García‐García et al. (2025) Intercomparison of Earth Observation products for hyper-resolution hydrological modelling over Europe

Identification

• Journal: Remote Sensing of Environment
• Year: 2025
• Date: 2025-11-20
• Authors: Almudena García‐García, Pietro Stradiotti, Federico Di Paolo, Paolo Filippucci, Milan Fischer, Matěj Orság, Luca Brocca, Jian Peng, Wouter Dorigo, Alexander Gruber, Bram Droppers, Niko Wanders, Arjen Haag, Albrecht Weerts, Ehsan Modiri, Oldřich Rakovec, Félix Francés, M. Dall’Amico, Martha C. Anderson, Christopher Hain, Luis Samaniego
• DOI: 10.1016/j.rse.2025.115131

Research Groups

• Helmholtz Centre for Environmental Research - UFZ, Department of Remote Sensing, Leipzig, Germany
• Institute for Earth System Science and Remote Sensing, Leipzig University, Leipzig, Germany
• Department of Geodesy and Geoinformation, TU Wien, Vienna, Austria
• Waterjade Srl, Pergine Valsugana (TN), Italy
• National Research Council of Italy, Research Institute for Geo-Hydrological Protection, Perugia, Italy
• Global Change Research Institute CAS, Brno, Czech Republic
• Helmholtz Centre for Environmental Research - UFZ, Department Computational Hydrosystems, Leipzig, Germany
• University of Potsdam, Institute of Environmental Science and Geography, Potsdam, Germany
• Department of Physical Geography, Utrecht University, Utrecht, The Netherlands
• Operational Water Management, Deltares, Delft, The Netherlands
• Hydrology and Environmental Hydraulics group, Wageningen University & Research, Wageningen, The Netherlands
• Faculty of Environmental Sciences, Czech University of Life Sciences Prague, Kamýcká 129, Praha – Suchdol, Czech Republic
• Research Institute of Water and Environmental Engineering (IIAMA), Universitat Politècnica de València, Valencia, Spain
• Hydrology and Remote Sensing Laboratory, USDA-ARS, Beltsville, United States
• Marshall Space Flight Center, Earth Science Branch, NASA, Huntsville, United States

Short Summary

This study comprehensively evaluated high-resolution Earth Observation (EO) products for precipitation, snow cover area, surface soil moisture, and evapotranspiration over Europe against observational references. It identified specific merged precipitation, MODIS/Sentinel-2 snow, and NSIDC SMAP soil moisture products as best performing for hyper-resolution hydrological modeling, while evapotranspiration products showed similar overall performance.

Objective

• To provide a thorough comparative assessment of high-resolution Earth Observation (EO)-based products for hydrological modeling over Europe, focusing on technical features (spatial/temporal resolution, availability, latency) required for integration into hyper-resolution (approximately 1 km²) hydrological and land surface models.

Study Configuration

• Spatial Scale: Europe (for precipitation, surface soil moisture, evapotranspiration) and the Alps mountain range (for snow cover area). EO product resolutions ranged from 20 meters to 25 kilometers, with a target for hyper-resolution modeling at approximately 1 square kilometer.
• Temporal Scale: Evaluation periods varied by variable and product, generally spanning from the early 2000s to 2022. Data resolutions were typically daily, with some products offering sub-daily or 8-day intervals.

Methodology and Data

• Models used: The study evaluated EO products for their suitability in hydrological models, rather than using hydrological models directly. However, various algorithms and frameworks were used to generate the EO products themselves, including:
  • SM2RAIN algorithm (for SM2RAIN-ASCAT precipitation).
  • Integrated Multi-satellitE Retrievals for GPM (IMERG) algorithm (for IMERG-LR precipitation).
  • Triple collocation technique (for merged precipitation products).
  • Radiance data from MODIS and Landsat, SAR data from Sentinel-1, multispectral images from Sentinel-2 (for snow cover).
  • Disaggregation techniques combining L-band SMAP, SMOS L3, ASCAT, Sentinel-1 SAR data with surface temperature and NDVI (for surface soil moisture).
  • Modeling frameworks combining EO retrievals and meteorological products (in-situ/reanalysis) with physical and/or machine learning methods (for evapotranspiration products like ALEXI, ETMonitor, GLEAM, HOLAPS).
• Data sources:
  • Earth Observation (EO) products evaluated:
    • Precipitation: CHIRPS, IMERG-LR, SM2RAIN-ASCAT, Merged IMERG-SM2A, Merged ERA5-IMERG-SM2A, ERA5 Land reanalysis, EMO-1 gridded meteorological dataset.
    • Snow Cover Area (SCA): MODIS SCA, Sentinel-2/Landsat-8 SCA, Sentinel-2 FSC, Sentinel GFSC, Sentinel-1 HS.
    • Surface Soil Moisture (SSM): NSIDC SMAP 1 km, BEC SMOS L4, CGLS SWI, CGLS SSM, UFZ-S1, ESACCISM 1km.
    • Evapotranspiration (ET): ALEXI, AVHRR, ETMonitor, GLEAM, HOLAPS, MODIS-Aqua, MODIS-Terra.
  • Observational references (benchmarks):
    • Precipitation: E-OBS v25.0e dataset (10 km resolution, in-situ based).
    • Snow Cover Area: In-situ measurements from over 300 meteorological stations in the Alps (Matiu et al., 2021).
    • Surface Soil Moisture: International Soil Moisture Network (ISMN) in-situ observations (from 492 stations across 20 networks).
    • Evapotranspiration: Eddy covariance measurements from the ICOS network (37 usable sites over Europe).

