摘要
利用增长模繁殖法(BGM),对2012年6月我国长江中下游及华南地区的一次强降水过程进行了不同扰动方案的集合预报试验研究。分别对集合预报中只考虑初值扰动、加入模式物理过程扰动和加入地表参量扰动的三种不同扰动方案进行了对比试验,并对最优试验方案的降水集合预报结果进行检验分析。(1)同时考虑风场、温度场、湿度场、高度场和地表参量的初值扰动,以及考虑模式物理过程扰动的降水集合预报效果最优,可有效提高灾害性天气的预报效果。(2)初值扰动和模式物理过程扰动对降水预报都有着重要的影响,相同的模式物理过程扰动会由于不同的初值扰动而产生较大差异,并且扰动的集合平均预报对降水有较好的指导意义。(3)对于位势高度场预报检验,集合预报平均要好于控制预报,并且随着预报时效延长,这种优势更加明显。(4)对此次强降水过程在长江中下游、华南地区和全国区域的各量级降水预报TS评分检验可知,集合预报较单一的决定性预报(控制预报)效果有明显改善;集合平均预报一般都好于控制预报,同时集合预报也为预报员提供了天气不确定性的警示作用。
Applying the method of BGM,an ensemble forecast has been done of a heavy rainfall in the Yangtze River and south China in June 2012.Three schemes are compared with each of them including only initial turbulence,turbulence of multi-physics and the earth＇s surface turbulence respectively.Verification is done of the correlation coefficient,RMS error of the circulation,TS score of precipitation among the best result of different schemes.It can be seen from the analysis that （1） a multi-physics turbulence scheme is the best of all.With initial turbulence of wind,temperature,humidity and height,as well as the multi-physics turbulence and the earth＇s surface turbulence taken into account,the result of heavy rainfall of ensemble forecast can be improved.（2） The initial turbulence and multi-physics turbulence are all important to rainfall forecast.The initial turbulence brings different result in spite of the same physics.The ensemble average has better result for rainfall forecast.（3） The ensemble forecast is better than the control in the verification of atmospheric circulation,and the result is more obvious for the forecasting time.（4） The precipitation verification with the TS in the Yangtze River and south China shows that there is obvious improvement for ensemble forecasting compared with single control forecasting.The average of ensemble is better than the control basically,and the ensemble forecasting cautions the forecaster of weather uncertainty.
出处
《热带气象学报》
CSCD
北大核心
2014年第4期663-674,共12页
Journal of Tropical Meteorology
关键词
集合预报
增长模繁殖法
初值扰动
模式物理过程扰动
华南暴雨
试验研究
ensemble forecasting
BGM method
initial turbulence
mutil-physics turbulence
rainfall in south China
numerical experiment
