Hydrological Sciences Journal

ISSN: 0262-6667 (Print) 2150-3435 (Online) Journal homepage: http://www.tandfonline.com/loi/thsj20

Regional climate change impacts: I. Impacts on water resources M. A. MIMIKOU & Y. S. KOUVOPOULOS To cite this article: M. A. MIMIKOU & Y. S. KOUVOPOULOS (1991) Regional climate change impacts: I. Impacts on water resources, Hydrological Sciences Journal, 36:3, 247-258, DOI: 10.1080/02626669109492507 To link to this article: http://dx.doi.org/10.1080/02626669109492507

Published online: 29 Dec 2009.

Submit your article to this journal

Article views: 237

View related articles

Citing articles: 9 View citing articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=thsj20 Download by: [37.44.207.102]

Date: 17 January 2017, At: 08:25

Ifydroloffcal Sciences - Journal - des Sciences Hydrolopques, 36,3, 6/1991

Regional climate change impacts: I. Impacts on water resources*

M. A. M M K O U & Y. S. KGUVOPOULOS National Technical University of Athens, Department of Civil Engineering, Division of Water Resources, Hydraulic and Maritime Engineering 5, boon Polytechniou, 157 73 Athens, Greece

Abstract Regional effects of greenhouse warming on water resources, and more specifically on surface runoff, are assessed for a mountainous region of central Greece comprising four drainage basins by using a conceptual model and plausible hypothetical scenarios of temperature and precipitation change. Results show considerable sensitivity of runoff characteristics to climate change and indicate certain basin morphociimatic characteristics such as snow cover, basin aridity and morphology, runoff coefficients etc., which modify considerably basin response. Research work is complemented by assessing the climate change impacts on water management works of the area. This second part of the work is published separately in the following paper in this issue. Impacts régionaux du changement climatique: L Impacts sur les ressources en eau Résumé On évalue les effets régionaux de l'augmentation de la température atmosphérique du à l'effet de serre sur les ressources aquatiques. Plus précisément on étudie les effets sur l'écoulement superficiel, pour une région montagneuse de la Grèce centrale comprenant quatre bassins de drainage. Un modèle conceptuel et des scénarios hypothétiques plausibles pour le changement de la température et des précipitations sont utilisés. Les résultats montrent une sensitivité considérable des caractéristiques de l'écoulement au changement du climat et ils indiquent clairement que certaines caractéristiques morphoclimatiques, comme par exemple la couverture neigeuse, l'aridité et la morphologie du bassin, le coefficient d'écoulement, changent considérablement la réponse du bassin. Ces recherches sont complétées par l'estimation de l'influence du changement de climat sur les opérations de gestion des eaux de la région. La deuxième partie de ce travail est publiée séparément dans ce même numéro dans un autre article.

•Paper presented in Open Session O.II.l: Hydrology at the 15th General Assembly of the European Geophysical Society, in Copenhagen, Denmark, 23-27 April 1990. Open for discussion until 1 December 1991

247

248

M. A. Mimikou & Y, S. Kouvopoulos INTRODUCTION

An increasing trend of the global annual temperature has been indicated by general circulation models (GCMs) which, on the average, predict an increase of the annual temperature (greenhouse warming) by 3 ± 1.5°C following a doubling of the atmospheric C0 2 . The regional effects of such a change on water resources and water management works are issues of primary socioeconomic importance (Gleick, 1989). This paper and the following paper originate from a major research programme that addresses regional climatic change impacts on water resources and on the design and operation parameters of a water management system comprising four multi-purpose reservoirs under construction in a mountainous region of central Greece. In this first paper, emphasis is given to the regional impacts on surface runoff under the assumption of climatically perturbed time series of temperature and precipitation. A more extended presentation regarding other forms of water resources (e.g. soil moisture) and a detailed description of the methodology adopted are given elsewhere (Mimikou et al, 1990). Conclusions are drawn on basin response to climatic forcing as expressed by assumed temperature and precipitation scenarios. Considerable distortions of surface water availability and distribution, both in space and time, are evident and are shown to depend on specific regional characteristics. In the following paper (Mimikou et al, 1991), the derived time series of climatically affected runoff is used for the assessment of climatic change impacts on the water management works in the study area.

