基于雷达基数据的风暴单体跟踪与预报

Storm cell tracking and forecasting based on radar raw data

多普勒雷达基数据对风暴单体的跟踪及预报具有十分重要的意义。针对雷达监测预报的原理和特点,建设性地提出了一种跟踪和预报方法。根据"体扫间隔,特征相似,近距离优先"三个匹配准则来匹配两时刻的风暴单体,再利用加权最小二乘法对风暴单体在下一时刻的位置进行预报。通过对天津市2005~2011年间74个天气过程的实验和评估,结果表明:该方法的可预报单体数更多,单体平均预报误差更小,能够更好地实现风暴单体的跟踪及预报。

It is of great significance to track and forecast storm ceil based on Doppler radar raw data. In view of the principle and characteristics of radar monitoring and prediction, a method of storm cell tracking and forecasting is constractively given. According to three matching criterions of＂ body scanning interval, similar characteristics, close first＂ , it can match the two moments before and after the storm cell. Weighted least squares method is used to forecast the position of the storm cell in the next moment. Through experiment and evaluation of 74 weather process of Tianjin in 2005-2011, evaluation results show that this method can forecast more storm cell, average forecasting error of storm cell is more smaller, and can achieve better tracking and forecasting of storm cell.