摘要
在用经验统计方法和降水判别函数进行 2 4h和 12h晴雨预报的基础上 ,再用BP人工神经网络建立降水量级预报模型。经 2 0 0 3年汛期试用 ,预报准确率高于上级指导预报准确率 ,12h预报准确率高于 2 4h预报准确率。
In terms of expirical statistical method and precipitation measurement function, the 24h and 12h clear-rain forecast is carried out first, then forecast made of precipitation grade is developed by means of BP artificial neural network. Experiment in 2003 flood season shows that the system has achieved a good effect.
出处
《河南气象》
2004年第3期12-13,共2页
Meteorology Journal of Henan
