Published June 23, 2026
| Version v2
Dataset
Open
Random Buildings dataset
Description
Dataset of paper "Anchored-Branched Steady-state WInd Flow Transformer (AB-SWIFT): a metamodel for 3D atmospheric flow in urban environments"
This is a dataset of microscale atmospheric flow around various randomly generated urban geometries and for different atmospheric stability classes.
See https://github.com/cerea-daml/abswift for associated code.
V2: includes large-scale examples, i.e. simulations with more buildings, spread over a larger area.
See https://github.com/cerea-daml/domain-extension
Files
random_buildings_dataset.zip
Additional details
Related works
- Is described by
- Preprint: arXiv:2603.25635 (arXiv)
- Preprint: arXiv:2606.14597 (arXiv)
Funding
- Association Nationale de la Recherche et de la Technologie
- 2023/1614
- Électricité de France (France)
- Commissariat à l'Énergie Atomique et aux Énergies Alternatives
Software
- Repository URL
- https://github.com/cerea-daml/abswift
- Programming language
- Python
- Development Status
- Unsupported