Firstly,the paper uses Euclidean distance to find the most recent forecasts from the point of observation stations,then correspond to rainfall error analysis,last sub-period statistics derived from these errors each time variance,thus more accurately evaluated rainfall forecasting methods.The rainfall measured directly corresponding forecast rainfall comprehensive data analysis,error caused by a combination of different public feelings,resentment degrees,are different,the degree of resentment into seven grades,then sub-period average of all grades statistics objectionable degree comprehensive variance and resentment of the two factors,and then to evaluate the rainfall forecast methods.Matlab software by simulating the model fit better with reality,can provide a reference for weather forecasting department.%文章首先利用欧氏距离,找出与观测站点距离最近的预报点,然后将它们对应的雨量进行误差分析,最后统计这些误差分时段得出各时段方差,进而精地评价雨量预报方法.全面分析直接将实测雨量对应的预报雨量数据,结合不同的误差造成公众感受(反感度)的不同,将反感度分为7个等级.再分时段统计出所有等级反感度的平均值,综合方差和反感度2个因素,进而评价雨量预报方法.通过Matlab软件的模拟,该模型与现实情况拟合较好,可为气象预测部门提供一定的参考.
展开▼