Inter-annual monitoring of land use land cover changes with emphasis on forest reserves under a spatial planning perspective

Abstract

Monitoring the inter-annual land cover evolution is considered of crucial importance at global level, since it may reveal significant changes across the Earth surface. Primary aim of the paper is to explore the spatiotemporal changes occurred in a highly touristic region which hosts a diversity of high-value natural resources (forests, coastal landscapes, croplands). To this end, a supervised land use image classification has been conducted for two years (2000, 2019) adopting the support vector machine algorithm. The outcomes of the classification have been verified by the overall accuracy (OA) index and kappa coefficient (KC). Both indices indicated high accuracy (over 80% and 0.8 respectively) of correct classification. Next, a matrix and a map of changes were developed to provide a quantitative and qualitative perspective of land use changes. The primary changes occurred were related to the transformation of forest to cropland and vice versa, followed by a mild urban expansion (especially for touristic premises).