Projecting Agricultural Land Abandonment and Expansion in Turkiye Using the Subspace KNN Algorithm (2030)

Abstract

One of the most important changes in agricultural land worldwide is the problem of land abandonment in underdeveloped and developing countries and in regions with low resilience to climate change. This problem is also an ongoing debate in Turkiye. This study first evaluates the process of abandonment and expansion of agricultural land in Turkiye based on the literature before presenting maps of the probability of abandonment and recovery of agricultural land. In this context, the abandonment and expansion of agricultural land due to changes in Coordination of Information on the Environment (CORINE) data for 2006-2018 are used as the dependent variables. Probability maps were constructed using the subspace KNN model with 13 independent variables. The immediate surroundings of cities are seen to come first as the places with a high probability of abandonment. For rural areas, this appears as the areas close to roads and places with forced natural environmental conditions. A very low recovery situation has emerged in terms of places that can be reclaimed. The areas that stand out among these are where hazelnut and tea plantations have expanded in the Black Sea region, as well as the areas with relatively favorable topography in the interior of the Western and Central Taurus Mountains in the Mediterranean region.