Possible Impacts of Climate Change on Soil Moisture ... - Springer Link

POSSIBLE IMPACTS OF CLIMATE CHANGE ON SOIL MOISTURE AVAILABILITY IN THE SOUTHEAST ANATOLIA DEVELOPMENT PROJECT REGION (GAP): AN ANALYSIS FROM AN AGRICULTURAL DROUGHT PERSPECTIVE ALI UMRAN KOMUSCU, AYHAN ERKAN and SUKRIYE OZ Turkish State Meteorological Service, Research Department, Ankara, Turkey E-mail: [email protected]

Abstract. This paper presents probable effects of climate change on soil moisture availability in the Southeast Anatolia Development Project (GAP) region of Turkey. A series of hypothetical climate change scenarios and GCM-generated IPCC Business-as-Usual scenario estimates of temperature and precipitation changes were used to examine implications of climate change for seasonal changes in actual evapotranspiration, soil moisture deficit, and soil moisture surplus in 13 subregions of the GAP. Of particular importance are predicted patterns of enhancement in summer soil moisture deficit that are consistent across the region in all scenarios. Least effect of the projected warming on the soil moisture deficit enhancement is observed with the IPCC estimates. The projected temperature changes would be responsible for a great portion of the enhancement in summer deficits in the GAP region. The increase in precipitation had less effect on depletion rate of soil moisture when the temperatures increase. Particularly southern and southeastern parts of the region will suffer severe moisture shortages during summer. Winter surplus decreased in scenarios with increased temperature and decreased precipitation in most cases. Even when precipitation was not changed, total annual surplus decreased by 4 percent to 43 percent for a 2 ◦ C warming and by 8 percent to 91 percent for a 4 ◦ C warming. These hydrologic results may have significant implications for water availability in the GAP as the present project evaluations lack climate change analysis. Adaptation strategies – such as changes in crop varieties, applying more advanced dry farming methods, improved water management, developing more efficient irrigation systems, and changes in planting – will be important in limiting adverse effects and taking advantage of beneficial changes in climate.

1. Introduction Long-term temperature records measured over landmasses indicate that our climate is changing. Global mean surface air temperature has risen between 0.3 ◦ C and 0.6 ◦ C in the last 100 years, and half of this increase has occurred since the mid-1970s (IPCC, 1996). The increase in the global mean surface temperatures is often attributed to an anthropogenic greenhouse enhancement (Mitosek, 1992). The atmospheric concentrations of greenhouse gases, mainly carbon dioxide (CO2 ), methane (CH4 ) and nitrous oxide (N2 O), have risen significantly since the preindustrial era. Current estimates indicate that the CO2 concentrations in the atmosphere have reached almost 360 ppmv3 , which is a 30 percent increase from its pre-industrial levels (IPCC, 1995). The IPCC has developed a range of scenarios Climatic Change 40: 519–545, 1998. © 1998 Kluwer Academic Publishers. Printed in the Netherlands.

RAPATI -- CLIM 1790 -- PIPS Nr. 161728 (climkap:spacfam) v.1.0 clim1790.tex; 9/10/1998; 11:58; p.1

520

ALI UMRAN KOMUSCU ET AL.

of future greenhouse gas and aerosol precursor emissions based on social and economic projections, including population and economic growth, land use, energy availability, and fuel mix, during the period 1990 to 2100 (IPCC, 1995). Estimates of carbon dioxide emissions under these scenarios vary from 6 GtC5 per year to as much as 36 GtC5 per year by the year 2100, assuming optimistic figures of low population and economic growth to 2100. A midrange IPCC emission scenario, IS92a, projects an increase of 2 ◦ C in global mean surface temperatures by 2100 compared to 1990 (IPCC, 1995). The highest IPCC emission scenario (IS92e), on the other hand, gives a warming about 3.5 ◦ C under a ‘high’ value of climate sensitivity. The recent IPCC estimates are lower than what the models predicted earlier due primarily to lower emissions scenarios, inclusion of the cooling effect of sulfate aerosols, and improvements in the treatment of the carbon cycle. The model simulations show that the enhanced greenhouse effect driven largely by increasing concentrations of carbon dioxide will have potential impacts on other climatic variables (Hare, 1985). The model simulations predict an increase in global average precipitation induced by the increasing concentrations of greenhouse gases (Legates and McCabe, 1991). The increase in precipitation is predicted on average by 5–15% in middle and high latitudes in winter. Several climate models, on the other hand, predict that summer precipitation will actually decline in mid-continental areas (Schneider, 1989). It is also predicted that the global warming will enhance evaporation, and the increase in evaporation is predicted to overwhelm any increases in precipitation (Manabe and Wetherald, 1980). Climate change, if it occurs as projected, might have significant implications for agricultural and water resources. Many scientists argue that if the global climate patterns are altered by global warming, the agricultural potential of nations would significantly change (Rao, 1987). A majority of the countries affected by agricultural drought are in arid and semi-arid areas and in the dry regions of the world. Crop production in these areas is largely determined by climatic features. It is estimated that the warming would increase the water requirements of most crops, thus placing more demand on the water supplies available for irrigation (Williams and Balling, 1993). Warrick (1988) argues that in the midlatitude cereal regions, potential yields may decrease by 3–17 percent if an average warming of 2 ◦ C occurs. Viable acreage in arid regions of the western states and the Great Plains is expected to drop by nearly a third under a 3 ◦ C warming combined with a 10 percent decrease in precipitation (Schneider, 1989). Model simulations for Western Europe indicate that yields of grains have been found to decrease substantially in the Mediterranean area due to increased drought resulting from the combination of increased temperature and precipitation decrease (Kenny et al., 1993). One of the most important consequences of future climatic changes may be alterations in regional hydrologic cycles and subsequent changes in the quantity and quality of regional water resources (Gleick, 1987). Of particular concern are changes in the timing and magnitude of critical hydrologic variables – specifically runoff and soil moisture. Relatively small changes in temperature and precipita-

