摘要
利用日本数值天气预报产品和地面天气图,集成多种统计学方法对沈阳市降水概率进行预报,经专家订正后对社会发布。本预报模式避免了单一统计学方法的局限性,并且实现了历史因子的自动追加,能够反映最新的天气信息。经在沈阳市气象台一年的试预报,根据预报结果计算得半贝里尔得分B=0.065;Bias得分为Bias=2%;预报技巧得分Bs=80.8%。同时准确率明显高于日本数值预报结果。
Based on Japanese numerical weather prediction product and ground synoptic chart, the of forecast model precipitation probability in Shenyang was applied. This model can avoid the shortages of single statistical method with integrated multistatistical ones, and obtain up-to-data weather information with factors adding to the database automatically. After one year probation in Shenyang Municipal Meteorological Bureau, Half Brier score B=0.065, bias score 2% and forecast skills score Bs=80.8% were gained according to the forecast results. The forecast accuracy using this model was obviously higher than Japanese numerical forecast results.
出处
《沈阳农业大学学报》
CAS
CSCD
北大核心
2005年第2期148-151,共4页
Journal of Shenyang Agricultural University
基金
沈阳市科委资助项目(2001305-01)
关键词
降水概率
天气预报
预报模型
precipitation probability
weather forecast
forecast model
