<![CDATA[Madagascar is one of the countries affected by sand. This phenomenon creates not only the change of land use but especially causes negative impacts to the development of this country on different sectors such as agriculture. Despite a classic methodology based on field surveys and site observations to collect informations as measure sand height and analyze soil conditions, it remains slown labor-intensive and time-consoming for monitoring the evolution of sand encroachment zone, and presents limitations in risk assessment. This article proposes the use of multi-sensor and multi-date Landsat satellite images for the study and analysis of the evolution of silting risks in Madagascar. A methodology has been developed not only to evaluate the evolution of land use but also to predict the risks of sand. It has been applied on the commune of Antanambao in Marovoay district of Madagascar. Thus, we have established a change map of the evolution on this study area during the ten (10) years as well as three (3) maps showing the forecast of the risks of sand. In addition, we also have deduced that the object-based classification gave a higher accuracy than the pixel-based classification in this study case, while the method based on the Digital Elevation Model is enormously interesting to the prediction of sand risks in an area.]]>
Copyrights © 2026
