【摘要】 目的探索上海地区三带喙库蚊春季首次出现日预测预报模式,用现场定期捕捉三带喙库蚊来验证预测预报模式的可行性。方法根据1982~1992年和2001~2003年每年3月1日起(冬季遇日平均气温≥10℃时)在动物养殖场挂紫外线诱蚊灯捕蚊资料以及三带喙库蚊首次出现日前后有关气象资料,用回归分析、相关分析、相关系数普查等统计方法,建立上海地区三带喙库蚊春季首次出现日预测预报模式,并于2005~2006年连续2年现场灯诱捕蚊进行验证预测预报模式的可行性。结果三带喙库蚊春季首现日与首现日温度日期两者之间的相关系数为0.9919,达到极显著水平(r≥r0.01=0.622 6),建立的多元回归预测模型为y2=55.766 4-62.508 3×x1+0.179 9×x2-0.291 4×x3,预测三带喙库蚊春季首现日期与实况值的拟合效果达到基本准确以上的历史拟合率为76.9%(10/13),预测值与实况值的绝对误差平均为4.5 d,从预测值和实况值两根曲线的变化趋势看,预测值和实况值比较吻合,用此模型预测2005、2006年三带喙库蚊春季首现日分别出现在4月3日和4月5日,实际为4月7日和4月1日,绝对误差均为4 d,预测基本准确。结论采用温度特征值———首现日温度日期诊断三带喙库蚊春季首现日和使用多元回归方程预测三带喙库蚊春季首现日的误差统计表明,其预测或判别结果精度较高,稳定性好,两种方法相互补充,用于乙脑媒介检测及相关疾病流行病学调查工作中进行应用。更多还原

【Abstract】 Objective To study the predictive model by the meteorological factors for the first appearance of Culex tritaeniorhynchus in Shanghai area. Methods All-night-catches of mosquito were made by light traps at pigsties in March 1 to 31 May in 1982 to 1998 and 2001 to 2003,when the daily mean temperature was ≥10℃.According to this date and analysing the meteorology of the first appearance of Culex tritaeniorhynchus in spring,by the regressive analysis correlation analysis and general survey of correlation coefficient.It had built a predictive model to test and verify the model possibility from 2005 to 2006.Results The correlation coefficient between the first appearance of Culex tritaeniorhynchus and the temperature of first appearance.day was 0.991 9(r≥r_0.01 =0.622 6),the predictive model of multiple regression was y₂=55.766 4-62.508 3×x₁+0.179 9×x₂-0.291 4×x₃,the historical goodness of fit coincidence rate over the basic acuracy between the estimated earliest appearing date and the actual first appearing date of Culex tritaeniorhynchus was 76.9%（10/13）.The absolute error average was 4.5d between the predictive date and the actual date;in addition from the two curves variety trend that predicts the value and actual conditions to be worth relatively fitting together.By this model to predictive that first appearance of Culex tritaeniorhynchus in 2005 and 2006 was on 3 April and 5 April,the actual appearance date was on 7 April and 1 April,the absolute error average was 4d,predictive model was of basic accuracy.Conclusion Adopting the temperature characteristic value-used day of temperature of the first appearance,Culex tritaeniorhynchus diagnosed the day of first appearance.Culex tritaeniorhynchus and multiple regression of usage returned the square distance to predict now in spring.Culex tritaeniorhynchus databased at the spring head day of error margin statistics the enunciation,it predicted or distinguished the result accuracy higher,stability good.Two kinds of methods added together,it could be working in the investigation of medium monitor,forecast and the related disease epidemiology.更多还原