The FuXi-CFD dataset accompanies the paper 'Reconstructing fine-scale 3D wind fields with terrain-informed machine learning'. It contains over 12,000 cases of large-scale CFD-generated wind field simulations used for training and evaluating the FuXi-CFD model. The dataset was authored by linchensen and last updated on Hugging Face in February 2026.
Use Cases
- Training terrain-informed machine learning models for wind field reconstruction based on CFD simulation data.
- Evaluating the performance of physics-informed ML models against high-fidelity CFD results.
- Studying the interaction between complex terrain and 3D wind flow patterns.
- Benchmarking new simulation or downscaling methods for computational fluid dynamics.
Strengths
- Contains over 12,000 simulation cases, providing a substantial corpus for model development.
- Data is generated from CFD simulations, offering high-fidelity ground truth for wind fields.
- Specifically designed for terrain-informed machine learning applications.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count per case and total data size are unknown, which may limit suitability assessment.
- Description metadata is limited; actual data quality requires manual inspection after download.
Provenance
- Source
- linchensen on Hugging Face, accompanying an academic paper.
- Collection Method
- Generated via Computational Fluid Dynamics (CFD) simulations.
- Time Range
- null
- Freshness
- Last updated 2026-02-25 10:57:13; freshness should be verified.
- Geography
- null