中文摘要：

The Newtonian relaxation is used to assimilate data from the Second Global Soil Wetness Project (GSWP2) into the initial field for soil moisture in a regional climate model (i.e., BCC_RegCM 1.0, hereinafter called RegCM). The results show that different nudging parameters can improve the simulation of the air temperature and the precipitation to some extent in summertime season over East China. When the nudging intensity is 0.5 and nudging duration is 5 days, the result is desirable. Introducing the data assimilation of soil moisture into RegCM, the amelioration is not only in the spatial distribution pattern but also in the magnitude of the temperature and the precipitation. More specifically, the assimilation experiments have corrected the systematical cold bias of the air temperature and the precipitation on the low side, which is much closer to the observation compared with the control experiment. On the whole, the assimilation experiments shed light on the fact that the simulated bias in the original RegCM may be affected by the description of soil moisture in the land surface process. Therefore, the effect of soil moisture in RegCM should be paid more attention in the future.

英文摘要：

The Newtonian relaxation is used to assimilate data from the Second Global Soil Wetness Project (GSWP2) into the initial field for soil moisture in a regional climate model (i.e., BCC_RegCM 1.0, hereinafter called RegCM). The results show that different nudging parameters can improve the simulation of the air temperature and the precipitation to some extent in summertime season over East China. When the nudging intensity is 0.5 and nudging duration is 5 days, the result is desirable. Introducing the data assimilation of soil moisture into RegCM, the amelioration is not only in the spatial distribution pattern but also in the magnitude of the temperature and the precipitation. More specifically, the assimilation experiments have corrected the systematical cold bias of the air temperature and the precipitation on the low side, which is much closer to the observation compared with the control experiment. On the whole, the assimilation experiments shed light on the fact that the simulated bias in the original RegCM may be affected by the description of soil moisture in the land surface process. Therefore, the effect of soil moisture in RegCM should be paid more attention in the future.