Prediction of Land Use Change through the Cellular Automata-Markov Model: A Case Study of the Upper Sangkae River Basin in Cambodia

Abstract

The expansion of agricultural land and the diminishment of forest cover in Battambang Province (Cambodia) has been recently reported. At the same time, while forest cover has decreased, the amount of water resources in the river basin has been variable. The aim of the current study was to forecast land use change in the upper Sangkae River basin of Battambang Province by 2030. For this purpose, remote sensing and geographic information systems (GIS) methods were used to analyze satellite data from 2014 to 2018, using the generated maps as data input in the cellular automata (CA)-Markov model. We also integrated the CA-Markov model and GIS spatial analyst tools to assess what will most likely occur in the presence of policy intervention from land use development planning by 2030. Additionally, the model simulated actual and predicted land use in 2022 for accuracy assessment, using the Kappa Index of Agreement for confirmation. Based on the findings, the modeled scenario predicted the increase in built-up land and the decrease in the natural forest cover by 2030 in the absence of a land use policy. Additionally, the findings suggested that in the absence of a land use policy, forest cover will suffer from continued deforestation until forest loss reaches the protected area boundary. Conversely, in the presence of a land use policy, the model shows an increase in forest cover by 2030, even though some areas would be allocated under economic land concessions for industrial agriculture. Moreover, non-forest cover, such as farmlands and paddy fields, is not expected to decline, whereas built-up land is forecast to dramatically increase, with or without policy intervention. This study sheds light on the use of practical evaluation tools for governmental land use policies and development planning.