Future Changes in Surface Temperature and Precipitation in Arid and Semi-arid Regions Revealed by the Multi-Model Ensemble (MME) Dataset for the IPCC AR4

Abstract

Characteristics of climate change in arid and semi-arid regions have been investigated using the Multi-Model Dataset (MMD), where data submitted for IPCC AR4 by many research centers are compiled under the project CMIP3, WCRP. Five arid and semi-arid regions (Central Asia, Middle East, South West United States, Sahara and Australia) are selected and future changes in surface temperatures and precipitation are investigated using MME data. A significant increase in surface temperature in the five regions is found, but precipitation change is noticed only in winter and in the central Asia and Sahel regions. As our understanding of the climate system in an arid region is limited and many unknown processes operate, it is still questionable as to whether these simulation results are reliable. Therefore, it is necessary to estimate the reliability, for which the relationship between the time-averaged error in the present climate and the future change due to a doubling of C02 is examined. The relationship to surface temperature has been examined over the Central Asia and the Sahara regions. After screening MME data, 14 models are selected, and a meaningful correlation in the annual and seasonal means is found in those regions. This is because the surface temperature is considered to be governed by the radiative energy balance. Tn contrast, a significant correlation to change of annual mean precipitation could not be found over these regions. This is because precipitation is governed by a rainfall event and convective systems. A relationship is found only in winter, because precipitation in winter may be controlled by disturbances in the mid-latitude. Key words: SIobal warming, climate model, TPCC, MMD.