北京市海淀区腹泻疾病气象指数预报研究及应用

Research and Application of Meteorological Index Forecast of Diarrhea Disease in Haidian District, Beijing

基于北京市海淀区2004--2007年夏季（6—9月）疾病预防控制中心的逐日腹泻疾病发病资料和同期地面气象观测资料，首先采用经典方差贡献分析技术分析了逐日发病人数与气象要素的相关性，选取方差贡献较大的要素作为预报因子，然后根据概率积分方法将发病人数划分为5个级别，采用回归方法分别建立了不同预报时效的预测模型，并检验历史样本的拟合结果。研究结果表明：（1）腹泻疾病发病人数与湿度、风速、气温均呈显著的正相关，其中，水汽压对腹泻疾病的方差贡献最大。（2）1~4d的预报效果较好，其中1～2d预报结果的完全正确率在40％以上，预报误差≤±1级的准确率可达98％。

Based on the synchronization data of diarrhea disease and the meteorological observation in Haidian District of Beijing from June to September during 2004-2007,the relationship between the incidence of diarrhea disease and weather conditions was analyzed in terms of Classical Variance Contribution Analysis technology. The meteorological factors which have significant variance contribution were selected as forecast factors. The daily incident number of disease was divided into five grades by using multiple regression probability grade analysis, and the forecast models for different forecast periods were established withRegression Analysis method and were examined with historical independent sample tests. The results show as follows. （1）There was significant correlation between incident number of diarrhea disease and meteorological condition, especially for the factors of humidity,wind speed and temperature, and the proportion of variance contribution of vapor pressure was the highest.（2）The forecasting for 1 ~4 days had a good effect, and the exact correct rate for 1-2 days forecast was more than 40%, if the difference between the forecasted grade and actual grade was no more than 1, the mean exact correct rate was 98%.