Water resources for agriculture in a changing
climate: international case studies
Cynthia Rosenzweig a, Kenneth M. Strzepek b, David C. Major c, Ana Iglesias c, David N. Yates d, Alyssa McCluskey b, Daniel Hillel c
a NASA/Goddard Institute for Space Studies, 2880 Broadway, New York, NY 10025, USA
b University of Colorado, Boulder, CO, USA
c Columbia University, Center for Climate Systems Research, New York, NY 10025, USA
d National Center for Atmospheric Research, Boulder, CO, USA
Global Environmental Change Volume 14, Issue 4, December 2004, Pages 345-360
Abstract
This integrated study examines the implications of changes in crop water demand and water availability for the reliability of irrigation, taking into account changes in competing municipal and industrial demands, and explores the effectiveness of adaptation options in maintaining reliability. It reports on methods of linking climate change scenarios with hydrologic, agricultural, and planning models to study water availability for agriculture under changing climate conditions, to estimate changes in ecosystem services, and to evaluate adaptation strategies for the water resources and agriculture sectors. The models are applied to major agricultural regions in Argentina, Brazil, China, Hungary, Romania, and the US, using projections of climate change, agricultural production, population, technology, and GDP growth.
For most of the relatively water-rich areas studied, there appears to be sufficient water for agriculture given the climate change scenarios tested. Northeastern China suffers from the greatest lack of water availability for agriculture and ecosystem services both in the present and in the climate change projections. Projected runoff in the Danube Basin does not change substantially, although climate change causes shifts in environmental stresses within the region. Northern Argentina’s occasional problems in water supply for agriculture under the current climate may be exacerbated and may require investments to relieve future tributary stress. In Southeastern Brazil, future water supply for agriculture appears to be plentiful. Water supply in most of the US Cornbelt is projected to increase in most climate change scenarios, but there is concern for tractability in the spring and water-logging in the summer.
Adaptation tests imply that only the Brazil case study area can readily accommodate an expansion of irrigated land under climate change, while the other three areas would suffer decreases in system reliability if irrigation areas were to be expanded. Cultivars are available for agricultural adaptation to the projected changes, but their demand for water may be higher than currently adapted varieties. Thus, even in these relatively water-rich areas, changes in water demand due to climate change effects on agriculture and increased demand from urban growth will require timely improvements in crop cultivars, irrigation and drainage technology, and water management. Published by Elsevier Ltd.
Keywords: Climate; Climate change; Water supply; Water demand; Agriculture; Irrigation; Adaptation
1. Introduction
Climate change, population growth, and economic development will likely affect the future availability of water resources for agriculture differently in different regions. The demand for and the supply of water for irrigation will be influenced not only by changing hydrological regimes (through changes in precipitation, potential and actual evaporation, and runoff at the watershed and river basin scales), but by concomitant increases in future competition for water with nonagricultural users due to population and economic growth. Here we compare future water availability for agriculture and its ability to provide ecosystem services under changing climate conditions in case study regions throughout the world. The study regions are in Northern Argentina, Southeastern Brazil, Northeastern China, the Hungarian and Romanian parts of the Danube Basin, and the US Cornbelt (Fig. 1). (Results from the US Cornbelt case study are discussed in detail in Strzepek et al., 1999.) We evaluate near-term adaptation strategies for both the water resources and agricultural sectors in the regions.
Fig. 1. Study countries and regions
The regions differ in socio-economic development, technological possibilities, and climatic regime, but all have relatively ample (less in Northeastern China) water supplies for agriculture in the current climate. Thus, one purpose of the study is to consider how major agricultural regions may fare under changing climate conditions, since they may become even more important as food-producing centers relative to agricultural areas in more marginal, semi-arid regions that have been found to be vulnerable to climate change (IPCC, 2001a). The objective here is to develop a comparative framework for regional water resources for agriculture that integrates water availability, agriculture, and technology with demographic, economic, and climate forecasts.
The framework was developed by an interdisciplinary team and uses publicly available and widely validated models of water supply and demand and of crop growth and irrigation management. The models are validated at both site and regional scales, permitting the evaluation of uncertainties at different spatial scales. Consistent modeling assumptions, available databases, and scenario simulations are applied to major agricultural regions, capturing a range of possible future conditions. We tested the sensitivity of the modeling framework to technological change and also observed the degree of environmental stress in each basin under alternative scenarios. For two of the case studies, those in China and the US, we examined additional forecasts and the use of different cultivars (varieties) for adaptation to changing seasonality. Among earlier uses of integrated model methodology are Strzepek et al. (1999); Strzepek et al. (1995); Rosenzweig and Parry (1994); and Major and Schwarz (1990).
The agricultural regions chosen for the comparative study include major corn and soybean growing areas and associated river basins (Table 1). The selection of these areas is based on their importance in current or potential corn and soybean production, and on their sensitivity to current and future climate regimes (see e.g., Lobell and Asner, 2003). The countries within which the study areas are located contain about one-third of global arable land and irrigated land, and currently account for about 70% of world corn production and over 90% of world soybean production (FAO, 1998).
