regional impacts of climate change on water resources quantity and quality indicators

15
Regional impacts of climate change on water resources quantity and quality indicators M.A. Mimikou * , E. Baltas, E. Varanou, K. Pantazis Department of Civil Engineering, Division of Water Resources, Hydraulic and Maritime Engineering, Laboratory of Hydrology and Water Resources Management, National Technical University of Athens, 5, Iroon Polytechniou, 157 73 Athens, Greece Received 29 July 1999; received in revised form 6 April 2000; accepted 6 April 2000 Abstract The aim of this paper is to assess the impacts of climate change on water resources (surface runoff) and on water quality. Two GCM-based climate change scenarios are considered: transient (HadCM2) and equilibrium (UKHI). A conceptual, physically based hydrological model (WBUDG) is applied on a catchment in central Greece, simulating the effect of the two climate scenarios on average monthly runoff. A newly developed in the stream model (R-Qual) is applied in order to simulate water quality downstream of a point source under current and climatically changed conditions. Simulated parameters include monthly concentrations of BOD, DO and NH 1 4 . Both scenarios suggest increase of temperature and decrease of precipitation in the study region. Those changes result in a significant decrease of mean monthly runoff for almost all months with a considerable negative impact on summer drought. Moreover, quality simulations under future climatic conditions entail significant water quality impairments because of decreased stream flows. q 2000 Elsevier Science B.V. All rights reserved. Keywords: Modelling; Stream; Climate change; Scenarios; Quality; Quantity 1. Introduction The research work presented herein originates from two EU funded research programmes (EUROTAS and CHESS, both funded under the fourth FP—DGXII) that address the impacts of ‘greenhouse’ warming on water resources quantity and quality on a catchment, regional and European scale. The study area is the Pinios river basin, situated in the central part of Greece. Because of an intensive cultivation, water needs are considerable. Addition- ally, the large application of pesticides and fertilisers can cause serious degradation of the quality of the surface and groundwater of the basin. Because of physical, chemical and biological processes occurring within the river, the river’s role is far from that of a conduit for the transport of inputs. These processes have a strong purifying effect leading to an improve- ment of water quality downstream from a source of pollution. A quality parameter of primary importance that accelerates these processes is the dissolved oxygen concentration. Hence purification is largely dependent on the continued oxygenation of the river water, which is enhanced under high flow conditions that encourage surface aeration. On the contrary the aggregated contribution of pollutants under low flow conditions can cause serious downstream problems regarding water quality. Eutrophication, Journal of Hydrology 234 (2000) 95–109 0022-1694/00/$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S0022-1694(00)00244-4 www.elsevier.com/locate/jhydrol * Corresponding author. Tel.: 130-1-7722878. E-mail address: [email protected] (M.A. Mimikou).

Upload: ma-mimikou

Post on 01-Nov-2016

215 views

Category:

Documents


3 download

TRANSCRIPT

Page 1: Regional impacts of climate change on water resources quantity and quality indicators

Regional impacts of climate change on water resources quantityand quality indicators

M.A. Mimikou*, E. Baltas, E. Varanou, K. Pantazis

Department of Civil Engineering, Division of Water Resources, Hydraulic and Maritime Engineering, Laboratory of Hydrology and WaterResources Management, National Technical University of Athens, 5, Iroon Polytechniou, 157 73 Athens, Greece

Received 29 July 1999; received in revised form 6 April 2000; accepted 6 April 2000

Abstract

The aim of this paper is to assess the impacts of climate change on water resources (surface runoff) and on water quality. TwoGCM-based climate change scenarios are considered: transient (HadCM2) and equilibrium (UKHI). A conceptual, physicallybased hydrological model (WBUDG) is applied on a catchment in central Greece, simulating the effect of the two climatescenarios on average monthly runoff. A newly developed in the stream model (R-Qual) is applied in order to simulate waterquality downstream of a point source under current and climatically changed conditions. Simulated parameters include monthlyconcentrations of BOD, DO and NH1

4 .Both scenarios suggest increase of temperature and decrease of precipitation in the study region. Those changes result in a

significant decrease of mean monthly runoff for almost all months with a considerable negative impact on summer drought.Moreover, quality simulations under future climatic conditions entail significant water quality impairments because ofdecreased stream flows.q 2000 Elsevier Science B.V. All rights reserved.

