RegCM4区域气候模式对新疆地区冬季地表状态的模拟分析

利用区域气候模式Reg CM4.3(Regional Climate Mode version 4.3)对新疆地区冬季的地表状态进行了模拟分析,通过与ERA40再分析资料的对比分析发现,温度分布形势模拟较好,地面热力状态受地形影响显著,陡峭地形附近由于热性质差异大和非均匀性强会导致较大模拟误差;模式较好模拟出降水和潜热通量北疆多南疆少,山区多盆地少的分布特征,模拟出通过反照率影响,地表吸收的短波辐射呈现出沙漠腹地吸收多而天山地区吸收少的分布,对北疆呈感热通量汇而南疆呈感热通量源的感热分布形势也模拟较好;模拟的雪水当量与降水分布有较好的一致性,春季融雪径流与冬季雪水当量分布及降水均有较好的对应关系。通过模拟分析也发现,现有方案实际感热通量计算中以地面温度代替地面位温,造成感热通量偏小,因此会低估南疆感热源效应和高估北疆感热汇效应。此外,积雪量和地面温度模拟偏高可能是春季北疆主要积雪区径流偏强的原因。

Analysis of Simulation on Wintertime Surface Condition in Xinjiang by RegCM4 Model

Based on Comparison with the ERA40 reanalysis, the regional climate model RegCM4.3 simulation on wintertime surface condition in Xinjiang is analyzed. The results show that RegCM4.3 can well simulate the spatial distribution of surface temperature in Xinjiang, which is greatly affected by terrain. The prominent differences in surface thermal properties and strong heterogeneity caused by large topography gradient lead to significant simulation errors. The model realistcally simulates the spatial distribution features of precipitaioon and latent heat flux, both are larger in northern Xinjing than in southern Xinjiang and larger in mountaineous region than in basins. The simulated surface net downward shortwave flux is much larger in central desert than in Tianshan mountain region. Southern and northern Xinjiang are the source and sink of sensible heat flux respectively, which is well reproduced by RegCM4.3. The simulated distribution of snow water equivalent agrees well with precipitation. Meanwhile, the spring snowmelt runoff is also consistent with both snow water equivalent and precipitation. Analysis of the simulation also reveals that the sensible heat flux is underestimated due to the fact that the ground potential temperature is replaced by ground temperature in RegcM4.3. As a result, the sensible heat source in southern Xinjiang is underestimated while the sensible heat sink in northern Xinjiang is overestimated. Additionally, the large simulated bias of spring snowmelt runoff in northern Xinjiang might be attributed to biases of simulated snow and ground temperature.