Roy et al. (2026) Assessment of Remote Sensing Precipitation Products for Hydrological Analysis in an Ungauged Watershed

Identification

• Journal: Water Resources Management
• Year: 2026
• Date: 2026-01-01
• Authors: Dibyandu Roy, J. Indu
• DOI: 10.1007/s11269-025-04459-1

Research Groups

• Department of Civil Engineering, Indian Institute of Technology, Bombay, India

Short Summary

This study proposes a novel Monsoon Index (MI) to quantify the temporal distribution and magnitude of monsoon rainfall for selecting suitable Remote Sensing Precipitation Products (RSPPs) in ungauged, high-altitude regions. It found that IMERG most accurately represents observed monsoon characteristics and significantly improves daily runoff estimation in the Ranikhola watershed compared to other RSPPs.

Objective

• To propose a novel monsoon-based index (Monsoon Index, MI) capable of objectively selecting the most suitable Remote Sensing Precipitation Product (RSPP) for hydrological analysis in Indian high-altitude regions.
• To evaluate the efficacy of different RSPPs for daily runoff estimation using an optimized lumped hydrological model.

Study Configuration

• Spatial Scale: Ranikhola watershed, Gangtok District, Sikkim, India (27°14′15″ N – 28°23′49″ N, 88°29′25″ E – 88°43′19″ E), covering an area of 254.46 km².
• Temporal Scale: 2000–2020 for Monsoon Index (MI) calculation and RSPP evaluation; 2015–2020 for hydrological model calibration and validation.

Methodology and Data

• Models used:
  • Hydrological Simulation Model (HYSIM) - a lumped conceptual hydrological model.
• Data sources:
  • Remote Sensing Precipitation Products (RSPPs):
    • Integrated Multi-satellite Retrievals for Global Precipitation Measurement (GPM) (IMERG) final run (0.1° × 0.1° spatial resolution, half-hourly temporal resolution).
    • Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN)-Climate Data Record (CDR) (0.25° × 0.25° spatial resolution, half-hourly temporal resolution).
    • Multi-Source Weighted-Ensemble Precipitation (MSWEP) (0.25° × 0.25° spatial resolution, 3-hourly temporal resolution).
    • Global Precipitation Climatology Project (GPCP) 1 Degree Daily (1dd) (2.5° × 2.5° spatial resolution, daily temporal resolution).
  • Observed Data:
    • Daily gridded rainfall (0.25° × 0.25°) from India Meteorological Department (IMD) for 2000–2020.
    • Daily gridded maximum and minimum temperatures (0.50° × 0.50°) from IMD for 2000–2020.
    • Daily observed discharge data from Central Water Commission (CWC) gauging station at Singtam for 2015–2020.
  • Ancillary Data: High-resolution Digital Elevation Model (DEM), FAO soil type data.

Main Results

• A novel Monsoon Index (MI) was proposed, integrating monsoon rainfall magnitude and its dynamically derived temporal distribution.
• The observed MI for the Ranikhola watershed (2000–2020) had a median value of 0.653, indicating a highly monsoon-dominated region.
• Among the assessed RSPPs, IMERG exhibited the strongest agreement with the observed MI, with a median MI of 0.633. It also showed the lowest average deviation (0.021), lowest error statistics (RMSE 0.048, MAPE 5.769), and highest R² (0.617) when compared to the observed MI.
• The HYSIM hydrological model, calibrated with observed data (2015–2018), performed strongly (calibration: R² 0.829, PBIAS -0.009, NSE 0.787, RSR 0.461; validation: R² 0.889, PBIAS -0.019, NSE 0.845, RSR 0.383).
• When used to simulate runoff, IMERG-driven simulations showed the best performance, with the highest R² (0.741), NSE (0.654), and lowest PBIAS (0.000) and RSR (0.588) compared to observed runoff.
• Other RSPPs performed less effectively: GPCP and PERSIANN underestimated runoff (median 6.57 m³/s and 9.16 m³/s, respectively), while MSWEP significantly overestimated it (median 23 m³/s). Observed runoff had a median of 15.2 m³/s.

Contributions

• Introduced a novel Monsoon Index (MI) that uniquely integrates both the magnitude of monsoon rainfall and its dynamically derived temporal distribution, providing a region-specific and physically meaningful criterion for evaluating rainfall products.
• Demonstrated an effective and transferable framework for identifying the most suitable RSPP for hydrological analysis in ungauged, topographically complex, and monsoon-dominated high-altitude watersheds.
• Confirmed that MI-selected RSPPs, specifically IMERG for the Ranikhola watershed, can substantially improve daily runoff estimation, offering a reliable and practical alternative for hydrological assessments in data-scarce regions.

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Citation

@article{Roy2026Assessment,
  author = {Roy, Dibyandu and Indu, J.},
  title = {Assessment of Remote Sensing Precipitation Products for Hydrological Analysis in an Ungauged Watershed},
  journal = {Water Resources Management},
  year = {2026},
  doi = {10.1007/s11269-025-04459-1},
  url = {https://doi.org/10.1007/s11269-025-04459-1}
}