Spatial and decadal prediction of land use/land cover using multi-layer perceptron-neural network (MLP-NN) algorithm for a semi-arid region of Asir, Saudi Arabia

• Published in: Earth science informatics Vol. 14; no. 3; pp. 1547 - 1562
• Main Authors: Alqadhi, Saeed, Mallick, Javed, Balha, Akanksha, Bindajam, Ahmed, Singh, Chander Kumar, Hoa, Pham Viet
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.09.2021; Springer Nature B.V
• Subjects: Absorption coefficient; Absorptivity; Algorithms; Arid regions; Arid zones; Earth and Environmental Science; Earth Sciences; Earth System Sciences; Information Systems Applications (incl.Internet); Land cover; Land use; Land use management; Landsat; Landsat satellites; Mathematical models; Multilayer perceptrons; Natural resources; Neural networks; Ontology; Prediction models; Regional planning; Remote sensing; Research Article; Satellite imagery; Semi arid areas; Semiarid lands; Semiarid zones; Simulation and Modeling; Space Exploration and Astronautics; Space Sciences (including Extraterrestrial Physics; Urban sprawl; Saudi Arabia; Land change modelling; Asir; Artificial intelligence; Landsat
• ISSN: 1865-0473; 1865-0481
• DOI: 10.1007/s12145-021-00633-2

The present study uses Landsat satellite images of 1990, 2000 and 2018 to identify the land-use changes. Multilayer perceptron-neural network based land change modelling (LCM) has been applied to model future land-use/land cover (LULC). The prediction model has been validated using simulated and classified LULC maps of 2018 which resulted into an overall accuracy of 88%. The results indicate 389.27% increase in built-up area as the prominent land-use change during 1990–2018 and an increase of 56.25% in built-up area is forecasted during the year 2018–2040. Land absorption coefficient and land consumption rate indices, used to characterize urban expansion, indicate continued compact built-up structure during 1990–2018 due to population increase. The observations derived from this study would be useful as it will help the regional planners with forecasted land-use beforehand in planning the built-up and abundantly available natural resources in the area according to the increasing future demands.