmain-Opinaca were

derived using a similar linear extrapolation procedure.

These extrapolated grid points were then averaged so as to derive

the mean basin precipitation and the mean land surface discharge on a

total annual basis. This averaging was weighted so as to reflect the

approximate proportion of the total basin that each extrapolated grid

point represented and so as to reflect the spatial variation of

precipitation and temperature (used to evaluate discharge) as shown

elsewhere (Singh et al, 1987). These results are shown in Tables 2a,

2b, 3a, 3b, and 4a, 4b.

Impact of climate change on hydro-electric generation 407

Table 2a Annual input parameters for deriving unadjusted net basin

supply (NBS) and correction factor from extrapolated grid

points for La Grande basin: Scenario A (GFDL); normals

data

Extrapolated

Grid Points

A6/10

A6/ll

Land

Discharge

(m3/S)

345.2

447.1

406.4*

Lake Preci-

pitation

(mm)

747.1

844.3

805.4"

Lake Evap

Soil Unit

D3023

R1037

orat ion

(mm)

708.5

459.7

559.2**

Unadjus ted

Mean NBS

(m3/S)

1163-8

Underest imate

of Mean

Discharge

%

34.0

Correction

Factor

1.514

Weighted mean = .6 (A6/11) + -4 (A6/10)

Weighted mean = .6 (R1037) + .4 (D3023)

Table 2b Annual input parameters for deriving unadjusted net basin

supply (NBS) and correction supply from extrapolated grid

points for La Grande basin: Scenario B (GISS); normals

data

Extrapolated

Grid Points

B12

B12/13-18

Land

Discharge

(m3/S)

313.2

382.4

354.8*

Lake Preci-

pitation

( mm)

705.6

782.7

752.2*

Lake Evap

Soil Unit

D3023

A1037

orat ion

( mm)

708.5

459.7

559.2**

Unadjusted

Mean NBS

(m3/S)

1002.3

Underesc imate

of Mean

Discharge

7.

43.1

Correction

Fac tor

1.757

*

Weighted mean = .6 (B12/

, - 18) + .4 (B )

** Weighted mean = .6 (R1037) + .4 (D3023)

Lake evaporation on the other hand was calculated according to the

Priestley-Taylor (1982) equation, written as

Elake = {et S/S + Y(Q * " QG)}/ L

(4)

where:

E^ake is tne

monthly lake evaporation (mm), Q* is the net

radiation (MJ/m2/month), QG is the soil heat flux (MJ/m2/month),S is

the air temperature-dependent slope of the saturation vapor pressure

curve (Pa/°C), y is the psychrometric constant (Pa/°C), a is a non

- dimensional surface evaporability factor and L is the latent heat

of evaporation (MJ/kg).

The value of a used here is 1.26, as recommended by Priestley

and Taylor (1972) and Stewart and Rouse (1977) for open water

408 B. Singh

Table 3a Annual input parameters for deriving unadjusted net basin

supply (NBS) and correction factor from extrapolated grid

points for Caniapiscau basin: Scenario A (GFDL); normals

data

Extrapolated

Grid Points

A 6/U

A 7/ll

Land

Discharge

(m3/S)

447.1

476.5

461.8*

Lake Préci-

pitât ion

(mm)

844.3

866.8

855.6*

Lake Evaporation

Soil Unit

(mm)

H3055

R1016

432.9

449.7

441.3**

Unadjusted

Mean NBS

(m3/S)

528.4

Underescima te

of Mean

Discharge

7.

32.9

Correct ion

Factor

1.491

*

Weighted mean = .5 (A,,,,) + -5 (A-/n)

** Weighted mean = .5 (H3055) + .5 (R1016)

Table 3b Annual input parameters for deriving unadjusted net basin

supply (NBS) and correction factor from extrapolated grid

points for Caniapiscau basin: Scenario B (GISS); normals

data

Extrapolated

Grid Points

B 13

B 13/18

Land

Discharge

(m3/S)

386.0

451.5

418.8*

Lake Preci-

pitation

(mm)

762.8

859.7

811.3*

Lake Evap

Soil Unit

H3055

R1016

orat ion

(mm)

432.9

449.7

441.3

Unadjusted

Mean NBS

(m3/S)

477.8

Under est imat e

of Mean

discharge

7.

39.4

Correct ion

Factor

1.649

*

Weighted mean = .5 (B,,) + .5 (B.,.„)

13

1 J / o

** Weighted mean = .5 (H3055) + .5 (R1016)

surfaces. However, as cautioned by Singh and Taillefer (1986), this

value of a can increase appreciably in the presence of horizontal

warm air advection.

The term (Q*-QG) was derived from the Canadian 1951-1980 normals

data of global solar radiation (K4-) as described by Stewart (1983).

Similarly the term S was calculated from the Canadian normals (1951-

1980) air temperature data. In consequence the normals values of

Elake a r e identical for both scenarios A(GFDL) and B(GISS).

For the projected changes in Eiaire, Stewart (1983) assumed that

(Q* - QG) remained unchanged and that only S changed as a function

of changes in air temperature as predicted by both the GFDL and GISS

scenarios.

However the Stewart (1983) calculations of ^iake a c e ideated in

the middle of the soil class units for Canada (Clayton et al, 1977).

Impact of climate change on hydro-electric generation 409

Table 4a Annual input parameters for deriving unadjusted net basin

supply (NBS) and correction factor from extrapolated grid

points for Eastmain-Opinaca basin: Scenario A (GFDL);

normals data

Extrapolated

Grid Points

A6-10/10

A6-11/10-11

Land

Discharge

(m3/S)

384.1

486.0

435.1*

^ake Preci-

pitation

( mm)

801.5

898.8

850.2*

Lake Evaporation

Soil Unit

(mm)

D3124

R1037

476.0

459.7

467.9**

Unadjusted

Mean NBS

(m3/S)

542.9

jnderes tima te

of Mean

Discharge

7.

