国家级极端高温短期气候预测系统的研制及应用  

Development and Application of National Prediction System for Extreme High Temperature

作　　者：刘绿柳[1] 孙林海[1] 廖要明[1] 朱艳峰[1] 邹旭恺[1] 王有民[1] 颜京辉[1] 

机构地区：[1]国家气候中心,北京100081

出　　处：《气象》2008年第10期102-107,共6页Meteorological Monthly

基　　金：中国气象局2007年多轨道业务建设项目"高温热浪监测预警业务系统"

摘　　要：目前,国家级旬、月极端高温预测的业务工作尚未展开,国家级极端高温短期气候预测系统的研制为中国高温预测业务提供了一个有效平台。该系统以百分位相对阈值和35°C、38°C绝对阈值作为高温阈值指标,应用月动力预测模式、动力预测与统计降尺度相结合、物理统计相似三种不同的方法,预测未来1～40天的旬、月极端高温发生概率及高温日数,定期滚动发布预测产品。并以相关系数为衡量指标,应用交叉检验方法对三种方法的预测效果进行分析,结果表明不同方法高相关区的空间分布存在差异,在进行高温预测时,应综合各方法高相关区的预测结果给出综合预测意见。对2007年7月极端高温预测个例分析表明三种方法均有一定的预测技巧。At present, national prediction system for dekad and monthly extreme high temperature has not been undertaken. Here, a national forecast system for extreme high temperature was developed. It is a powerful platform that regularly issued three kinds of predicting products at website with three methods of dynamical monthly forecast model, statistical downscaling and physical analogue using percentile values at 90 %, 95 % and 99 % as relative thresholds and 3512 or 3812 as absolute thresholds. The products consist of probability prediction, the number of day exceeding 35℃ or 38℃ involving 7 periods： 1 - 10 day, 11 - 20 day, 21 - 30 day, 1 - 30 day, 6 - 35day and 31 -40 day. The spatial distribution of correlation coefficient between hindcast and observation shows that high value areas are different among these methods. So it is hard to say which method is better than others. Thus three predicted results should be considered as well as distribution of correlation coefficient before final extreme high temperature prediction. The predictions by these methods for July, 2007 were generally close to the observation. But the result by statistical downscaling method is better, and that by DERF is worse relatively.

关 键 词：高温预测 月动力延伸预测模式 统计降尺度 物理统计相似 

分 类 号：P463.3[天文地球—大气科学及气象学] S165.27[农业科学—农业气象学]