STUDY REGION The study area lies in the central mountainous region of Greece, about the 39°30'N parallel. It comprises four drainage basins with the following characteristics:

Drainage basin

Mesohora Sykia Pyli Uouzaki

River

Upper Acheloos Upper Acheloos Portaikos Pliouris

Area 2 (km ) 633.0 1173.0 134.5 140.5

Mean elevation (m a.m.s.1.) 1390 1299 800 575

The two Acheloos basins are completely mountainous whereas the other two lie on the east slopes of the mountain range, adjacent to the Thessalia Plain. A major agricultural development plan for the Thessalia Plain has been initiated by the Greek Government. The main purpose is to transfer water from the humid western parts of Greece to irrigate the fertile but semiarid

249

Regional climate change impacts: I. Impacts on water resources

Thessalia Plain. This will be accomplished by diverting a significant portion of the Upper Acheloos water to the Mouzaki reservoir. Hypsometric differences en route will be exploited for power production. The basic scheme comprises four multipurpose reservoirs (for irrigation and power generation) at the outlets of the respective basins, a diversion tunnel from the Sykia reservoir to the Mouzaki reservoir, a tunnel connecting the Pyli reservoir with the Mouzaki reservoir, and four power plants. The study area along with the associated hydrometeorological stations and the water development scheme are depicted in Fig. 1.

Fig. 1

General plan of the study area.

HYDROLOGICAL SIMULATION The use of water balance models to estimate the hydrological impacts of an expected climate change is already prevalent (Nëmec & Schaake, 1982; Gleick,

M. A. Mimikou & Y. S. Kouvopoulos

250

1986). The underlying idea is that variations in key climatic parameters such as precipitation, P, and temperature, T, produce significant changes in the hydrological cycle of a basin. Scenarios of climatically changed P and T can either be deduced from running general circulation models (GCMs) or from reasonable hypotheses. A regionally and temporally robust water-balance model would then provide the integrated simulation of hydrological processes in the basin and, by accepting climatically affected time series of precipitation and temperature as inputs, would produce respective estimates of surface runoff and other important hydrological variables. In order to produce climatically affected time series of runoff for examining both the response of the study area drainage basins and the sensitivity of the associated water management works, a monthly water balance model was developed and an approach along the lines above was undertaken. The model The model, which has been presented in detail in a previous publication (Mimikou et al., 1990), was specially developed to make full use of the available data. In particular, it incorporates an évapotranspiration accounting algorithm that may accept as inputs even subjective estimates of monthly values of pertinent meteorological variables in the absence of measured values (FAO, 1977). Owing to this capability as well as to the fact that the operation of the hydropower plants is described on a monthly basis, the time step of one month was preferred although it is recognized that a finer time step would have demonstrated basin dynamics rather better. The model was calibrated and satisfactorily tested by using a variation of a differential split sample test (Gleick, 1989) which was selected as most appropriate given the limited extent of the available hydrometeorological records (16 years). It was beneficial that at least two major periods of differing climatic conditions could be identified in this short record. The spatial robustness of the model was tested in three basins within the study area. The Mouzaki basin was excluded from the calibration due to lack of sufficient runoff data and does not appear in the subsequent sections concerning basin response. However, since the Mouzaki reservoir plays an important role in the water management project, estimates of climatically affected runoff were also made for this catchment, based on a profound basin similarity to the well-gauged Pyli basin (PPC, 1986). Climatic scenarios The climatically affected time series of precipitation and temperature used as inputs to the model were obtained from reasonable hypotheses, well within the scope of current climatological predictions regarding equivalent C 0 2 doubling. Uniform increases throughout the year of 1, 2 and 4°C over the historical temperature time series were assumed, to obtain three temperature