clim1790.tex; 9/10/1998; 11:58; p.2

POSSIBLE IMPACTS OF CLIMATE CHANGE ON SOIL MOISTURE AVAILABILITY

521

tion, together with the non-linear effects on evapotranspiration and soil moisture, can result in significant changes in runoff, especially in arid and semi-arid lands (IPCC, 1995). Even in areas where models predict an increase in precipitation, higher evaporation rates may lead to reduced runoff. Moreover, drier conditions may trigger or accelerate desertification by reducing the growth of important plant species. Concerns about the potential effects of climate change on water resources impel hydrologists and resource planners to know what kinds of planning and management approaches are needed to mitigate such effects. As the water supply is a key factor in determining agricultural potential, effects of climatic variables on the hydrologic components need to be quantified. As semi-arid and arid areas sometimes lack direct evidence of drought, and dryness in these areas may occur naturally, it is necessary to examine the sensitivity of the hydrologic components with respect to changes in climate. Water resources planners often use hydrologic modeling techniques – particularly modified water-balance models – to assess the sensitivity of regional watersheds to the predicted climatic changes. Water-balance models offer significant advantages over other methods in accuracy, flexibility, and ease of use (Gleick, 1986). At present, general circulation models (GCMs) only provide projections on a large spatial scale and in most cases they do not agree on a likely range of changes in climatic parameters for a given region. Also, lack of detailed energy balance measurements and the associated difficulty with applying complex climate models lead many scientists to use existing simple water-balance models to predict the hydrologic consequences of climate change (Katwijk et al., 1993). This research documents a preliminary water-balance assessment of the implications of climate change for monthly variations in soil moisture deficit, soil moisture surplus (runoff), and actual evapotranspiration in the Southeast Anatolia Project (GAP) region of Turkey (Figure 1). The study links a water-balance model to a GCM-derived IPCC estimate of temperature and precipitation estimates and various climate-change scenarios to draw some conclusions about the sensitivity of the selected water-balance components to climate change. The aim here is to explore the links between the projected climate changes and chief manifestations of these changes as expressed through the water-balance components. The objective is to define range and temporal patterns of the changes in the selected hydrologic components resulting from different combinations of temperature and precipitation changes. The Southeast Anatolia Development Project (GAP) is probably the most ambitious project undertaken by the Turkish government in its history and is extremely vital both for the national and regional economy. The GAP is a massive and planned agricultural and water resources development program within the Turkish portions of the Euphrates and Tigris river basins and seeks a viable development for Southeast Anatolia. The project aims at an integrated development plan encompassing a wide range of physical, social, and economic infrastructures. The GAP will include 15 dams, 14 hydroelectric power plants, 19 irrigation projects,

clim1790.tex; 9/10/1998; 11:58; p.3

522


Figure 1. Location of the study area.

and infrastructure supporting agriculture, transportation, and nonfarm employment opportunities. Once completed, the GAP project will irrigate 2,032,203 ha of land (34% of the all possible irrigated lands in Turkey) and produce 43.5 percent of the nation’s hydropower. The Euphrates and Tigris river basins together provide 28 percent of the nation’s water supply. In this respect, the Turkish government views the project as a ‘comprehensive integrated regional development project’. A project of such magnitude inevitably carries a major importance with respect to the future of the region’s water resources and agricultural potential. The project also has international implications. Syria and Iraq are concerned about construction activities that could reduce the annual flow of the Euphrates and Tigris through their countries. The GAP area is located in the Continental Mediterranean rainfall region where the annual precipitation varies between 400 and 800 mm. The effect of Mediterranean climate, however, decreases from west to east and precipitation gets more variable in the eastern parts. The annual precipitation also decreases from north to south in the region. Low-lying plains of the region close to Syria exhibit arid conditions and the annual precipitation falls as low as 300 mm. The greatest portion of the annual precipitation falls in winter, December and January being the wettest months. Summers in the region are very dry with respect to precipitation conditions coupled with high temperatures. It should be noted that the dry conditions mentioned here do not strictly refer to ‘meteorologic drought’, but rather refer to ‘agricultural drought’ expressed by lack of moisture that plants suffer from in a growing season (roughly April through October). Because of the variable behavior of temperature and precipitation in the area, the demand of irrigation and water supply naturally available for plant growth

clim1790.tex; 9/10/1998; 11:58; p.4


523

varies from site to site. Because of the arid conditions dominating the growing season in the region, traditional dry farming methods – primarily idle fallow – are practiced. Crops which are resistant to drought and are suitable for dry farming, such as wheat, barley, and cereals, are sown in large areas in the region. For the two major cereals, the GAP produces about 10 percent of the nation’s wheat and 11 percent of the barley. It is a major concern that crop yields could be reduced if severe changes in the climate of the region occur. This decrease would primarily result from higher temperatures coupled with high heat stress and low rainfall that are already affecting the region. In this respect, new agricultural drought management strategies may need to be developed in response to possible changes in soil moisture in the arable lands of the region. A major drawback with the GAP is that the present planned agricultural and water resources development programs do not take climate change into account, and risk of dryness that might be caused by future climate change is not included in the project evaluations. In this sense, a climatic impact assessment of hydrologic components of the region becomes even more critical.

2. Methodology for Regional Hydrologic Studies of Climate Change One of the most widely used methods to assess the regional hydrologic implications of climatic change is the water-balance model. A major problem, however, is that, in most cases, it is not possible to determine the ways in which climate will be affected by human actions. Thus, much of the efforts were directed towards the development of large-scale computer models – typically referred as ‘general circulation models’ (GCMs). These models permit us to evaluate changes in global climate patterns with the increasing concentrations of greenhouse gases. A major limitation of these models, however, is that they are unable to provide a detail assessment of regional or local impacts due to resolution problems. Once a general assessment is obtained from the GCMs, the predicted changes then can be incorporated into the water-balance models so that soil moisture and runoff characteristics of a region can be studied under the modified inputs. Almost all water-balance models aim to quantify the relationship between precipitation and runoff and keep track of changes in soil moisture and runoff with a specified water input such as precipitation. The advantages of this approach are its relative simplicity and minimal data requirements. On the other hand, most waterbalance models lack temporal resolution required for assessing dynamic behavior of river basins (Schaake, 1989). Many currently used water-balance models have evolved from the simple water-balance methods introduced by Thornthwaite in the 1940s (Mather, 1978). This study uses a computerized version of the original Thornthwaite water-balance model, which was developed by Willmot (1977). The model used here is the modified version of the original model and includes a snowpack option. The model requires data on precipitation, temperature, soil-water

clim1790.tex; 9/10/1998; 11:58; p.5

524


holding capacity, heat index, and latitude of a given station and computes potential and actual evapotranspiration, soil moisture deficits, and soil moisture surplus for a predetermined time period. In its most basic form, the water-balance method developed by Thornthwaite is a monthly, weekly, or daily comparison of water supply (precipitation) with the climatic demand for water (Mather, 1978). The typical data requirements for the model are monthly-average temperature and precipitation, field capacity, and latitude of the study area. The procedure involves three independent variables – precipitation (P ), potential evapotranspiration (PE), and soil moisture storage (ST). Potential evapotranspiration is defined as the water loss from a large homogenous, vegetation-covered area that never suffers from a lack of moisture and is primarily a function of climatic conditions. Other variables in the Thornthwaite’s waterbalance method include change in soil moisture (ST), actual evapotranspiration (AE), water deficit (D), water surplus (S), and runoff (RO), of which the function depends on the three independent variables. The Thornthwaite method for estimating monthly PE may be written as N 10Ta a1 l1 , (1) PE = 16 12 30 I where l is actual day length (h), N is the number of days in a month, Ta is the mean monthly air temperature (◦ C), and a1 is defined as a1 = 6.75 × 10−7 I 3 − 7.71 × 10−5 I 2 + 1.79 × 10−2 I + 0.49 ,

