Mliyeh et al. (2026) Advancing hydrological modeling in the Mediterranean: Multi-objective calibration of the SWAT+ model using open-source data and tools

Identification

• Journal: Agricultural Water Management
• Year: 2026
• Date: 2026-03-02
• Authors: Mohammed Mouad Mliyeh, Mourad Aqnouy, Ismail Bouizrou, Marouane Laaraj, Aqil Tariq, Mahmoud Zemzami, Lahcen Benaabidate, Saro Lee
• DOI: 10.1016/j.agwat.2026.110263

Research Groups

• Laboratory of Geo-Resources and Environment, Faculty of Sciences and Techniques, Sidi Mohammed Ben Abdellah University, Fez, Morocco
• Geosciences and Technologies Research Team, Department of Geosciences, Faculty of Sciences and Techniques of Errachidia, Moulay Ismail University of Meknes, Errachidia, Morocco
• Institute of Ecology and Landscape, Hochschule Weihenstephan-Triesdorf University (HSWT), Freising, Germany
• Department of Wildlife, Fisheries and Aquaculture, College of Forest Resources, Mississippi State University, MS, USA
• Higher School of Education, Mohammed First University, Oujda, Morocco
• Earth Systems Science and Data Solutions Lab, Chapman University, Orange, CA, USA
• Schmid College of Science and Technology, Chapman University, Orange, CA, USA
• Department of Environmental Sciences, Faculty of Science, Alexandria University, Alexandria, Egypt

Short Summary

This study evaluated multi-variable calibration strategies for the SWAT+ model in the Upper Oum Er Rbia watershed, Morocco, integrating streamflow and remote sensing evapotranspiration data. The multi-variable approach achieved satisfactory and balanced performance for both streamflow (NSE = 0.75, KGE = 0.77) and evapotranspiration (NSE = 0.51, KGE = 0.64), highlighting the value of open-access remote sensing ET data in refining hydrological model parameters for data-scarce regions.

Objective

• To develop and calibrate a SWAT+ hydrological model for the Upper Oum Er Rbia watershed through a multi-objective approach that integrates streamflow (Q) and evapotranspiration (ET), relying exclusively on open-source datasets and tools.

Study Configuration

• Spatial Scale: Upper Oum Er Rbia watershed, Morocco, covering an area of 3560 square kilometers.
• Temporal Scale:
  • Simulation period: 30 years (1993–2022).
  • Warm-up period: 7 years (1993–1999).
  • Calibration period: 12 years (2000–2011).
  • Validation period: 11 years (2012–2022).
  • Time step: Monthly.

Methodology and Data

• Models used:
  • SWAT+ (Soil and Water Assessment Tool Plus)
  • R-SWAT (for parallel parameter calibration, sensitivity analysis, and uncertainty assessment)
  • QSWAT+ module within QGIS 3.34 interface (for model construction)
• Data sources:
  • Digital Elevation Model (DEM): Copernicus Global Digital Elevation Model (GLO-30), 30 meter resolution.
  • Land Use Map: GLCC map by USGS EROS.
  • Soil Maps: Food and Agriculture Organization (FAO) and UNESCO.
  • Rainfall and Streamflow: Oum Er Rabia Hydraulic Basin Agency (OERHBA), daily rainfall from 6 gauging stations, monthly streamflow from Ahmed El Hansali Station (2000-2022).
  • Temperature: ERA5 product (ECMWF reanalysis dataset), accessed via Climate Engine.
  • Evapotranspiration (ET): GLDAS-2 product, accessed via Google Earth Engine.
  • Humidity, Solar Radiation, Wind Speed/Direction: Generated by the wgncfsrworld weather generator to fill gaps.

Main Results

• Sensitivity analysis identified 12 parameters significantly influencing hydrological processes, with 8 impacting streamflow and 9 affecting evapotranspiration. Five parameters (Condition II curve number, moisture bulk density, available water capacity of the soil layer, land surface slope, and saturated hydraulic conductivity) were sensitive to both variables.
• Single-variable streamflow calibration (S_Q) achieved very good streamflow performance (calibration: NSE = 0.77, KGE = 0.79, PBIAS = 3.8%; validation: NSE = 0.76, KGE = 0.81, PBIAS = 15.4%), but resulted in limited evapotranspiration performance (NSE = 0.05 for calibration).
• Single-variable evapotranspiration calibration (S_ET) showed good ET performance (calibration: NSE = 0.63, KGE = 0.79, PBIAS = 11.1%; validation: NSE = 0.66, KGE = 0.81, PBIAS = 12.8%), but poor streamflow performance (NSE = -0.12 for calibration).
• The multi-variable calibration strategy (Multi_Q&ET) achieved satisfactory and balanced performance for both streamflow and evapotranspiration:
  • Streamflow (calibration): NSE = 0.75, KGE = 0.77, PBIAS = 19.2%.
  • Streamflow (validation): NSE = 0.67, KGE = 0.70, PBIAS = 9.1%.
  • Evapotranspiration (calibration): NSE = 0.51, KGE = 0.64, PBIAS = 5.0%.
  • Evapotranspiration (validation): NSE = 0.55, KGE = 0.71, PBIAS = 2.9%.
• The Multi_Q&ET scenario yielded an average annual evapotranspiration of 361 mm (61.3% of 589 mm precipitation) and a water yield of 233 mm, aligning with hydrological characteristics of semi-arid and Mediterranean catchments.

Contributions

• Development of a fully reproducible, open-source SWAT+ modeling framework.
• A novel comparative evaluation of single- and multi-variable calibration strategies for streamflow (Q) and evapotranspiration (ET) in Morocco, implemented through SWAT+ and R-SWAT.
• Provision of spatially explicit insights into the response of key water-balance components under different calibration configurations.
• Introduction of the earliest implementation of a SWAT+ multi-objective calibration framework in Morocco, supported by parallelized computation, addressing a documented methodological gap in national hydrological assessments.
• Demonstration of the significant contribution of open-access remote sensing evapotranspiration data in refining model parameters, offering a realistic and effective strategy to enhance hydrological model calibration, particularly valuable for improving water resource management in regions vulnerable to climate change and limited by observational data.

Funding

• US Department of Education award number P116Z220190, “Earth Systems Science and Data Solutions Lab (EssDs): Applying Data Science Techniques to Achieve the UN Sustainable Development Goals”.

Citation

@article{Mliyeh2026Advancing,
  author = {Mliyeh, Mohammed Mouad and Aqnouy, Mourad and Bouizrou, Ismail and Laaraj, Marouane and Tariq, Aqil and Zemzami, Mahmoud and Benaabidate, Lahcen and Lee, Saro},
  title = {Advancing hydrological modeling in the Mediterranean: Multi-objective calibration of the SWAT+ model using open-source data and tools},
  journal = {Agricultural Water Management},
  year = {2026},
  doi = {10.1016/j.agwat.2026.110263},
  url = {https://doi.org/10.1016/j.agwat.2026.110263}
}