Keywords: Modelling; Stream; Climate change; Scenarios; Quality; Quantity

1. Introduction

The research work presented herein originates fromtwo EU funded research programmes (EUROTAS andCHESS, both funded under the fourth FP—DGXII)that address the impacts of ‘greenhouse’ warming onwater resources quantity and quality on a catchment,regional and European scale.

The study area is the Pinios river basin, situated inthe central part of Greece. Because of an intensivecultivation, water needs are considerable. Addition-ally, the large application of pesticides and fertilisers

can cause serious degradation of the quality of thesurface and groundwater of the basin. Because ofphysical, chemical and biological processes occurringwithin the river, the river’s role is far from that of aconduit for the transport of inputs. These processeshave a strong purifying effect leading to an improve-ment of water quality downstream from a source ofpollution. A quality parameter of primary importancethat accelerates these processes is the dissolvedoxygen concentration. Hence purification is largelydependent on the continued oxygenation of the riverwater, which is enhanced under high flow conditionsthat encourage surface aeration. On the contrary theaggregated contribution of pollutants under lowflow conditions can cause serious downstreamproblems regarding water quality. Eutrophication,

Journal of Hydrology 234 (2000) 95–109

0022-1694/00/$ - see front matterq 2000 Elsevier Science B.V. All rights reserved.PII: S0022-1694(00)00244-4

www.elsevier.com/locate/jhydrol

* Corresponding author. Tel.:130-1-7722878.E-mail address:[email protected]

(M.A. Mimikou).

Page 2: Regional impacts of climate change on water resources quantity and quality indicators

which is caused by excessive levels of nutrients, is themost notable example.

The climate change scenarios are constructed bythe Climatic Research Unit (CRU) of the Universityof East Anglia (UK) and are the results of two climatechange experiments based on the General CirculationModels (GCM) (Hulme et al., 1994). More specifi-cally, one equilibrium experiment, UKHI andone transient experiment, HADCM2 are applied.Changes refer to rainfall, temperature and potentialevapotranspiration.

In order to estimate the hydrological effects ofclimate changes, a conceptual, physically basedwater balance model (WBUDG), developed for thispurpose, is applied. It allows the calculation of surfacerunoff, soil moisture and actual evapotranspiration.

Water quality parameters are estimated by using anin stream model (R-Qual), which in combination withthe WBUDG model allows a complete simulation offlows and water quality, under current and changedclimatic conditions up to the terminal year 2050.

2. Study region and data used

The Pinios river is located in the Thessaly district(central part of Greece) The total drainage area of theriver is 9.450 km2, with a varied topography fromnarrow gorges to wide flood plains.

The Pinios catchment area consists of 15 sub basinsdrained by the main river and its five most impor-tant tributaries. Owing to lack of sufficient data formost sub basins, the study focuses on the Ali Efentisub basin (Fig. 1), for which reliable hydrometeor-ological time series of adequate length are avail-able. The area of the catchment is 2.868,61 km2.Some general characteristics of this basin aregiven in Table 1.

The climate is humid and temperate with sub-stantial seasonal variations. The natural vegetationcomprises of grasslands and dense forests.

Monthly hydrometeorological data for a 36 year

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–10996

Fig. 1. The study area: the Ali Efenti basin.

Table 1General characteristics of the study area

Area (km2) 2868.61

Mean Elevation (m) 539.73Mean annual rainfall (mm) 933.3Mean annual storm runoff

coefficient0.428

Mean annual flow (m3/s) 39.05

Page 3: Regional impacts of climate change on water resources quantity and quality indicators

period (1960–1996) were acquired from 15 stationslocated within and near the study area, as seen in Fig.1. Those data refer to precipitation, temperature, windvelocity, relative humidity and sunshine durationvalues.

