时间序列自回归模型预测茶园小绿叶蝉种群动态的探讨

Discussion on forecasting on population dynamics of tea green leafhopper based on temporal sequence autoregression model

选定西湖龙井茶叶原产地的优质丰产茶园、普通丰产茶园、山林间茶园、种质资源圃和茶-林间作等5类代表性茶园,从2007年4月20日～12月26日,5 d 1次调查假眼小绿叶蝉种群数量,作为时间序列自回归模型的数据源。据此对自回归模型预测假眼小绿叶蝉种群动态的可行性、建模步骤及准确性验证进行了探讨。与其他相关预测模型相比,自回归模型显得简便易行实用。

The five typical tea gardens,i.e.high efficient fertility tea garden,common fertility tea garden,tea garden among mountain forest,tea plant idioplasm resource garden,and tea and forest intercrop tea garden,were chosen in Xihulongjing Tea provenance.The population abundance of tea green leafhopper,Empoasca vitis Gothe,was investigated at intervals of five days,from April 20 to December 26 in 2007,which acted as the data source of the temporal sequence autoregression model.Based on the data source,the feasibility,the modeling procedure and the accuracy on the autoregression model to be used to forecast leafhopper dynamics,was discussed.Compared with the other correlative models,the autoregression model appeared simple and practical.