陕西省旱涝季度、年度预报和集成预报方法  

RESEARCH OF FLOOD/DROUGHT FORECAST AND COMPOSITE FORECAST METHODS FOR THE SEASON AND YEAR IN SHAANXI PROVINCE

作　　者：王玉玺[1,2] 田武文[1,2] 杨文峰[1,2] 李智嘉[1,2] 董亚非[1,2] 杨月娟[1,2] 

机构地区：[1]陕西省气象科学研究所 [2]陕西省气象台

出　　处：《高原气象》1998年第4期427-436,共10页Plateau Meteorology

基　　金：中国科学院"八五"重大应用项目;"陕西灾害气候方法研究"课题

摘　　要：在大量天气预报实践的基础上，通过分析地理环境、气候背景、大型天气过程的转型及“突变”、太阳活动及天文原因选取预报因子，分析了降水特征，确定了旱涝指标，并采用人工神经元网络、车比雪夫多项式、方差分析等方法建立预报模型，最终用集成预报方法建立预报模型和预报系统。经过试报，准确率达到８３．３％，取得了良好的效果。Flood/drought is the main climitic disaster in Shaanxi province. As a matter of experiences in weather forecast,forecast factors are selected,precipitational characters are analysed and flood/drought indicatrixes are detemined by analysing geographical environment,climitic background,turning pattern and abrupt variation of large scale weather,solar activity and astronomical factor. Finally, using artificial nervous unit network,Chebyshev polynomial and variance analysis etc.. A composite forecast method forecasting the season and year flood/drought in Shaanxi province is found. By test forecast in 1994, accurate rate reaches to 83.3 percent and good effect is obtained.

关 键 词：旱涝 季度预报 年度预报 集成预报 

分 类 号：P426.616[天文地球—大气科学及气象学] S423[农业科学—植物保护]