36.2

Correct ion

Factor

1.567

*

Weighted mean = .5 (A6_10/10> +

.5 ( A 6 _ n/10 _ u )

** Weighted mean = .5 (D3124) + .5 (R1037)

Table 4b Annual input parameters for deriving unadjusted net basin

supply (NBS) and correction factor from extrapolated grid

points for Eastmain-Opinaca basin: Scenario B (GISS);

normals data

Extrapolated

Grid Points

B12/17

B12/18

Land

Discharge

(m3/S)

387.1

415.1

401.1*

Lake Preci-

pitation

(mm)

806.4

831.1

818.8*

Lake Eva

Soil Unit

D3124

R1037

>oration

(mm)

476.0

459.7

467.9**

Unadjusted

Mean NBS

(m3/S)

500.0

Underestimate

of Mean

Discharge

7.

41.2

Correct ion

Factor

1.702

*

Weighted mean = .5 (Bi-wjo) +

-5 ^B12/17'>

** Weighted mean = .5 (D3124) + .5 (R1037)

As is evident in equation (4) however, the soil characteristics have

no influence on the values of Elake

. Because our grid points for

E-t k

were tied to the soils classification, we chose the dominant

soilSclasses within each basin (Figure 1 and Figure 2) and calculated

a weighted mean for Elake

based on the relative areas of these soil

class units. Again it must be emphasized that the weightings do not

reflect soil characteristics but moreso changes in (Q*-QG) and in air

temperature.

The normals value of Eiake

as shown in Tables 2a, 2b, 3a, 3b and

4a, 4b however seem a bit high for this region when compared to

results found for the Great Lakes region (Cohen, 1986a; Howe et al,

1986).

As a verification of the Priestley - Taylor E]_ake calculations, we

calculated Eiàke using a mass-transfer type equation, that is based

410 B. Singh

on the temperature dependent mean monthly vapor pressure gradient

(Pa) between the lake surface and the air and the mean monthly wind

speed (km/hr) (Richards and Urbe, 1969; Quinn and Den Hartog, 1981)»

These were done for stations where the necessary lake water and air

temperature and wind speed were available, namely La Grande Rivière

(53°N and 77.5°W) in the La Grande basin and Nitchequon (54.5°N and

70.8°W) which is located within the Capiapiscau basin. Unfortunately

the data set is limited by the short time-period of water temperature

data: 1978-1984 for La Grande Riviere and 1980-1984 for Nitchequon,

which for all intents and purposes relates to the normals (1951-1980)

period.

Table 5 shows the total annual and average totals of E^g^g using

the mass - transfer method. The average totals of E^g^e , though

somewhat influenced by lower than average values in 1983, are in

general 1 ower than the Priestley-Taylor E^gj^g calculations shown in

Tables 2a, 2b, 3a, 3b,4a and, 4b.

Table 5 Mass-transfer evaporation calculations (mm/year)

La Grande Rivière

Total Annual

Year

Evaporation (mm)

1978

582.9

1979

523.5

1980

498.1

1981

489.7

1982

504.4

1983

291.9

1984

501.9

476.9

Nitchequon

Total Annual

Year

Evaporation (mm)

1980

495.6

1981

424.4

1982

458.7

1983

307.6

1984

299.5

397.2

Our calculations of NBS for the normals (1951-1980) period for

both scenarios using equations (1) and (3) however, were substantial-

ly lower than the mean measured discharge (SEBJ, 1986) for the 1949-

1980 period. These under-estimations ranged from 32.9% for the Cania-

piscau basin for scenario A to 43.1% for the La Grande basin for

scenario B (Tables 2a, 2b, 3a, 3b and 4a, 4b).

We suspect that these under-estimations are due partly to the fact

that our assumed water holding capacity (WHC = 100mm) is too high in

the calculation of land discharge by the Thornthwaite method since

substantial portions of these basins are covered by bare rock sur-

faces and partly due to the slight over estimation of Ei^g by the

Priestley-Taylor method. Also our linear method of extrapolating grid

points could have biased our results.

In order to correct for these aberrations, we calibrated the

values of annual NBS by introducing a correction factor, which is

simply the ratio of the mean measured discharge (1949-1980) to that of

Impact of climate change on hydro-electric generation 411

the non-adjusted NBS values for both scenarios (Tables 2a, 2b, 3a,

3b and 4a,4b). These correction factors ranged from 1.491 for scena-

rio A for the Caniapiscau basin to 1.757 for scenario B for the La

Grande basin (Tables 2a, 2b, 3a, 3b and 4a, 4b).

These same correction factors, derived using the normals (1951-

1980) data for both scenarios (A and B) were also applied to the

projected annual NBS values for both scenarios (A and B) so as to

derive changes in annual NBS resulting from climate changes (Tables

6a, 6b, 7a, 7b and 8a, 8b).

Table 6a Annual input parameters for deriving unadjusted and

adjusted mean net basin supply (NBS) from extrapolated

grid points for La Grande basin: Scenario A (GFDL)

projected data

Extrapolated

Grid Points

A6/10

A6/ll

Land

Discharge

(ra3/S)

398.4

493.9

455.9*

Lake Preci-

pitation

(mm)

947.2

1034.2

999.4*

Lake Evap

Soil Unit

D3023

R1037

oration

(mm)

806.9

493.9

619.1**

Unadjusted

Mean NBS

(m3/S)

1382.5

Correction

Factor as

per Normals

1.514

Adjusted

Bean NBS

(m3/S)

2073.7

Weighted mean = .6 (A6/11) + -4

(6/10)

Weighted mean = .6 (R1037) + .4 (D3023)

Table 6b Annual input parameters for deriving unadjusted and

adjusted man net basin supply (NBS) from extrapolated

grid points for La Grande basin: Scenario B (GISS)

projected data

Extrapolated

Grid Points

B12

B12/13-18

Land

Discharge

(m3/S)

368.5

442.7

413.1*

Lake Preci-

pitation

(mm)

837.2

921.8

888.0*

Lake Evap

Soil Unit

D3023

R1037

oration

(mm)

820.7

507.8

633.0**

Unadjus ted

Mean NBS

(m3/S)

1185.9

Correction

Factor as

PER Normals

1.757

Adjusted

mean NBS

(m3/S)

2083.5

*

Weighted mean = .6 ( B I-WIT IQ)+

-^ ^B i?^

** Weighted mean = .6 (R1037) + .4 (D3023)

Since the same correction factors are applied for the normals and

projected values for both scenarios, cateris paribus, the comparative

effects should not be affected.

