中国各省月预报与无技巧预报的评估分析  

The Quantitative Assessment of Random Forecast Ps Score Among the Provinces in China

作　　者：田武文[1] 

机构地区：[1]陕西省气候中心,西安710015

出　　处：《气象》2010年第9期100-105,共6页Meteorological Monthly

基　　金：陕西省气象局重点科研项目<省级滚动预测业务系统及国家动力气候模式统计降尺度解释应用研究>;陕西省气象局科技创新基金项目"气候变暖背景下低温冷害规律及对策研究"共同资助

摘　　要：在Ps评分中,预报技巧是由一段时间内实际业务的Ps评分与随机预报评分之差确定。通过对中国各省无技巧预报评分的评估,定量分析了在现行评分办法中影响无技巧评分存在差异的两个因素。统计结果及分析表明:(1)由于各省测站要素各等级气候概率分布的差异,决定了随机预报准确率存在差异。以省份为单位,随机预报评分降水最大差异为14分,气温最大差异为7分;(2)目前预报正确性判定标准的设计,使得在只做二级预报或持续预报时,可以大幅度提高无技巧预报准确率,给无技巧预报评分的定量评估带来不确定因素;(3)对比持续预报和随机预报的评估结果,全国平均降水提高了9分,温度提高了13分。研究结果表明,由于各省之间的无技巧预报评分存在明显差异,因此直接利用Ps评分比较预报技巧,不能客观反映出预报能力的差异。In the Ps score assessment,the forecast skills are determined by the difference between Ps score of operational forecast and random forecast accuracy over a period of time.The paper attempts to make quantitative evaluation of the random forecast accuracy among all the provinces and regions in China,and the statistical results are as follows.（1） Owing to the obvious difference of the random forecast accuracy about station elements determined by climate probability in each grade in every province,the maximum difference of precipitation score is 14,and the maximum difference of temperature score is 7.（2） Since the judging standard of forecast accuracy has been designed vulnerably,the accuracy of two grade forecast and continuous forecast may be higher than random forecast accuracy,it brings uncertainty to the quantitative evaluation of random forecast accuracy;（3） The score of random forecast might be higher than that of the actual forecast if these factors affecting the score are applied to quantitative assessment.The result shows that the current scoring methods could not access the forecast skills in various provinces properly.

关 键 词：Ps评分 各等级气候概率分布 随机预报 持续预报 预报准确率 

分 类 号：P456.3[天文地球—大气科学及气象学]