(2)

where I is a heat index derived from the sum of 12 monthly index values, i, obtained from 1.514 Ta (3) i= 5 (Thornthwaite and Mather, 1957). Whenever precipitation exceeds the climatic demand for water, the soil moisture begins to increase. When the soil moisture reaches field capacity, a water surplus develops resulting in increased runoff from the area. Thus, surplus is the excess of P–PE whenever ST is at field capacity. In other words, when P exceeds PE, AE is equal to PE, and the excess water of PE replenishes ST. It should, however, be noted that surplus does not always occur when P–PE is positive. No surplus can develop as long as soil moisture storage is below the field capacity. On the other hand, if precipitation fails to supply the climatic demand for water, the soil moisture will be depleted faster than it can be replenished and a moisture deficit may result. In other words, a water deficit (D) occurs when P and soil moisture withdrawal are less than PE. By comparing precipitation and potential evapotranspiration for a given time period, it is possible to obtain quantitative estimates of soil water storage, water surplus or excess water above climatic demands (which

clim1790.tex; 9/10/1998; 11:58; p.6


525

will be contributed to oceans and streams as runoff), and water deficit or climatic demands for water not supplied by available precipitation or stored as soil moisture. It should, however, be noted that Thornthwaite’s method tends to underestimate evapotranspiration compared to Blaney-Criddle’s method (Kolars and Mitchell, 1991). On the other hand, such low estimates may be taken to represent the absolute minimum amount of water a field crop needs. Complete details of water-balance computations are described by Thornthwaite and Mather (1955). The method has been used in various climates around the world and has been found to provide reasonable estimates of monthly and annual runoff (Thornthwaite and Mather, 1955; Mather and Feddema, 1988). The effects of global warming on water-balance components in 12 regions of the world with the outputs from the GCMs were evaluated by Mather and Feddema (1988). They found that most regions exhibited an increase in annual water deficit and a decrease in summer soil moisture, and that in turn enhanced the dryness that might decrease the water availability. Likewise, Gleick (1987) developed and tested water-balance modelling techniques for the Sacramento Basin in California and found large decreases in summer soil moisture levels and summer runoff volumes for all 18 climate change scenarios.

3. Climatic Impact Assessment In this study, the effects of climatic change are evaluated in two ways. The first way is to determine the sensitivity of the selected water-balance components to hypothetical changes in the magnitude and temporal distribution of precipitation and temperature. The advantage of this approach is to test the sensitivity of specific regions to climatological variations and provide an opportunity to evaluate possible transient climatic responses. Schneider and Thompson (1981) argue that the transient responses of the climate may be quite different. Secondly, IPCC estimates of regional changes in temperature and precipitation predicted by the GCMs will be used as input to estimate the impacts of the predicted climatic changes in water-balance components of the region. The prescribed changes in temperature and precipitation were obtained from reviews of climate change approaches by Gleick (1986). The scenarios are listed in Table I and include ten hypothetical cases involving combinations of plus 2 and plus 4 degrees Celsius (◦ C) and plus and minus 0, 10, and 20% precipitation changes. The precipitation scenarios compare well with precipitation changes generated from GCM data and are similar to assumptions made by previous hydrologic studies for climatic impact assessment (Gleick, 1987). None of the scenarios selected includes decreases in average monthly temperature since the models always demonstrated that increasing concentrations of atmospheric trace gases will lead to increases in the surface air temperature at both global and regional scales. The general circulation models (GCMs) estimate a warming of 2 ◦ C in winter and 2–

clim1790.tex; 9/10/1998; 11:58; p.7

526


TABLE I Hypothetical climate-change scenarios Change in temperature

Change in precipitation

T T T T T

plus 2 ◦ C plus 2 ◦ C plus 2 ◦ C plus 2 ◦ C plus 2 ◦ C

No change –10% –20% +10% +20%

T T T T T

plus 4 ◦ C plus 4 ◦ C plus 4 ◦ C plus 4 ◦ C plus 4 ◦ C

No change –10% –20% +10% +20%

3 ◦ C in summer months for Southeastern Europe and most parts of Turkey where a Mediterranean climate predominates (IPCC, 1990). The estimated changes by the IPCC are based on ‘business-as-usual’ emissions and generated by the highresolution models, scaled to be consistent with an estimate of global warming of 1.8 ◦ C by 2030. The models also predict a slight increase in winter precipitation, while in summer, precipitation decreases by 5%–15% and soil moisture by 15%–25% in the region if the predicted warming takes place. There are several high-resolution GCM simulation results that give a general pattern of temperature and precipitation changes for Turkey. For example, the Canadian Climate Center model predicts nearly 8 ◦ C warming due to doubling of CO2 for December through February (IPCC, 1990). GFDL, on the other hand, gives less pessimistic figures and predicts a 4 ◦ C warming for the same period. Both models also predict a 4 ◦ C warming for June through August when atmospheric CO2 is doubled. As stated earlier, although the GCMs are useful in identifying climatic effects on a global scale, they have drawbacks in regional assessments. Hansen (1988) warns that projections of future water availability on a regional scale by current GCMs may be unreliable. The GCMs use a large-scale grid system to represent global atmospheric circulation. For example, the GISS model uses a grid of 8◦ latitude by 10◦ longitude and the GFDL model uses a 4◦ by 8◦ grid. In this study, the GCM-derived IPCC temperature and precipitation changes projected for most parts of Southern Europe and parts of Turkey where a Mediterranean climate predominates are used. Since the subregions within the GAP region are found to be smaller than a grid box and cannot be accurately represented by the GCM data, the results were treated as one scenario for the entire region.

clim1790.tex; 9/10/1998; 11:58; p.8


527

Figure 2. Major agricultural and water resources project regions of GAP.

The study covers 13 major agricultural and water resources project regions (Figure 2). There are 28 climate stations within and nearby the study area. A grouping of the stations was made within each individual project area. Using the Thiessen method, representative values of precipitation, temperature, and moisture field capacity were constructed for each project area and thus, each project area was treated as a single unit of a study site.

