Utilisation of Alsat 2A image for the vegetation mapping of an anthropised forest: example of the Tamgout national forest (Tizi Ouzou, Algeria)

Abstract

Abstract: Forest vegetation in Algeria was reported to be affected by anthropic impact (fires, overgrazing). The aim of this study is to establish the land cover mapping of the national forest of Tamgout (Tizi Ouzou) using Alsat 2A image. This forest located on the coastal chain of Kabylia (3 800 ha) is known for its high anthropic activities resulting in strongest impacts sheltering a very important forest heritage. It is a cork oak forest slightly degraded, but its ecosystem services and socio-economic value are undeniable for local populations, especially non-timber forest products. The use ENVI 4.5 software processing of Alsat 2A satellite image, focused on supervised classification with minimal distance method, allowed land cover mapping of the Tamgout forest (classification performance: 93,94%, Kappa Index: 0.91). We determined the surface importance of vegetation units identified with a sampling mission on 4 ha of land (i.e. 0.1% of forest total area). Natural forest vegetation was estimated at 1263 ha (33%) of cork oak (Quercus suber), 760 ha (20%) of zen oak (Quercus canariensis), and 945 ha (25 %) of mixed formation. Maquis was about 302 ha (8%). Artificialized units such as bare land, agricultural land and agglomerations were evaluated at 530 ha (14% of forest total area). Our findings serve as a basis for the analysis of spatial and temporal land cover changes, and contribute to a better knowledge of the anthropisation phenomenon, following the recurrent fires, and to the fragmentation process in the Mediterranean forest environment.