中文摘要：

基于2005—2008年南京市呼吸系统疾病死亡人数监测资料、同期气象数据和NCEP／NCAR再分析数据，在风寒指数基础上研究气象要素与呼吸系统疾病死亡人数的相关关系，并进行合成分析和个例分析．结果表明，南京市呼吸系统疾病死亡人数存在季节性差异，高峰期出现在每年的1、2和12月,气温、气压与呼吸系统疾病死亡人数之间存在显著的相关，相关系数分别为-0．460、0．379，表明气温和气压对呼吸系统疾病死亡人数影响较大，主成分分析法结果也证明了这一点．风寒指数作为反映综合气象因子的指数，相关系数为-0．428，通过了0．01的显著性检验．气象因子和风寒指数对呼吸系统疾病的影响均存在明显的滞后效应，气压和气温在第5天达到最大，风寒指数则在第4天影响最大．利用合成分析发现，造成呼吸系统疾病死亡人数异常偏多的冷空气过程中，显著相关的气象因子是变压、变湿和风寒指数，其阈值分别为8．6hPa、i0．1和-1．83．利用2005年3月10-12日冷空气过程进行个例分析，结果显示冷空气过程会带来气象要素的一系列变化，该次个例过程地面增压22．1hPa，风速增加4．6m／s，24h变温达-14℃并带来降水，风寒指数3d平均为-2．3，均显著超过阈值范围，最终导致平均呼吸系统疾病日死亡人数远大于4a日平均值，进一步证实了冷空气过程会导致呼吸系统疾病日死亡人数显著增加，并存在3-4d的滞后效应．

英文摘要：

By using the death rate data on respiratory diseases monitored in Nanjing City, meteorological data and reanalysis data from NCEP/NCAR from 2005 to 2008, the wind chill index was calculated to an- alyze the relationship between meteorological factors and deaths from respiratory diseases. Composite analysis and a case study were also made to support the conclusion. The main results are shown as fol- lows： the peak of respiratory disease death rate was in January, February and December of each year in Nanjing, presenting seasonal variations. There was a significant correlation between temperature, air pres- sure and mortality of respiratory diseases, with a correlation coefficient of-0.46 and 0.379 respectively. It was shown that temperature and pressure had great influences on the mortality of respiratory system diseases, which was verified again by the results of a principal component analysis. As an index fusingsynthetic meteorological factors, the wind chill index was significantly related to the mortality of respi- ratory system diseases, with a coefficient of-0.428, which passed the significance test at the level of 0.01. The influences of meteorological factors and wind chill index on the respiratory system diseases also presented an obvious lag effect. The greatest impact of pressure, temperature and wind chill index on the respiratory system diseases lagged behind five days and four days, respectively. Composite analysis showed that during the cold air process that significantly increased respiratory system disease deaths rate, variations of the pressure, humidity, wind chill index had significant effects on the death toll in this re- spect, with the threshold being 8.6 hPa, ±0.16 and -1.83, respectively. Choosing a typical cold air process from March 10th to 12th 2005, we found that the cold air process brought a series of changes in meteoro- logical elements, including a ground pressure enlargement of 22.1 hPa, wind speed increase of 4.6 m/s, 24 h temperature decrease of-14 ℃, precipitation and an a