数值预报产品在宝鸡市温度预报中的应用

Application of Numerical Prediction Products in Temperature Forecast of Baoji

采用2008年4月1日至2012年12月31日宝鸡市自动站观测的逐日最高最低气温实况资料与天气在线、兰州、中央台的温度预报产品,通过对中央台、天气在线的数值预报产品中24—72h温度预报准确率进行检验,结果表明,三家产品均是最低气温预报准确率高于最高气温预报准确率,其中最高、最低气温预报准确率均较高的为9月份,三家数值月报产品横向比较,其中中央台预报准确率最高,其次是兰州,德国天气在线最低。并对数值预报产品进行回归分析,得出宝鸡市24—72h的最高最低温度预报方程,选取三个月实况资料对其准确率检验,预报效果较为良好。

Based on the daily maximum and minimum temperature recording data from automatic station during April 1st, 2008 and December 31th, 2012 in Baoji, the accuracy of 24-72 hour temperature prediction among CCTV, weather on line and Lanzhou numerical prediction products were tested. It is found that all the products show higher accuracy of the lowest temperature forecast than the highest temperature forecast and both of them are better in September. Crosswise comparison among the three monthly products show the accuracy of CCTV products is the highest, and Lanzhou is lower, German weather on line is the lowest. The 24-72 hour highest and lowest temperature forecast equation of Baoji is built by making regression analysis for the numerical prediction products. The accuracy of the equation is checked by using three months＂ recording data, and it shows that the forecast result is good.