作者: Stéphane de La Rocque , Hélène Guis , Frédéric Mauny , Jean-François Viel , Annelise Tran
DOI:
关键词: Cartography 、 Satellite 、 External validation 、 Sensitivity (control systems) 、 Landscape ecology 、 Vegetation 、 Computer science 、 Scale (ratio) 、 Set (abstract data type) 、 Digital elevation model
摘要: Summary Landscape ecology is seldom used in epidemiology. The aim of this study to assess the possible improvements that can be derived from use landscape approaches on several scales when exploring local differences disease distribution, using bluetongue (BT) Corsica as an example. environment BT-free and BT-infected sheep farms described a fine scale, high resolution satellite images digital elevation model. Land-coverage characterised by classifying image. metrics are calculated quantify number, diversity, length edge connectance vegetation patches. for three sizes buffers around farms. models tested with without see if such improve models. Internal external validation performed relative impact scale versus variables discriminatory ability explored. Results show all irrespective number parameters included, perform better than those without. 1-km buffer model combines both best application set variables. It has good discriminating sensitivity specificity.