Understanding the mechanisms that drive spatial pattern change over space and time is a central goal of landscape ecology, but these patterns are often complex. Multifractals, an extension of the concept of fractal geometry, are especially suited to the description of complex pattern. Comparing maps between periods, known as post-classification comparison, is one approach to deriving land cover change. In our paper we examine, using multifractal analysis, the forest spatial pattern over a decade. This analysis intends to pursue forest expansion and deforestation in the region of interest. This surveillance of forested lands is very important in context of climate change in our region, perceived by the appearance of extreme weather and desertification trend.