412 B. Singh

Table 7a Annual input parameters for deriving unadjusted and

adjusted mean net basin supply (NBS) from extrapolated

grid points for Caniapiscau basin: Scenario A (GFDL)

projected data

Extrapolated

Grid Points

A6/ll

A7/ll

Weighted mean

Land

Discharge

(m3/S)

493.9

552.5

523.2*

Lake Preci-

pitation

(mm)

1034.2

996.6

1015.4*

Lake Evaporation

Soil Unit

(mm)

H3055

R1016

483.3

503.2

493.3**

Unadjusted

mean NBS

(m3/S)

611.5

Correction

Factor as

per Normals

1.491

Adjusted

mean NBS

(m3/S)

911.8

*

Weighted mean = .5 (6/11) + .5 (7/11)

** Weighted mean = .5 (H3055) + .5'(R1016)

Table 7b Annual input parameters for deriving unadjusted and

adjusted mean net basin supply (NBS) from extrapolated

grid points for Caniapiscau basin: Scenario B (GISS)

projected data

Extrapolated

Grid Points

B13

B13/18

Eeighted mean

Land

Discharge

(m3/S)

429.4

516.9

473.2*

Lake Preci-

p'itation

(mm)

885.3

1006.4

945.9*

Lake Evaporation

Soil Unit

(mm)

H3055

R1016

497.7

517.9

507.8**

Unadjusted

mean NBS

(m3/S)

544.8

Correction

Factor as

per Normals

1.649

Adjusted

mean NBS

(m3/S)

898.4

*

Weighted mean = .5 (B ,) + .5 (B,,...)

** Weighted mean = .5 (H3055) + .5 (R1016)

Results and discussion

Ignoring the presence of reservoirs for the moment, Tables 6a, 6b,

7a, 7b, and 8a and 8b show significant increases in NBS for all three

drainage basins, for both scenarios A and B„ These increases, range

from 6*8% for scenario A to 20*3% for scenario B in the Opinaca-

Eastmain basin. For the other two drainage basins, namely La Grande

and Caniapiscau, the percent increases in NBS are quite similar for

both scenarios: 17,7% (GFDL) and 18.3%(GISS) for La Grande and 15.7%

(GFDL) and 14,0% (GISS) for Caniapiscau (Tables 9a and 9b),

By adding the additional free water surface created by the reser-

voirs, these results do not change significantly (Tables 9a and 9b),

For scenario A, the increase in NBS is reduced from 17,7% to 16,5%

Impact of climate change on hydro-electric generation 413

Table 8a Annual input parameters for deriving unadjusted and

adjusted mean net basin supply (NBS) from extrapolated

grid points for Eastmain-Opinaca basin: Scenario A (GFDL)

projected data

Extrapolated

Grid Points

A6-10/10

A6-ll/10-ll

Weighted mean

Land

Discharge

(m3/S)

409.0

504.6

456.8*

Lake Preci-

pitation

(mm)

929.8

1016.8

973.3*

Lake Evaporation

Soil Unit

(mm)

D3124

R1037

569.0

493.4

531.2**

Unadjusted

mean NBS

(m3/S)

579.8

Correct ion

Factor as

per Normal:

1.567

Adjusted

mean NBS

(m3/S)

908.5

*

Weighted mean = .5 (A6 _10/1Q ) +

.5 (Afi_n / 1 Q _ U )

** Weighted mean = .5 (D3124) + (R1037)

Table 8b Annual input parameters for deriving unadjusted and

adjusted mean net basin supply (NBS) from extrapolated

grid points for Eastmain-Opinaca basin: Scenario B (GISS)

projected data

Extrapolated

Grid Points

B12/17

B12/18

Weighted mean

Land

Discharge

(m3/S)

461.7

486.4

474.1

Lake Préci-

pitât ion

(mm)

963.8

982.4

973.1

Lake Evaporation

Soil Unit

(mm)

D3124

R1037

523.5

507.8

515.7

Unadjusted

Tiean NBS

(m3/S)

601.4

Correction

Factor as

per Normals

1.702

Adjusted

mean NBS

(m3/S)

1023.6

*

Weighted mean = .5 (B 2/17 ) + .5 (B^/18)

** Weighted mean = .5 (D3124) + .5 (R1037)

for La Grande basin, decreases from 6.8% to 6.7% for Eastman-Opinaca

basin and remains unchanged at 15.7% for Caniapiscau basin. On the

other hand for scenario B, the increase in NBS is reduced in all

basins: 18.3% down to 15.6% for La Grande basin, 14.8% down to 13.0%

for Caniapiscau basin and 20.3% down to 20.2% for Eastman-Opinaca

basin.

These minor reductions in NBS increases, caused by the addition of

reservoirs, is most likely due to the fact that the decreases in land

area discharge are greater than the increase in total lake and reser-

voir evaporation. This is not surprising for these drainage basins

that are covered by bare rock surfaces over significant parts of

their surface area.

The values of the change in NBS project an increase that ranges

414 B. Singh

Table 9a Normal and projected NBS and percent change in NBS for all

drainage basins: Scenario A (GFDL)

Drainage

Bas in

La Grande

Caniapiscau

Eastmain—Opinaca

Adjusted

(

Normals

1761.9

787.8

850.7

NBS without reservoirs

m3 /S)

Projected

2073.7

911.8

908.5

Percentage

Increase (V*)

17.7

15.7

6.8

Adjus te

(m

Norma Is

1761.9

787.8

850.7

d NBS with reservoirs

3/s)

Projected

2051.8

911.8

907.9

Percentage

Increase (%;

16.5

15.7

6.7

Table 9b Normal and projected NBS and percent change in NBS for all

drainage basins: Scenario B (GISS)

Drainage

Basin

La Grande

Caniapiscau

Eastmain-Opinaca

Adjusted

(

Normal s

1761.0

787.9

851.0

NBS without reservoirs

m3/S)

Pro jected

2083.5

898.4

1023.6

Percentage

Increase (7.)

