Astrophysics Simulation · polymathic-ai
supernova_explosion_64
Hydrodynamic simulation dataset of supernova blastwaves propagating through turbulent interstellar gas clouds.
What supernova_explosion_64 is
supernova_explosion_64 is an astrophysical simulation dataset modelling the expansion of core-collapse supernova blastwaves into cold, turbulent interstellar gas clouds mimicking the Milky Way. Created by researchers including Keiya Hirashima and distributed as part of The Well collection, it provides 3D time-series data capturing the compression of a monatomic ideal gas following a stellar explosion.
Supernova feedback—the injection of kinetic energy and momentum from dying massive stars into the surrounding interstellar medium—governs star formation and galactic structure. Because shockwaves accelerate gas to supersonic speeds across temperatures spanning seven orders of magnitude, direct numerical simulations require very small time-steps and extensive computing resources, making them difficult to incorporate directly into large-scale galaxy simulations. This dataset provides standardised trajectories designed to train and benchmark machine learning surrogate models for high-resolution galactic feedback.
How it was made
The dataset was produced using the smoothed particle hydrodynamics code ASURA-FDPS. Each simulation models a 60-parsec cubic domain with open boundary conditions, initialised with a virialised isothermal gas sphere having Burgers turbulence where the energy spectrum follows E(k) proportional to k^-2. An initial thermal energy of 10^51 erg is deposited at the centre of the domain, heating gas to approximately 10^7 K and driving an adiabatic, supersonic blastwave into the ambient medium.
The physical system is governed by hydrodynamic equations for an ideal gas with a specific heat ratio of 5/3, incorporating gravitational acceleration, viscous acceleration, and radiative heating and cooling terms. Trajectories are sampled onto a uniform Cartesian grid of 64 x 64 x 64 cells across 59 non-uniform time-steps spanning from 0 to 0.2 megayears, recording pressure, density, temperature, and velocity fields.
Key facts
- Grid resolution
- 64 x 64 x 64 uniform Cartesian grid
- Trajectory count
- 740 trajectories
- Temporal coverage
- 0 to 0.2 Myr
- Spatial domain
- 60 pc x 60 pc x 60 pc
- Physical variables
- Pressure, density, temperature, and velocity
- Dataset size
- 268.2 GB
- Simulation software
- ASURA-FDPS (Smoothed Particle Hydrodynamics)
What you can use it for
- Supernova Surrogate Modelling
- Training deep learning surrogate models and neural operators to emulate high-resolution supernova feedback for integration into large-scale galaxy formation codes.
- Spatiotemporal Fluid Dynamics Forecasting
- Benchmarking multivariate neural forecasting architectures on 3D compressible fluid dynamics with extreme density contrasts and shock discontinuities.
- Thermodynamic and Momentum Analysis
- Investigating how thermal energy converts into momentum in non-conservative conditions governed by radiative heating and cooling.
Explore the data
No queryable preview
This dataset is not available in a queryable form — usually an unusual file format or a repository too large to convert. The original source may still be readable.Strengths and limits
Good at
- Captures complex supersonic shock propagation and radiative processes over seven orders of magnitude in physical quantities.
- Includes 740 diverse trajectories generated from varied turbulence seeds across realistic interstellar cloud profiles.
- Features standard benchmark evaluation scores across architectures including U-Net, CNextU-net, and Fourier Neural Operators.
Watch out for
- Spatial resolution is restricted to a 64-cubed grid across the 60-parsec domain.
- Initial conditions use a fixed ambient temperature of 100 K and a constant baseline hydrogen number density of 44.5 per cubic centimetre.
- Radiative heating and cooling explicitly prevent global conservation of thermal energy and momentum.
Questions
Can I look inside supernova_explosion_64 without downloading it?
Not here: it is not available in a queryable form. Its description and statistics are still shown.
What licence is supernova_explosion_64 under?
Its authors declare cc-by-4.0. Read the terms on the original page before reusing the data.
What equations govern the fluid dynamics in this dataset?
The simulations solve hydrodynamic equations for a monatomic ideal gas with an adiabatic index of 5/3, accounting for pressure gradients, viscous acceleration, gravitational potential, and radiative heat influx and outflux.
How well do baseline neural networks perform on this benchmark?
On the volume root mean square error (VRMSE) metric, U-Net achieves the lowest error at 0.3063, followed by CNextU-net at 0.3181, FNO at 0.3783, and TFNO at 0.3785.
How is each simulation explosion triggered?
A total thermal energy of 10^51 erg is injected into the centre of the 60-parsec box at the initial time-step, heating the local gas to roughly 10^7 K to drive the blastwave.
How to cite it
Hirashima, Keiya, Kana Moriwaki, Michiko S. Fujii, Yutaka Hirai, Takayuki R. Saitoh, and Junichiro Makino. '3D-Spatiotemporal forecasting the expansion of supernova shells using deep learning towards high-resolution galaxy simulations.' Monthly Notices of the Royal Astronomical Society 526, no. 3 (2023): 4054–4066.
Written from supernova_explosion_64's own dataset card. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.