The basic input variable of the model, monthlyareal precipitation, was calculated by the Thiessenmethod and corrected for the elevation on the basisof a precipitation–elevation relationship. Meanmonthly air temperature of the basin was calculatedby associating the mean monthly temperature of eachstation with its elevation and by establishing a relationbetween the mean temperature of the stations and themean temperature at the mean elevation of the basin.Mean monthly values of daily wind are accepted asinput by the model in the form of three classes ofmagnitude and thus strict conditions of accuracywere not required. In the same manner, mean monthlyminimum relative humidity was classified into threeclasses and monthly relative sunshine duration intofive classes.

Mean monthly runoff values at the outlet of thebasin for the same time period were used for the cali-bration of the hydrological model. Monthly valueshave been derived from a 36 year record of meandaily values and were then expressed in mm ofequivalent depth.

Concerning water quality, 12 stations located alongthe Pinios river are measuring physical and chemicalparameters. Monitoring the quality of the river startedat 1988. The available data are monthly values of thefollowing: Temperature, Color, Turbidity, Hardness,Conductivity, pH, DO, Chlorides, Bromides, Sulfates,Nitrites, Nitrates, Ammonium, Total Nitrogen andTotal Phosphorus. In order to simulate water quality,two gauging stations were used located upstream anddownstream of a point source discharge (e.g. a wastewater treatment plant, WWTP), at a distance of 4 kmapproximately.

Parameters required for modelling the water qualityof the river, were mainly the discharges from pointsources produced by industry and sewage treatmentworks, and pollution entering the stream from nonpoint sources. According to data provided by localauthorities, 43% of the organic pollution is derivedfrom wastewater treatment plants and industrialareas while 16% of total phosphorus and 64% oftotal nitrogen are contributing stream pollution fromdiffuse sources.

General sources of pollution within the study areaare presented in Table 2.

3. Climate change scenarios

The climate scenarios used have been constructedby the Climatic Research Unit (CRU) of the Univer-sity of East Anglia, England.

The methodology adopted used the CRU 1961–1990 baseline climatologies for Europe, the resultsfrom two GCM (General Circulation Models) climatechange experiments (UKHI and HadCM2) and arange of projections of global warming calculatedby MAGICC (Model for the Assessment of Green-house gas Induced Climate Change), a simple upwel-ling-diffusion energy balance climate model (Hulmeet al., 1994).

The primary difference between UKHI (UKMeteorological Office High Resolution Model) andHadCM2 (Hadley Centre Coupled Model v2)(Hulme et al., 1999) is that the former is an equili-brium climate change experiment performed using anatmospheric GCM coupled to a single representationof the ocean. HadCM2 is a transient climate changeexperiment performed using fully coupled ocean andatmosphere GCMs.

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–109 97

Table 2Estimated pollution sources on the study area

Pollutant Quantity (t/year)

BOD5 17.24Total Nitrogen 12.64Total Phosphorus 1.25

Table 3Ali Efenti basin, grid codes and corresponding weightingcoefficients

Grid coordinatesLongitude× latitude

Code numbers Weighting coefficients

21800× 39800 20101290 0.01221800× 39800 20101295 0.180218300 × 39800 20151290 0.208218300 × 398300 20151295 0.55722800× 398300 20201295 0.043P

1

Page 4: Regional impacts of climate change on water resources quantity and quality indicators

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–10998

Fig. 2. Flow chart of the Water Balance Model (WBUDG).

Page 5: Regional impacts of climate change on water resources quantity and quality indicators

The GCM output of the variables precipitation,temperature and evapotranspiration was inter-polated to 0:508 × 0:508 from the original GCMresolution �3:758 × 2:508 for both UKHI andHadCM2).

