资源环境科技发展态势分析平台

This study evaluates the sensitivity of precipitation in Southeast Asia (SEA) to the choice of land-surface schemes available in RegCM4, that is BATS1e (BATS) and CLM4.5 (CLM), over the CORDEX SEA domain. Two similar simulations using different land-surface schemes were conducted for the period 1989 to 2007, with a horizontal resolution of 25 x 25 km. Four different types of observational data were used to evaluate the performance of these simulations. It was found that both of these simulations reproduced the SEA precipitation climatology spatial patterns reasonably well. However, the BATS simulation systematically produced higher frequency and more intense precipitation over land. The CLM simulation, although overestimating the precipitation frequency, produced a lower amount of on-land precipitation. The precipitation annual cycle and interannual variability were both insensitive to the choice of land-surface scheme. The errors in the precipitation simulation were found to originate from the simulated convective precipitation (CPr). Large-scale precipitation was found to be less sensitive to the choice of land-surface scheme. The influence of the land-surface scheme on SEA precipitation simulation took place through the interaction between soil moisture content and CPr. The soil moisture content was consistently higher in the BATS simulation, as was the CPr simulated. During December to February (DJF), the difference in the precipitation amount simulated by BATS and CLM simulations was lower over the SEA mainland, even though the soil moisture content in both simulations was very different. This difference was due to the inactivity of the convective activities over the SEA mainland during DJF. This study concludes that the simulation of SEA precipitation amount was sensitive to the choice of land-surface scheme, and recommend the use of CLM4.5 land-surface scheme for simulation of SEA precipitation.