| 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,可以随时更新预报模式。 |
| 关键词: 时变 门限自回归 建模参数 |
| DOI： |
| 分类号: |
| 基金项目:"云南省灾害性天气预测预报的研究"课题 |
|
| A TIME-DEPENDENT PARAMETRIC THRESHOLD AUTOREGRESSIVE MODEL AND ITS APPLICATION ON CLIMATE FORECAST |
|
Yang Ming1, Zhang Wancheng2, Sha Wensheng2
|
|
1.Meteorological Bureau of Yunnan Province, Kunming, 650034;2.Honghe Prefecture Meteorological Bureau, Yunnan
|
| Abstract: |
| 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. |
| Key words: Time-dependent Threshold autoregressive Model build-up Parameter |
