| 摘要: |
| [目的]为提高国家级智能网格指导预报（SCMOC）对河西走廊干旱区汛期不同量级降水预报的订正水平。[方法]对2019—2022年汛期SCMOC每日08时和20时起报的逐12h格点降水预报产品和同水平分辨率的三源融合网格实况降水分析产品（CMPAS）进行概率匹配，得到河西走廊不同量级降水的预报订正值，并利用面向对象检验方法（MODE）检验订正前后评分变化。结果 SCMOC对河西走廊小雨和中雨的预报能力强于大雨和暴雨，对中雨及以下量级降水预报偏大，对大雨及以上量级降水预报偏小；(2) 预报—观测概率匹配订正法在河西走廊对中雨和大雨订正效果较好，小雨因SCMOC本身预报能力强而订正效果不明显。[结论]该方法对河西走廊干旱区的降水有一定的订正能力，可为降水预报业务提供一定的技术支持。 |
| 关键词: 河西走廊；格点降水预报；概率匹配；模式订正 |
| DOI： |
| 投稿时间：2024-01-09修订日期：2024-09-23 |
| 基金项目:国家自然科学基金面上项目（41975015）；甘肃省气象局面上项目（Ms2022-11）；武威市自然科学基金项目（WW2201RPZ025） |
|
| Application of Forecast-observation Probability Correction Method in Grid Precipitation Forecasting in Hexi Corridor |
| LiTianjiang,Li Lingping,LI Yanying,WU Wen |
| (Wuwei National Climate Observatory of Gansu Province) |
| Abstract: |
| To improve the correction ability of national level intelligent grid guidance forecast (SCMOC) for different levels of precipitation forecasts during the flood season in the arid areas of the Hexi Corridor, we used the forecast-observation probability matching correction method in the probability matching between the 12 h grid precipitation forecast product of SCMOC initiated at 08:00 and 20:00 every day in the flood season of 2019—2022 and the CMA Multi-source merged Precipitation Analysis System (CMPAS) grid precipitation observation data with the same horizontal resolution, and obtained the forecast correction values for different levels of precipitation in the Hexi Corridor. Moreover, the changes in scores before and after correction were tested by the Method for Object-Based Diagnostic Evaluation (MODE). The results show that: (1) SCMOC has a stronger ability to predict light rain and moderate rain than heavy rain and torrential rain in the Hexi Corridor. The forecast of moderate rain and below is biased towards higher value, but that of heavy rain and above is biased towards lower value. (2) The forecast-observation probability matching correction method has a good correction effect on moderate and heavy rain in the Hexi Corridor, while the correction effect on light rain is not significant due to the strong forecasting ability of SCMOC itself. Therefore, this method has a certain correction ability for precipitation in the arid areas of the Hexi Corridor and can provide some technical support for precipitation forecasting. |
| Key words: Hexi Corridor; grid precipitation forecast; probability matching; model correction |
