| 摘要: |
| 该文利用常规观测资料（实况探空资料、自动站雨量数据）以及ECMWF、SWC-WARMS（西南区域模式）、GRAPES-MESO、GRAPES-GFS对比分析了2019年在副高边缘环流形势下的2次四川盆地区域性暴雨天气过程。结果表明：①初始场相似的2次过程，导致强降水落区不同的主要原因在于副高的动态变化不同，其次西风带槽脊的细微差异以及低层水汽条件的差异也与降水落区息息相关。②ECMWF在这2次过程中对副高位置动态变化的预报，西进较东退更为准确,且数值模式对此类大范围强降水的预报，ECMWF较其它模式更具有优势，但局限性在于范围偏小、量级偏弱，可结合SWC-WARMS进行订正，而GRAPES-GFS、GRAPES-MESO在模式稳定性和预报准确率方面都不及上述两者。 |
| 关键词: 暴雨；预报误差；模式稳定性 |
| DOI： |
| 投稿时间：2021-04-28 |
| 基金项目:高原与盆地暴雨旱涝灾害四川省重点实验室重点专项（SCQXKJZD2020002）：基于网格实况的模式降水偏差定量检验及订正试验 |
|
| Forecast Bias Analysis and Model Test Comparison of Two Subtropical High Edge Rainstorms in Sichuan Province |
| CONG Fang,CHEN Chaoping |
| (Sichuan Meteorological Observatory, Chengdu 610072 , China ;Key Laboratory of Heavy Rain and Drought/Flood Disaster in Plateau and Basin of Sichuan Province, Chengdu 610072 , China) |
| Abstract: |
| Conventional observation data (real sounding data, rainfall data of automatic stations with quality control) and numerical prediction model products such as ECMWF, SWC-WARMS (Southwest Region Model), GRAPES-MESO, and GRAPES-GFS are used to compare and analyze two regional rainstorm processes in the Sichuan Basin in 2019 under the peripheral circulation situation of the subtropical high.The results show that:①The two processes with similar initial fields are mainly caused by the different dynamic changes of the subtropical high, and the slight differences of trough ridges in the westerly belt and the differences of water vapor conditions in the lower layer are also closely related to the precipitation fall areas.②ECMWF is more accurate in predicting the dynamic change of the subtropical high position in the two processes,westward advance is more accurate than to retreat to the east. ECMWF has more advantages than other models in predicting such large-scale heavy rainfall, but its limitation lies in the small range and weak magnitude, which can be corrected by combining with SWC-WARMS, while GRAPES-GFS, GRAPES-MESO are inferior to the above two models in terms of model stability and prediction accuracy. |
| Key words: rainstorm； prediction error； mode stability |
