Spatially downscaling GCMs outputs to project changes in extreme precipitation and temperature events on the Loess Plateau of China during the 21st Century

A simple transfer function method is used to spatially downscale extreme precipitation and temperature indices of six GCMs (CNRM-CM3, GFDL-CM2.1, INM-CM3.0, IPSL-CM4, MIROC3.2_M and NCAR-PCM) under three emission scenarios (A1B, A2, B1) from grid outputs to target station to project their potential spatiotemporal changes on the Loess Plateau of China during the 21st century. GCMs project that extreme climate events will keep the present change trend, i.e. longer heatwave duration and growing season length, less cold extremes, smaller annual extreme temperature range, more frequent and intense precipitation, and longer drought duration; and climate models suggest that the changes during the 21st century are a further amplification of those at present. Spatial variations exist in the changes of extreme indices, the greatest changes will occur in the southeast and northwest region for extreme precipitation indices while in the north and southeast region for extreme temperature events. The projected changes in extreme climate will possibly bring great impacts on the soil losses and agriculture on the Loess Plateau and countermeasures should be planned.