Multi-temporal Image Analysis for Land Cover Classification and Change Detection of Kuching Division, Sarawak

Abstract

Kuching and its surrounding area are vital in Sarawak’s economic growth. Over the years, the transformation of its land cover (LC) has brought significant ecological, physical, and socioeconomic consequences. Updated and precise LC maps are essential for urban planning, sustainable development, and environmental forest degradation monitoring. This study mainly focuses on LC classification using the Support Vector Machine Classifier (SVM) algorithm. The change has been identified for five general classes, which are Urban Land, Barren Land, Forest Land, Agriculture Land, and Water Bodies, for 35 years using Landsat 5 TM image dated 26 June 1988 and Landsat 9 OLI image dated 16 April 2023 in ArcGIS 10.6.1 software. Change detection analysis indicated that from 1988 to 2023, the LC patterns changed significantly. The most substantial changes were urban land, which increased significantly from 6,200.71 ha in 1988 to 22,144.76 ha in 2023, which represents a net increase of 15,944.05 ha or 3.88 per cent change (%), followed by agriculture land, barren land, forest land, and water bodies categories. The results show good classification performance because the user’s accuracy for every class is above 85%. LC change, which displays the spatial expansion of the urban land in Kuching, indicates the development of Sarawak is in line with the aspiration to be a developed state by 2030 (Post COVID-19 Development Strategy 2030).