用动力相似方法预报广东省区域暴雨试验

DYNAMIC ANALOGUE METHODS IN EXPERIMENTING REGIONAL HEAVY PRECIPITATION IN GUANGDONG

应实际业务预报需要,利用最新43年欧洲中期数值预报中心(ECMWF)逐日再分析资料,采用多种动力相似方法研究了广东省短期区域暴雨预报,发现了预报员预报经验成长曲线模式,找到了影响暴雨预报的决定性因子850hPa风场,证明了双时次相似的优越性,并研制了广东短期区域暴雨预报的动力相似法,预报评分明显高于业务平均水平。该方法的应用软件实现了短期区域暴雨的自动、客观、定时、定点、定量预报,目前已经投入业务使用。

To meet the urgent need of operational rainstorm forecast in Guangdong province, we use newly-released ECMWF reanalyzed daily data（45 years） to study the dynamic analogue methods. We got the experience-growth curve model of ideal forecaster, which appears to be a powerful function. We found that a decisive factor in the forecast is the 850-hpa wind and statistics show the advantage of the two-time-levels analogue method. We developed Guangdong short-term regional rainstorm forecast system based on these results and had remarkably better forecast score than the real operational forecast in recent 3 years. The applied software has been put into operational run and is making rainstorm forecast automatically, objectively, timingly, locally and quantitatively.