International Journal on Advanced Science, Engineering and Information Technology, Vol. 12 (2022) No. 1, pages: 237-243, DOI:10.18517/ijaseit.12.1.12384

Prediction of Bandung District Land Use Change Using Markov Chain Modeling

Abraham Suriadikusumah, Asep Mulyono, Muhammad Hilda Rizki Maulana

Abstract

Land conversion has occurred and converted into non-agricultural purposes in Bandung district, Indonesia. Consequently, the availability of agricultural land has decreased. The tendency of accessibility to infrastructure, natural resources, and population growth contribute to land-use change. This study aims to identify the relationship between distance accessibility factors such as distance to road, river, city center, slope and population density, and the land conversion in Bandung district. In this study, three multitemporal land use data (2007, 2012, and 2017) were used to determine transition probability matrix land-use change and predict land use in 2027. ArcGIS and Idrisi Selva software were used to simulate the land-use change in 2007–2017 and land use projection in 2027 using the Markov Chain model. The study resulted in Bandung district during 2007 to 2017 there was a decline in the areas of forest (0.03%), garden/plantation (7.02%), dryland (3.48%), paddy field (3.07%), and bushes (3.56%). Meanwhile, 34.57% of settlement and 11.44% of water bodies area increased in 2017. Land use was converted into settlement/built-up (18%), paddy fields (2.6%), water bodies (0.3%) and forest (0.1%). Distance to road, river, city center, and slope factor tended a negative correlation, while population density factor obtained positively correlated to the extent of land-use change. Land use prediction in 2027 resulting the most extensive land use was for paddy fields (22.58%), followed by forest (19.00%), garden/plantation (18.13%), settlement/built-up area (16.38%), dryland (12.32%), bushes (11.21%), and water bodies area (0.38%).

Keywords:

Land use; conversion; projection; markov chain; Bandung district.

Viewed: 1187 times (since abstract online)

cite this paper     download