251

Regional climate change impacts: I. Impacts on water resources

scenarios, whereas for precipitation, five scenarios were constructed by assuming uniform changes of the historical time series throughout the year by +20%, +10%, 0% -10%, and -20%. Relying on GCM outputs was considered improper in this case, mainly due to the very coarse resolution of the grid they operate on. No matter how informative GCM predictions may be over a much greater region, they are presently incapable of accounting for local climate variations which are very marked in this region of Greece even within distances of 50 km. On the other hand, working with hypothetical scenarios suits the purpose of a sensitivity analysis of water resources and basin response which is pursued rather than a prediction of changes. The simplest assumption of constant precipitation and temperature change throughout the year for each scenario is intended to provide the grounds for the examination of more complex distributions in the future. Moreover, it facilitates an insight into the basin morphoclimatic characteristics as affecting basin response to climate change. The temperature and precipitation scenarios thus derived combine into 3 x 5 = 15 climatic scenarios yielding inputs to the water balance model to produce an equal number of climatically affected time series of runoff / for each drainage basin. Subsequently, the derived runoff series / may be considered as hypothetical, climatically affected inflows into the respective reservoirs in order to check their sensitivity to climate change. As mentioned before, especially for the Mouzaki basin, the climatically affected series I were derived from the corresponding runoff series for Pyli by analogy, i.e. assuming the same value of mean annual specific runoff reduced to unit depth of precipitation.

SENSITIVITY OF RUNOFF TO CLIMATIC CHANGES Basin characteristics A comparative presentation of basin characteristics that have determined how temperature and precipitation changes affect runoff is given below. (a) The Mesohora and Sykia basins are completely mountainous, cool and prone to snow. Their mean annual temperatures are 7.9°C and 8.8°C respectively. This reflects into high percentages of snow content in winter precipitation as calculated by the model: 27% and 23% respectively. In other words, a significant retention of precipitation occurs in snowpack storage. On the other hand, the Pyli basin is considerably lower and warmer (11.2°C). Although it receives some snow, its snowpack storage is almost negligible. (b) Model calibration produced very high monthly coefficients of direct runoff (model parameters) for the Pyli basin as compared to Mesohora and Sykia basins (about twice as much). For the latter two basins, these coefficients attain identical values. The peculiarity of the Pyli basin complies with high runoff coefficients detected there by a previous study (PPC, 1986) and is explained by the local physiographic characteristics of the Pyli basin (i.e. steep slopes and a thin soil layer).

252

M. A. Mimikou & Y. S. Kouvopoulos (c)

The Pyli basin is more humid than the other two basins. Although lower, it receives almost the same amount of yearly precipitation, approximately 2000 mm, and its specific discharge exceeds those of the other two basins by approximately 20%. The above characteristics are tabulated in Table 1. Table 1 Basin characteristics Yearly precipitation

(m)

Mean annual Winter. mow percent temperature in precipitation (model output) (°C) (%)

(mm)

Mean annual specific runoff ? -7 -2 (m?s lkm z)

1390 1299 800

7.9 8.8 11.2

1951 2084 2023

0.0385 0.0413 0.4840

Mean elevation

Mesohora Sykia Pyli

27 23 0.3

Direct runoff coefficients (calibration adjusted) O N D J F M A Mesohora Sykia Pyli

M

J

J

A

S

0.2 0.3 0.3 0.2 0.2 0.3 0.3 0.3 0.2 0.1 0.05 0.05 0.2 0.3 0.3 0.2 0.2 0.3 0.3 0.3 0.2 0.1 0.05 0.05 0.4 0.5 0.65 0.7 0.7 0.8 0.8 0.7 0.7 0.65 0.45 0.4

Soil moisture holding capacity (calibration adjusted) (mm) 300 300 225

Basin response In order to obtain a basis for the comparative examination of basin response to the different climatic scenarios, a run of the model was performed for each basin using the historical time series of temperature and precipitation as inputs. The runoff series thus derived, referred to as base run results in the following, were used for the calculation of the percent deviation from normal of climatically affected quantities. Extreme annual quantities Maximum and minimum annual flows are examined in the time span of 16 years of the available records. Those extremes represent runoff during the wettest and driest years of the period and their percent changes with respect to base run results are tabulated in Table 2. It is obvious that minimum annual runoff is much more affected than maximum runoff by temperature increases in the two similar basins of Mesohora and Sykia. The reduction of runoff is connected to increased évapotranspiration and the explanation for the differing impact between maximum and minimum runoff is that annual évapotranspiration does not deplete water in proportion to annual precipitation, as Gleick has also pointed out in explaining similar results (Gleick, 1986). On the contrary, temperature increases in the Pyli basin have a slightly more accented effect on maximum

