低频天气图在中期延伸期天气预报的应用

Application of Low-Frequency Weather Maps in Mid-Term Extension Weather Forecast

摘要: 低频天气图是一种从大气低频振荡的角度来做延伸期预报的新方法，是未来延伸期预报的新方向。本文采用2012~2017年4月至6月逐日500 hpa流场数据，选取的区域范围为北纬10˚~70˚、东经80˚~140˚之间，另有湖南省97个地面观测站的降水资料。根据这些数据绘制出低频天气图，确定出适合湖南省降水天气过程的天气关键区，分析出低频气旋(C)和低频反气旋(A)的移动路径和变化周期，建立湖南省低频天气图降水预报的模型，该模型可以提前30天左右对湖南地区进行降水预报。最后对湖南地区2017年4~6月份的强降水过程进行了延伸期预报应用，根据本方法利用建立的预报模型总共进行了8次预报，这8次预报中有7次的强降水过程是预报出来了，有1次强降水过程是空报了，另外还有3次漏报的情况，综合预报准确率在60%以上，从应用的准确率可以看出使用低频天气图进行湖南地区的延伸期天气预报有一定的可行性。对空报和漏报的原因进行分析，分析了2017年4~6月平均降水量和降水日数的空间分布图，得出漏报和空报的原因可能是与湖南的地形有关，降水分布不均匀，湘西和湖南北部降水较多，以后研究的方向可以根据实际降水范围划分区域分别预报。

Abstract: The low-frequency weather map is a new method for extended weather forecasting from the perspective of low-frequency oscillations in the atmosphere, and is a new direction for future extension forecasting. This article uses daily 500 hpa flow data from April to June 2012-2017, the selected area ranges from latitude 10˚ - 70˚ and longitude 80˚ - 140˚, and precipitation data from 97 ground stations in Hunan Province. Based on these data, a low-frequency weather map was drawn to determine a weather-critical area suitable for the precipitation process in Hunan Province, and the movement path and change period of the low-frequency cyclone (C) and the low-frequency anticyclone (A) were analyzed, and the low-frequency synoptic precipitation in Hunan Province was established. The model for forecasting can predict precipitation in Hunan in 30 days or so. Finally, an extended period forecast application was made for the heavy precipita-tion process in April-June 2017 in Hunan, and a total of eight forecasts were performed using the established forecast model according to the method. Among these eight forecasts, seven heavy precipitation processes were forecasted. One heavy precipitation process was empty and there were three missed cases. The integrated forecast accuracy rate was over 60%. From the accuracy of the application, it can be seen that the use of low-frequency weather maps for the extended period weather forecast in Hunan is feasible. The analysis of the reasons for the empty and missing reports was conducted, and the spatial distribution maps of the average precipitation and precipitation days from April to June in 2017 were analyzed. The reasons for the missing reports and the empty reports may be related to the topography of Hunan, and the uneven distribution of precipitation. There are many precipitations in Xiangxi and northern Hunan, and the direction of future research can be predicted based on the actual precipitation area.