Fiche du document
12 octobre 2022
- ISIDORE Id: 10670/1.817158...
- hal: hal-04380501
info:eu-repo/semantics/OpenAccess
Sujets proches
Selection, ArtificialCiter ce document
Thibault Catry et al., « Predicting tiger mosquito breeding sites through earth observation and statistical analysis », HAL SHS (Sciences de l’Homme et de la Société), ID : 10670/1.817158...
Métriques
Partage / Export
Résumé
Reunion Island is a densely populated French region of the Indian Ocean. There, the Asian Tiger mosquito, Aedes albopictus, is the main vector of arboviroses such as dengue and chikungunya. Recently, vector control services have been collecting in situ data on potential breeding sites covering about 10% of the districts of the island. This study investigates the capability of environmental features extracted from earth observation data to be used in predicting the distribution of potential breeding sites of Ae. albopictus in Reunion Island. Spectral indices (NDVI, NDWI and Brightness index) were calculated from a very-high spatial resolution Pléaides mosaic of 2020 images over the island, while data on the annual cumulative rainfall and the annual average temperature were obtained from the Météo-France / CIRAD network. Textural information over urban areas was also extracted from the Pléaides mosaic. Corresponding textural indices highlight different district typologies based on spatial arrangements between buildings and vegetation. For each district, we analyzed the correlations between i) the number of potential observed breeding sites and ii) the mean values of textural and spectral indices, as well as the mean precipitation and annual temperature, through both univariate and multivariate analysis; the most significant variables were eventually combined using a generalized linear model. We found that the correlation between those combined variables and in situ data on breeding sites of Ae. albopictus is significant. This statistical relationship makes it possible to use earth observation environmental variables as predictors of the number of potential breeding sites for all areas in the island.
