Increasing population caused by migration from rural to urban leads the cities, whose ownership is primarily public, to spread over forest areas and agricultural lands in their periphery. Today, simulation models created by geospatial data mining (GDM) techniques are frequently used to model the effects of uncontrolled urban sprawl on land use/land cover (LULC) changes. However, most of these models have yet to consider the impact of ownership diversities. The study’s main purpose is to investigate the interactions and interdependencies between Spatiotemporal LULC changes and private land ownership and monitor the precipitous urban sprawl through simulation models created by GDM techniques. In this context, the Sancaktepe district of Istanbul, where the population growth is rapid and the LULC change is very excessive, was chosen as the study area. CORINE LULC data for 1990, 2000, and 2018 were used in the study. Two hybrid prediction models were produced for 2040 by integrating multi-layer perception (MLP) and support vector machine (SVM) algorithms with Markov Chains (MC) method. A base model was prepared with the intersection of these two to take protection precautions in the short and medium term. According to the results, it was determined by MLP that private land ownership significantly affects LULC. In addition, the pressure of urbanization on forests and agricultural lands, most of which are publicly owned, increases, and urban sprawl causes informal settlement.
Read full abstract