利用双偏振雷达相态资料改进数值模式云初始场研究

“Spin-up” is one of the most serious problems for short-range precipitation forecast in Numerical weather prediction model. It is directly related to the cloud information deficiencies in the first guess field, especially to the less information of hydrometeors classification state and the corresponding balance relationship of kinetic motion. Hence, in this project the hydrometeor type products of dual-polarimetric radar is applied to improve the cloud initial field. On the base of the existing cloud initial field construction module in GRAPES_meso model, we will develop a data blending and assimilating scheme for the classified hydrometeors of the dual-polarimetric radar through the combination of the physical initialization theory and the Nudging method of data assimilation. With the cloud analysis system in hand, the objective analysis of the dual-polarimetric radar hydrometeors is processed on the cloud droplet, rain, snow and graupel in NWP model by optimum interpolation method, as well as the latent heat can be evaluated. The analyzed microphysical cloud variables and thermodynamic information are assimilated by newtonian-relaxation approximation algorithm. Therefore, the obtained initial field closed to the atmospheric real state would decrease the “spin-up” time, and improve of GRAPES-MOSO operational short-range precipitation forecasting.

“Spin-up”问题是限制短时临近数值预报水平的急待解决问题，初始云微物理场分析不准或缺失是导致该问题的关键原因，尤其是缺乏准确的降水相态微物理初值及与之平衡的动力信息。为此本项目提出应用一种新的资料双偏振雷达相态识别资料来改进数值模式云初始场。拟采取现有GRAPES中尺度模式云初始场形成模块作为研究平台，将物理初始化理论与Nudging同化方法相结合，建立双偏振雷达相态识别资料在模式微物理初始场形成中的融合同化方案。在现有云分析基础上，以最优插值方案实现双偏振雷达相态识别资料对雨水、雪水、霰的客观分析；并据此计算潜热释放信息；对于融合分析出的云微物理变量及热动力信息拟采用牛顿松弛逼近算法进行同化。通过改进数值模式云微物理及与之平衡的热动力初值，从而为缩短“spin up”时间、提高短时临近预报准确率提供科学依据。

数值预报中的“spin-up”问题严重限制了短时临近预报水平的提高。针对此问题的主要原因，初始场中云微物理场分析不准或缺失，尤其是缺乏准确的降水相态微物理初值及与之平衡的动力信息本项目展开研究。建立了应用双偏振雷达相态识别资料改进数值模式云初始场中的降水微物理（雨水、雪水、霰等）初值及热动力信息的同化方案，个例应用模拟显示可提高短临预报降水水平。具体为分析了双偏振雷达相态资料质量，并据此和研究需要进行个例选择，建立了适用研究的三个个例资料数据集；梳理确认了可开展研究的基础云初始场形成方案，并在此方案基础上建立了双偏振雷达相态识别资料改进数值模式云初始场的算法，将原有基于其他资料的经验关系分析降水微物理变量优化成应用双偏振雷达资料的客观分析，并完成了双偏振雷达相态识别资料在改进数值模式微物理初始场形成中的融合同化研究，应用增量松弛逼近算法改进了原云初始场形成方案；利用GRAPES新的包含双偏振雷达相态资料的云微物理初始场形成方案进行数值模拟个例试验，对模拟结果进行了分析，并引入了卫星云图模拟技术。个例结果显示，GRAPES新的包含双偏振雷达相态资料的云微物理初始场形成方案可有效分析出云微物理和热动力信息初始场，应用于预报后，显著缩短spin-up时间，提升强对流等短时临近降水预报结果。