Monitoring and Predicting of Land Use and Cover Change for the period 2000-2030 Using Remote Sensing Data and Cellular Automata Approach

Understanding and characterize land use and land cover changes are crucial for informed decision-making in various management disciplines, including forestry, agriculture, industrial development, urban planning, rural and urban administration, and natural resource management. In this study, the land use and land cover (LULC) changes in İzmit province and its adjacent areas, undergoing rapid industrialization, were analyzed for the periods 2000-2010 and 2020 using Remote Sensing (RS) and Artificial Neural Network (ANN) methodologies. Additionally, a LULC projection for the year 2030 was generated and mapped. Within the scope of this study, land use changes across four categories (forest, water, agricultural, and built-up areas) were simulated utilizing elevation and slope variables derived from satellite imagery. Landsat 5 Thematic Mapper, Landsat 7 Enhanced Thematic Mapper Plus, and Landsat 8 Operational Land Imager satellite imagery were employed as data sources for the simulation. As a result of classified images Kappa values were calculated as 91% for 2000, 87% for 2010 and 94% for 2020. The validation value of the 2030 simulation was determined as 89.2%. This study project that, forest areas will decrease by 0.41%, agricultural areas by 4.38%, and water areas by 0.04%, while built-up areas in the industrial city of İzmit are expected to increase by 37.06% from 2020 to 2030. It is projected that forest and aquatic ecosystems are experiencing gradual spatiotemporal decline, whereas agricultural lands are undergoing a more rapid rate of reduction, a trend anticipated to persist.