Application of Convolutional Neural Networks to Predict Magnetic Fields Directions in Turbulent Clouds

Abstract

We adopt the deep learning method CASI-3D (Convolutional Approach to Structure Identification-3D) to infer the orientation of magnetic fields in sub-/trans- Alfvenic turbulent clouds from molecular line emission. We carry out magnetohydrodynamic simulations with different magnetic field strengths and use these to generate synthetic observations. We apply the 3D radiation transfer code RADMC-3d to model 12CO and 13CO (J = 1-0) line emission from the simulated clouds and then train a CASI-3D model on these line emission data cubes to predict magnetic field morphology at the pixel level. The trained CASI-3D model is able to infer magnetic field directions with low error (< 10deg for sub-Alfvenic samples and <30deg for trans-Alfvenic samples). We furthermore test the performance of CASI-3D on a real sub-/trans- Alfvenic region in Taurus. The CASI-3D prediction is consistent with the magnetic field direction inferred from Planck dust polarization measurements. We use our developed methods to produce a new magnetic field map of Taurus that has a three-times higher angular resolution than the Planck map.

ApJ Accepted

References in corpus (21)

• Theory of Star Formation
• Rotation-invariant convolutional neural networks for galaxy morphology prediction
• The COMPLETE Survey of Star-Forming Regions: Phase I Data
• Inefficient star formation through turbulence, magnetic fields and feedback
• FOREST Unbiased Galactic plane Imaging survey with the Nobeyama 45-m telescope (FUGIN) I: Project Overview and Initial Results
• Radiative torque alignment: Essential Physical Processes
• Effective destruction of CO by cosmic rays: implications for tracing H${}_2$ gas in the Universe
• Searching for Exoplanets Using Artificial Intelligence
• The Five College Radio Astronomy Observatory CO Mapping Survey of the Taurus Molecular Cloud
• Velocity Gradients as a Tracer for Magnetic Fields
• Tracing interstellar magnetic field using velocity gradient technique: Application to Atomic Hydrogen data
• Magnetic Field Morphology in Interstellar Clouds with the Velocity Gradient Technique
• The Physical Nature of Neutral Hydrogen Intensity Structure
• Formation of HI Clouds in Shock-compressed Interstellar Medium: Physical Origin of Angular Correlation Between Filamentary Structure and Magnetic Field
• Photodissociation Region Diagnostics Across Galactic Environments
• Striations in the Taurus molecular cloud: Kelvin-Helmholtz instability or MHD waves?
• The relative orientation between the magnetic field and gradients of surface brightness within thin velocity slices of 12CO and 13CO emission from the Taurus molecular cloud
• Application of Convolutional Neural Networks to Identify Stellar Feedback Bubbles in CO Emission
• Velocity Gradients: Magnetic Field Tomography towards the Supernova Remnant W44
• Velocity Anisotropy in Self-Gravitating Molecular Clouds. I: Simulation
• A Census of Protostellar Outflows in Nearby Molecular Clouds

Cited by in corpus (3)

• Nature of Striation in 21 cm Channel Maps: Velocity Caustics
• Denoising Diffusion Probabilistic Models to Predict the Density of Molecular Clouds
• Modeling turbulent and self-gravitating fluids with Fourier neural operators