The study area lies in the central region of Greecebetween 398000 and 408150 North Latitude. To applythe scenarios, the whole area is divided into five gridswith 0:58 × 0:58 dimensions, according to the geogra-phical map. Each grid is represented by a code numberspecific to the longitude and latitude of its south westcorner and calculated using the following formula:

Code� { �longitude1 180:0� × 100 000:0

1�latitude1 90:0� × 10:0} �1�A weighting coefficient was calculated for each cellaccording to the percentage contribution of the cell’ssurface to the entire area of the basin.

Having in mind that, rainfall and potential evapo-

transpiration are expressed in percent (%) changesand the temperature in degrees, the same coefficientis applied to calculate the contribution of each grid tothe climate change.

The Ali Efenti grid code numbers and their corre-sponding weighting coefficients are listed in Table 3.

4. Climate change impacts on regional waterresources quantity and quality

4.1. Water quantity

The assessment of climate change impacts onseveral water resources quantity indicators wasbased on the use of a monthly water balance model(WBUDG), developed for this purpose. The flowchart of the model structure is shown in Fig. 2. Themain input parameters are precipitation, temperature,relative humidity, sunshine duration and wind speed.The main output parameters are evapotranspiration,soil moisture and stream runoff. Details on themodel operation can be found in a series of previouspublications (Mimikou and Kouvopoulos, 1991;Mimikou et al., 1991a).

The model was calibrated using the 36 year histor-ical hydrometeorological and hydrometric data,described previously. Model parameters which havebeen estimated by calibration are the monthly stormrunoff coefficient (SRC), the maximum soil moisture(SMAX), the lower temperature parameter (T0), theupper temperature parameter (T1), the minimum raincontent coefficient (a ), the melt rate factor (DF), thewatershed lag coefficient (K1) and the groundwater lagcoefficient (K2) (Mimikou et al., 1991b). The para-meters and their values obtained by calibration arepresented in Table 4. The mean Nash number(WMO, 1986) for the 36 year time period was foundto be satisfactorily and equal to 0.60.

For the application of the climate change scenariosto the hydrometeorological data, synthetic series(from 1996 up to 2050) of precipitation and tempera-ture based on the historical data were produced. Thestochastic autoregressive models AR(1) and AR(2)were applied and tested in order to check their effi-ciency. AR(1) was found to perform better for thegeneration of precipitation series and AR(2) for thegeneration of temperature series. For the period from

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–109 99

Table 4Summary of WBUDG model parameters

Storm runoff coefficient (SRC)

October 0.10November 0.22December 0.45January 0.52February 0.48March 0.70April 0.68May 0.48June 0.30July 0.18August 0.13September 0.12

Maximum Soil Moisture(SMAX, mm)

225

Lower temperature parameter(T0, 8C)

23.5

Upper temperature parameter(T1, 8C)

2.0

Minimum rain contentcoefficient (a)

0.30

Melt-rate factor (DF,mm/degree/day)

0.20

Watershed lag coefficient (K1) 0.35Groundwater lag coefficient(K2)

0.50

Nash number (NTD) 0.60

Page 6: Regional impacts of climate change on water resources quantity and quality indicators

1996 up to the terminal year 2050, 50 synthetic seriesof each variable (precipitation and temperature) weregenerated.

Runs of the water balance model were performedfor all 50 synthetic series considering zero climatechange (no application of climate change scenarios)

and are referred as base runs. The outputs of the baseruns are used for comparison reasons.

Later, the synthetically generated series were modi-fied in order to take into account properly the changesprovided by each of the two climate change scenarios.Further runs of the water balance model were

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–109100

Fig. 3. Flowchart of NTUA in Stream Model (R-Qual).

Page 7: Regional impacts of climate change on water resources quantity and quality indicators

performed (50 runs accounting UKHI changes and 50runs accounting HadCM2 changes). Finally, theresults were averaged over the 50 output series for eachcase (base run, UKHI change and HadCM2 change).