4. Climatic Impact Assessment Results The ten hypothetical cases of temperature and precipitation changes, involving combinations of plus 2 and plus 4 degrees Celsius (◦ C) and plus and minus 0, 10, and 20% precipitation changes, were applied to the monthly water balance of the thirteen sub-regions in GAP, while the GCM-produced IPCC temperature and precipitation changes were used for the entire GAP region. The results then were plotted as differences from the present long-term monthly water-balance components, namely actual evapotranspiration, soil moisture deficit, and soil moisture surplus. The subregions which exhibited similar temporal patterns and range of changes were categorized under one group and a representative subregion from each group was illustrated. Percent changes in the selected water-balance components on an annual basis are also listed in Table II.

clim1790.tex; 9/10/1998; 11:58; p.9

Soil moisture deficit (mm) Ad-Göksu 539 627 Ad-Kahta 555 644 Batman 590 676 Bat-Silvan 527 614 Boz-Urfa 646 738 Cizre 792 883 Dic-Kralkızı 540 626 Garzan 561 648 Gaziantep 517 612 Mar-Ceylan 675 761 Siv-Hilvan 599 688 Sur-Baziki 544 639 Urf-Harran 675 768 16.3 16.0 14.6 16.5 14.2 11.5 15.9 15.5 18.4 12.7 14.9 17.5 13.8

–4.2 –8.0 –7.3 –5.4 –10.8 0.0 –7.0 –5.4 –12.1 –15.8 –8.8 –21.1 –43.5

Present 2 ◦ C %

Soil moisture surplus (mm) Ad-Göksu 406 389 Ad-Kahta 199 183 Batman 220 204 Bat-Silvan 296 280 Boz-Urfa 102 91 Cizre 0 0 Dic-Kralkızı 227 211 Garzan 297 281 Gaziantep 116 102 Mar-Ceylan 101 85 Siv-Hilvan 182 166 Sur-Baziki 57 45 Urf-Harran 23 13

Sub-regions

719 725 758 697 827 975 709 730 713 848 768 741 859

373 166 188 264 82 0 194 265 91 74 149 34 2

33.4 30.6 28.5 32.3 28.0 23.1 31.3 30.1 37.9 25.6 28.2 36.2 27.3

–8.1 –16.6 –14.5 –10.8 –19.6 0.0 –14.5 –10.8 –21.6 –26.7 –18.1 –40.4 –91.3

4 ◦C %

617 631 667 604 728 854 616 638 602 755 680 629 759

453 230 249 337 125 0 260 335 140 123 212 72 40

14.5 13.7 13.1 14.6 12.7 7.8 14.1 13.7 16.4 11.9 13.5 15.6 12.4

11.6 15.6 13.2 13.9 22.5 0.0 14.5 12.8 20.7 21.8 16.5 26.3 73.9

2◦ −P + 10 %

607 618 658 594 722 825 607 628 595 749 673 620 751

517 277 294 395 163 0 310 390 181 160 257 100 68

12.6 11.4 11.5 12.7 11.8 4.2 12.4 11.9 15.1 11.0 12.4 14.0 11.3

27.3 39.2 33.6 33.4 59.8 0.0 36.6 31.3 56.0 58.4 41.2 75.4 195.7

2◦ −P + 20 %

636 658 686 625 749 912 635 657 624 771 696 648 792

326 136 159 223 58 0 161 227 66 52 121 18 0

18.0 18.6 16.3 18.6 15.9 15.2 17.6 17.1 20.7 14.2 16.2 19.1 17.3

–19.7 –31.7 –27.7 –24.7 –43.1 0.0 –29.1 –23.6 –43.1 –48.5 –33.5 –68.4 0.0

2◦ −P − 10 %

646 671 695 635 760 942 645 667 636 783 705 667 828

262 90 114 165 25 0 111 173 30 20 77 0 0

19.9 20.9 17.8 20.5 17.6 18.9 19.4 18.9 23.0 16.0 17.7 22.6 22.7

–35.5 –54.8 –48.2 –44.3 –75.5 0.0 –51.1 –41.8 –74.1 –80.2 –57.7 0.0 0.0

2◦ −P − 20 %

709 712 749 687 816 946 699 720 701 837 761 732 850

437 213 233 321 115 0 244 319 128 106 195 62 30

31.5 28.3 26.9 30.4 26.3 19.4 29.4 28.3 35.6 24.0 27.0 34.6 25.9

7.6 7.0 5.9 8.4 12.7 0.0 7.5 7.4 10.3 5.0 7.1 8.8 30.4

4◦ −P + 10 %

699 699 739 677 805 917 690 711 689 830 753 722 842

500 260 278 379 148 0 294 373 164 142 240 89 58

29.7 25.9 25.3 28.5 24.6 15.8 27.8 26.7 33.3 23.0 25.7 32.7 24.7

23.2 30.7 26.4 28.0 45.1 0.0 29.5 25.6 41.4 40.6 31.9 56.1 152.2

4◦ −P + 20 %

TABLE II Percent changes in annual soil moisture deficit and surplus with the prescribed scenarios

729 738 767 707 838 1004 718 740 725 860 779 751 893

309 119 143 206 48 0 144 211 55 42 106 7 0

35.3 33.0 30.0 34.2 29.7 26.8 33.0 31.9 40.2 27.4 30.1 38.1 32.3

738 752 776 717 849 1034 727 750 736 871 793 781 930

36.9 35.5 31.5 36.1 31.4 30.6 34.6 33.7 42.4 29.0 32.4 43.6 37.8

–39.4 –63.8 –55.5 –49.7 –85.3 0.0 –58.1 –47.5 –83.6 –90.1 –63.2 0.0 0.0

4◦ −P − 20 %

–23.9 246 –40.2 72 –35.98 –30.4 149 –52.9 15 0.0 0 –36.6 95 –29.0 156 –52.6 19 –58.4 10 –41.8 67 –87.7 0 0.0 0

4◦ −P − 10 %

528 ALI UMRAN KOMUSCU ET AL.

clim1790.tex; 9/10/1998; 11:58; p.10


A.

+2 ◦ C WARMING

529

WITH PRESCRIBED CHANGES IN PRECIPITATION

A 2 ◦ C warming with no precipitation change and prescribed changes in precipitation (+10 and –10 percent) under a 2 ◦ C warming are applied to monthly waterbalances of the thirteen sub-regions of the GAP. Figure 3 illustrates changes in actual evapotranspiration (AE) from the present. With respect to timing and magnitude of the changes in the actual evapotranspiration, the subregions were classified roughly into three groups. The first group included Cizre, which showed the most extreme response to the given climate change scenarios. It portrayed a well-marked response particularly to the changes in precipitation. The next group consisted of the regions which indicate extreme response to the given climate change scenarios to a lesser degree and included the Urfa-Harran and Adıyaman-Kahta subregions, which are very sensitive particularly to the decrease in precipitation. The rest of the subregions were placed in the third group where changes in AE to the given climate change scenarios gain less variable character. Response to the change in temperature and precipitation portrayed very little variations from one scenario to the next, and the range of the changes remained minimal. A noticeable characteristic of changes in AE, in general, was the increase in water use in spring and early summer throughout the region. The actual water use decreased severely during the summer due to insufficient precipitation and lack of moisture in the soil. In other words, the available water through precipitation and soil moisture does not meet the moisture demand of crops, resulting in a serious agricultural drought, if not supplemented by irrigation. As the region receives most of its rain during the winter months, increases in precipitation cannot meet the substantial demand for moisture which arises in summer. The demand for water under the projected climate change will be higher in the south-central and southeastern parts of GAP, reaching its maximum in the Cizre subregion. In Figure 4, changes in the soil moisture deficit (D) from the present are presented for the 2 ◦ C warming and prescribed precipitation changes. A grouping was made in order to identify the subregions which portrayed similar trends and changes. As a result, four different groups of subregions were determined. The first group included the Cizre which exhibited the greatest increase in soil moisture deficit. Even a 20 percent increase in precipitation was insufficient to overcome the summer soil moisture deficit in this subregion. In the second group of subregions, response to the given climate change scenarios was not as dramatic as that observed in the first group but it was still significant. In particular, a 20 percent decrease in precipitation coupled with 2 ◦ C warming caused a significant increase in summer soil moisture deficit. The third group had the largest geographical coverage and included 7 subregions. The changes in soil moisture deficit in this group did not vary considerably from one subregion to the next, and the magnitude of the changes remained relatively small compared to the first two groups. The final group had

