NeuralPDR: Neural Differential Equations as surrogate models for Photodissociation Regions
arXiv:2506.14270 · doi:10.1088/2632-2153/ade4ee
Abstract
Computational astrochemical models are essential for helping us interpret and understand the observations of different astrophysical environments. In the age of high-resolution telescopes such as JWST and ALMA, the substructure of many objects can be resolved, raising the need for astrochemical modeling at these smaller scales, meaning that the simulations of these objects need to include both the physics and chemistry to accurately model the observations. The computational cost of the simulations coupling both the three-dimensional hydrodynamics and chemistry is enormous, creating an opportunity for surrogate models that can effectively substitute the chemical solver. In this work we present surrogate models that can replace the original chemical code, namely Latent Augmented Neural Ordinary Differential Equations. We train these surrogate architectures on three datasets of increasing physical complexity, with the last dataset derived directly from a three-dimensional simulation of a molecular cloud using a Photodissociation Region (PDR) code, 3D-PDR. We show that these surrogate models can provide speedup and reproduce the original observable column density maps of the dataset. This enables the rapid inference of the chemistry (on the GPU), allowing for the faster statistical inference of observations or increasing the resolution in hydrodynamical simulations of astrophysical environments.
Accepted for publication in Machine Learning: Science and Technology. Focus on ML and the Physical Sciences, Mach. Learn.: Sci. Technol (2025)
References in corpus (27)
- Decoupled Weight Decay Regularization
- HEALPix -- a Framework for High Resolution Discretization, and Fast Analysis of Data Distributed on the Sphere
- On the difficulty of training Recurrent Neural Networks
- Mamba: Linear-Time Sequence Modeling with Selective State Spaces
- The Relation Between Optical Extinction and Hydrogen Column Density in the Galaxy
- A PDR-Code Comparison Study
- STARFORGE: Toward a comprehensive numerical model of star cluster formation and feedback
- Photodissociation and X-Ray Dominated Regions
- Latent ODEs for Irregularly-Sampled Time Series
- SILCC-Zoom: The dynamical and chemical evolution of molecular clouds
- On Neural Differential Equations
- 3D-PDR: A new three-dimensional astrochemistry code for treating Photodissociation Regions
- xLSTM: Extended Long Short-Term Memory
- PDRs4All III: JWST's NIR spectroscopic view of the Orion Bar
- SILCC-Zoom: H and CO-dark gas in molecular clouds -- The impact of feedback and magnetic fields
- Parameterized Neural Ordinary Differential Equations: Applications to Computational Physics Problems
- Reducing the complexity of chemical networks via interpretable autoencoders
- Chemulator: Fast, accurate thermochemistry for dynamical models through emulation
- PDFchem: A new fast method to determine ISM properties and infer environmental parameters using probability distributions
- Incorporating Astrochemistry into Molecular Line Modelling via Emulation
- Neural networks: solving the chemistry of the interstellar medium
- Efficient Marginalization of Discrete and Structured Latent Variables via Sparsity
- Mapping Synthetic Observations to Prestellar Core Models: An Interpretable Machine Learning Approach
- Speeding up astrochemical reaction networks with autoencoders and neural ODEs
- Reduced Order Model for Chemical Kinetics: A case study with Primordial Chemical Network
- CODES: Benchmarking Coupled ODE Surrogates
- 3D-PDR Orion dataset and NeuralPDR: Neural Differential Equations for Photodissociation Regions