Journal
ASTROPHYSICAL JOURNAL
Volume 910, Issue 2, Pages -Publisher
IOP Publishing Ltd
DOI: 10.3847/1538-4357/abe46d
Keywords
Interstellar medium; Magnetohydrodynamical simulations; Non-Gaussianity; Convolutional neural networks; Astronomy data analysis
Categories
Funding
- National Science Foundation Graduate Research Fellowship [DGE-1745303]
- NSF [AST-1614941]
- Simons Foundation
- DIRAC Institute in the Department of Astronomy at the University of Washington
- FAS Division of Science Research Computing Group at Harvard University
Ask authors/readers for more resources
This paper demonstrates the sensitivity of wavelet scattering transform (WST) combined with linear discriminant analysis (LDA) to non-Gaussian structures in 2D interstellar medium dust maps. The WST-LDA method shows a high true positive rate in classifying magnetohydrodynamic (MHD) turbulence simulations and is robust to observational artifacts. Further applications in 3D and potential use on all-sky dust maps for extracting hydrodynamic parameters are discussed.
The complex interplay of magnetohydrodynamics, gravity, and supersonic turbulence in the interstellar medium (ISM) introduces a non-Gaussian structure that can complicate a comparison between theory and observation. In this paper, we show that the wavelet scattering transform (WST), in combination with linear discriminant analysis (LDA), is sensitive to non-Gaussian structure in 2D ISM dust maps. WST-LDA classifies magnetohydrodynamic (MHD) turbulence simulations with up to a 97% true positive rate in our testbed of 8 simulations with varying sonic and Alfvenic Mach numbers. We present a side-by-side comparison with two other methods for non-Gaussian characterization, the reduced wavelet scattering transform (RWST) and the three-point correlation function (3PCF). We also demonstrate the 3D-WST-LDA, and apply it to the classification of density fields in position-position-velocity (PPV) space, where density correlations can be studied using velocity coherence as a proxy. WST-LDA is robust to common observational artifacts, such as striping and missing data, while also being sensitive enough to extract the net magnetic field direction for sub-Alfvenic turbulent density fields. We include a brief analysis of the effect of point-spread functions and image pixelization on 2D-WST-LDA applied to density fields, which informs the future goal of applying WST-LDA to 2D or 3D all-sky dust maps to extract hydrodynamic parameters of interest.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available