Vishwakarma et al. (2025) Mapping of Vegetation Responses and Impacts on Groundwater in a Drought Afflicted Data-Scarce River Basin using Remotely Sensed Information

Identification

• Journal: Journal of the Indian Society of Remote Sensing
• Year: 2025
• Date: 2025-11-24
• Authors: Amit Vishwakarma, Ajanta Goswami
• DOI: 10.1007/s12524-025-02358-x

Research Groups

• Department of Earth Sciences, IIT Roorkee, Uttarakhand, India
• Centre of Excellence in Disaster Mitigation and Management, IIT Roorkee, Uttarakhand, India

Short Summary

This study mapped vegetation responses and groundwater impacts in the drought-afflicted, data-scarce Bhadar River basin in India using an integrated approach of remote sensing and hydrological modeling. It successfully simulated drought propagation patterns, identified five major drought periods between 1996 and 2022, and quantified their varying effects on vegetation vigor and groundwater recharge.

Objective

• To evaluate the performance of a SWAT-based hydrologic model in a drought-afflicted river basin in India.
• To record and characterize spatial drought responses on vegetation vigor and groundwater health during fluctuating drier and wetter periods.
• To determine if remotely sensed information provides an acceptable mechanism for understanding surplus and deficit periods of groundwater recharge over past years in the region.
• To outline the impacts of drought events in response to scanty rainfalls and arising water scarcity.

Study Configuration

• Spatial Scale: Bhadar River basin, Gujarat, India (21°31′ N to 22°12′ N latitudes and 69°42′ E to 71°19′ E longitudes). Data resolutions varied: Digital Elevation Model (DEM) at 30 meters, Landsat 7/8 products at 30 meters, MODIS products at 250 meters, GLDAS at 0.25° x 0.25°, and GRACE data (coarser, rescaled for the basin).
• Temporal Scale: Overall study period from 1996 to 2022. SWAT model simulations covered 1993–2021 (29 years, including 3 years warm-up). Vegetation indices (NDVI, NDWI, EVI) were analyzed for 1999–2022. GRACE data covered 2002–2016, and GLDAS data covered 2003–2022.

Methodology and Data

• Models used:
  • Soil and Water Assessment Tool (SWAT) hydrological model.
  • SWAT-Calibration and Uncertainty Programming (CUP) with the SUFI-2 algorithm for sensitivity, calibration, and validation.
• Data sources:
  • Satellite/Remote Sensing: Landsat-7, Landsat-8 (Collection-1 Tier-1 8-Day top-of-atmosphere reflectance Composites for NDVI, NDWI, EVI), MODIS products (MCD43A4, MOD09Q1.061 for NDVI, EVI), NASA’s GRACE Tellus Monthly Mass Grids (for Terrestrial Water Storage (TWS) and Groundwater Storage (GWS) anomalies), GLDAS (Global Land Data Assimilation System) assimilated data products (Version 2, for TWS, GWS, soil moisture, fluxes), SRTM (Shuttle Radar Topography Mission) Digital Elevation Model (DEM).
  • Meteorological/Hydrometeorological: IMD (India Meteorological Department) gridded datasets (83 years of climatic data for SWAT parametrization), NASA POWER data (precipitation, temperature, relative humidity, solar radiation, winds for 24 stations, 1981–2021).
  • Ancillary/Thematic: Land Use Land Cover (LULC) data, FAO-classified soil layer, slope layers, river discharge data from Kamadhiya and Ganod stations.
  • Tools: Google Earth Engine (GEE) for data processing and ArcGIS-V10.3 for spatial analysis and visualization.

Main Results

• The SWAT model demonstrated good performance with R² ≥ 0.59, NSE ≥ 0.67, and PBIAS ranging from 8.63 to -14.32.
• Five distinct drought periods were identified between 1999 and 2022: 2000–2001, 2002–2003, 2009–2010, 2012–2013, and 2018–2019, with 2000–2003 being the longest vegetative drought.
• Years 2000 and 2010 were identified as the driest based on selected indicators.
• A long-term trend (1996-2022) showed an increase in the greenness index, correlating with an increasing trend in rainfall.
• GRACE data indicated 2012 had the lowest vegetative growth due to deficit groundwater recharge.
• Four major periods of groundwater recharge decline were observed between 2000 and 2022, with 2000–2006 being the most prolonged.
• Groundwater droughts often preceded vegetative droughts, suggesting a temporal lag in drought propagation.
• Rabi (winter) season droughts were found to be more intense and critical than Kharif (monsoon) season droughts, primarily due to the winter crops' reliance on recharged groundwater.
• Spatially, drought severity increased from the southern to the northern parts of the basin, with areas further from the main river experiencing more pronounced impacts.
• Linear regression models showed acceptable correlations: SWAT-based water yield versus satellite-based TWS (R² = 0.62), GLDAS root zone soil moisture versus modelled soil water content (R² = 0.72), and deep percolation versus remotely sensed GWS (R² = 0.495).
• The overall drought frequency for the Kharif season was 1 in every 4 years, and for the Rabi season, it was 1 in every 3.42 years.
• Approximately 75% of the basin's surface flows contribute to groundwater for natural aquifer replenishment.

Contributions

• Developed and validated an efficient hydrological model (SWAT) for a data-scarce river basin, demonstrating its applicability for drought impact assessment.
• Provided a comprehensive, integrated assessment of drought propagation by combining remote sensing-derived vegetation indices (NDVI, NDWI, EVI) with hydrological model outputs and gravity-based water storage data (GRACE/GLDAS).
• Quantified the spatio-temporal dynamics of vegetation and groundwater droughts, including the identification of specific drought periods and their severity.
• Highlighted the critical role of groundwater in sustaining Rabi crops and the varying seasonal responses to drought.
• Established key relationships between different drought variables through linear regression, enhancing the understanding of drought dynamics in data-scarce regions.
• Offered a framework for early conceptualization and reliable forecasting of droughts, providing valuable insights for water resource management and policy-making in similar regions.

Funding

This research was not funded.

Citation

@article{Vishwakarma2025Mapping,
  author = {Vishwakarma, Amit and Goswami, Ajanta},
  title = {Mapping of Vegetation Responses and Impacts on Groundwater in a Drought Afflicted Data-Scarce River Basin using Remotely Sensed Information},
  journal = {Journal of the Indian Society of Remote Sensing},
  year = {2025},
  doi = {10.1007/s12524-025-02358-x},
  url = {https://doi.org/10.1007/s12524-025-02358-x}
}