Simulation of historical and projected climate change in arid and semiarid areas by CMIP5 models

Based on Climatic Research Unit Time Series 3.1 temperature and Global Precipitation Climatology Center full data reanalysis version 6 precipitation data, the abilities of climate models from the fifth phase of the Coupled Model Intercomparison Project to simulate climate changes over arid and semiarid areas were assessed. Simulations of future climate changes under different representative concentration pathways (RCPs) were also examined. The key findings were that most of the models are able to capture the dominant features of the spatiotemporal changes in temperature, especially the geographic distribution, during the past 60 years, both globally as well as over arid and semiarid areas. In addition, the models can reproduce the observed warming trends, but with magnitudes generally less than the observations of around 0.1-0.3 A degrees C/50a. Compared to temperature, the models perform worse in simulating the annual evolution of observed precipitation, underestimating both the variability and tendency, and there is a huge spread among the models in terms of their simulated precipitation results. The multi-model ensemble mean is overall superior to any individual model in reproducing the observed climate changes. In terms of future climate change, an ongoing warming projected by the multi-model ensemble over arid and semiarid areas can clearly be seen under different RCPs, especially under the high emissions scenario (RCP8.5), which is twice that of the moderate scenario (RCP4.5). Unlike the increasing temperature, precipitation changes vary across areas and are more significant under high-emission RCPs, with more precipitation over wet areas but less precipitation over dry areas. In particular, northern China is projected to be one of the typical areas experiencing significantly increased temperature and precipitation in the future.