具有时变参数的门限自回归模型及其在气候预报中的应用  

A TIME DEPENDENT PARAMETRIC THRESHOLD AUTOREGRESSIVE MODEL AND ITS APPLICATION ON CLIMATE FORECAST

作　　者：杨明[1] 张万诚[1] 沙文生[1] 

机构地区：[1]云南省气象局,云南省红河州气象局

出　　处：《热带气象学报》1996年第3期224-229,共6页Journal of Tropical Meteorology

基　　金："云南省灾害性天气预测预报的研究"课题

摘　　要：将时变参数模型引入门限自回归模型中，提出具有时变参数的门限自回归模型，并对昆明、蒙自、河口等地区３月温度序列进行预报，结果表明：这种模型比门限自回归模型的预报准确率有明显提高，这是因为用这种模型进行预报时。In this paper, A time dependent parametric threshold autoregressive model is given and is appllied to forecasting the average temperature of March in Kunming, Mengzi, Hekou et al. Computational results show that the prediction accuracy of time dependent parametric threshold autoregressive model has been obviously improved. This is because the time dependent parametric threshold autoregressive model can update predicting equation at any time of need.

关 键 词：时变 门限自回归 建模参数 气候预报 

分 类 号：P456[天文地球—大气科学及气象学]