Monthly water balance modeling for hydrological impact assessment of climate change in the Dongjiang (East River) Basin, South China

Monthly water balance models are important tools for hydrological impact assessment of climate change. Traditionally monthly models adopt a conceptual, lumped-parameter approach. Based on an extensive survey and review of existing monthly water balance models, six models with different conceptualization and structure, i.e., Thomthwaite-Mather, Belgium, Xinanjiang, Guo, WatBal and Schaake, were compared through calibration and validation using observed data of hydrology and climate of 1960-1988 in the Dongjiang basin. The model comparison offered insights for the development of a monthly distributed model which integrates the spatial variations of basin terrain and rainfall into runoff simulation. An innovative feature of the new model is that the spatial distribution of soil moisture capacity which is described as a parabolic curve in Xinanjiang model is represented by a cumulative frequency curve of index of relative difficulty of runoff generation based on the concept of topographic index in TOPMODEL. The calibration and validation results show that the developed model with only three parameters is suitable for monthly runoff simulation in the Dongjiang basin.; The developed model was applied to evaluate the changes in water availability in the Dongjiang basin under hypothetical climate change scenarios and those derived from projections of three General Circulation Models (GCM), i.e., CGCM1, CSIRO and ECHAM4. Sensitivity analyses based on hypothetical scenarios suggest that climatic change has greater effects on runoff than on soil moisture and greater effects on water availability in dry months than in wet months. The effects of precipitation changes on the amount of runoff and soil moisture can be characterized by a magnification factor whereas temperature increases alone produce negligible effects. Hydrological simulation with inputs of three GCM-generated scenarios indicates that annual and rain-season runoff will increase by 0.3°io to 13.9% and 7.6% to 12.0%, respectively, by the 2050s. Dry-season runoff will change between -23.2% and +26.4%. Average annual and dry-season soil moisture will decrease by 1.3% to 6.9% and 1.0% to 8.1%, respectively. Soil moisture will demonstrate little change in rain-season. Increase in annual runoff and reduction in annual soil moisture will be apparent over the whole basin, but there is relatively little consistency among the three GCM-generated scenarios as to the magnitudes of spatial change in runoff and soil moisture. Although these results are not definitive statements as to what will happen to runoff and soil moisture in the Dongjiang basin, they rather have significant implications for the study of response strategies of water supply and flood control to climate change.