CLOVER: Convnet Line-fitting Of Velocities in Emission-line Regions

Abstract

When multiple star-forming gas structures overlap along the line-of-sight and emit optically thin emission at significantly different radial velocities, the emission can become non-Gaussian and often exhibits two distinct peaks. Traditional line-fitting techniques can fail to account adequately for these double-peaked profiles, providing inaccurate cloud kinematics measurements. We present a new method called Convnet Line-fitting Of Velocities in Emission-line Regions (CLOVER) for distinguishing between one-component, two-component, and noise-only emission lines using 1D convolutional neural networks trained with synthetic spectral cubes. CLOVER utilizes spatial information in spectral cubes by predicting on $3\times 3$ pixel sub-cubes, using both the central pixel's spectrum and the average spectrum over the $3\times 3$ grid as input. On an unseen set of 10,000 synthetic spectral cubes in each predicted class, CLOVER has classification accuracies of $\sim 99\%$ for the one-component class and $\sim 97\%$ for the two-component class. For the noise-only class, which is analogous to a signal-to-noise cutoff of four for traditional line-fitting methods, CLOVER has classification accuracy of $100\%$. CLOVER also has exceptional performance on real observations, correctly distinguishing between the three classes across a variety of star-forming regions. In addition, CLOVER quickly and accurately extracts kinematics directly from spectra identified as two-component class members. Moreover, we show that CLOVER is easily scalable to emission lines with hyperfine splitting, making it an attractive tool in the new era of large-scale NH${}_3$ and N${}_2$H${}^{+}$ mapping surveys.

arXiv commentAccepted for publication in ApJ

References in corpus (41)

• Adam: A Method for Stochastic Optimization
• Astropy: A Community Python Package for Astronomy
• Identifying Exoplanets with Deep Learning: A Five Planet Resonant Chain around Kepler-80 and an Eighth Planet around Kepler-90
• Filamentary Accretion Flows in the Embedded Serpens South Protocluster
• Low Virial Parameters in Molecular Clouds: Implications for High Mass Star Formation and Magnetic Fields
• The COMPLETE Survey of Star-Forming Regions: Phase I Data
• SDC13 infrared dark clouds: Longitudinally collapsing filaments?
• Molecular gas kinematics within the central 250 pc of the Milky Way
• A Survey for Infall Motions toward Starless Cores. I. CS (2-1) and N${}_2$H${}^{+}$ (1-0) Observations
• Direct observation of a sharp transition to coherence in Dense Cores
• Fast Radio Burst 121102 Pulse Detection and Periodicity: A Machine Learning Approach
• A Large Catalog of Accurate Distances to Molecular Clouds from PS1 Photometry
• Hipparcos distances of Ophiuchus and Lupus cloud complexes
• The Green Bank Ammonia Survey (GAS): First Results of NH3 mapping the Gould Belt
• The JCMT Gould Belt Survey: First results from the SCUBA-2 observations of the Ophiuchus molecular cloud and a virial analysis of its prestellar core population
• Trigonometric Parallaxes of Massive Star Forming Regions: VIII. G12.89+0.49, G15.03-0.68 (M17) and G27.36-0.16
• An Application of Deep Neural Networks in the Analysis of Stellar Spectra
• LZIFU: an emission-line fitting toolkit for integral field spectroscopy data
• Filamentary Accretion Flows in the Infrared Dark Cloud G14.225-0.506 Revealed by ALMA
• Complex, Quiescent Kinematics in a Highly Filamentary Infrared Dark Cloud
• Autonomous Gaussian Decomposition
• Synthetic C${}^{18}$O observations of fibrous filaments: the problems of mapping from PPV to PPP
• GaussPy+: A fully automated Gaussian decomposition package for emission line spectra
• The Green Bank Ammonia Survey: Dense Cores Under Pressure in Orion A
• Abundant cyanopolyynes as a probe of infall in the Serpens South cluster-forming region
• Droplets I: Pressure-Dominated Sub-0.1 pc Coherent Structures in L1688 and B18
• Probing Inward Motions in Starless Cores Using The HCN J = 1-0 Hyperfine Transitions : A Pointing Survey Toward Central Regions
• The Herschel view of star formation in the Rosette molecular cloud under the influence of NGC 2244
• An Ammonia Spectral Map of the L1495-B218 Filaments in the Taurus Molecular Cloud : I. Physical Properties of Filaments and Dense cores
• The JCMT 12CO(3-2) Survey of the Cygnus X Region: I. A Pathfinder
• Temperature structure and kinematics of the IRDC G035.39-00.33
• KFPA Examinations of Young STellar Object Natal Environments (KEYSTONE): Hierarchical Ammonia Structures in Galactic Giant Molecular Clouds
• The JCMT Gould Belt Survey: First results from SCUBA-2 observations of the Cepheus Flare Region
• The Green Bank Ammonia Survey: A Virial Analysis of Gould Belt Clouds in Data Release 1
• CASI: A Convolutional Neural Network Approach for Shell Identification
• Using an Artificial Neural Network to Classify Multi-component Emission Line Fits
• The Green Bank Ammonia Survey: Observations of Hierarchical Dense Gas Structures in Cepheus-L1251
• The Radio Ammonia Mid-Plane Survey (RAMPS) Pilot Survey
• Correlating Infall with Deuterium Fractionation in Dense Cores
• The discrimination between star-forming and AGN galaxies in the absence of Hα and [NII]: A machine learning approach
• Infall/Expansion Velocities in the Low-Mass Dense Cores L492, L694-2, and L1521F: Dependence on Position and Molecular Tracer

Cited by in corpus (3)

• Velocity-coherent Filaments in NGC 1333: Evidence for Accretion Flow?
• A lack of constraints on the cold opaque HI mass: HI spectra in M31 and M33 prefer multi-component models over a single cold opaque component
• Machine Learning Approach to Integral Field Unit Spectroscopy Observations: I. HII Region Kinematics