Coupling long-term prospection data and remote-sensing vegetation index to help in the preventative control of Desert Locust
2011
Piou, Cyril | Salem Benahi, Ahmed | Bonal, Vincent | El Hacen Jaavar, Mohamed | Lebourgeois, Valentine | Lecoq, Michel | Vassal, Jean-Michel | Bioagresseurs, analyse et maîtrise du risque (UPR Bioagresseurs) ; Centre de Coopération Internationale en Recherche Agronomique pour le Développement (Cirad) | Centre National de Lutte Antiacridienne ; CNLA | Territoires, Environnement, Télédétection et Information Spatiale (UMR TETIS) ; Centre de Coopération Internationale en Recherche Agronomique pour le Développement (Cirad)-AgroParisTech-Centre national du machinisme agricole, du génie rural, des eaux et forêts (CEMAGREF) | CIRAD scientific workshop
Prospection data are generally collected in oriented manner and toward the immediate needs of pest management. Despite the evident statistical bias these data present, when coupled with external indicators of environmental status, prospection data can help in characterizing interesting relationships between the focused pest and its environment. Desert Locust management is generally done through a preventative control avoiding population to reach high and uncontrollable densities. The areas of potential start of gregarization process for Desert Locust are large and preventative management teams need to prospect all these areas to be efficient. A challenge of ongoing research is to be able to guide on where prospection surveys should be done depending on meteorological and vegetation conditions. An analysis of relationship between long-term prospection data of Desert Locust observations from 2005 to 2009 and spatio-temporal statistics of a vegetation index gathered by remote-sensing was conducted using logistic regressions. The vegetation index was a composite Normalized Difference Vegetation Index (NDVI) given every 16 days and at 250m spatial resolution (MOD13Q1 from MODIS satellite). The statistics extracted from this index were: 1) spatial means at different scales around the prospection point, 2) relative differences of NDVI variation through time before the prospection and 3) large scale summary of vegetation quality. Identical statistics could potentially be computed for actual NDVI. By extrapolation of the chosen logistic regression model, maps of probability of presence of locust could be constructed. This methodology should help in focusing prospection toward sensible parts of the gregarization areas at specific times.
显示更多 [+] 显示较少 [-]