CAJ | 학술논문

目的分析海南省疟疾流行空间分布特征及其与自然环境因素的相关性,构建海南省疟疾流行空间分布预测模型.方法获取2000年海南省雨季(5-10月)各县(市)疟疾发病资料以及气象、土地利用类型构成比、地表温度(LST)和各地平均高程等数据,分析发病率与上述各环境因素的相关性,应用负二项回归分析建立发病率的预测模型,并应用预测模型预测疟疾流行风险的空间分布.结果海南省2000年雨季各市(县)疟疾发病率与各地的海拔高度、林地面积构成比、草地面积构成比呈显著正相关;与耕地面积构成比、城乡、工矿、居民用地面积构成比、LST呈显著负相关;与水域面积构成比、未利用土地面积构成比、平均气温、平均最高气温、平均最低气温、平均极温差、平均相对湿度及降雨量无明显相关性.负二项回归分析引入方程的因子为LST,回归方程为:Ⅰ(月发病率,单位:1/100万)=exp(-1.672-0.399×LST).结论海南省疟疾流行空间分布与多种环境因素有关,可以利用遥感技术获取有关环境指标来预测疟疾流行风险的空间分布.
목적 분석해남성학질류행공간분포특정급기여자연배경인소적상관성,구건해남성학질류행공간분포예측모형.방법 획취2000년해남성우계(5-10월)각현(시)학질발병자료이급기상、토지이용류형구성비、지표온도(LST)화각지평균고정등수거,분석발병솔여상술각배경인소적상관성,응용부이항회귀분석건립발병솔적예측모형,병응용예측모형예측학질류행풍험적공간분포.결과 해남성2000년우계각시(현)학질발병솔여각지적해발고도、임지면적구성비、초지면적구성비정현저정상관;여경지면적구성비、성향、공광、거민용지면적구성비、LST정현저부상관;여수역면적구성비、미이용토지면적구성비、평균기온、평균최고기온、평균최저기온、평균겁온차、평균상대습도급강우량무명현상관성.부이항회귀분석인입방정적인자위LST,회귀방정위:Ⅰ(월발병솔,단위:1/100만)=exp(-1.672-0.399×LST).결론 해남성학질류행공간분포여다충배경인소유관,가이이용요감기술획취유관배경지표래예측학질류행풍험적공간분포.
Objective To better understand the characteristics of spatial distribution of malaria epidemics in Hainan province and to explore the relationship between malaria epidemics and environmental factors, as well to develop prediction model on malaria epidemics. Methods Data on Malaria and meteorological factors were collected in all 19 counties in Hainan province from May to Oct. , 2000, and the proportion of land use types of these counties in this period were extracted from digital map of land use in Hainan province. Land surface temperatures (LST)were extracted from MODIS images and elevations of these counties were extracted from DEM of Hainan province. The coefficients of correlation of malaria incidences and these environmental factors were then calculated with SPSS 13.0, and negative binomial regression analysis were done using SAS 9.0. Results The incidence of malaria showed (1) positive correlations to elevation, proportion of forest land area and grassland area; (2) negative correlations to the proportion of cultivated area, urban and rural residents and to industrial enterprise area, LST; (3) no correlations to meteorological factors, proportion of water area, and unemployed land area. The prediction model of malaria which came from negative binomial regression analysis was: Ⅰ(monthly, unit:1/1 000 000) = exp( - 1. 672 - 0. 399 × LST). Conclusion Spatial distribution of malaria epidemics was associated with some environmental factors, and prediction model of malaria epidemic could be developed with indexes which extracted from satellite remote sensing images.