Natoo et al. (2025) Potential of multi-source Geospatial data in Accurately Estimating the Live Storage Capacity of Reservoir

Identification

• Journal: ISPRS annals of the photogrammetry, remote sensing and spatial information sciences
• Year: 2025
• Date: 2025-12-19
• Authors: Nilima Ghosh Natoo, Prasun Kumar Gupta, Bhaskar R. Nikam
• DOI: 10.5194/isprs-annals-x-5-w2-2025-451-2025

Research Groups

• Geoinformatics Department, Indian Institute of Remote Sensing (ISRO), Dehradun, India
• Centre for Space Science and Technology Education in Asia and the Pacific (IIRS campus, ISRO), Dehradun, India
• Indian Space Research Organisation Headquarters, Bengaluru, India

Short Summary

This study develops a novel, purely geospatial methodology to accurately estimate the live storage capacity (LSC) of gauged reservoirs, eliminating the need for field data by fusing multi-source satellite imagery for water surface area and interpolating altimetry data for water levels. The method, applied to the Ukai reservoir, achieved high accuracy (Root Mean Square Error of 91.0 m³) compared to observation-based estimates.

Objective

• To develop and validate a novel geospatial-based methodology for remotely estimating the live storage capacity (LSC) of gauged reservoirs, relying solely on multi-source satellite data and advanced computational techniques, thereby eliminating the need for traditional field gauge observations.

Study Configuration

• Spatial Scale: Regional, focusing on the Ukai (Vallabh Sagar) reservoir in Gujarat, India (73.7073 °E, 21.3235 °N).
• Temporal Scale: At least one water-year, with specific data analysis for water surface area (WSA) from 2021 to 2023 and water elevation from 2019 to 2020. Daily water level and area estimates were derived from coarser satellite data.

Methodology and Data

• Models used:
  • Trapezoidal rule (integral relationship) for volume computation, specifically the Prismoidal formula for higher accuracy.
  • Cubic spline interpolation for temporal gap-filling and enhancing the resolution of altimetry water level data.
  • Modified Normalized Difference Water Index (MNDWI) for water body delineation.
  • Automatic Otsu thresholding method for separating water pixels.
• Data sources:
  • Satellite Imagery:
    • Optical: Sentinel-2 (10 m, 20 m spatial resolution, 10-day revisit for S-2A), Landsat-8 (30 m spatial resolution, 16-day revisit).
    • Synthetic Aperture Radar (SAR): Sentinel-1 (10 m spatial resolution, 12-day revisit for S-1A, C-band, VV polarization).
    • Processed and analyzed using Google Earth Engine (GEE) platform.
  • Altimetry Water Level: Sentinel-3A (27-day revisit) data downloaded from the Database for Hydrological Time Series of Inland Waters (DAHITI).
  • Field Gauge Data (for validation): Daily water level data from the India Water Resources Information System (WRIS portal).
  • Baseline Elevation-Capacity Data: Hydrographic Survey of Ukai (2003) from Narmada Water Resources, Water Supply and Kalpasar Department, Gandhinagar, Gujarat.
  • Computational Environment: Python platform with libraries including Pandas, SciPy, and Matplotlib.

Main Results

• A novel geospatial-based methodology was successfully developed to estimate the live storage capacity (LSC) of gauged reservoirs without relying on field gauge data.
• For the Ukai reservoir, the geospatial-based LSC estimates demonstrated excellent agreement with observation-based estimates, achieving an accuracy level up to the 5th decimal digit.
  • Mean Error: 8.79 m³
  • Standard Deviation of Error: 91.1 m³
  • Root Mean Square Error (RMSE): 91.0 m³
  • Bias: 8.79 m³
• The fusion of optical (Sentinel-2, Landsat-8) and SAR (Sentinel-1) satellite imageries significantly reduced revisit gaps and missing data points, especially during monsoon periods, providing a temporally richer dataset for water surface area delineation.
• Advanced cubic spline interpolation effectively pre-processed coarse altimetry water level data (e.g., Sentinel-3A's 27-day revisit) to generate a regular, temporally richer daily time series, which was crucial for accurate LSC estimation.
• The Prismoidal formula (Method-1) for volume calculation was found to outperform the Average-End-Area method (Method-2) for the Ukai reservoir case study.

Contributions

• Development of a purely geospatial, field data-free methodology for accurate live storage capacity estimation in gauged reservoirs, addressing the challenge of concurrent availability of water level observations and updated Area-Elevation-Capacity curves.
• Demonstration of the effectiveness of multi-source satellite data fusion (optical and SAR) to overcome temporal gaps and cloud cover limitations in water surface area delineation.
• Validation of advanced interpolation techniques (cubic spline) for transforming coarse altimetry data into high-resolution water level time series, crucial for precise LSC calculations.
• Quantification of uncertainty and error propagation in LSC estimation, highlighting the importance of temporally rich geospatial inputs and robust computational platforms (GEE, Python) for achieving high accuracy.
• The established methodology provides a foundation for future applications, including the potential for refining LSC computation for natural (ungauged) glacial lakes and integrating machine learning for long-term monitoring.

Funding

The authors acknowledge the following for data and resources: * Water Resources Authority of Ukai Reservoir (Narmada, Water Resources, Water Supply and Kalpasar Department, Gandhinagar, Gujarat) for providing elevation-capacity data. * India Water Resources Information System (WRIS) portal for water elevation gauge records. * GREALM and DAHITI portals for altimetry data sources. * Spatial Thoughts for online/offline technical and programming tutorials.

Citation

@article{Natoo2025Potential,
  author = {Natoo, Nilima Ghosh and Gupta, Prasun Kumar and Nikam, Bhaskar R.},
  title = {Potential of multi-source Geospatial data in Accurately Estimating the Live Storage Capacity of Reservoir},
  journal = {ISPRS annals of the photogrammetry, remote sensing and spatial information sciences},
  year = {2025},
  doi = {10.5194/isprs-annals-x-5-w2-2025-451-2025},
  url = {https://doi.org/10.5194/isprs-annals-x-5-w2-2025-451-2025}
}