Land use balance determination using satellite imagery and geographic information system: case study in South Sulawesi Province, Indonesia

This study was conducted to produce a GIS-based land use/land cover (LULC) balance map for a certain period as a reference for policymakers in planning their future regional development. This study also measures supervised classification accuracy based on remote sensing and geographic information system (GIS) integration with field conditions. In June 2005 satellite imagery 7 ETM+ was used as asset maps to assess land-use changes (LUC). Although in March 2019, the liability maps used satellite imagery 8 OLI/TIRS. Methods analysis consists of pre-image processing, image interpretation, random point, field check, and accuracy assessment. The image processing results were overlaid with an Indonesian topographic map to draw a LULC balance map. The findings indicate that in June 2005 and March 2019, each LULC had an assessment accuracy value of 82% and 86%, with a predicted assessment accuracy value of 18.05% and 20.50%, respectively. These findings are checked to determine the suitability performance of field-based imaging approaches based on the Cohen Kappa coefficient criteria of 0.45 and 0.48 for June 2005 and March 2019. Based on these results, the image processing precision and suitability were excellent since they are more than 80% and satisfy the Cohen Kappa performance criterion. Furthermore, geospatial data on the LULC balance map is essential as a guide for planners and decision-makers to plan their regional development. (C) 2021 Editorial office of Geodesy and Geodynamics. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd.