The impact of land use and land cover on groundwater fluctuations using remote sensing and geographical information system: Representative case study in Afghanistan

As urbanization continues to accelerate, managing groundwater resources sustainably has become a momentous challenge for many cities, including Kabul. This study seeks to investigate the profound impact of land use and land cover (LULC) on groundwater fluctuations in the capital of Afghanistan. The present research relied on remote-sensed images and groundwater table data analyzed using an advanced geographic information system (GIS) environment. To develop LULC maps, two distinct models of Landsat image classification (supervised and unsupervised) were employed. The research practiced three time periods (2000, 2013, and 2020) and four crucial themes (bare land, built-up area, vegetation, and water bodies) to develop LULC maps. Moreover, the current study used two time intervals (2016 and 2020) for groundwater level maps within the region. The results indicate the superiority of the supervised Landsat image classification model by putting to use the support vector machine (SVM) technique. This approach yielded noticeably higher accuracies, with outcomes of 94.23%, 90.09%, and 88.18% for 2000, 2013, and 2020, respectively. The unsupervised Landsat image classification model, conversely, revealed results of 89%, 82.5%, and 84.26% for the same periods. To validate the accuracy of LULC maps and interpolation for groundwater table maps, taking advantage of the confusion matrix and cross-validation techniques, respectively. Findings clearly support the significant impact of LULC on groundwater fluctuations in Kabul city. The preeminent accuracy of the supervised Landsat image classification model provides powerful evidence for its effectiveness in precisely assessing the impacts of LULC on groundwater fluctuations.