Devkota et al. (2025) Hydrological performance of ERA5 land in the data-scarce Himalayan region of the Langtang catchment

Identification

• Journal: Theoretical and Applied Climatology
• Year: 2025
• Date: 2025-12-13
• Authors: Ashish Devkota, Pawan Kumar Bhattarai, Vishnu Prasad Pandey, Subbarao Pichuka, Ananta Man Singh Pradhan, Susen Shrestha
• DOI: 10.1007/s00704-025-05933-x

Research Groups

• Department of Civil Engineering, Pulchowk Campus, Institute of Engineering, Tribhuvan University, Lalitpur, Nepal
• Department of Civil Engineering, Indian Institute of Technology Madras (IITM), Chennai, India
• Water Resources Research and Development Centre, Government of Nepal, Kathmandu, Nepal
• Center for Climate Change and Transformation (CCT), Eurac Research, Bolzano, Italy

Short Summary

This study evaluates the hydrological performance and bias correction sensitivity of ERA5-Land reanalysis data for runoff simulation in the data-scarce Langtang catchment in the Himalayas. It finds that correcting only ERA5-Land temperature data significantly improves runoff simulations (Nash-Sutcliffe Efficiency = 0.73, Root Mean Square Error = 2.89 m³/s), indicating temperature is more critical for hydrological modeling accuracy in snow-fed Himalayan regions.

Objective

• To evaluate the hydrological performance and bias correction sensitivity of ERA5-Land (ERA5L) reanalysis data for runoff simulation in the data-scarce, snow- and glacier-dominated Langtang catchment in the Himalayas, using the Integrated Catchment Hydrological Model (ICHYMOD).
• To assess the impact of bias-correcting ERA5L precipitation and temperature data on runoff simulation accuracy and examine the sensitivity of hydrological modeling to these input data adjustments.

Study Configuration

• Spatial Scale: Langtang catchment, central Nepalese Himalaya. Area: 352.6 km². Elevation range: 3651 m to 7223 m above sea level. ERA5L data used at 0.1° spatial resolution, covering the catchment with 16 grid cells.
• Temporal Scale:
  • Observed hydro-meteorological data: 2008–2013 (hourly).
  • ERA5L reanalysis data: 2001–2021 (hourly temperature, 24-hour accumulated precipitation).
  • Model calibration period: 2009–2010.
  • Model validation period: 2011–2013.
  • MODIS snow cover data: 2002–2022 (8-day version 6.1 Terra–Aqua cloud-free snow cover).

Methodology and Data

• Models used:
  • Integrated Catchment Hydrological Model (ICHYMOD): A lumped hydrological model for converting surplus precipitation and snowmelt into runoff.
  • TOPMELT (Topographic melt): An enhanced temperature index snowpack model integrated with ICHYMOD, accounting for snow and glacier dynamics across elevation and radiation classes.
• Data sources:
  • Reanalysis: ERA5-Land (ERA5L) from the Copernicus Climate Data Store (CDS) ECMWF. Variables: 2-meter temperature, total precipitation.
  • Observation: Daily runoff data from Langtang hydrological station and daily meteorological data (precipitation, temperature) from Kyanging AWS, collected from the Department of Hydrology and Meteorology (DHM) and ICIMOD.
  • Satellite: Moderate Resolution Imaging Spectroradiometer (MODIS) 8-day version 6.1 Terra–Aqua cloud-free snow cover (ICIMOD) for fractional snow cover area (fSCA) validation.
  • Ancillary: Randolph Glacier Inventory (RGI) 6.0 for high-mountain Asia (ICIMOD), Digital Elevation Model (DEM).
  • Bias Correction: Quantile-Quantile Mapping (QM) for precipitation and linear correction for temperature, using observed station data as reference.

Main Results

• ICHYMOD Performance (with observed data):
  • Calibration (2009-2010): Nash-Sutcliffe Efficiency (NSE) = 0.72, Root Mean Square Error (RMSE) = 2.79 m³/s.
  • Validation (2011-2013): NSE = 0.84, RMSE = 2.50 m³/s.
• Water Balance (from model): Glacier melt contributes 13.2% and snowmelt contributes 29.4% to the total runoff, summing to 42.3%.
• Snow Cover Area (SCA) Validation: Comparison of TOPMELT-based fractional snow-covered area (fSCA) with MODIS data yielded NSE = 0.73 and RMSE = 0.05.
• ERA5L Biases:
  • Precipitation: ERA5L overestimates observed precipitation by 139% (average annual ERA5L: 1812 mm vs. observed: 755.8 mm), particularly for lower values.
  • Temperature: ERA5L exhibits a cold bias (average annual ERA5L: -4.7 °C vs. observed: 4.1 °C), attributed to elevation mismatch.
• Impact of Bias Correction on Runoff Simulation (using ERA5L inputs):
  • Uncorrected Temperature & Precipitation: NSE = -0.45, RMSE = 7.72 m³/s (unsatisfactory).
  • Uncorrected Temperature, Corrected Precipitation: NSE = -0.75, RMSE = 8.46 m³/s (unsatisfactory).
  • Corrected Temperature, Uncorrected Precipitation: NSE = 0.88, RMSE = 2.22 m³/s (strong performance). Glacier contribution 12.3%, snow contribution 27.2%.
  • Corrected Temperature & Precipitation: NSE = 0.88, RMSE = 2.24 m³/s (strong performance). Glacier melt 12.9%, snowmelt 28.1%.
• Key Finding: Correcting only ERA5L temperature data (NSE = 0.73, RMSE = 2.89 m³/s) resulted in better runoff simulation performance compared to correcting both temperature and precipitation (NSE = 0.70, RMSE = 3.02 m³/s), highlighting the higher sensitivity of hydrological modeling to temperature correction in snow-fed Himalayan regions.

Contributions

• First application and validation of the integrated TOPMELT-ICHYMOD glacio-hydrological modeling framework in a data-scarce Nepalese Himalayan catchment (Langtang), demonstrating its transferability beyond its original Italian Alpine test cases.
• Provides a comprehensive evaluation of ERA5-Land reanalysis data for runoff simulation in a high-altitude, snow- and glacier-dominated region, addressing critical data scarcity challenges.
• Identifies the dominant role of temperature correction in improving ERA5L-driven hydrological modeling accuracy in snow-fed Himalayan environments, offering crucial insights for bias correction strategies.
• Develops a transferable framework that can enhance regional hydrological modeling, water resource management, infrastructure planning, and climate adaptation strategies in similar poorly gauged mountain basins by effectively incorporating cryospheric contributions.

Funding

Not explicitly mentioned in the provided text.

Citation

@article{Devkota2025Hydrological,
  author = {Devkota, Ashish and Bhattarai, Pawan Kumar and Pandey, Vishnu Prasad and Pichuka, Subbarao and Pradhan, Ananta Man Singh and Shrestha, Susen},
  title = {Hydrological performance of ERA5 land in the data-scarce Himalayan region of the Langtang catchment},
  journal = {Theoretical and Applied Climatology},
  year = {2025},
  doi = {10.1007/s00704-025-05933-x},
  url = {https://doi.org/10.1007/s00704-025-05933-x}
}