18.3

14.0

20.3

Adjuste

(m

Normals

1761.0

787.9

851.0

d NBS with reservoirs

3/s)

Pro jected

2035.0

890.6

1022.7

Percentage

Increase (%)

15.6

13.0

20.2

from 6.7% (Eastman-Opinaca: Scenario A) to 20.2 percent (Eastmain-

Opinaca: Scenario B). On the other hand comparable studies in the

Great Lakes region (Cohen, 1986, 1987; Sanderson et al, 1985; Howe

et al., 1986) project a decrease in NBS of the order of 10 percent.

Th"ë explanation for this difference seems to be related to the very

high projected increases in precipitation that are of the order of 30

to 40 percent in July and January for both scenarios in these basins.

This order of precipitation increase masks and even over-rides the

projected increases in temperature and hence lake evaporation. In the

Great Lakes area both precipitation and temperature increases were

smaller, and in the case of precipitation there was even a decrease

in some areas. Also our methods and results, though lacking in some

areas seem well founded and consistent. The consistency of our re-

sults are shown in Table 10 where it is shown that there is little

change in the ratio of net basin supply to total basin precipitation.

This reflects the concurrent increases in discharge, resulting from a

net increase in precipitation minus evaporation, and in

precipitation.

Impact of climate change on hydro-electric generation 415

Table 10 Annual Net Basin Supply/total annual precipitation ratios

(%) from measured, normals and projected data

Drainage Basin

La Grande

Caniapiscau

Opinaca-Eastmain

Measured

Mean Dis-

charge/

Prec. (7.)

77*

77**

84*

Adjusted Nee Basin Supply / Precipitation ratio (%)

Scenario A (GFDL)

Scenario B (GISS)

Normals [Projected

70.4

78-8

78.4

66.7

76.8

73.1

Change

-3.7

-2.0

-5.3

Normal s

75.4

79.7

81.4

Projected

75.5

81.2

82.4

Change

+ 0.1

+ 1.5

+ 1.0

Socio-economic costs/benefits

According to Mclntyre (1984) and Sanderson et al (1986) there seems

to exist a matching relationship between NBS and the average amount

of energy generated in the Great Lakes - St. Lawrence River Basin. If

these same relationships were to apply to the James Bay area, then

according to Tables 9a and 9b, the hydro-electric generating poten-

tial should increase by 15.6 percent (GISS) to 16.5 percent (GFDL)

for the La Grande basin, by 13.0 percent (GISS) to 15.7 percent

(GFDL) for the Caniapiscau basin and by 6.7 percent (GFDL) to 20.2

percent (GISS) for the Eastman-Opinaca basin.

The present potential generating capacity of the three drainage

basins, namely the La Grande, the Caniapiscau and the Opinaca-

Eastmain, and their sum total are shown in Table 11. If one were to

assume an equivalent percentage change between net basin supply NBS

and generating capacity, then according to Table 11, the hydro-

electric generating potential of the La Grande basin would increase

by 5.0 TW/h (GISS) to 5.3 TW/h (GFDL), that of the Opinica basin by

2.7 TW/h (GISS) to 3.3 TW/h (GFDL) and that of the Caniapiscau basin

by 0.6 TW/L (GFDL) to 1.8 TW/h (GISS). The total hydro-electric

production potential of all three basins together should therefore

increase by 9.2TW/h (GFDL) to 9.5TW/h (GISS).

Since warmer conditions and higher air-conditioning demands are

projected for south-eastern Canada and for the Northeast United

States, and since air conditioners generally function on electricity,

the potential increased supply of hydro-electricity from the James

Bay area will go towards satisfying this increased demand to the

south. Of course, this will be offset by a decrease in winter heating

requirements.

Also, since electricity is generally one of the cheaper forms of

energy, increased hydro-electricity generating potential as caused

by C02

- induced climatic change in the James Bay area would allow

for substitution of more expensive forms of energy such as coal and

oil by less expensive hydro-electricity. These different supply and

demand changes and consequent substitution effects will have to be

416 B. Singh

Table 11 Change in net basin supply (NBS) for the sum total of all

three drainage basins and change in hydro-electric

generating capacity

Drainage Basin

La Grande

Caniapiscau

Eastmain-Opinaca

Total

'resent

aenera-

ng ca-

pacity

(TU/hJ*

32.2

20.9

9.0

62.2

3

NBS (m / S ) S c e n a r i o A (GFDL)

Normals

1761.9

787.8

850.7

3400.4

Pre iected

2051.8

911.8

907.9

3871.5

Change

an NBSX

16.5

15.7

6.7

13.9

PnjectBJ

gen. cap.

(TW/h)

37.6

24.2

9.6

71.4

NBS (m / S ) S c e n a r i o B ( G I S S )

Change in

gen. cap.

(TW/h)

5.3

3.3

0.6

9.2

Normals

1761.0

787.9

851.0

3399.9

Projected

2035.0

890.6

1022.7

3948.3

Change

iMBS*

15.6

13.0

20.2

16.1

'rejected

gen. cap.

(TW/h)

37.3

23.6

10.8

71.7

Change in

gen. cap.

(TW/h)

5.0

2.7

1.8

9.5

Source: M. Tram-Van, SEBJ, personnal communication.

quantified so as to derive economic cost/benefit figures. The

comparative advantage of the James Bay area, with respect to other

supply regions of hydro-electricity such as the Great Lakes region

and to changes in demand patterns in the south will also have to be

taken into account.

Conclusions and recommendations

C02

- induced climate change as projected by both the GFDL (scenario

A) and GISS (scenario B) models will supposedly increase NBS by the

order of 6.7 to 20.2 percent for the three drainage basins considered

in the James Bay territory, namely the La Grande, the Caniapiscau and

Opinaca-Eastmain basins.

This increase in net basin supply should increase the hydro-

generating potentials of these drainage basins by about 9.2TW/h

(GFDL) to 9.5 TW/h (GISS). These increased generating potentials

should provide important economic advantages for the province of

Quebec, since hydro-electricity, being a relatively cheap energy

form, would replace the other more conventional and costly forms of

energy such as coal and oil.

