A NUMERICAL STUDY OF TROPICAL DEEP CONVECTION USING WRF MODEL

The Weather Research Forecast model (WRF) configured with high resolution and NCEP 1°×1° reanalysis data were used to simulate the development of a tropical deep convection over the Tiwi Islands,northern Australia,and to investigate the sensitivity of model results to model configuration and parameterization schemes of microphysical processes.The simulation results were compared with available measurements.The results show that the model can reproduce most of the important characteristics of the observed diurnal evolution of the convection,including the initiation of convection along the sea-breeze front,which is then reinforced by downdraft outflows,merging of cells and the formation of a deep convective system.However,further improvement is needed to simulate more accurately the location and the time for initiation of the deep convective system.Sensitivity tests show that double-nesting schemes are more accurate than the non-nesting schemes in predicting the distribution and intensity of precipitation as far as this particular case is concerned.Additionally,microphysical schemes also have an effect on the simulated amount of precipitation.It is shown that the best agreement is reached between the simulation results and observations when the Purdue Lin scheme is used.     

WRF模式对澳洲一次热带深对流系统的模拟研究

利用WRF(Weather Research Forecast)模式和1 °×1 °NCEP全球分析资料对2005年11月16日澳大利亚北部Tiwi岛上的一次热带深对流个例进行了数值模拟,对模式网格分辨率和微物理参数化方案的影响进行了敏感性试验,并与实测资料进行了对比。试验结果显示：模式较好地再现了这次热带对流云的日变化特征,即早期降水由沿岸的海风锋初始对流形成,之后出现单体合并现象并最终形成成熟的深对流系统。但是,模式在模拟深对流系统出现的时间及位置方面仍然需要改进。敏感性试验显示,就本次个例而言,二重嵌套方案在降水分布以及降水强度方面的模拟结果比非嵌套方案更接近实际探测。另外,微物理过程参数化方案对降水量的模拟也有一定影响,采用Purdue Lin方案的模拟结果更接近于实测情况。