多源观测数据逐半小时同化对一次区域持续性暴雨的敏感性试验

Sensitivity Experiments of Half-Hourly Assimilation of Multi-Source Data in One Regional Persistent Rainstorm Process

利用基于WRF模式(Weather Research and Forecasting model)的逐半小时快速更新预报系统,对2020年8月13—14日鲁南一次区域持续性特大暴雨过程进行逐小时同化和逐半小时同化敏感性试验,以探讨高频资料同化应用问题,其中,逐半小时同化敏感性试验方案包括在半点同时同化地面自动气象站(AWS)观测资料和飞机报(AMDAR)资料及分别“拒绝”其中一种同化资料共三种方案。结果发现:半点同时同化AWS资料和AMDAR资料的试验对此次持续性暴雨24 h累计降水和短时强降水预报较逐时同化试验有显著改进作用,说明提高同化频次以提高高频资料的利用率对数值预报有正效果;在半点增加高频资料同化过程中,同时同化AWS资料和AMDAR资料的试验,均比“拒绝”其中一种同化资料的试验准确;“拒绝”同化AMDAR资料的试验比“拒绝”同化AWS资料的试验预报效果差,说明半小时同化试验中AMDAR资料同化起到了更为重要的作用;快速更新循环同化对初始场的改善是一个逐步调整的过程,不同同化资料和频次的试验对分析场的影响各有不同,同时同化AWS资料和AMDAR资料的半小时同化试验初始场与实况最接近;不同同化方案地面要素场高温高湿区相对辐合线的位置不同和高空冷暖空气强弱及配置不同,造成飑线系统的运动方向不同,是导致各试验区域持续性降水落区存在较大差异的主要原因;而飑线系统位置的差异导致其与西南暖湿气流中的新生系统结合程度不同,造成了未来降水强弱的差异。

The hourly and half-hourly data assimilation sensitivity experiments for one regional persistent rainstorm process in southern Shandong from August 13 to 142020 are carried out using the half-hourly update and prediction system based on WRF with the purpose of exploring the application of high frequency data assimilation.The half-hourly data assimilation experiments include three assimilation schemes:assimilating of the automatic weather station(AWS)data and the aircraft meteorological data relay(AMDAR)data at the half-hour frequency,and“rejecting”one of them respectively.The results show that the half-hourly cooperative assimilation experiment of AWS observation and AMDAR data can significantly improve the numerical prediction of 24 h accumulated precipitation and 1 h short-term severe precipitation during this persistent rainstorm relative to the hourly assimilation experiment.This indicates that increasing the assimilation frequency to increase the utilization of high-frequency data has a positive effect on the numerical prediction.In the process of half-hourly data assimilation,the experiment of assimilating AWS data and AMDAR data at the same time is more accurate than the experiment of“rejecting”one of them respectively.The result of assimilation experiment“rejecting”AMDAR data is worse than that of“rejecting”AWS data,indicating that the assimilation of AMDAR data plays a more import role than that of AWS data in half-hourly assimilation.The improvement of initial field by rapid updating cyclic assimilation is a process of gradual adjustment.The experiments with different assimilation data and frequencies have different effects on the analysis field.The initial field of half-hourly assimilation experiment of collaborative assimilating AWS data and AMDAR data is the closest to the observation.The different positions of the high temperature and high humidity area of the ground element field relative to the convergence line under different assimilation schemes,and the different strength and configuration of high-altitude cold and warm air lead to different movement directions of the squall line system,which is the main reason for the great difference of precipitation areas in each experiment.However,the location difference of squall line system leads to the different degree of integration with the newborn system in the warm and humid airflow in the southwest,resulting in the difference of precipitation intensity in the future.