Malone J B, Yilma J M, McCarroll J C, Erko B, Mukaratirwa S, Zhou X
Pathobiological Sciences, School of Veterinary Medicine, Louisiana State University, Baton Rouge, LA 70803, USA.
Acta Trop. 2001 Apr 27;79(1):59-72. doi: 10.1016/s0001-706x(01)00103-6.
Annual and seasonal composite maps prepared from the normalized difference vegetation index (NDVI) and earth surface maximum temperature (T(max)) satellite data from the archives of the Global land 1-km program of the United States Geological Survey (USGS) were studied for. their potential value, using geographic information system (GIS) methods, as surrogates of climate data in the development of environmental risk models for schistosomiasis in Ethiopia. Annual, wet season and dry season models were developed and iteratively analyzed for relationships with Schistosoma mansoni distribution and infection prevalence rates. Model-predicted endemic area overlays that best fit the distribution of sites with over 5% prevalence corresponded to values of NDVI 125-145 and T(max) 20-33 degrees C in the annual composite map, NDVI 125-145 and T(max) 18-29 degrees C for the wet season map, and NDVI 125-140 and T(max) 22-37 degrees C for the dry season map. The model-predicted endemic area was similar to that of a prior model developed using an independent agroecologic zone data set from the United Nations Food and Agriculture Organization (FAO). Results were consistent with field and laboratory data on the preferences and limits of tolerance of the S. mansoni-Biomphalaria pfeifferi system. Results suggest that Global 1-km NDVI and T(max), when used together, can be used as surrogate climate data for development of GIS risk assessment models for schistosomiasis. The model developed for Ethiopia based on global 1-km satellite data was extrapolated to a broader area of East Africa. When used with FAO agroecologic zone climate data limits of <27 degrees C for average annual mean temperature and annual moisture deficits (annual rain-annual potential evapotranspiration) of <-1300 mm, the model accurately represented the regional distribution of the S. mansoni-B. pfeifferi system in the East Africa extrapolation area.
研究了利用美国地质调查局（USGS）全球陆地1公里计划档案中的归一化植被指数（NDVI）和地表最高温度（T(max)）卫星数据编制的年度和季节合成地图，运用地理信息系统（GIS）方法，探讨其作为气候数据替代指标在埃塞俄比亚血吸虫病环境风险模型开发中的潜在价值。开发了年度、雨季和旱季模型，并对其与曼氏血吸虫分布及感染流行率的关系进行了迭代分析。模型预测的流行区叠加图与流行率超过5%的地点分布最匹配，在年度合成地图中对应NDVI值为125 - 145、T(max)为20 - 33摄氏度；雨季地图中NDVI值为125 - 145、T(max)为18 - 29摄氏度；旱季地图中NDVI值为125 - 140、T(max)为22 - 37摄氏度。模型预测的流行区与先前使用联合国粮食及农业组织（FAO）独立农业生态区数据集开发的模型相似。结果与关于曼氏血吸虫-费氏双脐螺系统偏好和耐受限度的实地及实验室数据一致。结果表明，全球1公里的NDVI和T(max)结合使用时，可作为气候数据替代指标用于血吸虫病GIS风险评估模型的开发。基于全球1公里卫星数据为埃塞俄比亚开发的模型被外推到东非更广泛地区。当与FAO农业生态区气候数据结合使用时，即年平均温度<27摄氏度且年水分亏缺（年降雨量-年潜在蒸散量）<-1300毫米，该模型准确地反映了东非外推区曼氏血吸虫-费氏双脐螺系统的区域分布。