4.2. Water quality

Concerning water quality, a recently developed instream model (R-Qual) was applied to the study area(Mimikou et al., 1999b). This model is a one-dimen-sion finite differences mathematical model that hasbeen derived by applying the mass conservation prin-ciples to small completely mixed control volumes(Beck and Young, 1976; Demuynck et al., 1997). Itis based on the classical advection dispersion equation

and it is able to simulate up to 10 water quality para-meters which are: Temperature, BOD, DO, N–NO3,N–NO2, N–NH4, N-Organic, P-Dissolved, P-Organicand Algae. It allows for multiple waste discharges,tributary inflows, non-point source pollution loadingand it can operate in both steady state and dynamicmode. The model can be used to study the impact ofwaste loads on in-stream water quality or to identifythe magnitude and quality characteristics of non-pointwaste loads. Detailed description of the model isbeyond the scope of this paper. A schematic flowchartof R-Qual is presented in Fig. 3.

The in stream model was applied to the study areaagainst measured quality values for a time period ofthree years (1988–1991) by using water quality

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–109 101

Fig. 4. R-Qual application to Pinios River. BOD simulation.

Fig. 5. R-Qual application to Pinios River. DO simulation.

Page 8: Regional impacts of climate change on water resources quantity and quality indicators

measurements at the two gauging stations describedabove. Although the model is able to simulate severalquality parameters as described previously, thecurrent study included only the simulation of BOD,DO and ammonium ions NH41. The objective was tomodel stream quality downstream of this point sourceunder known flow conditions and subsequently toassess climate change impact in terms of dilutioncapacity of the river, because of decreased flows andtemperature increase.

The water quality model applied to the Pinios riverhas given generally good simulations as it is shown inFigs. 4–6. In Fig. 7 the flow profile for the total simu-lated period is presented. It can be seen that for thelimited flow periods (summer months) major waterquality impairments were observed. Model para-

meters obtained from calibration against observedvalues are well within the range of other model appli-cations (Eatherall et al., 1998) and in accordance withliterature referred values (Chapra, 1997). Calibratedparameters and values obtained are presented inTable 5.

After the model was calibrated under currentclimate conditions, further runs of the model wereperformed to account for climate change, as intro-duced by the two aforementioned scenarios. Climatechange induced in the water balance model gave theclimatically changed flow profile at the outlet of theAli Efenti basin. This new flow profile served as inputto the in stream model. Temperature inputs used in thequality model were the values given by the scenarios,as estimated by the two GCMs.

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–109102

Fig. 6. R-Qual application to Pinios River. NH41 simulation.

Fig. 7. Flow profile of Pinios river at gauging station located at the outlet of Ali Efenti basin upstream of the WWTP discharge.

Page 9: Regional impacts of climate change on water resources quantity and quality indicators

5. Results

Table 6 presents the mean monthly runoff for theterminal year 2050, regarding the base run, UKHIchange and HadCM2 change. Additionally and forcomparison, the percent increase or decrease of theclimatically changed runoff from the base run is alsogiven.

It has been estimated that there will be a significantdecrease of the mean monthly runoff for all months bytaking into account climate change from the transientexperiment. A significant decrease of the meanmonthly runoff for almost all months is estimated bytaking into account climate change from the equili-brium experiment as well, with the exception of runoffduring November, where almost no change from thebase run is observed.

Both climate change scenarios suggest that thehighest decrease in mean monthly runoff will beduring summer and especially during June. A graphi-cal comparison of the aforementioned results can beseen in Fig. 8.