253

Regional climate change impacts: I, Impacts on water resources

Table 2 Percent changes of extreme annual flows under temperature change Mesohora Maximum Minimum

Sykia Maximum

Minimum

Pyli Maximum

Minimum

Base run (mm)

1611

707

1882

711

1808

1035

Ar =•- re AT =-- 2°C AT =•- 4°C

-2 -4 -P

- 3 - 6 -13

-2 -4 -7

- 3 - 7 -14

-1 -2 -4

0 -2 -3

runoff than on minimum runoff. This is so due to the lack of adequate water retention to provide for évapotranspiration, whose actual value thus remains limited especially during dry years. The lack of sufficient water retention is associated with local physiographic characteristics in the Pyli basin such as high direct runoff and low moisture holding capacity of the soil (see Table 1). Mean annual surface runoff For each climatic scenario, the mean annual surface runoff was calculated over the 16 year period. Results are shown in Fig. 2 in graphs of percentage change of the base run value of runoff vs percent (climatic) change of precipitation. The Mesohora basin appears slightly more sensitive than the Sykia basin to temperature changes alone, but definitely more so compared with the Pyli basin. This is attributed directly to those morphoclimatic characteristics that favour water retention in the basin, such as snow cover, soil moisture, and moderate direct runoff coefficients. The lack in the Pyli basin results in considerably lower sensitivity to temperature changes. It is interesting to see that runoff goes up in the Pyli basin by a slight 1% under the 1°C change. This is explained by the normally very dry soil conditions in this basin during the summer. Actual évapotranspiration is therefore limited and its increase -AT=1°C AT=2°C AT=4°C

precipitation change (%> ) (c)

Fig. 2 Percent changes of mean annual runoff as a function of precipitation change for: (a) the Mesohora basin; (b) the Sykia basin; and (c) the Pyli basin.

M. A. Mimikou & Y. S. Kouvopoulos

254

under a 1°C temperature rise is not enough to compensate for the winter runoff increase, leaving thus an overall 1° increase in annual runoff. Precipitation changes produce effects on mean annual runoff characterized by a magnification factor. That is, the ratio of the resulting change in runoff over the causative change of precipitation, both expressed in percentages, is greater than unity. This has also been reported in previous studies (Gleick, 1986; Dooge, 1989). Values of the magnification factor thus defined are given in Table 3 for each successive change of precipitation and for all temperature increases. It is obvious that the magnification factor (regarding annual runoff) is practically independent of both temperature and precipitation scenarios in all basins. A second comment is that the magnification factor, which can be thought of as a measure of basin sensitivity to precipitation change, is almost the same in the similar Mesohora and Sykia basins and larger than in the Pyli basin. Therefore, the magnification factor appears to depend on local characteristics. On this subject, Dooge (1989) has commented that basin sensitivity to precipitation changes depends on its aridity. This is confirmed here since basin sensitivity appears to vary in the same sense with basin aridity among the three basins examined. The mean annual specific runoff appearing in Table 1 is considered to be an indication of basin humidity.

Table 3

Magnification factor for mean annual surface runoff

Precipitation scenarios

-20% to -10%

-10% to 0%

0% to 10%

10% to 20%

Mesohora basin AT = rc AT = 2°C AT = 4°C Sykia basin AT = rc 2°C Ar = 4°C Ar = Pyli basin AT = rc AT = 2°C AT = 4°C

1.4 1.4 1.4 1.4 1.4 1.4 1.2 1.2 1.2

1.5 1.4 1.4 1.5 1.4 1.4 1.3 1.2 1.2

1.5 1.5 1.4 1.4 1.4 1.4 1.1 1.3 1.2

1.5 1.5 1.5 1.5 1.5 1.5 1.4 1.2 1.3

Mean seasonal runoff For each climatic scenario, the mean winter (December to February) and mean summer (June to August) runoff were calculated over the 16 year period. The results are shown in Fig. 3 in graphs of percent change of the base ran value of seasonal runoff vs percent (climatic) change of precipitation. The effects of temperature changes alone are strongly influenced by snow cover and secondly by évapotranspiration. The warm Pyli basin, lacking winter snow accumulation, barely follows the behaviour of the other two basins. Moreover, in the Pyli basin, the change of winter runoff even reverses direction for an additional warming beyond AT = 2°C, because temperature in this case controls runoff through évapotranspiration rather than through snow cover. On the contrary, the cool and snowy Mesohora and Sykia basins

