一种场对场的复合门限统计预报模型及其应用

A FIELD TO FIELD COMPOUND THRESHOLD REGRESSION MODEL AND ITS APPLICATION

本文根据对数据总体非线性特征可用逐段线性化来逼近的原理，同时在考虑预报系统中预报对象场和预报因子场的整体性和突跳性的基础上，获得一种场对场的复合门限统计预报模型。应用该模型对具有明显突跳性的云南五月雨量场进行了模拟和预报，结果较为理想。

In this paper,based on the principle that the nonlinear characteristics of data can be described gradually by linear characteristics of data and according to main consideration of the entirety and catastrophic attribute between predicted object field and factors field, a field to field compound threshold regression model is presented, and then this model is used to forecast the May rainfall field of Yunnan province. The results are satisfactory.