Warmer conditions that are projected for south-east Canada and

north-east United States should allow for increased electricity

demand for cooling in the summer but for less electricity demand for

heating in the winter. However milder winters should favor electrici-

ty as a heating agent.

For subsequent phases of this segment of the research, it would

however be advisable to increase the spatial resolution of the global

circulation models (GCM's) so as to be able to better assess climate

change impacts over smaller spatial distances. Also, if extrapolation

Impact of climate change on hydro-electric generation 417

has to be resorted to, a more objective method than mere linear ex-

trapolation should be devised.

A more accurate estimate of the water holding capacity (WHC) of

the soils of this region should also be made for calculating land

dischargee For the calculation of evaporation, more valid assumptions

should be made with respect to the possible changes in net radia-

tion (Q*) resulting from global climate change.

Finally the economic variables involved such as increased

electricity generating potential, higher summer-time and lower winter

time energy demand together with substitution effects and regional

changes in comparative advantage should be quantified and modelled so

as to derive more precise economic costs or benefits.

ACKNOWLEDGEMENTS

This paper eminates from contract funds from the

Quebec region of the Atmospheric Environment Service (SEA) of

Environment Canada. I also wish to acknowledge M. Richard Gilbert and

M. Gerald Vigeant of SEA for their prompt and generous support in

matters relating to contract details and data sources. I also

sincerely thank Dr. S. Cohen (Environment Canada) for providing

guidance and copies of the Thornthwaite Monthly Water Budget and of

the mass transfer programs; Dr. R.B. Stewart and his assistant

R. Muna for providing the Priestley-Taylor evaporation calculations,

M. Thack Tran-Van of SEBJ for the hydrologie and hydro-electric

generation data, Mme Lynn Gregorie for data analysis, M. Guy

Frumignac for the cartography and Mme Joëlle Casamajou for typing the

manuscript.

References

Bruce, J.P. (1984) Great Lakes levels and flows: past and future.

Journal of Great Lakes Research, 10 (1), 126-134.

Clayton, J.S., W.A. Ehrlich, D.B. Cann, J.H. Day, I.B. Marshall

(1977) Soils of Canada. Vols 1 and 2. Research Branch. Canada

Department of Agriculture, Ottawa, 239 p.

Cohen, S.J. (1986) Impacts of C02_induced climatic change on water

resources in the Great Lakes basin. Climatic change, 8 , 135-153.

Cohen, S.J, (1987) Climatic change, population growth, and their

effects on Great Lakes water supplies Professional Geographer, 38

(4), pp. 317-323.

Hengeveld, H.G. and R.B. Street (1986) Development of C02 climate

change scenarios for Canadian regions. Unpublished Paper.

Howe, D.A., D.S. Marchand. C. Alpagh and P.K. Stokoe (1986) Socio-

economic assessment of the implications of climatic change for

commercial navigation and hydro-electric power generation in Great

Lakes-St Lawrence River System. Special Report for Atmospheric En-

vironment Service, Environment Canada,122p.

Johnstone, K.J. and P.Y.T. Louie (1984) An operational water budget

for climate monitoring. Unpublished Manuscript, Canadian Climate

Centre Report No. 84-3, Downsview, Ontario, 52 p.

Mclntyre, J. (1984) Water-existing and possible future economic value

for hydro-electric generation. In proceedings, Ontario Water

Resources Conference, Futures in Water, June 12-14, 1984, Toronto,

418 B. Singh

Ontario.

Priestley, C.H.B. and R.J. Taylor (1972) On the assessment of surface

heat flux and evaporation using large-scale parameters. Monthly

Weather Review, 100, 81-92.

Quinn, F.H. and F. den Hartog (1981) "Evaporation Synthesis" in E.J.

Albert and T.L. Richards(eds) IFYGL- The International Field Year

for the Great Lakes, U.S. Dept. of Commerce, Ann Arbor, 221-245.

Richards, T.L. and J.G. Irbe (1969) Estimates of monthly evaporation

losses from the Great Lakes, 1950 to 1968, based on the Mass

Transfer Technique. Proceedings Twelfth Conference on Great Lakes

Research, International Association of Great Lakes research, 469-

WT.

Sanderson, M.E., T. Choi, D.A. Howe, D.S. Marchand and P.K. Stokoe,

(1985) Socio-economic assessment of the implications of climatic

change for future water resources in the Great Lakes/St. Lawrence

river system. Hydro-electric power generation and commercial

navigation. DSS Contract NO. 02 SE.KM147-4-1414. 127 p.

SEBJ. La Grande Complex-Hydrological Studies. Unpublished paper.

Singh, B., R. Taillefer and J. Poitevin (1984) Les échanges radiatifs

et énergétiques et le bilan thermique du sol en Jamesie. Canadian

Geotechnical Journal, 21, 223-240.

Singh, B. and R. Taillefer (1986) The effect of synoptic-scale advec-

tion on the performance of the Priestley-Taylor evaporation

formula. Boundary Layer Meteorolgy, 36, 267-282.

Singh, B., L. Gregorie, D. Skiadas, G. Renaud, A. Viau and B. Cairns.

(1987) Prospective d'un changement climatique our les ressources

naturelles du Quebec. DSS Contract No. S-102-1-5510-4210-0000-101-

1100, 291p.

Stewart, R.B. and W.R. Rouse (1977) Substantiation of the Priestley

and Taylor parameter a = 1.26 for potential evaporation in high

latitudes. Journal of Applied Meteorology, 16, 649-650.

Stewart, R.B. (1983) Modelling methodology for assessing crop

production potential in Canada. Research Branch, Agriculture

Canada,, Contribution 1983-12E, 29p.

Tran-Van, T.

(1987) SEBJ, Service

Hydraulique, Personal

Communication.

~p Climatic change and

- H water resources

T J systems

The Influence of Climate Change and Climatic Variability on the Hydrologie

Regime and Water Resources (Proceedings of the Vancouver Symposium,

August 1987). IAHSPubl. no. 168, 1987.