clim1790.tex; 9/10/1998; 11:58; p.11

530


Figure 3. Changes in actual evapotranspiration from the present with +2 ◦ C warming and prescribed precipitation changes.

clim1790.tex; 9/10/1998; 11:58; p.12


531

Figure 4. Changes in soil moisture deficit from the present with +2 ◦ C warming and prescribed precipitation changes.

clim1790.tex; 9/10/1998; 11:58; p.13

532


similarities with the third group, but it greatly differed from it in the sense that the soil moisture deficit peaked in summer but then declined gradually. A general behavior observed in all the groups is that part of the spring and the entire summer in the region will be affected by severe water shortage. Enhancement in the soil moisture deficit even extends into early autumn in most cases. The severity of the dryness increases during this period since the moisture demand is not satisfied by precipitation. Regardless of the increase in precipitation, a 2 ◦ C warming will be sufficient to cause an extended period of drought throughout the year in the region. In all the subregions, 2 ◦ C of warming coupled with a 20 percent decrease in precipitation causes a major soil moisture deficit in the region. The extent of the drought also varies geographically. The effect of warming in the summer deficit is most evident in the Cizre subregion. The timing of the deficit period shifts one month ahead in the southeastern parts and starts in April. June is the month in which the demand for moisture peaks in all the subregions. Interestingly, soil moisture surplus also increases on an annual basis with an increase in precipitation, but the enhancement in the soil moisture deficit seems to overwhelm that increase, still creating a demand for moisture (Table II). Response of the soil moisture surplus (S) to various combinations of 2 ◦ C of warming and precipitation changes is illustrated in Figure 5 for individual groups of subregions. The Cizre subregion is not included in the grouping as it exhibited no quantitative change due to a lack of soil moisture surplus throughout the year. Roughly four groups of subregions were determined with respect to the changes in soil moisture surplus under the given climate change scenarios. The first group included Adıyaman-Kahta where a significant response was marked at both 20 percent increase and decrease in winter months. The response indicates that precipitation seems to be the major factor in changes in soil moisture surplus in the region. The second group included three subregions and exhibited a similar pattern to the first group, except that no peak was observed in February when the precipitation increased 20 percent. The third group subregions were characterized by a more variable response in winter months and then a declining trend in the following months. The third group also had the largest areal coverage, including 7 of the subregions. The final group included two subregions where the peak response shifted to March when precipitation was decreased by 10 and 20 percent. Interestingly, soil moisture surplus exhibited nearly the same range of change to both 10 and 20 percent decrease in precipitation. This indicates that decrease in precipitation below a certain threshold value does not cause a significant change in the soil moisture surplus. In general, the largest changes in the soil moisture are seen with the 20 percent increase and decrease in precipitation. Seasonal variations in the soil moisture surplus changes exhibit noticeable differences among the subregions. In the Adıyaman-Kahta, Bozova-Urfa, Gaziantep, and Mardin-Ceylanpınar subregions, the changes in the surplus from the present exhibit a rapid decline in February, when the precipitation is decreased by 10 and 20 percent. In other words, even

clim1790.tex; 9/10/1998; 11:58; p.14


533

Figure 5. Changes in soil moisture surplus from the present with +2 ◦ C warming and prescribed precipitation changes.

clim1790.tex; 9/10/1998; 11:58; p.15

534


the winter precipitation increase in these regions is not in sufficient quantity to yield a high moisture surplus. Except in the cases of 10 and 20 percent increase in precipitation, much of the winter precipitation in these regions replenishes soil moisture deficits, leaving less water available for winter surplus. Where precipitation was not changed, total annual surplus decreased by 4 percent to 43 percent in the region (Table II). The largest decrease in the annual soil moisture surplus was observed in the Urfa-Harran subregion with 43 percent. B.

+4 ◦ C WARMING

WITH PRESCRIBED CHANGES IN PRECIPITATION

A grouping based on the temporal pattern and range of changes in actual evapotranspiration resulted in four different groups of subregions. Cizre comprised the first group which was characterized by the most profound response to the given climate change scenarios. The changes in AE in Cizre were larger in magnitude and peaked earlier than the other subregions. In the second group, in which UrfaHarran was placed, response to change in temperature and precipitation was less but still significant, particularly for the 20 percent decrease in precipitation. The Adıyaman-Kahta subregion, which was placed in the third group, exhibited interesting features. While the response of AE to different scenarios remained nearly the same in winter and spring months, the magnitude of the response in summer differs noticeably from one scenario to the next. All the other 10 subregions were placed in the fourth group and temporal trends in the changes in AE nearly resembled each other in this group. The largest changes are observed in May and June. The response of the actual evapotranspiration to +4 ◦ C increase in temperature and combined changes in precipitation exhibited several distinct features (Figure 6). A gradual increase in AE during the winter months, a rapid decline with the beginning of summer, and then again a gradual rise toward the late summer and early autumn are common in almost all the subregions. It is obviously seen that winter precipitation in the region is in sufficient quantity to meet climatic water demand in early spring, even with 4 ◦ C increase in temperature. The changes in AE from the present is higher in early spring, reaching a peak in May. The largest are changes in AE are observed when precipitation is either increased or decreased by 20 percent. This is especially pronounced in the Cizre and Urfa-Harran subregions. Figure 6 indicates that, regardless of the changes in precipitation, actual water used by crops will decrease in summer in the region since a greater portion of the precipitation falls in winter. In other words, a severe moisture shortage in summer months is inevitable in the region if not supplemented by irrigation. It also should be stressed that in the case of a +4 ◦ C increase in temperature, even winter precipitation, in most cases, fails to recharge the soil moisture storage to meet the moisture demands of plants in summer months. As the drier conditions dominate the region in summer, soil moisture will be depleted even faster. Coupled with the antecedent moisture conditions, this will put more pressure on the demand for water for a healthy crop growth.

clim1790.tex; 9/10/1998; 11:58; p.16


535

Figure 6. Changes in actual evapotranspiration from the present with +4 ◦ C warming and prescribed precipitation changes.

clim1790.tex; 9/10/1998; 11:58; p.17

536


Figure 7. Changes in soil moisture deficit from the present with +4 ◦ C warming and prescribed precipitation changes.