The assessment of the regional hydrological effects

of the examined climate change is based on thechanges observed on eight mean hydrological indica-tors, such as the mean annual precipitation, meanannual potential evapotranspiration, mean annualactual evapotranspiration, mean annual runoff, meansummer runoff, mean winter runoff, annual maxrunoff and annual min runoff (Arnell, 1998; Mimikou,1995). Table 7 presents all final results of the esti-mated mean hydrological indicators from the applica-tion of the equilibrium and transient scenarios, for theAli Efenti basin up to the terminal year of analysis2050. A graphical display of these results can be seenin Fig. 9. From the percentage change of the meanannual values it can be concluded that the resultsfrom the application of the two scenarios behave simi-larly. The scenarios suggest a decrease in mean annual

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–109 103

Table 5Summary of R-Qual model parameters

Calibration parameter Calibrated value (1/day)

BOD decompositionand settling rates

0.2–0.4

N-organic hydrolysisand settling rates

0.25–0.35

N–NH4 oxidation rate 0.25–0.35

Table 6Estimated mean monthly runoff (mm) for the terminal year

Month Base run (mm) UKHI HADCM2

mm % mm %

October 11.8 10.6 210.2 9.6 218.6November 26.1 26.2 0.3 21.5217.6December 69.5 62.9 29.5 59.9 213.8January 60.0 51.2 214.7 52.0 213.3February 50.8 38.5 224.2 38.4 224.4March 72.5 59.1 218.5 62.1 214.3April 65.1 53.9 217.2 53.7 217.5May 32.3 26.8 217 23.2 228.2June 10.2 7.5 226.5 5.5 246.1July 6.2 4.6 225.8 4.5 227.4August 3.6 2.9 219.4 2.1 241.66September 5.3 4.4 217 4.0 224.5

Fig. 8. Comparison of mean monthly runoff as estimated with base run, UKHI and HadCM2 scenarios.

Page 10: Regional impacts of climate change on water resources quantity and quality indicators

precipitation, which leads to reduced runoff values.As expected, Potential Evapotranspiration isincreased because of increase of the temperature.The more significant decrease is observed in themean summer runoff (May–October) and especiallyfor the transient scenario (HadCM2). These results arein accordance with the outputs of previouslycontacted research in other regions of Greece (Mimi-kou et al., 1991a, 1999a), during which the impact ofclimate change on various forms of water resourceshas been examined.

Water quality impairments because of climatechange have been estimated and are presented interms of mean monthly concentrations of BOD, DOand NH41 in mg/lt in Tables 8–10 respectively. Agraphical representation can be seen in Figs. 10, 12and 14.

Regarding water quality downstream of a pointsource pollution load, basic conclusions derivedfrom the above results are that climate change has astrong impact on water quality parameters. Bothscenarios are inducing decreased stream flows

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–109104

Table 7Estimated mean hydrological indicators for all scenarios

Hydrological indicators Base run (mm) UKHI HADCM2

mm % mm %

Mean annual precipitation 906.1 814.7 210.0 778.1 214.1Mean annual pot. evap/tion 859.1 943.6 9.8 948.6 10.4Mean annual act. evap/tion 492.5 459.7 26.7 441.4 210.4Mean annual runoff 413.6 349.1 215.6 337.5 218.4Mean summer runoff 68.9 56.8 217.6 48.9 229.1Mean winter runoff 346.5 292.3 215.6 287.4 217.1Annual max runoff 808.2 679.4 215.9 641.8 220.6Annual min runoff 194.2 175.7 29.5 166.5 214.3

Fig. 9. Mean Hydrological Indicators for year 2050.

Page 11: Regional impacts of climate change on water resources quantity and quality indicators

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–109 105

Table 8Mean monthly BOD concentrations (mg/l) for all scenarios for the terminal year

Base run UKHI % Change HADCM2 % Change

October 6.1 6.3 2.8 6.5 6.2November 5.1 5.1 20.6 5.3 3.1December 4.4 4.4 1.2 4.5 1.5January 4.4 4.5 1.5 4.4 0.2February 4.5 4.7 3.7 4.6 3.1March 4.3 4.5 1.9 4.4 1.8April 4.4 4.5 2.2 4.5 2.3May 4.9 5.1 3.5 5.2 6.4June 6.2 6.8 8.9 7.3 17.2July 7.3 8.1 11.1 8.4 14.8August 8.9 9.5 6.3 10.5 17.9September 8.0 8.4 5.1 8.8 10.1