255

Regional climate change impacts: I. Impacts on water resources

exhibit increased sensitivity to temperature changes alone, with corresponding increases of winter runoff due to snow melting and decreases of summer runoff due to both a shift towards winter of spring snowmelt runoff and enhanced évapotranspiration. The latter is conducive to rendering the two basins more sensitive to warming during the dry season than during the wet season. An analogous observation has been made regarding extreme annual quantities which are observed in the driest and wettest years. A continuous increase of temperature diminishes the importance of snow cover as a controlling factor (no more snow would be left to melt) and évapotranspiration then becomes important. For example, in Fig. 3(b), the winter runoff increase due to temperature increase in the Sykia basin appears bounded at AT = 2°C. Regarding precipitation changes, once again the magnification factor is apparent, mostly in winter runoff as can be seen in Table 4 where respective values are tabulated for all climatic scenarios. The magnification factor is again slightly larger in the Mesohora basin than in the Sykia basin and definitely larger than in the Pyli basin regarding winter runoff and partly so regarding summer runoff. The magnification factor for the Pyli basin assumes values in a comparatively narrower range, close to unity. The explanation is

winter

AT=1°C AT=2°C AT=4«C

-20

summer

20

/ oy?, 2?

10 -20

0

,

0

,

20

-2p

0

J /

-10

/

"2i°

precipitation change (%>)

/

-20 -30 -40

:

-50

•

-SO -70

/

yy

/

/

/s (b)

(c)

Fig. 3 Percent changes of seasonal (winter & summer) runoff as a function of precipitation change for: (a) the Mesohora basin; (b) the Sykia basin; and (c) the Pyli basin.

M. A. Mimikou & Y. S. Kouvopoulos Table 4

256

Magnification factor for mean seasonal runoff

Precipitation scenarios

-20% to -10% Winter Summer

-10% to 0% 0% to 10% 10% to 20% Winter Summer Winter Summer Winter Summer

Mesohora basin AT AT AT Sykia basin AT AT AT Pyli basin AT AT AT

1.8 1.9 1.9 1.6 1.8 1.9 1.3 1.3 1.3

1.7 1.8 2.0 1.6 1.7 1.7 1.4 1.4 1.3

= rc

= 2°C = 4°C = 1°C = 2°C = 4°C = 1°C = 2°C = 4°C

1.1 0.9 0.8 1.1 1.0 0.9 1.1 1.1 1.1

1.1 0.9 0.7 1.2 1.0 0.9 1.0 1.0 1.0

1.7 1.8 1.8 1.6 1.6 1.7 1.1 1.4 1.4

1.3 0.9 0.9 1.2 1.0 1.0 1.2 1.1 1.1

1.6 1.7 1.8 1.5 1.6 1.7 1.7 1.4 1.4

1.3 1.0 0.9 1.3 1.1 1.0 1.0 1.1 1.0

that runoff in the Pyli basin, due to its high coefficient, is much more closely correlated with precipitation, and this accounts for the largely proportional manner of variations between precipitation and runoff, that is, a magnification factor close to unity. The magnification factor of seasonal runoff depends on temperature increases through the controlling factor of snow cover, increasingly so for winter runoff and decreasingly so for summer runoff, where it assumes values even less than unity for AT = 2°C in the more sensitive Mesohora basin. In the absence of snow cover, it is much more stable under temperature increase (Pyli basin). A characteristic of runoff affecting the storage capacity of reservoirs is the range of the inter-annual variation of runoff. Assuming as a measure the ratio of mean winter to mean summer runoff, the serious implications of warming on reservoir design are evident. In Table 5 the above ratios are given for all climatic scenarios. It is obvious that precipitation changes do not affect this ratio which is strongly dependent on temperature increases with the exception of the Pyli basin (independent of warming, marginally dependent on precipitation changes). Once again the local morphoclimatic characteristics render the Mesohora basin as the most sensitive with a range from

