【摘要】 利用欧洲中心数值预报产品,采用完全预报方法(PP法)分季节建立广东省分县第4、5天最高、最低气温预报方程,并进行误差检验和分析。结果表明:误差大小有明显的季节差异,并具有一定的地域分布特点;最低气温的预报效果明显好于最高气温,夏秋季最低气温、夏季最高气温的平均绝对误差均小于2℃,具有较高的参考价值。逐日误差与天气密切相关,并且随季节和天气的不同有一定的偏差规律,较大的误差主要出现在秋冬季降雨天气或强冷空气影响、夏季热带气旋影响下出现明显降水、春季明显回暖期等特殊天气时段。更多还原

【Abstract】 Based on numerical weather predictions of ECMWF,the forecast equations were established by season with Perfect Prediction Method to predict maximum and minimum temperature prediction for day 4 to 5 in Guangdong.The verification shows that the mean error varied obviously with season and its distribution is associated with the geography.The accuracy of the minimum temperature prediction is significantly higher than that of the maximum.The absolute mean error of the maximum temperature prediction in summer and the minimum in summer and autumn are all less than 2 ℃,which shows a good value for reference.The daily temperature error is closely related to weather pattern as well as season.There are several situations that may cause considerable errors such as rainy weather,strong cold air in autumn and winter,tropical cyclones weather in summer,and rapid warm-up weather in spring.更多还原