The effect of land surface processes on the 1991 Flood Over Yangtze-Huai River Valley as Simulated From Sunya Regional Climate Model

In this presentation, we investigated the effect of land surface process on the simulation of East Asia summer monsoon (Meiyu). The effect was studied by comparing climate simulations using two versions of State University of New York at Albany (SUNYA) regional climate model (ReCM0 and ReCM1). The ReCM0 uses a simple bucket surface model and Blackadar's planetary boundary layer (PBL) scheme as were adopted in MM5, while the ReCM1 is implemented with Simplified Simple Biosphere model (SSiB) and Mellor-Yamada's PBL scheme. Several-month long simulations, using both the ReCM1 and ReCM0, were conducted on the severe precipitation event occurred in Yangtze-Huai River Valley (YHRV) in East Asia in summer of 1991. The use of detailed land surface process in ReCM1 significantly improves the diurnal range of surface air temperature. The smaller diurnal cycle of ReCM0 was found due to the smaller sensible flux and larger ground heat flux that is related with the fixed deep soil temperature during the integration. It was also shown that the heavy precipitation over the YHRV was well reproduced in ReCM1, due to the enhanced water vapor mixing ratio simulated with SSiB and vegetation-soil scheme and its associated PBL scheme, while the simulated precipitation was rather weak in ReCM0.

The memory and influence of the initial soil moisture were also studied using ReCM1. In the control experiment, the soil moisture was initialized with the climatological soil moisture. Two sensitivity experiments (dry and wet) were conducted by applying ±30% of fractional soil moisture in the control experiment. For the dry case the memory of initial soil moisture is much longer and the influence on the simulation is much larger than that for the wet case. For ReCM1, initial soil moisture affects the simulated precipitation in the sub-divided regions, but has little effect on the precipitation in the combination of sub-regions. The results also show that initial soil moisture affects the simulated surface air temperature substantially. It is suggested that the initial soil moisture needs to be verified against the observation before it is used in the ReCM.