三种土壤导热率模型对中国北方地表温度的模拟

Simulation of Land Surface Temperature in Northern China by Three Soil Thermal Conductivity Models

为了提高模式CLM4.5(The Community Land Model Version 4.5)对我国北方地表温度的模拟性能,本文利用近30年实况地表温度站点观测资料,开展了Johansen、Côté-Konrad及Lu-Ren三个常用土壤导热率模型对地表温度模拟影响的研究.结果表明,三种导热率方案均能较好地模拟出中国北方地表温度的时空特征,但Lu-Ren方案的模拟误差、均方根误差更小,与实况间的相关性最显著.实况分析表明,近30年来北方地表温度处于快速上升通道中,尤其2003年以来增温趋势更加显著;三种方案均能模拟出这种变化趋势,与实况变化间相关显著,但模拟的增温幅度偏小,其中Lu-Ren方案模拟的变化趋势与实况值最接近.基于Taylor图和降水模拟评估表明,三种方案中Lu-Ren方案模拟效果最好.由此可见,在CLM4.5模式中增加适合我国北方气候模拟的土壤导热率新方案(Lu-Ren方案),提高了CLM模式的模拟效果,促进了该陆面模式的发展.

In order to improve the simulation performance of NCAR CLM4.5(The Community Land Model Version 4.5)for land surface temperature(LST)in northern China,the ability of three commonly used soil thermal conductivity(STC)models were tested in this paper.The results show that these schemes can simulate the temporal and spatial characteristics of land surface temperature in northern China well.The LST in northern China has risen in recent 30 years,especially since 2003.The trend can be simulated by all these three schemes and shows a significant correlation with the observed LST,but the simulated warming amplitude is small than that of observation.The overall evaluation based on the taylor diagram and precipitation simulation shows that LR scheme has the best simulation effect.Therefore,a new scheme of STC that more suitable for climate simulation in northern China is added to CLM4.5,which promotes the development of the land surface model.