Severity of the water shortage in the GAP region is more pronounced in Figure 7, which shows the monthly changes in soil moisture deficit from the present with the +4 ◦ C increases in temperatures. The grouping was made in order to classify the subregions, which are characterized by similar temporal trends and quantitative change. Cizre is still the most sensitive to the changes in climate change scenarios. In the second group, less responsive subregions, such as SuruçBaziki and Urfa-Harran, are placed. The most noticeable aspect of the changes in soil moisture deficit in those regions is the high response to the 20 percent decrease in precipitation in spring and early summer. The rest of the 10 subregions are

clim1790.tex; 9/10/1998; 11:58; p.18


537

classified under one group and exhibit similar trends in monthly changes of the soil moisture deficit. The subregions in this group typically have two peaks during the year and do not portray marked differences to the prescribed changes in climate from one scenario to the next. In general, an extended period of water deficit, lasting almost 7 months, dominates throughout the region. Western and southeastern parts of the region are affected most by the water shortage. As winter precipitation in the region is in sufficient quantity, +4 ◦ C warming is not able to cause a major soil moisture deficit during the period of November through March. On the other hand, as the climatic demand for water increases more than precipitation during spring and summer, a +4 ◦ C warming decreases water availability in all the subregions, causing a dry period lasting 6–7 months out of each year. This extended period of soil moisture deficit is attributed to the increased spring and summer evapotranspiration and lack of precipitation. At least a 20–30 percent increase in precipitation would be necessary to maintain the summer soil moisture content at its present level with a warming of 4 ◦ C. As expected, the largest changes in D occur with +4 ◦ C warming combined with 20 percent decrease in precipitation. In the Cizre subregion, the change from the present D reached nearly 90 mm in May. Note that the beginning of the water shortage period in the region varies, too. In most parts of the region, soil moisture deficit begins in May and peaks in June. On the other hand, in the most southeastern parts where the annual precipitation falls as low as 300 mm in a year, soil moisture deficit starts in April and the change in D from the present is marked with larger differences. In other words, the timing of D shifted one month ahead in the southeastern portion of the GAP region. Another interesting conclusion is that with a warming of 4 ◦ C, the effect of increasing precipitation on reducing the soil moisture deficit is insignificant. In most subregions, increasing precipitation even 20 percent had little impact on reducing the summer soil moisture deficit. It is possible to conclude that in areas where summers are very dry, additional precipitation increases would not eliminate the adverse impact of warming on the soil moisture content. Changes in the soil moisture surplus from the present with 4 ◦ C of warming and prescribed precipitation changes are presented in Figure 8. It is interesting that the same grouping identified for the 2 ◦ C of warming scenarios are formed with the scenarios applied here. Monthly trends in the changes were quite similar to those observed with 2 ◦ C of warming and did not present significant differences. The change in the soil moisture surplus from the present clearly reflects the effect of warming on how much water would be available for runoff. Even in the case of a 20 percent increase in precipitation, no moisture surplus occurs June through December in most parts of the region (Figure 8). Only an additional increase in winter precipitation causes slight rise in the winter surplus. Especially in the central and northern parts of the region, some increase in winter surplus occurs, but the subsequent spring and summer runoff would be greatly reduced even with a 20 percent increase in precipitation. Where precipitation was not changed, the total

clim1790.tex; 9/10/1998; 11:58; p.19

538


Figure 8. Changes in soil moisture surplus from the present with +4 ◦ C warming and prescribed precipitation changes.

clim1790.tex; 9/10/1998; 11:58; p.20


539

annual surplus decreased by 8 percent to 91 percent in the region (Table II). The decrease in the annual soil moisture surplus is particularly pronounced in the UrfaHarran subregion. While the effect of increased soil moisture lasts until May in the northern parts of the region, the winter moisture surplus is depleted earlier in the south-central and southeastern parts where the drier conditions predominate 6–7 months out of each year, causing less water to be available for crops during the growing season. C . RESULTS OF THE GCM SCENARIOS

The IPCC Business-as-Usual scenario based on the GCMs predicts a 2 ◦ C warming in winter and 2–3 ◦ C in summer, some precipitation increase in winter, and a 5–15% precipitation decrease in summer for Southeastern Europe and the Mediterranean Region. The same models also predict a 15–25% decrease in summer soil moisture. Changes in the water-balance components from the present with the GCM estimates exhibit marked differences compared to the changes observed with the hypothetical scenarios. It is interesting to note that soil moisture surplus increased by nearly 10 percent on an annual basis (Table III). By contrast with the hypothetical scenarios, the changes in precipitation had a considerable impact on the soil moisture changes. The summer deficit enhanced 15 percent, but this was not mainly because of the increased temperatures in summer, but rather a decrease in the already low precipitation, which had less effect on summer soil moisture in the region. Similarly, the effect of slight increase in winter precipitation is also reflected in soil moisture surplus in the first few months. A 10 percent increase in the soil moisture surplus was observed in the January–April period for the entire region. The GCM results indicate that the actual water use peaks in May and continues to peak in June, thus there is a slight shift in the timing of peak of actual water use. Another significant difference with timing and magnitude of the actual water use with this scenario is seen in the spring. While the GCM results indicate less water use in the spring, the hypothetical scenarios exhibit a rising demand for water for the same period. It is likely that a 5 percent increase in winter precipitation meets the moisture demands of crops to some degree and supplies some moisture for use by crops, causing lower AE rates. It also should be reminded that the GCM-derived IPCC estimates do not project a continuous temperature decrease throughout the year in contrast to the hypothetical scenarios.

5. Conclusions Results of this study suggest that the projected climate change can have major impacts on the timing and magnitude of soil moisture changes in agricultural areas of the GAP region. Among the most significant results obtained in this study are

clim1790.tex; 9/10/1998; 11:58; p.21

540


TABLE III Changes in the water-balance components after GCM predictions IPCC based on business-as-usual scenario Month Temp. Baseline

UPE

APE

Prec.

Diff.

ST

DST

1 2 3 4 5 6 7 8 9 10 11 12

4 7 21 45 82 131 163 158 119 66 27 8

3 6 21 49 99 161 202 184 123 63 23 7

82 74 75 70 43 5 1 1 2 28 58 77

78 68 53 21 –56 –155 –201 –183 –121 –35 35 70

162 162 162 162 106 26 4 1 0 0 35 105

57 0 0 0 –56 –80 –22 –3 –1 0 35 70

941

515

Changes after IPCC estimates 1 5.7 6 5 2 7.2 10 8 3 9.5 17 18 4 14.5 40 44 5 20.2 78 95 6 29.2 157 192 7 33.1 177 221 8 32.4 175 204 9 24.9 118 121 10 17.9 6.2 59 11 11.0 23 20 12 7.7 11 9

86 78 75 70 43 5 1 1 2 28 58 81

Total

Total

3.7 5.2 9.5 14.5 20.2 26.2 30.1 29.4 24.9 17.9 11.0 5.7

996

526

80 69 57 26 –52 –187 –220 –203 –119 –31 38 71

162 162 162 162 110 20 3 0 0 0 39 110

52 0 0 0 –52 –90 –17 –2 0 0 38 71

AE

Def.

