高分辨率区域模式不同时间步长对一次强降水预报的影响

Influence of different time steps of high resolution regional model on heavy precipitation forecast

使用华南短临预报模式本地化改造后的广西对流尺度数值预报模式系统,对2022年6月3-4日广西强降水过程进行数值模拟,研究不同时间步长对模式预报效果的影响。结果表明:(1)时间步长对位势高度、温度场和地面要素场的影响较小,对风场影响较大;(2)强降水区中低层物理量、t2m、u10m预报误差较小,模式预报偏差主要来源于高层物理量和v10m;(3)时间步长对模式降水预报有较大影响,时间步长越小(大),降水预报敏感度越大(小),大暴雨以上量级和范围预报越大(小);(4)无论24h还是逐小时ETS评分都反应出小时间积分步长对大量级降水的预报效果更好,且对强对流天气发展趋势有一定指示作用,大的时间积分步长预报效果则相反。

The Guangxi convective scale numerical prediction model system which was localization reconstructed by the South China Short-term approach Prediction Model(CMA_GZ_R1)is used to simulate a heavy rainfall process in Guangxi on June 3-4,2022,during the dragon-boat rainy period,and the effects of different time steps on the model prediction are also studied.The results show that:(1)The time step has less influence on the geopotential height,temperature and surface element field,but a greater influence on the wind field;(2)The prediction errors of the physical quantities in the middle and lower layers,t2mand ul0m in the heavy precipitation area are small,and the model prediction errors mainly come from the high-level physical quantities and v1Om;(3)The time step has a greater influence on the model precipitation forecast,the smaller(larger)the time step,the greater(smaller)the precipitation forecast sensitivity,and the larger(smaller)the magnitude and range forecast above large rainstorms;(4)Both 24-hour and hourly ETS scores reflect that the small time integration steps are more effective in forecasting large-magnitude precipitation,and are indicative of strong convective weather trends,while large time integration steps have the opposite effect.