Climate change and water resources

A.J. Askew

World Meteorological Organization

Geneva, Switzerland

ABSTRACT

The consideration of the impact of climate on

water resources is placed in a wider context: the chain

reaction associated with such an impact crosses the

interfaces between climate, hydrology, water-resource

systems and society. Each interface demands collaboration

and exchange of information between specialists and

assurance that current tools and approaches are

appropriate for the new environment that may evolve. The

chain reaction must be studied from start to finish if the

final impact on water resources is to be assessed. Various

points are presented for consideration at each stage.

Impact du climat sur les ressources en eau

RESUME

La considération de l'impact du climat sur les

ressources en eau est placée dans un contexte plus large:

la réaction en chaîne associée à un tel impact comprend

les domaines communs au climat, à l'hydrologie, aux

systèmes de ressources en eau et à la société. Chaque

domaine commun nécessite la collaboration et l'échange

d'informations entre les spécialistes et l'assurance que

les outils et approches actuels soient appropriés au

nouvel environnement susceptible d'évoluer. La réaction en

chaîne doit être étudiée du de'but à la fin si l'impact

final sur les ressources en eau doit être évalué.

Différents points sont présentés pour être considérés a

chaque stade.

Introduction

The last ten to twenty years have seen a major change in the general-

ly accepted view of climate and of climatology. In the past,

considerations of climate concentrated on long-term steady-state

conditions. Variations in space were studied and explained, but

variations in time were seen more as interesting historical phenomena

than as anything of particular importance for mankind in the present

age.

This all changed in a remarkably short period of time when, in the

1970's, a greatly improved capability for modelling the global

atmosphere combined with increasing evidence and concern for the

421

422 A.J. Askew

consequences for the atmosphere of man's activities. Some preliminary

findings hinted at major changes in climate under certain conditions,

others led to quite different conclusions. At times the former

received considerable publicity and the scientific community

recognized the need for broadly-based programmes of research and

evaluation on the whole subject of climate and the potential impact

of its variability and change on society. Many countries have

established national climate programmes for this purpose. At the

international level, the World Meteorological Organization (WMO)

convened the World Climate Conference in February 1979 (WMO, 1979),

with the support of UNEP, FAO, Unesco and WHO. This led later in the

same year to the establishment of the World Climate Programme whose

overall objectives are:

(a) to apply existing climate information for the benefit of

mankind,

(b) to improve the understanding of climate processes,

(c) to monitor significant climate variations and changes.

The existence of climate variability has long been recognized,

including even the possibility of quite significant variations over

comparatively short periods of time. However, without any real means

of predicting such variations, they were not seen as being of any

great relevance to regional or national planning. Our ability to

predict variations is still very limited and any predictions that are

made are hotly debated. It is interesting to speculate on the cause

of the current interest in climate variability and change. Certainly,

the fact that man himself may be the cause of such variations, even

changes, has given the whole question far greater importance and

urgency.

If we had perfect knowledge and foresight of the situation, we may

find that variations and changes in climate over the next 200 years

are neither greater nor more sudden than those experienced during the

last 200 years. It is theoretically possible, therefore, that current

analytical and planning techniques are adequate for generations to

come and there is no real need for concern or for major changes in

the way we manage affairs. However, we have neither perfect knowledge

nor perfect foresight. What is more, there is a very real possibility

that man's actions will have significant negative effects on the

climate. These we should try to predict, plan for and, above all,

prevent wherever possible. Therefore, while the reasons behind the

current concern over climate change and variability may be studied by

those interested in the history and philosophy of science, society

has every right to expect scientists, engineers and planners to take

the subject itself with the utmost seriousness.

The basic needs of mankind are commonly taken to include food and

drink, clothing and shelter, and security against physical harm. The

provision of food and drink requires an adequate supply of water and

security demands protection from flooding as well as from other

threats to safety. Some of the most important, in fact probably the

most important, impacts on society of climate variability and change

are introduced through the water cycle. The hydrologist and the

water-resource engineer therefore have a major role to play in

studying and planning for these impacts and it is very important that

such work be undertaken in a co-ordinated fashion at national and

international levels.

Climate change and water resources 423

Climate variability and change

A whole range of definitions can be presented with reference to

climate variability and change. The following have been found useful

in developing plans for water-related projects under the WCP (WMO,

1985a):

"Weather is associated with the complete state of the atmosphere

at â particular instant in time and with the evolution of this

state through the generation, growth, and decay of individual

disturbances."

"Climate is the synthesis of weather over the whole of a period

essentially long enough to establish its statistical ensemble

properties (mean values, variances, probabilities of extreme

events, etc.) and is largely independent of any instantaneous

state."

"Climate change defines the difference between long-term mean

values of a climate parameter or statistic, where the mean is

taken over a specified interval of time, usually a number of

decades."

"Climate variability includes the extremes and differences of

monthly, seasonal and annual values from the climatically expected

value (temporal mean).

The differences are usually termed

anomalies."

Water-resource systems

The simple existence of a body of water does not define it as a

"water resource". For it to be a "resource" it must be available, or

capable of being made available, for use in sufficient quantity and

quality at a location and over a period of time appropriate for an

identifiable demand. There is therefore an important distinction to

be drawn between hydrology and the study of water resources.

A water resource may already be used or it may represent only a

potential for the future. In either case, its current and future

reliability are important factors and a firm prediction of a future

reduction in quantity or reliability may deny the use of the term

"water resource". In this sense, therefore, a future climate change

would not only affect the magnitude or reliability of existing water

resources but it would also make available resources that had not

previously been considered as such or result in the total loss of

many existing resources.

The hydrologie cycle is an integral part of the climate system and

is therefore involved in many of the interactions and feed-back loops

which give rise to that system's complexity. Water-resource systems

represent man's intervention in and use of the hydrologie cycle for

his own benefit. Even the simplest are subject to a number of

external influences each offering avenues for the impact of climate

change. They contain many interactions within them and, on a local

scale, can significantly modify the hydrologie cycle. Therefore,

while water-resource systems may be viewed principally as the

recipients of climate impacts, through the intermediary of the

hydrologie cycle, they may themselves have an impact on climate,

particularly where they are very large in scale or in number or where

424 A.J. Askew

the hydrologie cycle is in a delicate state of balance.

