GNSS/PWV在一次梅雨期暴雨中的模式预报应用

Application of GNSS/PWV in the Numerical Weather Prediction of a Meiyu Rainstorm

关键词:  GNSS/PWV  数据同化  快速更新同化系统  梅雨  TS评分

Key words:  GNSS/PWV  data assimilation  rapid update cycle  Meiyu  threat score

摘要:

地基全球导航卫星系统 (Global Navigation Satellite System，GNSS)监测大气可降水量(PrecipitableWater Vapor，PWV)是一种连续获取大气水汽信息的有效手段，对于区域天气尤其是灾害性天气观测与预报有重要作用。基于长三角 GNSS 应用示范网，开展 GNSS/PWV 资料在数值天气预报模式中的三维变分同化应用试验，通过设计 5个试验方案对 2010年 7月的一次梅雨期暴雨进行中尺度分析，考查区域地基 GNSS/PWV资料在梅雨期降水过程中对初始场和预报结果的改进能力。通过预报分析表明：本次梅雨期暴雨数值模式预报加入 GNSS/PWV的同化方案相比于常规资料同化方案 24小时降水预报效果改善超过 20%，48小时后也可提高约12%。可见 GNSS/PWV资料可以很好改进观测区域内的水汽分布，提供与暴雨天气紧密联系的水汽信息，有效改善了数值天气模式中的中尺度系统移动速度的48小时预报结果，进而提高降水落区预报。

Abstract:

The monitoring of atmospheric precipitable water vapor (PWV) via the Global Navigation Satellite System (GNSS) is a potent technique for the continuous collection of atmospheric moisture data, which is vital for regional weather observation and forecasting, especially in the context of severe weather events. Based on the GNSS application demonstration network in the Yangtze River Delta, the present study conducted five three-dimensional variational assimilation experiments incorporating GNSS/PWV data in a numerical weather prediction model. The objective was to refine the mesoscale analysis of a Meiyu rainstorm event that occurred in July 2010 and evaluate the potential of regional ground-based GNSS/PWV to enhance the model’s initial conditions and predictive accuracy for Meiyu precipitation events. The analysis showed that the assimilation of GNSS / PWV data into the numerical forecast model for this specific Meiyu rainstorm improved the accuracy of 24-hour precipitation forecast by over 20%, with a 12% improvement observed in the 48-hour forecast. These findings suggest that GNSS / PWV data may help effectively refine water vapor distribution in the observation area, provide water vapor information closely related to rainstorm weather, and significantly improve 48-hour forecast accuracy for the movement speed of mesoscale systems in numerical weather models, thereby enhancing precipitation area forecasts.