摘要
本文针对预报量(降水)与预报因子之间的非线性特点,增添了幂函数型和交叉乘积型预报因子,同时采用“两段筛选”方法建立MOS判别方程。业务预报试验结果表明:它的击中率达87.5％,Hedij技巧得分为0.34,特别是对大—暴雨的预报能力明显提高。这对一般统计与MOS天气预报具有普遍意义。
The exponential functions and cross-products of the atmospheric variables are considered as predictors based on the nonliner interrelation between the predictand and the predictors. In the meanwhile the MOS discriminate equati ons are de veloped by use of ' both phase screening ' . The rusults of the operational prediction experiment show that the hitting percentage is 87.5%, the Hedis skill score is 0.34 and the forecast accuracy for heavy-storm rainfall is improved obviously. It has universal significance for general statistical forecasts and MOS forecasts.
出处
《气象科学》
CSCD
北大核心
1991年第4期437-443,共7页
Journal of the Meteorological Sciences
