ANALISIS PREDIKSI PERUBAHAN TUTUPAN LAHAN TAHUN 2033 MENGGUNAKAN METODE CELLULAR AUTOMATA DAN LOGISTIC REGRESSION

Abstract

Land cover change is an important phenomenon that illustrates the interaction between human activities and land resources. Kemiling District, experienced an increase in land needs due to significant population growth from 2013 to 2023, which was 23,147 people or 36.64%, which prompted the need for research to predict land cover changes until 2033. Remote sensing technology and Geographic Information Systems (GIS) allow analysis of land cover changes by considering driving factors such as roads, settlements, soil types, slopes, and population density. The Cellular Automata (CA) and Logistic Regression (LR) methods were chosen because of their accuracy in spatial-temporal simulations. This study uses Landsat 8 imagery data for 2013, 2018, and 2023. The data is processed using a supervised classification method with the Support Vector Machine (SVM) algorithm. The results of the study showed significant changes in land cover, especially an increase in residential areas by 47.24% in 2023. Predictions for 2033 indicate that built-up land will continue to increase, while agricultural and open land will decrease. The CA and LR methods proved effective with kappa values in the good category. This study provides knowledge for decision-making on spatial planning and sustainable development in Kemiling District.