不同微物理参数化方案对我国北方一次大范围暴雪天气过程的数值模拟研究

本文以ERA5（ECMWF Reanalysis v5）再分析资料为初始场，利用WRF（The Weather Research and Forecasting）模式对2020年4月19～20日的一次大范围暴雪天气过程进行数值模拟研究。模式采用不同云微物理参数化方案进行敏感性试验，并与实测数据（自动站降水数据、雷达基数据）进行对比，分析了此次暴雪天气过程不同阶段的降水、雷达反射率、动热力和水凝物的时空演变和三维细致结构特征。研究表明:Morrison方案更好的模拟出了本次暴雪天气过程，表现在模拟的雷达回波强度、范围及形态更与实况一致，模拟出的降水量的相关系数和均方根误差都优于其他方案；其微物理细致结构表现为强上升运动和低层正涡度的长时间维持，以及7 km以上高层较多的冰晶、中低层较少的霰粒子和雨水粒子。从热动力场上看，bin（SBM fast）方案在600 hPa高度以下存在明显的涡度波列，这主要是因为bin方案将粒子群分档处理，没有捆绑不同粒子类型运动，更能细致描述出不同粒子的下沉拖曳作用。从云微物理特征上看，不同方案模拟的雪、霰、云水以及雨水粒子的比质量都较为接近，而对冰晶比质量的模拟不管在量级还是在分布范围上都存在很大的差异，这种差异决定了不同微物理方案模拟的雷达回波和降水量级和相态的差异。

Numerical Simulation of a Large-Scale Snowstorm Process in Northern China Using Different Cloud Microphysical Parameterization Schemes

Using ERA5 reanalysis data as the initial field, the WRF model was used to conduct a numerical simulation study of a large-scale snowstorm event from April 19–20, 2020. This work adopted different microphysical parameterization schemes for sensitivity experiments, and the capability of the model for simulation was evaluated based on observation data (precipitation data collected at automatic weather stations, radar base data). Temporal and spatial evolution characteristics of precipitation, radar reflectance, dynamic thermodynamics, and water condensate in heavy snow weather were also analyzed. Results reveal that the Morrison scheme can better simulate the snowstorm weather event, which shows that the simulated radar echo intensity, range, and shape are more consistent with the observation data, and the correlation coefficient and root mean square error of the simulated precipitation are better than other schemes. The detailed microphysical structure of the proposed scheme is characterized by a strong ascending motion and long-term maintenance of positive vorticity in the lower layer, more ice crystals in the upper layer above 7 km, and less graupel and rain particles in the middle and lower layers. From the perspective of the thermal dynamic field, there is an obvious vorticity wave train below the height of 600 hPa in the bin scheme. This is mainly because the bin scheme grades the particle swarm, does not bind different particle types to move, and can describe the sinking and dragging effect of different particles in a more detailed way. Cloud microphysical characteristics show that the specific masses of snow, graupel, cloud water, and rain particles simulated by different schemes are close to each other. However, the simulation of the specific mass of ice crystals has great differences in both magnitude and distribution range, which determine the difference in the magnitude and the phase state simulation of the radar echo and precipitation by different microphysics schemes.