Any investigation of the interaction between the climate system,

hydrologie cycle and water-resource systems must recognize the

complexity of the relationships involved. As regards water-resource

systems, one approach would be to first identify the various elements

involved in each system and then to study the climatologie and

hydrologie factors influencing the design and efficiency of operation

of each such element. A systematic presentation of the various types

of water-resource system and the hydrologie factors and techniques

involved in the design and operation of each type was compiled under

the auspices of the International Hydrological Programme of Unesco

(1982). The emphasis of WMO's Operational Hydrology Programme is on

the requirements of each type of project for hydrologie and

climatologie data. These were considered by Andrejanov (1975) over

ten years ago and a recent report (WMO, 1987) contains extensive

tabulations setting out such requirements.

At this point in time, the principal demand is not for precise

estimates of the potential impact of climate change on specific water

projects, but for indications of the general nature and extent of

such impacts on various types of water project. Each hydrologie

factor of relevance and each data requirement represent an avenue by

which climate change might have an impact on the type of project in

question. Therefore, reports such as those referred to above offer a

good basis for identifying not only the types of project to be

considered but also the climatologie elements and their characteris-

tics which are likely to be important in a study of the impact of

climate change.

Needs for information on climate change

The distinction drawn above between "variability" and "change", while

clear in principle, is by no means easy to apply in practice.

Variability is not so difficult to recognize and assess, but in order

to study the effect of climate change we must first be able to

distinguish change from variability. The progress being made in this

regard by both climatologists and hydrologists is to be noted and

they should be encouraged to continue with their important work.

Likewise, encouragement should be offered to those climatologists

and atmospheric physicists who study the climate system and seek to

explain its variability and past changes and to predict its future

behaviour. Their advice is vital to those concerned with the impact

on water resources for, without some indication of likely future

changes in climate, any discussion of impacts will be restricted to

theory and be of limited practical value. However, the water-resource

engineers and planners must be prepared to state clearly what advice

they require. The climatologists and atmospheric physicists have

legitimate interests of their own and, in addition, many other groups

make requests of them for specific information to satisfy their own

particular needs. It is not enough to express dissatisfaction with

the form or content of current climate predictions, the hydrologist

and water-resource engineer must clearly define their own needs and

make them known by appropriate means.

This list of needs is likely to encompass such parameters as:

Climate change and water resource? 425

(a) Space scale: global; hemispherical; continental; regional;

national; sub-national.

(b) Time scale: hundreds of years; tens of years; annual; seasonal;

monthly.

(c) Elements: radiation (at ground surface); temperature; wind

speed; precipitation; humidity.

(d) Characteristics: mean; variance; skew; probability of extreme

values; spatial and temporal correlation; cycles; trends; abrupt

discontinuities.

At present climate predictions concentrate on changes in the mean

values of a few selected elements over periods of ten to twenty years

on a global or hemispherical basis. It is not at all clear at first

sight what characteristics of what elements are likely to be of

greatest significance with regard to the impact on water resources.

Precipitation is a prime candidate but, unfortunately, it is the

element about which climate predictions say the least and say it with

the least confidence. Similarly, while abrupt changes or trends in

mean values are important, increased or decreased dispersion and/or

probabilities of extreme values are likely to be far more critical.

The mean precipitation and temperature may remain unchanged even

within each season, but a modest increase in their variability could

throw doubt on the viability of certain rain-fed crops or run-of-the

river hydropower schemes and could leave major reservoirs either

empty or threatened by floods which are greater than those for which

their spillways were designed.

If the types of water-resource projects are identified together

with the climate characteristics that are significant for each then a

list of climate prediction requirements might be drawn up. It may

not be such a difficult task for an individual with a particular

interest, but it will not be easy to obtain consensus on the matter

among a group of experts viewing the problem from various national or

regional perspectives. It is important that any such list be as

short as possible and those concerned must realize that it will be

many years before even a preliminary response can be made on some

items. Despite these difficulties, it is important that this task be

undertaken in an appropriate context at an early date. If it is not,

then the climatologists and atmospheric physicists will not be able

to take due account of hydrology and water resources in their

investigations and the most important channel for climate impact may

be poorly treated.

One last, but important, comment on this question: the identifica-

tion of needs is likely to be a dynamic and iterative process. As

each climate prediction is studied and as the likely impact on a

particular type of water-resource system is assessed, new informa-

tion about the sensitivity of the system to such changes will be

acquired which will often lead to a request for more detailed

predictions or predictions concerning additional parameters. The

list of needs will therefore evolve and change with time, but until a

first draft list is established, little progress can be made.

The impact of climate change on hydrology

The impact of climate change follows a chain reaction. As already

426 A.J. Askew

noted, the hydrologie cycle is part of the climate system and so the

first link in the chain is the impact on hydrologie processes. Here

the obvious approach is to use the hydrologie model as the basic

tool, adjust various parameters and inputs to simulate climate change

and study the model's response both as regards its state variables

and output. Even given a specific prediction scenario, it is not

easy to decide what adjustments to make to what; it is even more

difficult to interpret the results when the scenarios used are as

speculative as they are at present. Nevertheless the work that has

already been done in this regard (e.g. Nemec and Schaake, 1982) is of

great importance in that it has awakened interest in the hydrologie

community to the whole subject and has laid the groundwork for future

studies.

The more precise one wishes to be in any investigation, the more

one should question the qualities and appropriateness of the tools

used. Where the tool is inadequate for the task, the results can be

of little value. Klemes (1985) has set very exacting standards for

hydrologie models if they are to be used in this context. It is to be

hoped that model developers will apply such tests so that any results

obtained by using a model may be judged against the results of his or

similar tests.

Alternative approaches which do not make use of hydrologie models

have also been proposed. For example, studies in comparative

hydrology are seen as offering a potential source of information. As

with all analyses which trade variations in space for variations in

time, these should be approached with caution. They can illustrate

the types of equilibrium states that have been established in the

past under various external influences. The question to be asked is

how much they can tell us about the dynamic response of hydrologie

systems when they are subjected to changes in time in climatic

factors.

