【摘要】 以黑河流域一次降水过程为研究个例,利用中尺度数值预报模式WRF 3.1进行物理过程参数化方案选择的分组敏感性试验,分别讨论了WRF模式中的微物理、积云对流、陆面过程和边界层参数化方案对黑河流域降水模拟的敏感性,并利用该流域内12个气象站的日降水资料进行了验证。研究表明:没有确定的某个物理过程参数化方案在所有试验方案中的模拟效果一致好,不同的参数化方案配合使用模拟效果更佳。对比分析模拟结果发现,要有效地模拟出降水中心位置和降水量、降水区分布形态和范围,应同时选用MYJ边界层参数化方案与Noah陆面过程参数化方案;微物理参数化方案中选择Ferrier方案时,对降水的模拟效果要好于选用WSM3、WSM5、Lin方案;积云对流参数化方案对降水模拟的敏感性明显弱于其他物理过程参数化方案。更多还原

【Abstract】 A rainfall event in the Heihe River Basin was taken as an experimental case,and the mesoscale WRF3.1 was applied to study the sensitivity of different parameterized schemes to rainfall simulation.Effects of the microphysics,cumulus convection,land surface processes and the planetary boundary layer in the WRF parameterized schemes on the sensitivity of rainfall simulation in the Heihe River Basin were discussed,and the simulated results were validated with the observation data from 12 meteorological stations in the basin.The results showed that there was no single parameterized scheme to perform well in the simulation,and a combination of various parameterized schemes would be better.The results suggested that the center,value,distribution and extent of rainfall could be simulated well if the Mellor-Yamada-Janjic planetary boundary layer parameterized scheme and the Noah land surface process parameterized scheme were applied simultaneously.The results also suggested that the results simulated with Ferrier scheme were better than those with WSM3,WSM5 and Lin schemes.The sensitivity of cumulus convection parameterized scheme to the rainfall simulation was significantly lower than that of other schemes.更多还原