Table 5

Mean winter to mean summer runoff ratios

Precipitation scenarios

-20%

-10%

0%

10%

20%

Mesohora basin AT AT AT Sykia basin AT Ar AT Pyli basin AT AT AT

1.49 1.93 2.27 1.39 1.71 1.93 0.96 0.97 0.96

1.52 1.96 129 1.40 1.72 1.96 0.99 1.00 0.99

1.52 1.97 2.37 1.39 1.72 1.95 1.03 1.04 1.02

1.49 1.97 2.31 1.38 1.71 1.91 1.02 1.06 1.05

1.45 1.94 2.27 1.36 1.67 1.89 1.07 1.08 1.08

= 1°C = 2°C = 4°C = PC = 2°C = 4°C = 1°C = 2°C = 4°C

257

Regional climate change impacts: I. Impacts on water resources

1.52 for AT = 1°C to 2.37 for AT = 4°C. In general, the Mediterranean characteristic of high seasonality of runoff tends to become more intense under warming conditions. CONCLUSIONS The conclusions drawn from this research are the following: (a) mountainous, seasonally snow-covered Mediterranean basins with effective water retentive mechanisms exhibit under temperature increase serious reduction of mean annual runoff, even more serious reduction of summer runoff and minimum annual runoff along with a significant increase of winter runoff and of maximum annual runoff. In general reductions are more intense than increases, due to the conducive effect of évapotranspiration. A warmer, humid basin with local characteristics that inhibit water retention exhibits minimal sensitivity of runoff to temperature increase; (b) in general, a magnification factor characterizes the effects of precipitation change on runoff, and can be thought of as a measure of basin sensitivity to precipitation change. Regarding annual runoff, the magnification factor seems to be independent of temperature and positively associated with basin aridity, whereas on a seasonal basis it depends on the presence of winter snow cover and consequently on temperature increases; (c) the already high seasonality of runoff of the Mediterranean type is considerably intensified in basins sensitive to temperature increase, with apparent consequences on reservoir storage capacity; and (d) snow is distinguished as the most significant and determining factor of basin response to climatic change. This factor, in turn, depends on the orographic characteristics of the basin besides the general climatic ones. Additional characteristics are shown to be important, for example the coefficient of runoff which, when unusually high, does not permit other hydrological processes, sensitive to climatic change, to be accomplished or even to take place at all. Acknowledgements The work presented here is part of a major research project jointly funded by the General Secretariat of Research and Technology and the Greek Ministry of the Environment, Physical Planning and Public Works. The authors wish to thank the Public Power Corporation of Greece for providing the necessary hydrometeorological data for the accomplishment of this work. REFERENCES Dooge, J. C. I. (1989) Effect of C 0 2 increases on hydrology and water resources. In: Carbon Dioxide and Other Greenhouse Gases: Climatic and Associated Impacts, eds. R. Fantechi & A. Ghazi, Kluwer Academic Publishers, London, 204-213.

M. A. Mimikou & Y. S. Kouvopoulos

258

FAO (1977) Crop water requirements. Irrigation and Drainage Paper, No. 24, Food and Agriculture Organization, Rome, Italy. Gleick, P. H. (1986) Regional Water Availability and Global Climatic Change: The Hydrologie Consequences of Increases in Atmospheric CO* and Other Trace Gases, PhD Dissertation, University of California, Berkeley, California, USA. Gleick, P. H. (1989) Climate change, hydrology and water resources. Rev. Geophys. 27(3), 329344. Mimikou, M., Kouvopoulos, Y., Cavadias, G. & Vayiannos, N. (1990) Regional hydrological effects of climate change. /. Hydrol. (in print). Mimikou, M. A., Hadjisawa, P. S., Kouvopoulos, Y. S. & Afrateos, H. (1991) Regional climate change impacts: II. Impacts on water management works. Hydrol, Sci. J. 36 (3), 259-270. NSmec, J. & Schaake, J. (1982) Sensitivity of water resource systems to climate variation. Hydrol. Sci. J. 27 (3), 327-343. PPC (1986) Hydrologikos Schediasmos ton Potamon Portaikos kai Pliouris sûn Thessalia (Hydrological design on Portaikos and Pliouris Rivers in the Thessaly Plain). Public Power Corporation of Greece, Athens, Greece (in Greek). Received 24 July 1990; accepted 17 January 1991