Land-use transitions and hydrological futures in a monsoon-fed basin: insights from the Krishna River Basin, India

Graphical abstract illustrating the integrated framework of the study. Landsat images are processed in Google Earth Engine and classified using a CART algorithm to produce LULC maps for 1985, 1995, and 2005. LULC prediction is then performed using driver variables including distance to roads, distance to rivers, slope, and elevation through a multilayer perceptron-based approach to generate the 2050 LULC map. These historical and projected LULC scenarios are then coupled with hydrological modelling in SWAT using climate, terrain, and soil information. The final outputs show changes in surface runoff, percolation, and evapotranspiration under the different LULC scenarios. Land use and land cover (LULC) change is increasingly altering hydrology in large monsoon-fed basins, yet its basin-scale effects remain insufficiently quantified. This study develops an integrated LULC–hydrology framework to assess how historical and projected land transitions modify the water balance of the Krishna River Basin, India. Multi-temporal Landsat imagery (1985, 1995, 2005) was classified into five LULC classes using a CART classifier, with overall accuracies of 78–89%. These maps, together with spatial drivers, were used in a Multi-Layer Perceptron–Markov framework to project LULC for 2050. A SWAT model was calibrated and validated at the basin outlet (NSE ≈ 0.8; PBIAS within ±10%), and then applied under identical climatic forcing while varying only the LULC layer across 1985, 1995, 2005, and 2050. Results show a ∼21% decline in vegetation and a nearly fivefold increase in urban area by 2050, while agriculture remains broadly stable. Under these changes, mean annual runoff increases modestly (∼0.6%), percolation declines more clearly (∼10.4%), and evapotranspiration changes only slightly. The findings show that LULC change in the Krishna Basin affects hydrology more strongly through weakening recharge pathways than through large increases in basin-wide runoff, highlighting the need for hydrologically informed basin planning.