Spatial sensitivity of predicted soil erosion and runoff to climate change at regional scales.

Impacts of climate change on natural resources need to be evaluated in a range of geography and agricultural systems for better conservation planning. The objectives of this paper were to evaluate spatial sensitivity of predicted soil loss and runoff to climate change at large scales, and to simulate the ‘regional’ impacts of climate change on soil erosion on the southern Great Plains. Relative climate changes at three spatial scales between 1950 to 99, and 2070 to 99, projected using HadCM3 under the A2a, B2a, and GGa1 emission scenarios, were used to generate changed climates for Chickasha, Oklahoma. The Water Erosion Prediction Project (WEPP) model was run for each climate scenario at three spatial scales with three tillage systems. The HadCM3 predicted a general decrease in precipitation and an increase in the east-west precipitation gradient on the Southern Great Plains over the century. The decrease in precipitation resulted in general decreases in predicted runoff and soil loss. Percent changes, compared to the present climate, ranged from -33 to -3 percent for runoff and from -33 to 0 percent for soil loss. Variability of predicted runoff and soil loss over spatial scales was comparable to that between climate change scenarios. This finding suggests that the climatic impacts should be assessed at the spatial resolution at which global climate models have predictive skill, or at multiple spatial scales if the skillful resolution is unknown. The latter would yield additional information on uncertainty of predicted impacts due to uncertainty of spatial scales of climate input. [ABSTRACT FROM AUTHOR]