escaduto/Tree-Species-Detection

Satellite-based detection of Tree Species Argania spinosa in Morocco from 1984 to 2018 using Random Forest Supervised Classification # Satellite-based detection of Tree Species Argania spinosa in Morocco from 1984 to 2018 using Random Forest Supervised Classification ## Table of Contents - Project Description - Study Region - Methods - Key Findings - Challenges ## Project Description Following the commodification of its extracted oil, there has been a rapid appreciation of argan prices in the 1990’s. Likewise, an increased conservation effort in the management of the natural resources has been implemented within the region (Lybbert et al., 2011). A widespread interest in the sustained health of the argan forest, coupled with incentives for poverty alleviation of the Berber people in sourthwestern Morocco has been prevalent between various stakeholders. Considering that the Argania spinosa species is endemic to Morocco and the dominant role argan forests play in sustaining the livelihoods of rural households (e.g. grazing, timber harvesting) it is imperative to monitor the historical and present day extent of the forest. Therefore, this study aims (1) to quantify the extent of argan forest using a pixel-based supervised Random Forest classification approach and (2) measure the argan forest canopy density through various techniques and vegetation indices throughout 1984 and 2018. By identifying trends seen in multi-decadal timescales, such data and analysis can help aid in understanding the regional land-use land-cover change (LULC) as it pertains to the argan forest and climate, as well as anthropogenic factors that determine its longevity. ## Study Region The argan forest extent in 1994, as determined by the Morrocan forest inventory, was used as the primary ground-truth and reference data during the collection of the training dataset (Fig. 1). Unlike various other vegetation types in the region such as certain grassess and shrubs, argan tree presence is typically fixed unless external drivers such as overgrazing, logging, and/or intense climatic stressors are acted upon the forest community. …