The Well: Compressible MHD Turbulence (64 Cubed)
- Provider
- Polymathic AI (The Well); Blakesley Burkhart, Flatiron Institute CCA and Rutgers University
- License
- CC-BY-4.0 (source)
- Size
- 39 files
- Format
- HDF5 (in sampled files)
- Stored at
- Hugging Face (https://huggingface.co/datasets/polymathic-ai/MHD_64)
- Access
- Free and public. No account with the data provider is needed.
- Category
- Physics
- Source
- https://arxiv.org/abs/2412.00568
About The Well: Compressible MHD Turbulence (64 Cubed)
3D isothermal magnetohydrodynamic (MHD) turbulence without self-gravity, as in the diffuse interstellar medium, spanning subsonic to supersonic and sub-Alfvenic to super-Alfvenic regimes; MHD turbulence matters for the solar wind, galaxy formation and ISM dynamics. Run at 256 cubed with a Fortran MPI code, then anti-aliased with an ideal low-pass filter and downsampled to 64 by 64 by 64; each trajectory has 100 time steps on a periodic box. Fields are density, velocity and magnetic field. The datasheet lists 100 trajectories: 10 initial conditions times 10 combinations of sonic Mach number (0.5 to 7) and Alfvenic Mach number (0.7 or 2). Listing data/ (measured 2026-11-13) shows train/ (10 files, 60.8 GB, one per Mach number pair), valid/ and test/ (10 files, 7.3 GB each), plus an old_and_problematic_data/ folder at the root; about 159 GB in the repository. Files are HDF5 in the common The Well format; open them with the_well package (WellDataModule with base path hf://datasets/polymathic-ai/) or directly with h5py. Part of The Well (Ohana et al., NeurIPS 2024 Datasets and Benchmarks, arXiv:2412.00568); released under CC-BY-4.0, cite The Well and the simulation's source papers.
How to load The Well: Compressible MHD Turbulence (64 Cubed) in Python
from huggingface_hub import hf_hub_download
path = hf_hub_download(repo_id="polymathic-ai/MHD_64", filename="data/test/MHD_Ma_0.7_Ms_0.5.hdf5", repo_type="dataset")
print(path)Data structure
File types in sampled files: HDF5
Files
Example file: data/test/MHD_Ma_0.7_Ms_0.5.hdf5
Analyze this dataset without downloading it
On Scigantic, The Well: Compressible MHD Turbulence (64 Cubed) opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.
This page is also available as Markdown and as JSON.
Related Physics datasets
- Kolmogorov Flow at Re ~10^4, Lyapunov-Calibrated Trajectories for Probabilistic Neural PDE Forecasting (JAX-CFD)
- Sonkyo Benchmark Wind Turbine Blade Vibration under Varying Temperature: Shaker Tests with Added Masses and Cut Cracks, -15 to +20 C (derived copy)
- HiLiftAeroML: Wall-Modeled LES of the NASA CRM High-Lift Aircraft for AI Surrogates
- Quantum Physics Corpus Rebel Ready
- NeuralConvection 3D: Boussinesq-to-Compressible Natural Convection in a Cube for Fourier Neural Operators (Menglik, Shao, Hyde)
- TIDE Strongly Rotating Turbulence (Coriolis Force): 3D Incompressible Navier-Stokes DNS at 256^3 (Dai et al.)
- TIDE Passive Scalar Mixing in Forced Turbulence: 3D Incompressible Navier-Stokes DNS at 256^3 (Dai et al.)
- WAKESET: 1,091 High-Reynolds-Number RANS Simulations of Turbulent Wakes behind an XLUUV