Résumé
Future sustainable management of oasis is related to good selected archives of historical dynamics, assessment of actual stats by using earth observation data, and GIS tools. In this study, we combined remote sensing and landscape metrics to monitor the dynamics of Tunisian oasis landscapes, in particular, to characterize land use in the Nefzaoua and Djerid regions. The methodology is based on the analysis of NDVI maps and then through the metrics index analysis using Patch Analyst for oasis class. The application covers a period of thirty-six years from 1979 to 2015 from the processing of Landsat images series. Results reveal an unequally distributed oasis progression in space, with a slightly different rate over the period considered. The increased oasis area was mainly due to uncultivated land cover. However, the analysis of spatial metrics suggests a transition from an accelerated fragmentation phase of the oasis landscape, to an expansion phase by the continuous spreading of existing oasis surfaces. A novelty of this study is that the results observed for landscape metrics can be correlated with a human dimension. The use of Geographical Information System based metric information system and spatial analysis with Inverse Distance Weighted interpolation enabled the future mapping of oasis extension in South Tunisia.