采用门限自回归模型预测环境空气质量

应用汤家豪博士提出的门限自回归模型，以上海市环境空气质量时间序列监测数据和相应的气象数据，建立门限自回归大气污染预测模型，对上海市环境空气质量进行预测预报。结果表明：门限自回归模型计算较为简便且便于计算机自动建模。显著性检验表明门限自回归预报方程高度显著，在实际应用中，采用最新的数据建模和采用分站建立模型的方法可以使预测精度进一步提高。

Adopting Threshold Autoregressive Model to Predict Ambient Air Quality

Threshold Autoregressive (TAR) method was proposed by Dr. Tand Jiahao, and has been applied to ambient air quality prediction in Shanghai.TAR model was based on time series monitoring data of air quality in Shanghai and the relevant meteorological data. TAR model is convenient to calculate and can be built by computer automatically. The result of significant test showed that the prediction equations were high in actual application, using new data to build the model and using the method of building models of each site respectively could increase the accuracy of the prediction.