Surp.

3 6 21 49 99 85 23 4 2 28 23 7

0 0 0 0 0 75 179 180 120 35 0 0

22 68 53 21 0 0 0 0 0 0 0 0

351

590

165

5 8 18 44 95 95 18 3 2 28 20 9

0 0 0 0 0 97 202 201 119 31 0 0

28 69 57 26 0 0 0 0 0 0 0 0

346

650

180

clim1790.tex; 9/10/1998; 11:58; p.22


541

Figure 9. Adaptation strategies to climate change in GAP region.

decreases in soil moisture availability during the critical parts of the year across the region. Most parts of the region showed an increase in annual water deficit and a decrease in annual water surplus. Subregional variations in the selected components also exhibited interesting spatial patterns. The south-central and southeastern parts of the region will be affected by dry conditions most. Particularly the Cizre subregion, which is located at the most southeastern end of the region, is characterized by dramatic responses to all the hypothetical scenarios. The variations are generally consistent in the other parts with respect to timing and range of the changes. Both precipitation and temperature seem to be the principal factors that influence water availability in the drier parts of the region, and the evapotranspiration demand overwhelms the precipitation demand most of the year. As the precipitation increases from south to north, water demand decreases in the same direction as well. In the south-central part of the region, particularly in the Urfa-Harran, precipitation is the only determinant factor of soil moisture availability. On the western parts of the region, on the other hand, soil-water capacity overwhelms the influence of precipitation and of temperature on water availability.

clim1790.tex; 9/10/1998; 11:58; p.23

542


As the region receives most of its precipitation during the winter, an increase in precipitation in other months will not eliminate the adverse impact of warming on soil moisture. More than a 20 percent increase in precipitation in most parts of the regions will be necessary to counteract a 4 ◦ C warming, although the magnitude will be different. Where precipitation was not changed, total annual surplus decreased by 4 percent to 43 percent for a 2 ◦ C warming and by 8 percent to 91 percent for a 4 ◦ C warming. Largest decreases in the soil moisture surplus were observed in the Urfa-Harran subregion when the precipitation was not changed. This is a very important implication for the agricultural activities in the Urfa-Harran as most of the arable lands in the area are planned to be irrigated. Summer soil moisture deficits enhanced substantially with all the scenarios. It was observed that even a 20 percent increase in precipitation had very little effect on preventing soil moisture from being depleted faster if the predicted warming would take place. In other words, the projected temperature changes would be responsible for the greater portion of the enhancement in the summer deficits in the GAP region as a result of increased evapotranspiration rates. The severity of enhancement of the soil moisture deficit increases in the south-central and southeastern parts which already suffer from low precipitation. It was also observed that the least effect on the soil moisture deficit enhancement is seen with the GCM temperature and precipitation changes. This can be attributed to the fact that the GCMs do not project a continuous temperature decrease throughout the year in contrast to the other scenarios. Winter surplus decreased in scenarios with increased temperature and decreased precipitation in all subregions. Only when the winter surplus was enhanced were there 10 and 20 percent increase in precipitation. Interestingly, the Cizre sub-region responded neither to any of the scenarios nor to the GCM projected changes. It was apparent that more than a 20 percent increase in precipitation was necessary to yield even a little moisture surplus in Cizre due to very low precipitation coupled with high evapotranspiration rates. In central western parts of the region, with the increase in precipitation, winter surplus peaks in February, but then shows a rapid decline in the subsequent spring months. Actual evapotranspiration (ET ) increased in the first few months of the growing season and then exhibited a rapid decline as the moisture became very limited during the summer with all the scenarios. The higher rates of actual evapotranspiration in the spring months can be attributed to winter precipitation and moist soil conditions during this period. This, in turn, means crops can only use water supplied through precipitation and what is available in the soil. The demand for moisture need enhances with the 4 ◦ C warming as the moisture supply becomes limited with high evapotranspiration rates. With the 2 ◦ C warming, the demand for water by crops relatively decreases. Timing of the high and low peaks of actual evapotranspiration shifts slightly in the GCM predicted temperature and precipitation changes. Due to a non-increase in spring temperatures, the demand for actual water use is not significant in spring with the GCM projection. On the other hand,

clim1790.tex; 9/10/1998; 11:58; p.24


543

a 10 percent decrease in summer precipitation limits the moisture supply for crops and the demand suddenly increases in summer with the IPCC projection.

6. Concluding Remarks The results stressed here may have significant implications for the future of the GAP as the present project evaluations do not include climate change concerns for resource-use and development decisions concerning the project area. Even a 2 ◦ C warming alone indicates a certain degree of impact on the soil moisture conditions in all the subregions. Soil moisture variations may become a major limiting factor in crop production in agricultural areas of the GAP region. Intensification of the dryness during the growing season in the region means a higher demand for water by crops. A prolonged deficiency of soil moisture may retard the growth of the plant and lead to the incidence of agricultural drought. This, in turn, means enhancing the water supplies in the region and preventing crops from being subjected to extended periods of drought and resulting crop failures. Our conclusion from this study is that various adaptation strategies are needed to combat the agricultural drought in the region. Adaptation to climate change via new crops and crop varieties, applying more advanced dry farming methods, improved water management, developing more efficient irrigation systems, and changes in planting will be important in limiting the negative effects of the warming and taking advantage of beneficial changes in climate (Figure 7). As stated earlier, due to present dryness in the region, traditional dry farming methods are practiced. Crops which are resistant to drought such as wheat, barley, and cereals are sown in the region. It is then necessary to examine the response of these crops to even drier conditions in terms of crop yield. Continued and substantial improvements in crop yields would be needed to offset the adversity of the warming and decreased summer soil moisture in the region. A new dry farming method may need to be applied to counteract the anticipated dryness in the region as a result of the predicted warming. Another solution to mitigate the adverse impacts of drought on crops would be to switch to more heat- and drought-resistant crop varieties. Farmers may also try to plant two crops during a growing season, and plant and harvest earlier. Technological improvements, such as improved crop varieties from bioengineering, would be needed to keep up with the changes in climate. Soil can be protected by applying proper water conservation practices to reduce the potential impacts of drought in the region. Of course, irrigation practices are strongly advised in the drought-prone areas for higher crop yields. But, irrigation will be the first activity affected in the GAP region where water availability will be diminished as a result of decreased summer precipitation and soil moisture. Higher temperatures and enhancement in the summer soil moisture deficit certainly will boost the demand for irrigation. This, in turn, means more irrigation equipment will be installed in drylands of the region and farmers already irrigating would need to extract more water from

