Mapping Tree-Covered Forest Formations (Case Study: Beni-Mellal Forest): Evaluation Study of the Potential of the Mohammed VI-B Satellite Image

Abstract

The use of satellite imagery provides valuable information on vegetation (agriculture, forest, etc.). Among the recently launched satellites are the Mohammed VI satellite. The Mohammed VI satellite imagery has a very high spatial resolution (VHR) (50 cm), which enables the study of vegetation at a very fine scale, such as the mapping of tree species. However, to date, few studies have been published on vegetation analysis using Mohammed VI satellite imagery. For this reason, this study presents an attempt to use Mohammed VI satellite imagery for forest vegetation mapping. In this project, we employed three main approaches to map tree species, namely a pixel-based approach to eliminate the “non-vegetation” class, a photo-interpretation approach to create samples (training and validation), and an Object-Based Image Analysis (OBIA) to classify tree cover and tree species. This study demonstrated that the Mohammed VI satellite imagery allowed for the extraction of vegetation and tree cover with an accuracy exceeding 90%. However, the accuracy for mapping tree species is around 60%. Therefore, this work has shown that the Mohammed VI satellite imagery presents a real potential for mapping vegetation and, more specifically, forest tree species.