时间序列的自回归模型在杨凌地区降水量预报中的应用

Application of the Autoregressive Model of Time Series on the Rainfall Forecast of Yangling District

对平稳时间序列的理论和方法,自回归模型[1]的概念,定阶,模型建立,预报,平稳性进行了综述。通过对杨凌张家岗气象站1935—1955年的四季降水量建立自回归模型,利用所求得的模型对1956年的四季降水量进行了预报。结果表明,该模型对冬季降水量预报效果最好,接下来依次为秋季、夏季、春季。因此平稳时间序列预报方法适用于预报变化比较缓和、持续性和周期性比较好的天气,对变化剧烈的天气和转折性的天气的预报能力较差。

Base on the review of the theory and method of time series and autoregressive model, the autoregressive model concept, settling the rank, model establishment, forecast, the steady, and the four seasons rainfall value. The autoregressive model were given by using 1935--1955 twenty years of the four seasons rainfall of Yangling, Zhang Jiagang as an example to establish autoregressive models, and make use of the model to carry on the simple forecast to the four seasons rainfall of 1956 year. The result shows： the winter rainfall is forecasted best by this model, next to autumn, summer, spring. The model is more applicable to the tempered, durative, periodicly weather than the acute and transition weather.