The methods are an improvement over other approaches in several respects. Potential evapotranspiration is calculated consistently for both the supply of and the demand for crop irrigation water in the agricultural regions of each water basin. A weather generator is used to develop monthly time series for the climate change scenarios with spatial autocorrelation that permits analysis of potential changes in interannual variability. Trajectories of change in competing demands for water and demand for agricultural commodities based on population growth and economic development, and improvements in irrigation technology are included, as well as potential changes in climate. The use of water for ecosystem services is considered explicitly. The importance of using several forecasts of climate change in analyses is bolstered by considering the work of Stone et al. (2001), whose single forecast (albeit converted to mesoscale, a computationally intensive procedure), does not provide water managers with a range of possibilities for the same region considered in Strzepek et al. (1999). Finally, adaptation measures are evaluated from the point of view of both water resource and agronomic management.
Table 1 Description of study areas
a And parts of Catamarca, Formosa, Jujuy, Salta, and Tucuman.
b Parts of water regions located in Austria, Bosnia-Herzogovina, Bulgaria, Croatia, Czech Republic, Germany, Macedonia, Moldavia, Slovakia, Slovenia, Ukraine, Yugoslavia.
c All or parts of Colorado, Illinois, Indiana, Iowa, Kansas, Kentucky, Maryland, Minnesota, Missouri, Montana, North Carolina, New York, North Dakota, Ohio, Pennsylvania, South Dakota, Tennessee, Virginia, West Virginia, Wisconsin, Wyoming
2. Methods The modeling structure of the study is shown in Fig. 2, which illustrates the linkages of the water supply, crop demand, and water management models. The linked models include WATBAL for water supply (Yates, 1996; Kaczmarek, 1993); CERES-Maize, SOYGRO, and CROPWAT for crop yield and irrigation demand (Jones and Kiniry, 1986; Jones et al., 1988; CROPWAT, 1995); and WEAP for water demand forecasting, planning and evaluation (Stockholm Environment Institute, 1997; Sieber et al., 2002). These models are applied to the study regions for current conditions and for a set of scenarios projecting future changes in climate, agricultural production, population and GDP. For further information about models and methods, see
http://www.giss.nasa.gov/research/impacts.
2.1. Study regions Major maize and soybean growing areas and associated river basins in Hungary and Romania, Argentina and Brazil, China, and the United States were selected as case studies (see Fig. 1 and Table 1). The boundaries of the study areas were defined by the location of major river basins and agricultural production areas. Each case study was subdivided into smaller water regions, generally comprising a single river with its tributaries, but sometimes including additional streams. The major maize and soybean growing areas, crop modeling sites, and associated water regions for each study area are shown in Fig. 3.
Fig. 2. Model structure and interactions
2.2. Climate change scenarios The system of models is driven by climate change scenarios that represent a range of plausible future climate changes as projected by global climate model (GCM) simulations available at the initiation of the interdisciplinary study. The GCMs are those of the Geophysical Fluid Dynamics Laboratory (GFDL) (R30 version) (Manabe et al., 1992), Goddard Institute for Space Studies (GISS) (Hansen et al., 1984), and the Max Planck Institute (MPI) (Cubasch et al., 1992), with sensitivities to doubled CO2 forcing of 3.7 1C, 4.2 1C, and 2.6 1C, respectively. In order to compare results from the original set of GCMs with more recent ones, we also used two additional GCMs: those of the United Kingdom Met Office Hadley Centre (HadCM2) (HC) (Mitchell et al., 1995; Johns et al., 1997) and the Canadian Climate Centre (CGCM2) (CCC) (Flato and Boer, 2001). These GCM transient simulations were forced with a 1% per year increase of CO2 concentrations, which represents a mid-range of non-intervention CO2 emissions trajectories available from surveyed literature (Nakicenovic and Swart, 2000).
Fig. 3. Study areas, water regions, and crop modeling sites for Argentina, Brazil, China, Hungary and Romania, and the United States.
Time periods for the analysis reflect key periods relevant to water resource management decision-makers: near-term (the 2020s) and medium-term (the 2050s). The climate change scenarios for the 2020s and the 2050s are constructed from the 30-year period centered on the given time-slice. The GCM scenarios selected did not include the effect of sulfate aerosols; temperature projected for combined greenhouse gas and sulfate aerosol effects would be slightly reduced. The projections used fall within the range of forecasts reported in the 2001 IPCC Third Assessment Report (IPCC, 2001b).
Monthly mean temperature, precipitation and solar radiation changes are obtained for the study regions and are used to create climate change scenarios for the water supply and crop models. For the crop modeling sites, daily climate data are modified by monthly changes taken from the GCMs. For the water supply modeling, the monthly changes are applied to the monthly Stochastic Analysis and Modeling System (SAMS) time-series (United States Bureau of Reclamation, 1999). The SAMS model provides a stochastic, spatially explicit technique for generating regional weather time series preserving auto- and cross-correlations.
Full Text
ليست هناك تعليقات:
إرسال تعليق