This first link in the chain also involves the interface between

the atmosphere and the land surface. Inadequate modelling of the

water and energy balances at this interface are held to be one of the

factors which currently limit the performance of general atmospheric

circulation models. It is on these models that many of our hopes

depend for improved climate predictions. As hydrologists seek better

predictions of climate variability and change, they should therefore

be prepared to make a substantial input to the work of those who

model the atmosphere. The basis for such collaboration has already

been established and the work commenced (Eagleson, 1982; WMO, 1985b):

a most welcome sign.

Hydrologie processes and water-resource systems

The second link in the chain is that between hydrologie processes and

water-resource systems. As the systems are man-made it is easy to

see this as a simple matter. The anticipated variations and changes

in the climatologie and hydrologie processes can be introduced in the

relevant parameters and time series inputs to mathematical models of

the systems. The impact on their performance may then be investigated

on the basis of the outputs and general system response. However, as

mentioned earlier, the interaction between water-resource systems,

Climate change and water resources 427

climate and hydrology may be more complex than at first expected. It

may involve feed-back mechanisms and, in particular, significant

changes in certain processes may cause the system to operate in a

manner totally different from that experienced to date and may even

result in the system being unable to serve any useful purpose. For

example, a moderate increase in the probability of below-freezing

temperatures would have a negligable influence in the hydrologie

regime but could make it impossible to sustain the production of

citrus fruits. Unless an alternative crop could be cultivated, the

irrigation scheme serving the orchards would then be of little future

value. In another situation, a shift in timing of the wet season may

permit rain-fed agriculture where irrigation was previously essential.

The more obvious examples are long term trends in mean values or

changes in probabilities of extremes which could result in empty

reservoirs or increases in water logging or the threat of flooding.

The interface between water-resource systems and the natural

processes within which they are embedded is also complicated by the

fact that, while the systems are man-made and hence reasonably

clearly defined in physical terms, the manner in which they are

operated is rarely so well understood and can change or be changed

with considerable ease. The operating policy is one of the principal

characteristics of a water-resource system. For some systems it may

be possible to amend the policy so as to counteract or even take

advantage of climate variability and change. For others, however,

the freedom to change the policy may be very limited. The actual

manner in which water-resource systems are currently operated is the

result of a complex, even stochastic, balance of factors: physical,

socio-economic, political and human. There is no reason to believe

that this will not continue to be the case in the future and this

complexity and uncertainty should always be borne in mind when

evaluating theoretical optimum policies. Despite this, it is vital

that a consideration of operating policies be included as a part of

any water-resource system study.

Water-resource systems and society

The response to predictions and indications of climate change could

be structural, such as an increase in the height of levee banks, the

construction of new storage reservoirs or the installation of

additional turbines. It could also be non-structural: the implementa-

tion of hydrologie forecasting systems to allow more optimal use of

water supplies, revised operating policies for existing systems or

changes in social habits and economic activity.

The third link in the chain, that between water-resource systems

and society, is therefore of great importance and one that is

dominated by feed-back mechanisms. Both the physical characteristics

and the operating policy of a system are designed to meet the

perceived needs of society in one way or another. A change in climate

or a change in its variability could greatly affect these needs.

Domestic and agricultural demands for water would be expected to

increase if temperatures increase, but if this is accompanied by an

increase in precipitation then agricultural demand may fall,

depending on seasonal factors. If a significant change in climate is

428 A.J. Askew

predicted, farmers, industrial managers and the general public may

respond in a manner analogous to the response seen during recent oil

crises, and the net impact on society may be far less than anticipa-

ted. Conversely, a negative or ill-judged reaction could aggravate

the situation and amplify the impact.

In one sense, the water-resource system, with all its imprecision,

sits as a relatively well-defined entity linked on the one hand with

the natural environmental systems of climate and hydrology and on the

other with the socio-economic and political systems of man. This

latter interface, the third in the series, is complex, dynamic,

multi-faceted and, above all, difficult if not impossible to predict

as regards its future characteristics and performance. Much valuable

work has been done on the multi-objective planning of water-resource

systems and various techniques have been developed for rationally

accounting for competing demands expressed in financial terms or in

various measures of public safety and welfare. In theory, these

techniques should hold for the consideration of the impact of climate

change, but in practice it is likely to prove vastly more difficult

to express in concrete terms the desires and limitations of each

sector under the predicted future conditions than under present

circumstances. What will be the priorities of a society which is

faced with a change in climate where this might lead to marked

increases or decreases in temperature and precipitation, in food and

water supplies, in health and safety risks and in the general quality

of life? Where our hydrologie models need to be tested to ensure that

they will yield valid results under conditions beyond those for which

they were originally derived, so too will our socio-economic models.

Strictly speaking, this is outside the subject of this paper, but it

is certainly of relevance to the subject.

Hydrologists have long been concerned that hydrologie forecasts

are adequately disseminated and correctly interpreted so as to ensure

that they are of greatest value. Those who predict the response of

water-resource systems to climate change should be equally concerned

that their predictions could affect the validity of the predictions

themselves. It is essential, therefore, that current efforts to

involve the appropriate water users and decision makers in the

planning and design of water-resource systems be taken much further

in the study, planning and design of systems to respond to the impact

of climate change. Without true dialogue, the predictions could

include grave errors and the plans could prove very ineffective.

Concluding remarks

The purpose of this paper is not to review the current state-of-the-

art in the study of the potential impact of climate change on water

resources. Up until mid-1986 there were not so many published papers

on the subject (Beran, 1986), but the field is gaining in interest

and the papers presented at the current symposium should provide a

good indication as to what has been achieved and what studies are

planned for the future. The aim of the author has been to put the

whole field of study into its wider context, to raise certain

questions that need to be answered and to propose, in some instances,

what approaches might be taken.

Climate change and water resources 429

The past concentration on thirty-year normals has been replaced in

clim

  翻译：

查看mercari原页面

收藏该商品

立即下单