Tabassum et al. (2026) SWAT-based analysis of LULC change impacts on hydrology of the Subarnarekha River Basin, India

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2026
• Date: 2026-09-25
• Authors: Farhin Tabassum, Akhouri Pramod Krishna, Virendra Singh Rathore, C. Jeganathan
• DOI: 10.1016/j.ejrh.2026.104012

Research Groups

• Department of Remote Sensing and Geoinformatics, Birla Institute of Technology, Mesra, Ranchi, Jharkhand, India

Short Summary

This study investigated the impacts of historical (2012-2022) and future (2030-2050) land use and land cover (LULC) changes on the hydrology of the Subarnarekha River Basin (SRB), India. It found that increasing built-up, barren, and agricultural land, at the expense of forests and water bodies, leads to increased surface runoff and water yield, while reducing groundwater flow, lateral flow, and evapotranspiration, with these trends projected to intensify.

Objective

• To assess the impacts of historical and future land use and land cover (LULC) dynamics on the hydrology of the Subarnarekha River Basin (SRB) using an integrated remote sensing, machine-learning, and hydrological modeling framework.

Study Configuration

• Spatial Scale: Subarnarekha River Basin (approximately 21,000 km²), including analysis at sub-basin scale (25 sub-basins).
• Temporal Scale: Historical LULC analysis for 2012, 2017, and 2022. Future LULC projections for 2030, 2040, and 2050. Hydrological simulations and calibration/validation for 1991-2020 (streamflow) and 1996-2022 (groundwater levels).

Methodology and Data

• Models used:
  • LULC Classification: Support Vector Machine (SVM) classifier with Object-Based Image Analysis (OBIA) (Mean Shift algorithm).
  • Future LULC Prediction: Multi-Layer Perceptron-Markov Chain (MLP-MC) model (Land Change Modeler (LCM) module of TerrSet liberaGIS software).
  • Hydrological Modeling: Soil and Water Assessment Tool (SWAT) model (ArcSWAT 2012 interface).
  • SWAT Calibration & Uncertainty Analysis: SWAT Calibration and Uncertainty Procedures (SWAT-CUP) with Sequential Uncertainty Fitting version 2 (SUFI-2) algorithm.
  • Groundwater Recharge Estimation: Water Table Fluctuation (WTF) method.
  • Baseflow Separation: Eckhardt digital filter method.
• Data sources:
  • LULC Imagery: Indian Remote Sensing (IRS) Resourcesat Linear Imaging Self Scanner (LISS-IV) multispectral imagery (5.8 m spatial resolution) for 2012, 2017, 2022 (National Remote Sensing Centre - NRSC).
  • Topographic Data: Shuttle Radar Topography Mission (SRTM) Digital Elevation Model (DEM) (30 m resolution).
  • Anthropogenic Influence: Night-time light intensity data (DMSP-OLS and VIIRS, 30 arc-seconds spatial resolution), Population density data (WorldPop, 100 m spatial resolution), Road network data (OpenStreetMap).
  • Geological/Hydrogeological Data: Geological and lineament maps (National Geoscience Data Repository), Major principal aquifer map (Central Ground Water Board - CGWB via National Water Data Portal).
  • Soil Data: State Agricultural Management and Extension Training Institute (SAMETI), Jharkhand, and Indian Council of Agricultural Research (ICAR) - National Bureau of Soil Survey and Land Use Planning (ICAR-NBSS & LUP) (1:50,000 scale).
  • Meteorological Data: Daily gridded precipitation (Indian Meteorological Department - IMD, 0.25° × 0.25° spatial resolution), Minimum/maximum temperature, solar radiation, wind speed (ERA5-Land Daily Aggregated dataset - ECMWF), Relative humidity (Agrometeorological ERA5 dataset). All meteorological data cover 1991–2024.
  • Observed Streamflow: Daily data from Muri, Adityapur, Jamshedpur, and Ghatshila gauging stations for 1991–2020 (Central Water Commission - CWC).
  • Groundwater Levels: Continuous records from 39 monitoring wells for 1996–2022 (CGWB via India Water Resources Information System - WRIS).

