关键词: 蚊虫灯诱资料, 混合回归分析, 气象因素

Abstract: Multiple stepwise regression was used to analyse the relationship between meteorological factors and successive daily light trap catches of mosquitoes sampling from Aug. 28th to Sep. 16th in a dairy located in the suburb of Shanghai. Among wind speed, temperature, humidity and air pressure, wind speed affected light trap very significantly, so correcting coefficient of light trap for wind speed was calculated. The integrated regressive models of light trap catches were also contructed by including the catch of previous one or two days into multiple regression model as independent variables, and results showed that the accuracy and residual mode of the integrated models were improved significantly, indicating the integrated regressive models are more suitable for describing mosquito population dynamics.