Abstract:The confirmed climate change trends, such as the increase of extreme rainfall events, may have great impacts on soil erosion through altering driving forces of soil erosion and influencing the growth condition of plants. The Loess Plateau of China is one of the most erodible areas in the world, the potential response of soil erosion to future climate change need to be assessed specifically for this information is very useful for decision-making. The objective of this study was to assess the site-specific impacts of climate change during 2010-2039 on soil erosion in Changwu County in the tableland-gully region of the Loess Plateau. A new simple transfer function method and weather generator CLIGEN (CLImate GENerator) were used to downscale the spatial-temporal scales of four General Circulation Models (CCSR/NIES, CGCM2, CSIRO-Mk2 and HadCM3) under three emission scenarios (A2, B2 and GGa). The WEPP (Water and Erosion Prediction Project) model was employed to simulate the responses of agro-ecosystems. Two tillage systems, i.e. conventional tillage and conservation tillage, were used to discuss the adaptive measures to climate change. Results showed that climate would change significantly during 2010-2039 compared with 1957-2005. The annual mean precipitation would change -2.6% to 17.4%, the variance of annual precipitation would be 0.984 to 1.389 times of the present, and it is very likely that the precipitation amount and storm frequency would increase. Annual mean maximum and minimum temperature would increase 0.6 to 2.6 ℃ and 0.6 to 1.7 ℃, and the variance ratios of 2001-2039 to 1957-2005 for maximum and minimum temperature were 0.748-1.155 and 0.736-1.387, respectively. The probability of variance increase was great. The predicted climate changes would cause great changes in runoff and soil loss in slope field of the Loess Plateau. Under conventional tillage, WEPP predicted a change of 10% to 130% for runoff, -5% to 195% for soil loss during 2010-2039. However, WEPP predicted a change of -34.2% to 71% for runoff, -76.5% to -25.8% for soil loss during 2010-2039 under conservation tillage. The changed magnitudes of runoff and soil loss under conservation tillage were far less than those under conventional tillage. The results indicated that adoption of conservation tillage had great potential to reduce the adverse effects of climate changes on runoff and soil loss and thus protect agro-ecosystems. The new spatial downscaling method, emphasizing on matching the probability distributions, can be widely used in climate scenarios development based on GCMs as it relaxed the demands upon the existence of strong correlationships between station-measured and GCM-projected monthly precipitation and temperature.