Main Results

• LULC Change (2012-2022): Built-up land increased by 79.39%, barren land by 14.81%, and agricultural land by 0.89%. Forest cover declined by 5.54% and water bodies by 17.45%.
• Projected LULC Change (2022-2050): Built-up land is projected to increase by 70.81%, barren land by 32.09%, and agricultural land by 4.88%. Forest cover is expected to decline by 17.76% and water bodies by 21.51%.
• Hydrological Impacts (Basin Scale):
  • Historical (2012-2022): Annual surface runoff (SURQ) increased by 2.56%, and water yield (WYLD) by 1.07%. Lateral flow (LATQ) decreased by 5.87%, groundwater flow (GWQ) by 1.29%, and evapotranspiration (ET) by 0.77%.
  • Projected (2022-2050): SURQ is projected to increase by 6.07%, and WYLD by 2.62%. GWQ is expected to decline by 2.73%, LATQ by 4.49%, and ET by 1.77%.
• Hydrological Impacts (Sub-basin Scale): Sub-basins with intensive urbanization (e.g., SB1, SB10, SB11) showed the largest increases in surface runoff (up to 33.79% by 2050) and water yield (up to 9.45% by 2050). Lateral flow and groundwater flow generally declined across most sub-basins, with reductions up to 18.94% and 13.20% respectively by 2050 in some areas. Evapotranspiration also declined, with up to 4.79% reduction by 2050.
• Groundwater Recharge and Baseflow:
  • WTF-estimated groundwater recharge averaged 128.31 mm/year (10.5% of rainfall), comparable to SWAT-derived recharge of 124.79 mm/year (10.1% of rainfall), with a moderate R² of 0.58.
  • Eckhardt-derived baseflow (mean 55.90 m³/s) showed good agreement with SWAT-simulated baseflow (mean 64.66 m³/s) at Ghatshila station, with an R² of 0.63 for monthly estimates.
• Model Performance: SWAT model calibration (1994-2012) and validation (2013-2020) showed strong performance at four gauging stations, with Nash-Sutcliffe Efficiency (NSE) values ranging from 0.72 to 0.90 and R² values from 0.77 to 0.92. Percentage bias (PBIAS) values were within acceptable limits (±20%).

Contributions

• This study provides a comprehensive assessment of the combined impacts of historical and future LULC changes on both surface and subsurface hydrological processes in the Subarnarekha River Basin, addressing a critical gap in previous research for this monsoon-dominated region.
• It utilizes high-resolution IRS LISS-IV satellite imagery (5.8 m) for LULC classification, enabling a finer representation of spatially heterogeneous land-cover changes compared to studies using moderate-resolution datasets.
• The integrated framework, combining remote sensing, machine learning (MLP-MC for LULC prediction), and hydrological modeling (SWAT), offers a robust and transferable approach for assessing land-use-induced hydrological changes in similar rapidly transforming river basins.
• The analysis goes beyond commonly assessed surface runoff and water yield, also evaluating evapotranspiration, lateral flow, groundwater contribution, groundwater recharge (using WTF method), and surface water-groundwater interaction (using baseflow separation), providing broader hydrological insights.
• By maintaining identical meteorological forcing across LULC scenarios, the study effectively isolates the hydrological responses specifically attributable to land-cover transformation, distinguishing them from climatic variability.
• The findings offer crucial support for informed decision-making, strategic watershed planning, and the implementation of appropriate management measures for sustainable water resources in the SRB under ongoing and future land-use change.

Funding

• CSIR SRF-Direct program (File No. 09/0554(18412)/2024-EMR-I) of the Council of Scientific and Industrial Research (CSIR), India.

Citation

@article{Tabassum2026SWATbased,
  author = {Tabassum, Farhin and Krishna, Akhouri Pramod and Rathore, Virendra Singh and Jeganathan, C.},
  title = {SWAT-based analysis of LULC change impacts on hydrology of the Subarnarekha River Basin, India},
  journal = {Journal of Hydrology Regional Studies},
  year = {2026},
  doi = {10.1016/j.ejrh.2026.104012},
  url = {https://doi.org/10.1016/j.ejrh.2026.104012}
}