摘要
本文以T42数值预报产品为因子，用卡尔曼滤波方法建立可更新系数的预报方程作南京和芜湖的温度预报。从1994年1月到4月共84次试报中，用完全相同的因子卡尔曼滤波方法比回归分析方法预报结果精度提高0．5度。
The temperature forecasting equation with renewed coefficients are set up by using Kalman filter technique based on the factors of T42 productions for forecasting in Nanjing and Wuhu. The result shows that precision of Kalman filter method is by 0. 5℃ higher than that of linear regression by using identical factores from January to April in 1944 (in total 84 times of forecasting).
出处
《气象科学》
CSCD
北大核心
1995年第1期85-88,共4页
Journal of the Meteorological Sciences
关键词
T42产品
卡尔曼滤波
温度预报
Kalman filter
temperature forecasts
T42 productions
