Exploring the dynamics and future projections of land use land cover changes by exploiting geospatial techniques; A...

Global land cover change has caused significant environmental degradation and biodiversity loss. It affects ecosystem functions, livelihoods, and climate variation and has drawn substantial attention in recent decades. In the Kabul River Basin (KRB), there are limited studies on the historical Land Use/Land Cover (LULC) pattern, transition, intensity and future perspective. Therefore, this study aims to investigate long-term LULC changes and major drivers of LULC in the KRB over the past thirty years (1990-2020) and then to project the future LULC pattern for the years 2030, 2040 and 2050. Landsat Imageries of (1990-2020) were used as input data by utilizing the Random Forest Classifier algorithm (RF) in the Google Earth Engine (GEE) to classify the LULC. The LULC was then projected for the future, using the Cellular Automata Markov Chain Model (CA-MCM). The results demonstrated drastic LULC changes, controlled primarily by urbanization and agriculture expansion, which expanded from 467 Km² (0.7 %) to 2312 km² (3.4 %) and 6528 km² (9.6 %) to 10812 (15.9 %), between 1990 and 2020. In contrast, bare land decreased from 70606 km² (82.1 %) to 48212 km² (70.9 %) between 1990 and 2020. In addition, the study depicts that the expansion in built-up and vegetation areas in the KRB during the study period were at the utilization of bare land. Future LULC predictions indicated that between 2020 and 2050, bare land would trend downward from 48212 km² (70.9 %) to 46172 km² (67.9 %), while vegetation and built-up areas would trend upward from 2312 km² (3.4 %) to 3640 km² (5.3 %), 10812 km² (15.9 %) to 11622 km² (17.1 %), and water bodies and snowcover would slightly vary from 1.2 % to 0.9 % and 7.9 %-9.0 %. In addition, the results of LULC dynamics reveal a significant strong positive correlation between population and built, as well as population and vegetation. Conversely, there is a strong negative correlation between population and bare land. Our results provide precise insights on LULC patterns and trends in the KRB, which could be employed to design a sustainable framework for land use and ecosystem protection.