Zonato et al. (2026) GLIDE-SOL: a GPU-accelerated global lightweight infrastructure for diagnostic environmental modeling with SOLWEIG

Identification

• Journal: Geoscientific model development
• Year: 2026
• Date: 2026-08-11
• Authors: Andrea Zonato, Harsh Kamath, Naveen Sudharsan, Luca Monaco, Jonas Kittner, Luise (M. Sc.) Wolf, Matthias Demuzere, Ariane Middel, Benjamin Bechtel, Massimo Milelli
• DOI: 10.5194/gmd-19-7389-2026

Research Groups

• CIMA Research Foundation (Savona, Italy)
• Jackson School of Geosciences, University of Texas at Austin (Austin, TX, USA)
• Bochum Urban Climate Lab, Ruhr University Bochum (Bochum, Germany)
• B-Kode VOF (Ghent, Belgium)
• Global Facility for Disaster Reduction and Recovery, World Bank Group (Washington, DC, USA)
• The GAME School, Arizona State University (Tempe, AZ, USA)

Short Summary

This paper introduces GLIDE-SOL, a GPU-accelerated global lightweight infrastructure for diagnostic environmental modeling with SOLWEIG. GLIDE-SOL operationalizes SOLWEIG for rapid and repeatable thermal-comfort mapping across diverse urban environments.

Objective

• To develop a globally deployable workflow that integrates SOLWEIG with globally available datasets and executes efficiently on GPUs.
• To improve realism in urban climate modeling by incorporating diagnostic physical modules to compensate for limitations of coarse meteorological forcing.

Study Configuration

• Spatial Scale: Global, with applications at neighborhood-scale analyses to city-wide and multi-city experiments.
• Temporal Scale: Hourly simulations over long periods (e.g., months).

Methodology and Data

• Models used: SOLWEIG radiative balance libraries, with GPU-accelerated execution strategy.
• Data sources: Globally available datasets, including ESA WorldCover, MERIT DEM, GlobalBuildingAtlas, ERA5 reanalysis.

Main Results

• GLIDE-SOL improves UTCI performance (RMSE reduced from 8.09 to 2.84°C) and simulations of T mrt , air temperature, and wind speed.
• The workflow incorporates diagnostic physical modules: a Röckle-style directional wind-attenuation scheme and a simple urban heat island adjustment.

Contributions

• GLIDE-SOL provides a globally deployable, reproducible, and scalable framework for urban climate modeling.
• The workflow integrates harmonized global inputs with physics-based diagnostics, GPU acceleration, and explicit domain tiling.

Funding

• This research was funded by the CIMA Research Foundation, the Jackson School of Geosciences at the University of Texas at Austin, and the Bochum Urban Climate Lab at Ruhr University Bochum.

Citation

@article{Zonato2026GLIDESOL,
  author = {Zonato, Andrea and Kamath, Harsh and Sudharsan, Naveen and Monaco, Luca and Kittner, Jonas and Wolf, Luise (M. Sc.) and Demuzere, Matthias and Middel, Ariane and Bechtel, Benjamin and Milelli, Massimo},
  title = {GLIDE-SOL: a GPU-accelerated global lightweight infrastructure for diagnostic environmental modeling with SOLWEIG},
  journal = {Geoscientific model development},
  year = {2026},
  doi = {10.5194/gmd-19-7389-2026},
  url = {https://doi.org/10.5194/gmd-19-7389-2026}
}