Main Results

• Precipitation: The two merged products at 1 km resolution (Merged IMERG-SM2A and Merged ERA5-IMERG-SM2A) are recommended for hyper-resolution hydrological modeling over Europe, with Merged ERA5-IMERG-SM2A showing RMSE values less than 4 mm and correlation coefficients greater than 0.5 over most areas. Pure remote sensing products generally performed worse than merged or reanalysis-integrated products.
• Snow Cover Area (SCA): MODIS (250 m resolution) and Sentinel-2/Landsat-8 (20/30 m resolution) SCA products showed the highest classification accuracy (high recall, precision > 0.95) and are recommended. MODIS offers daily revisit time, while Sentinel-2/Landsat-8 provides higher spatial resolution.
• Surface Soil Moisture (SSM): The NSIDC SMAP product at 1 km resolution is recommended, yielding correlation coefficients greater than 0.6 at most stations and the lowest median unbiased RMSE (approximately 0.06–0.08 m³/m³). Other disaggregated products (ESACCISM, CGLS SWI) also performed well.
• Evapotranspiration (ET): All evaluated ET products showed similar performance against eddy covariance measurements, with correlation coefficients generally greater than 0.8 and median RMSE values between 0.3 and 0.4 mm/d. The ensemble mean of products achieved the highest correlation (0.92). While MODIS-Terra/Aqua (500 m) offers higher spatial resolution, its 8-day temporal resolution is a limitation; daily products like ETMonitor (1 km), ALEXI, and HOLAPS (5 km) could be valuable for daily constraints.

Contributions

• Provided a comprehensive, multi-variable intercomparison of high-resolution Earth Observation products (precipitation, snow cover area, surface soil moisture, evapotranspiration) specifically tailored for hyper-resolution hydrological modeling over Europe.
• Delivered concrete recommendations for the best-performing EO products for each hydrological variable, considering their technical specifications and performance against diverse observational benchmarks.
• Highlighted the benefits of merging different EO products and integrating reanalysis data for improved precipitation estimates, and the effectiveness of disaggregation techniques for soil moisture.
• Identified critical challenges and future research directions for the effective integration of high-resolution EO data into hydrological models, such as addressing scale mismatches, data gaps, and ensuring physical consistency.

Funding

• European Space Agency (ESA) 4DHydro initiative (contract 4000141141/23/I-EF).
• AdAgriF - Advanced methods of greenhouse gases emission reduction and sequestration in agriculture and forest landscape for climate change mitigation (for Milan Fischer, Matěj Orság, and Oldrich Rakovec).
• EU-FP6 project UERRA (for E-OBS dataset).
• Copernicus Climate Change Service.
• ECA&D project.

Citation

@article{GarcíaGarcía2025Intercomparison,
  author = {García‐García, Almudena and Stradiotti, Pietro and Paolo, Federico Di and Filippucci, Paolo and Fischer, Milan and Orság, Matěj and Brocca, Luca and Peng, Jian and Dorigo, Wouter and Gruber, Alexander and Droppers, Bram and Wanders, Niko and Haag, Arjen and Weerts, Albrecht and Modiri, Ehsan and Rakovec, Oldřich and Francés, Félix and Dall’Amico, M. and Anderson, Martha C. and Hain, Christopher and Samaniego, Luis},
  title = {Intercomparison of Earth Observation products for hyper-resolution hydrological modelling over Europe},
  journal = {Remote Sensing of Environment},
  year = {2025},
  doi = {10.1016/j.rse.2025.115131},
  url = {https://doi.org/10.1016/j.rse.2025.115131}
}