Ouassanouan et al. (2025) Downscaled ERA5 Land addresses agrometeorological data scarcity in North African basins

Identification

• Journal: Scientific Reports
• Year: 2025
• Date: 2025-11-04
• Authors: Youness Ouassanouan, Vincent Simonneaux, Mohamed Hakim Kharrou, Younes Fakır, Mohamed Wassim Baba, Bouchra Aït Hssaine, Chouaib El Hachimi, Laura Sourp, Abdelghani Chehbouni
• DOI: 10.1038/s41598-025-20552-2

Research Groups

• Center for Remote Sensing Applications (CRSA), Mohammed VI Polytechnic University (UM6P), Ben Guerir, Morocco
• CNES/CNRS/INRAE/IRD/UPS, Centre d’Etudes Spatiales de la Biosphère (CESBIO), Université de Toulouse, Toulouse, France
• International Water Research Institute (IWRI), Mohammed VI Polytechnic University (UM6P), Ben Guerir, Morocco
• Department of Geology, Faculty of Sciences–Semlalia, Cadi Ayyad University, Marrakech, Morocco

Short Summary

This study evaluates the accuracy of original and MicroMet-downscaled ERA5Land agrometeorological variables and their impact on reference evapotranspiration (ET0) estimation in a data-scarce North African basin, revealing that while downscaling improves some variables, it often degrades overall ET0 accuracy, yet ERA5Land remains valuable for long-term ET0 trend analysis.

Objective

• To evaluate the effectiveness of ERA5_Land reanalysis data in providing daily agrometeorological variables (air temperature, solar radiation, relative humidity, wind speed) for the period 2003–2021 at 10 study sites in a North African basin.
• To investigate whether downscaling the ERA5_Land data (10 km resolution) to station scale (250 m resolution) using the MicroMet model could improve the reliability of these meteorological variables.
• To compare the performance of the original ERA5Land data and the MicroMet-disaggregated ERA5Land data as potential sources for accurate daily estimation of reference evapotranspiration (ET0).
• To assess the long-term spatiotemporal changes in ET0 across the Tensift basin over the period 1950–2021, and the influence of climate variables and topography on ET0 variability.

Study Configuration

• Spatial Scale: Tensift basin, Morocco (3.0 × 10¹⁰ m²), encompassing both plain and mountainous areas. Data from 10 ground stations with altitudes ranging from 506 m to 3230 m. ERA5_Land data at 10 km spatial resolution, downscaled to 250 m.
• Temporal Scale:
  • Evaluation of ERA5Land and downscaled data: 2003–2021 (18 years) on a daily time step (aggregated from hourly ERA5Land and 30 min AWS data).
  • Long-term ET0 trend analysis: 1950–2021 (72 years) on an annual time step.

Methodology and Data

• Models used:
  • MicroMet: Quasi-physical based model for downscaling meteorological variables.
  • FAO56-Penman-Monteith (FAO56-PM) equation: Standardized method for ET0 estimation.
  • Mann–Kendall (MK) statistical test: Non-parametric test for trend analysis.
  • Sen Slope estimator: Non-parametric method for estimating the magnitude of a trend.
  • Pettitt test: Non-parametric test for detecting abrupt changes in time series.
• Data sources:
  • ERA5_Land reanalysis data: Hourly meteorological variables (air temperature, relative humidity, solar radiation, wind speed at 10 m) at 10 km spatial resolution.
  • Ground observations: Daily meteorological variables (air temperature, relative humidity, solar radiation, wind speed at 2 m) from 10 Automatic Weather Stations (AWSs) in the Tensift basin, with 30 min temporal resolution, covering various periods between 2002 and 2021.

Main Results

• Original ERA5Land performance: Air temperature (Tair) estimates were the most accurate (R² 0.93–0.98, RMSE 0.98–4.84 °C), followed by solar radiation (Rs) (R² 0.57–0.69, RMSE 51.49–62.55 W m⁻²), relative humidity (RH) (R² 0.33–0.86, RMSE 4.78–19.3%), and wind speed (u2) (R² < 0.5, RMSE 0.47–2.48 m s⁻¹). Tair and RH showed higher percent bias (PBIAS) in mountainous areas. ERA5Land generally underestimated ET0 (PBIAS -37.06% to -2.02%, RMSE 0.88–2 mm d⁻¹).
• MicroMet downscaling performance: Downscaling improved Tair and u2 biases, particularly in mountainous regions. RH showed no significant improvement. Rs was generally degraded, with increased RMSE and PBIAS, especially at mountainous sites. ET0 estimates using downscaled data did not generally improve; R² and RMSE remained stable, but PBIAS increased for most sites, shifting from underestimation to overestimation (e.g., Agdal PBIAS from -2.02% to 28.59%). Wind speed was identified as the most influential variable contributing to ET0 overestimation by MicroMet.
• Long-term ET0 trends (1950–2021) in Tensift basin (using original ERA5_Land): Annual ET0 exhibited three distinct phases: a decrease from 1950 to 1970, a nearly stable period from 1970 to 2000, and a significant increase from 2000 to 2021. The overall trend was predominantly increasing. The increase in Tair was identified as the main driver of the ET0 increase from 2000 to 2021, supported by increases in u2 and, to a lesser extent, Rs.

Contributions

• Provides a comprehensive evaluation of the reliability of original and MicroMet-downscaled ERA5_Land agrometeorological products (Tair, RH, Rs, u2) against ground observations in a data-scarce North African basin with complex topography.
• Assesses the impact of both original and downscaled meteorological variables on daily reference evapotranspiration (ET0) estimation using the FAO56-PM method.
• Offers a detailed analysis of long-term (1950–2021) spatiotemporal ET0 trends and their climatic and topographic drivers in the Tensift basin.
• Delivers critical insights into the potential and limitations of reanalysis data for irrigation and water management strategies in arid North African regions facing climate variability and data scarcity.

Funding

• Mohammed VI Polytechnic University (UM6P) for the first author's Ph.D. research.
• ASSIWAT project, funded by the OCP Group S.A. (Office Chérifien des Phosphates) under grant agreement no: AS_71.
• Tensift Observatory in the frame of the SudMed project.
• TREMA International Joint Laboratory (LMI-TREMA), funded by the University Cadi Ayyad (UCA, Morocco) and the French Research Institute for Development (IRD, CESBIO Laboratory, France).

Citation

@article{Ouassanouan2025Downscaled,
  author = {Ouassanouan, Youness and Simonneaux, Vincent and Kharrou, Mohamed Hakim and Fakır, Younes and Baba, Mohamed Wassim and Hssaine, Bouchra Aït and Hachimi, Chouaib El and Sourp, Laura and Chehbouni, Abdelghani},
  title = {Downscaled ERA5 Land addresses agrometeorological data scarcity in North African basins},
  journal = {Scientific Reports},
  year = {2025},
  doi = {10.1038/s41598-025-20552-2},
  url = {https://doi.org/10.1038/s41598-025-20552-2}
}