Table 9Mean monthly DO concentrations (mg/l) for all scenarios for the terminal year

Base run UKHI % Change HADCM2 % Change

October 10.1 10.0 20.54 9.9 21.14November 10.8 10.9 0.38 10.8 20.56December 11.4 11.3 20.43 11.3 20.58January 11.4 11.3 20.52 11.3 20.26February 11.3 11.2 21.12 11.2 20.93March 11.4 11.3 20.67 11.3 20.73April 11.3 11.3 20.70 11.3 20.81May 11.0 10.9 21.01 10.8 21.70June 10.0 9.7 23.16 9.5 25.55July 9.2 8.8 24.77 8.7 24.96August 8.0 7.8 22.32 7.4 27.33September 8.7 8.5 21.41 8.5 22.38

Table 10Mean monthly NH41 concentrations (mg/l) for all scenarios for the terminal year

Base run UKHI % Change HADCM2 % Change

October 0.06 0.07 7.462 0.08 14.935November 0.04 0.04 22.381 0.05 9.523December 0.02 0.03 8.695 0.03 8.695January 0.02 0.03 4.166 0.03 4.166February 0.03 0.03 15.384 0.03 15.384March 0.02 0.03 9.090 0.03 9.090April 0.03 0.03 8.333 0.03 12.500May 0.04 0.04 11.111 0.05 22.222June 0.07 0.08 20.000 0.10 38.571July 0.10 0.12 21.428 0.13 28.571August 0.14 0.15 10.071 0.18 28.776September 0.12 0.13 9.482 0.14 18.103

Page 12: Regional impacts of climate change on water resources quantity and quality indicators

because of a gradual decrease in precipitation values.This in terms of quality suggests that the stream willincreasingly lose its capacity to dilute pollutantsentering from diffuse or point sources.

The temperature rise that is also predicted by theclimatic change theoretically acts in favour of thestream self-purification as the biological degradationrate and settling rates which are temperature depen-dent increase but this benefit seems to be retractedfrom the decreased stream flows. Decreased flowsalso result in low water velocities. Oxygen re-aerationcoefficient is strongly related with water velocity thus

stream oxygenation is proportional affected. Highwater temperatures also result in low oxygen satura-tion concentration, which is the driving force forwater reaeration. Hence, less oxygen remains avail-able for bioxidation reactions. This is especially soduring summer months where the greatest change inprecipitation and the greatest increase in temperatureare expected. A further investigation (sensitivityanalysis) would be necessary in order to assesscompletely the exact impact of temperature increasein stream quality. Nutrient fluxes in the stream fromdiffuse pollution sources were not taken into account

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–109106

Fig. 10. Mean monthly BOD concentration for Base run, UKHI and HadCM2 scenarios.

Fig. 11. Comparison of % change of mean monthly BOD concentration for UKHI and HadCM2 scenarios.

Page 13: Regional impacts of climate change on water resources quantity and quality indicators

in this study. Therefore, no simulations for nitratesand phosphorus were performed. The use of a catch-ment delivery model that will be able to simulatenutrients transport in the stream is necessary in order toachieve a more holistic approach on the effect of climatechange on surface water quality in a catchment scale.

The impact of changed climatic conditions onstream quality is clearly indicated in Figs. 11, 13and 15, where % change for the three quality para-meters (BOD, DO and NH14 � is presented for the term-inal year for both scenarios. Reasonably forHADCM2 scenarios, which predict the greatest

change in precipitation, the major quality impairmentsare exhibited.

6. Conclusions

Basic conclusions drawn from this research are thefollowing:

• Both scenarios seem to give reasonable andquite consistent results. The transient scenarios(HadCM2) give more significant changes than theequilibrium (UKHI).

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–109 107

Fig. 12. Mean monthly DO concentration for Base run, UKHI and HadCM2 scenarios.

Fig. 13. Comparison of % change of mean monthly DO concentration for UKHI and HadCM2 scenarios.