clim1790.tex; 9/10/1998; 11:58; p.25

544


surface and groundwater sources. On the other hand, if large areas are planned to be irrigated in the region, then the problem of optimum water use rises as the demand for water by other sectors is anticipated as well. Also, additional water use from ponds and reservoirs for irrigation would have a significant impact upon the downstream river system. The efforts then should be directed toward reducing the demand for water in drought-prone areas of the region and taking comprehensive measures to adjust water divisions within the subregions. It should, however, be noted that the incremental cost of these adaptation strategies suggested here could create a serious burden for Turkey which has already invested billions of dollars in the GAP. But, we eventually believe that improved water management in the GAP that integrates future changes in climate may protect the water user at minimal cost. Otherwise, there will be substantial economic, social, and environmental costs, particularly in parts of the region that are already water-limited and where there is considerable competition among users.

References Gleick, P. H.: 1986, ‘Methods for Evaluating the Regional Hydrologic Effects of Global Climate Changes’, J. Hydrol. 88, 97–116. Gleick, P. H.: 1987, ‘Regional Hydrologic Consequences of Increases in Atmospheric CO2 and other Trace Gases’, Clim. Change 10, 137–161. Gribbin, J.: 1984, ‘What on Earth is Happening?’ Sunday Telegraph Magazine, 23–28. Haan, C. H.: 1977, Statistical Methods in Hydrology, Iowa State University Press. Ames. Iowa. Hansen, J., Fung, I., Lacis, A., Rind, D., Lebedeff, S., Ruedey, R., and Russel, G.: 1988, ‘Global Climate Changes as Forecast by Goddard Institute for Space Studies Three-Dimensional Model’, J. Geophys. Res. 93 (D8), 9341–9364. Hare, F. K.: 1985, ‘Climatic Variability and Change’, in Climatic Impact Assessment, Wiley, London, pp. 37–68. IPCC: 1990, Climate Change: The IPCC Scientific Assessment, WMO/UNEP/IPCC, Cambridge University Press, Cambridge, 365 pp. IPCC: 1995, Climate Change 1995 – Impacts, Adaptations and Mitigation of Climate Change: Scientific-Technical Analyses, Contribution of Working Group II to the Second Assessment Report of the Intergovernmental Panel on Climate Change, WMO/UNEP/IPCC, Cambridge University Press, Cambridge, 878 pp. Katwijk, V. F., Rango, A., and Childress, A. E.: 1993, ‘Effects of Simulated Climate Change on Snowmelt Runoff Modelling in Selected Basins’, Water Res. Bull. 29 (5), 755–765. Kenny, G. J., Harrison, P. A., Parry, M. L. (eds.): 1993, The Effect of Climate Change on Agricultural and Horticultural Potential in Europe, Environmental Change Unit, University of Oxford, Oxford, U.K., 224 pp. Kolars, J. F. and Mitchell, W. A.: 1991, The Euphrates River and the Southeast Anatolia Development Project, Southern Illinois University Press, Carbondale and Edwardville, 323 pp. Legates, D. R. and McCabe, G. J. Jr.: 1991, Reliability of Precipitation Estimates for DoubledCO2 Scenarios Simulated with Two General Circulation Models, Proceedings of the American Meteorological Society Special Session on Hydrometeorology, Salt Lake City, Utah, pp. 200–203. Manabe, S. and Wetherald, R. T.: 1980, ‘On the Distribution of Climate Change Resulting from an Increase in CO2 -Content of the Atmosphere’, J. Atmos. Sci. 37, 99–118.

clim1790.tex; 9/10/1998; 11:58; p.26


545

Mather, J. R.: 1978, The Climatic Water-Balance in Environmental Analysis, D.C. Heath and Company, Lexington, Massachusetts, 239 pp. Mather, J. R. and Feddema, J.: 1988, ‘Hydrologic Consequences of Increases in Trace Gases and CO2 in the Atmosphere’, in Effects of Changes in Stratospheric Ozone and Global Climate, Climate Change, U.S. Environmental Protection Agency, 3, pp. 251–271. McCabe, G. J. and Ayers, M. A.: 1989, ‘Hydrologic Effects of Climate Change on Delawere River Basin’, Water Res. Bull. 25 (6), 1231–1241. Mitosek, H.: 1992, Occurrence of Climatic Variability and Change Within the Hydrological Time Series, A Statistical Approach, IIASA, Laxenburg, Austria. Morison, J. I. L.: 1987, ‘Intercellular Carbon Dioxide Concentration and Stomatal Response to Carbon Dioxide’, 229–251, Stomatal Function, Stanford University Press, California, 503 pp. Rao, G. Appa: 1987, Drought Probability Maps, CAgM Report No. 24, WMO, Geneva, Switzerland, 45 pp. Schaake, J. C.: 1989, ‘Climate Change and U.S. Water Resources’, in Waggoner, P. E. (ed.), Climate to Flow (Chapter 8), Wiley, New York, pp. 177–218. Schneider, S. H. and Thompson, S. L.: 1981, ‘Atmospheric CO2 and Climate: Importance of the Transient Response’, J. Geophys. Res. 86, 3135–3147. Schneider, S. H.: 1989, ‘The Changing Climate’, Scient. Amer. 9, 70–79. Thornthwaite, C. W. and Mather, J. R.: 1955, ‘The Water Balance, Publications in Climatology’, Laboratory of Climatology 8 (1). Thornthwaite, C. W. and Mather, J. R.: 1957, ‘Instructions and Tables for Computing Potential Evapotranspiration and the Water Balance’, Publications in Climatology, Laboratory of Climatology 10 (3). Warrick, R. A.: 1988, ‘Carbon Dioxide, Climatic Change and Agriculture’, Geogr. J. 154, 221–233. Williams, M. A. J. and Balling, R. C.: 1993, Interactions of Desertification and Climate, World Meteorological Organization, Geneva, Switzerland, 230 pp. Willmott, C. J.: 1977, Watbug: A Fortran Algorithm for Calculating the Climatic Water Budget, Water Resources Center, University of Delaware, Newark, 55 pp. World Meteorological Organization: 1994, WMO Statement on the Status of the Global Climate in 1993, Technical Note 809, WMO, Geneva, Switzerland, 20 pp. World Meteorological Organization: 1995, WMO Statement on the Status of the Global Climate in 1994, Technical Note 826, WMO, Geneva, Switzerland, 20 pp. World Meteorological Organization: 1996, WMO Statement on the Status of the Global Climate in 1995, Technical Note 838, WMO, Geneva, Switzerland, 18 pp. (Received 29 July, 1996; in revised form 29 December, 1997)

clim1790.tex; 9/10/1998; 11:58; p.27