Advances in Earth Science
1, Liu Jian
1,2,3*, Wang Zhiyuan
Wang Xiaoqing, Liu Jian, Wang Zhiyuan. Comparison of Simulated and Reconstructed Temperature in China during the past 2000 years[J]. Advances in Earth Science, 2015, 30(12): 1318-.
Studying climate changes over the past 2 000 years has important scientific significance in exploring climate variability on decadal to centennial timescales, assessing the natural and anthropogenic contribution to the climate warming, and understanding the effects of human activities in the past and future climate changes. Due to the scarcity of observation and uncertainty of reconstruction in this period, climate model is developed as a useful tool for studying paleoclimate.The Community Earth System Model (CESM) is one of the state-of-the-art climate models, but its performance in simulating the temperature in China has not been examined.The temperature datasets of observation/reanalysis (GHCN_CAMS) and reconstructions during the past 2000 years in China were used to examine the performance of CESM. The comparison between the annual average temperatures of GHCN_CAMS reanalysis and simulation showed that the model well reproduced the spatial distributions and upward trend of the annual average temperature in China, and the comparison with reconstructions in five sub-regions of China indicated that the simulation were in good consistent with the average temperature changes of reconstructionson decadal time scales. On the centennial time scale, the average temperature fluctuations of simulation in the regions of China were in accord with reconstructions generally except for Central East and Tibet.There existed three warm periods of simulated temperature variation in China over the past 2000 years, including 0-540AD, 800-1250AD and 1901-2000AD, and two cold periods involving the 551-721AD and 1400-1850AD, which had some discrepancies with reconstructions. And the discrepancies between simulation and reconstructions might be related to uncertainties of the resolution, external forcing and parameterization of the subgrid-scale process in the model.