Chawanda et al. (2025) CoSWAT Model v1: A high-resolution global SWAT+ hydrological model

Identification

• Journal: Hydrology and earth system sciences
• Year: 2025
• Date: 2025-12-02
• Authors: Celray James Chawanda, Ann van Griensven, Albert Nkwasa, Jose Pablo Teran Orsini, Jaehak Jeong, Soon-Kun Choi, Raghavan Srinivasan, Jeffrey G. Arnold
• DOI: 10.5194/hess-29-6901-2025

Research Groups

• Department of Water and Climate, Vrije Universiteit Brussel, Belgium
• Blackland Research & Extension Center, Texas A&M Agrilife Research, USA
• Institute for Water Education (IHE) Delft, the Netherlands
• International Institute for Applied Systems Analysis (IIASA), Austria
• National Institute of Agricultural Sciences (NAS), Rural Development Administration, Republic of Korea
• Grassland Soil and Water Research Laboratory, USDA Agricultural Research Service (ARS), USA

Short Summary

This study developed CoSWAT Model v1, a high-resolution global SWAT+ hydrological model, and an open-source framework for its reproducible setup and execution. It demonstrates the feasibility of global SWAT+ modeling at 2 km resolution, showing reasonable evapotranspiration patterns but limited river discharge performance without calibration or reservoir implementation.

Objective

• Develop a high-resolution global SWAT+ hydrological model.
• Establish a reproducible framework (CoSWAT) for large-scale SWAT+ applications, ensuring reproducibility and scalability.
• Evaluate the model's performance against observed data for evapotranspiration and river discharge.

Study Configuration

• Spatial Scale: Global (excluding Greenland), with a 2 km resolution for Digital Elevation Model (DEM), land use, and soil maps. Climate data was at 0.5 decimal degrees (approximately 55.66 km along the equator).
• Temporal Scale: Simulation period from 1977 to 1990, including a 5-year warm-up period.

Methodology and Data

• Models used:
  • SWAT+ (version 60.5.7)
  • Community SWAT (CoSWAT) modeling framework (open-source, Python-based scripted workflow for automated data retrieval, preprocessing, and model configuration, maximizing parallel processing).
• Data sources:
  • Digital Elevation Model (DEM): ASTER global DEM.
  • Land Use Data: European Space Agency (ESA) 2007 land use map (300 m native resolution).
  • Soil Data: FAO soil data from the Harmonized World Soil Database (HWSD) (1 km² native resolution).
  • Climate Data: GSWP3-EWEMBI reanalysis dataset from the Inter-Sectoral Impact Model Intercomparison (ISIMIP) project (daily minimum/maximum air temperature, precipitation, relative humidity, solar radiation, near-surface wind speed).
  • River Flow Data (for evaluation): The Global Runoff Data Centre (GRDC).
  • Evapotranspiration (ET) Data (for evaluation): GLEAM v4 dataset (0.1° resolution).

Main Results

• The Community SWAT (CoSWAT) modeling framework was successfully developed, automating data retrieval, preprocessing, and model configuration for large-scale SWAT+ applications, leveraging parallel processing.
• The first global SWAT+ model was set up at a 2 km resolution, resulting in 2.63 million Hydrologic Response Units (HRUs).
• Evapotranspiration (ET) simulations for 1982–1990 showed reasonable spatial patterns compared to the GLEAM v4 dataset, with 78.54 % of sampled points exhibiting differences within ± 100 mm. Discrepancies were noted in regions like East Africa, Central Asia, and the Andes mountains.
• River discharge performance for 1982–1990 was limited without calibration or reservoir implementation, with only 23.02 % of gauging stations showing positive Kling–Gupta Efficiency (KGE) values. However, 85.31 % of stations showed a positive correlation (r) in flow values.

Contributions

• Development of the first high-resolution global SWAT+ model, demonstrating the feasibility of such an application.
• Creation of the CoSWAT modeling framework, an open-source, reproducible, and scalable solution that automates data handling and leverages parallel processing for large-scale SWAT+ simulations.
• Provides a robust foundation for future advancements in global hydrological modeling, enabling more detailed analysis of global water resources and climate change impacts, and supporting UN Sustainable Development Goals.
• Highlights the critical need for incorporating water management practices (e.g., reservoirs) and calibration in global hydrological models to improve river discharge simulations.

Funding

• Research Foundation – Flanders (FWO) International Coordination Action (ICA) “Open Water Network: Open Data and Software tools for water resources management” (project code G0E2621N)
• Open Water Network: impacts of global change on water quality (project code G0ADS24N)
• Fonds Wetenschappelijk Onderzoek (grant no. 2022-F2812650-228938)
• Rural Development Administration of South Korea (Award number PJ017314)

Citation

@article{Chawanda2025CoSWAT,
  author = {Chawanda, Celray James and Griensven, Ann van and Nkwasa, Albert and Orsini, Jose Pablo Teran and Jeong, Jaehak and Choi, Soon-Kun and Srinivasan, Raghavan and Arnold, Jeffrey G.},
  title = {CoSWAT Model v1: A high-resolution global SWAT+ hydrological model},
  journal = {Hydrology and earth system sciences},
  year = {2025},
  doi = {10.5194/hess-29-6901-2025},
  url = {https://doi.org/10.5194/hess-29-6901-2025}
}