Page 14: Regional impacts of climate change on water resources quantity and quality indicators

• Mean annual runoff values are reduced.• Mean winter runoff values (November–April) are

reduced.• The most significant reduction is expected in mean

summer runoff values (May–October).• The results on runoff change are in accordance with

results drawn from previous relevant research workin other catchments in Greece.

• Water quality impairment has been observed(increased BOD and NH41 values, decreased DOvalues) because of the loss of stream dilution capa-city and to low water velocities derived from thereduced stream flows.

• Major quality impairment is observed during

summer months where both climate scenariospredict the greatest precipitation decrease.

Acknowledgements

This research was supported by the Environmentand Climate Programme of the EU, DGXII, in theframework of the contracts ENV4-CT97-0535(EUROTAS: European River Flood Occurrence andTotal Risk Assessment) and ENV4-CT97-0440(CHESS: Climate, Hydrochemistry and Economicsof Surface—Water Systems).

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–109108

Fig. 15. Comparison of % change of mean monthly NH41 concentration for UKHI and HadCM2 scenarios.

Fig. 14. Mean monthly NH41 concentration for Base run, UKHI and HadCM2 scenarios.

Page 15: Regional impacts of climate change on water resources quantity and quality indicators

References

Arnell, N.W., 1998. The effect of climate change on hydrologicalregimes in Europe: a continental perspective. Global Environ-mental Change 9, 5–23.

Beck, M.B., Young, P.C., 1976. A dynamic model for DO-BODrelationships in a non-tidal stream. Water Resources andResearch 23, 1393–1442.

Chapra, S.C., 1997. Surface Water-Quality Modelling, McGraw-Hill, New York.

Demuynck, C., Bauwens, W., De Pauw, N., Dobbelaere, I., Poel-man, E., 1997. Evaluation of pollution reduction scenarios in ariver basin: application of long term quality simulations. WaterScience Technology 35 (9), 65–75.

Eatherall, A., Boorman, D.B., Williams, R.J., Kowe, R., 1998.Modelling in-stream water quality in LOIS. Science and TotalEnvironment 210–211, 499–517.

Hulme, M., Conway, D., Brown, O., Barrow, E., 1994. A 1961–1990 Baseline Climatology and Future Climate Change Scenar-ios for Great Britain and Europe. Climatic Research Unit,University of East Anglia.

Hulme, M., Barrow, E., Arnell, N., Harrison, P., Johns, T., Down-ing, T., 1999. Relative impacts of human—induced climatechange and natural climate variability. Nature 397, 688–691.

Mimikou, M., 1995. Climate Change in Environmental Hydrology.In: Singh, V.P. (Ed.), Kluwer Academic, Drodrecht, pp. 69–106.

Mimikou, M., Kouvopoulos, Y., 1991. Regional climate changeimpacts. I. Impacts on water resources. Hydrological SciencesJournal 36 (3 and 6), 247–258.

Mimikou, M., Hadjissava, P., Kouvopoulos, Y., Afrateos, H.,1991a. Regional climate change impacts. II. Impacts on watermanagement works. Hydrological Sciences Journal 36 (3 and 6),259–270.

Mimikou, M., Kouvopoulos, Y., Cavvadias, G., Vayiannos, N.,1991b. Regional hydrological effects of climate change. Journalof Hydrology 123, 119–146.

Mimikou, M., Kanellopoulou, S., Baltas, E., 1999a. Human impli-cation of changes in the hydrological regime due to climatechange in Northern Greece. Global Environmental Change 9,139–156.

Mimikou, M.A., Baltas, E., Varanou, E., Pantazis, K., 1999b.Impacts of climate change on the water resources quantity andquality. International Conference on Water, Environment, Ecol-ogy, Socio-economics, and Health Engineering, 18–21 October1999, Seoul, Korea.

WMO, 1986. Operational Hydrology Report, No. 23: Intercompar-ison of Models of Snowmelt Runoff. WMO No 646.

M.A. Mimikou et al. / Journal of Hydrology 234 (2000) 95–109 109