Robust profile decomposition for large extragalactic spectral-line surveys

Research output: Contribution to journalArticle

Abstract

We present a novel algorithm that is based on a Bayesian Markov Chain Monte Carlo (MCMC) technique for performing robust profile analysis of a data cube from either single-dish or interferometric radio telescopes. It fits a set of models comprised of a number of Gaussian components given by the user to individual line-of-sight velocity profiles, then compares them and finds an optimal model based on the Bayesian Inference Criteria computed for each model. The decomposed Gaussian components are then classified into bulk or non-circular motions as well as kinematically cold or warm components. The fitting based on the Bayesian MCMC technique is insensitive to initial estimates of the parameters, and suffers less from finding the global minimum in models given enough sampling points and a wide range of priors for the parameters. It is found to provide reliable profile decomposition and classification of the decomposed components in a fully automated way, together with robust error estimation of the parameters as shown by performance tests using artificial data cubes. We apply the newly developed algorithm to the H I data cubes of sample galaxies from the Local Volume HI galaxy Survey (LVHIS). We also compare the kinematically cold and warm components, and bulk velocity fields with previous analyses made in a classical method.

Original languageEnglish
Pages (from-to)5021-5034
Number of pages14
JournalMonthly Notices of the Royal Astronomical Society
Volume485
Issue number4
DOIs
Publication statusPublished - Jun 2019

Fingerprint

line spectra
decomposition
Markov chain
profiles
Markov chains
velocity distribution
galaxies
velocity profile
parabolic reflectors
performance tests
radio telescopes
inference
line of sight
radio
sampling
parameter
estimates
cold

Cite this

@article{e645ab8da50f40f68c9875b5b1b654f6,
title = "Robust profile decomposition for large extragalactic spectral-line surveys",
abstract = "We present a novel algorithm that is based on a Bayesian Markov Chain Monte Carlo (MCMC) technique for performing robust profile analysis of a data cube from either single-dish or interferometric radio telescopes. It fits a set of models comprised of a number of Gaussian components given by the user to individual line-of-sight velocity profiles, then compares them and finds an optimal model based on the Bayesian Inference Criteria computed for each model. The decomposed Gaussian components are then classified into bulk or non-circular motions as well as kinematically cold or warm components. The fitting based on the Bayesian MCMC technique is insensitive to initial estimates of the parameters, and suffers less from finding the global minimum in models given enough sampling points and a wide range of priors for the parameters. It is found to provide reliable profile decomposition and classification of the decomposed components in a fully automated way, together with robust error estimation of the parameters as shown by performance tests using artificial data cubes. We apply the newly developed algorithm to the H I data cubes of sample galaxies from the Local Volume HI galaxy Survey (LVHIS). We also compare the kinematically cold and warm components, and bulk velocity fields with previous analyses made in a classical method.",
keywords = "Galaxies: kinematics and dynamics, Galaxies: structure, Methods: data analysis",
author = "Oh, {Se Heon} and Lister Staveley-Smith and For, {Bi Qing}",
year = "2019",
month = "6",
doi = "10.1093/mnras/stz710",
language = "English",
volume = "485",
pages = "5021--5034",
journal = "Monthly Notices of the Royal Astronomical Society",
issn = "0035-8711",
publisher = "OXFORD UNIV PRESS UNITED KINGDOM",
number = "4",

}

Robust profile decomposition for large extragalactic spectral-line surveys. / Oh, Se Heon; Staveley-Smith, Lister; For, Bi Qing.

In: Monthly Notices of the Royal Astronomical Society, Vol. 485, No. 4, 06.2019, p. 5021-5034.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Robust profile decomposition for large extragalactic spectral-line surveys

AU - Oh, Se Heon

AU - Staveley-Smith, Lister

AU - For, Bi Qing

PY - 2019/6

Y1 - 2019/6

N2 - We present a novel algorithm that is based on a Bayesian Markov Chain Monte Carlo (MCMC) technique for performing robust profile analysis of a data cube from either single-dish or interferometric radio telescopes. It fits a set of models comprised of a number of Gaussian components given by the user to individual line-of-sight velocity profiles, then compares them and finds an optimal model based on the Bayesian Inference Criteria computed for each model. The decomposed Gaussian components are then classified into bulk or non-circular motions as well as kinematically cold or warm components. The fitting based on the Bayesian MCMC technique is insensitive to initial estimates of the parameters, and suffers less from finding the global minimum in models given enough sampling points and a wide range of priors for the parameters. It is found to provide reliable profile decomposition and classification of the decomposed components in a fully automated way, together with robust error estimation of the parameters as shown by performance tests using artificial data cubes. We apply the newly developed algorithm to the H I data cubes of sample galaxies from the Local Volume HI galaxy Survey (LVHIS). We also compare the kinematically cold and warm components, and bulk velocity fields with previous analyses made in a classical method.

AB - We present a novel algorithm that is based on a Bayesian Markov Chain Monte Carlo (MCMC) technique for performing robust profile analysis of a data cube from either single-dish or interferometric radio telescopes. It fits a set of models comprised of a number of Gaussian components given by the user to individual line-of-sight velocity profiles, then compares them and finds an optimal model based on the Bayesian Inference Criteria computed for each model. The decomposed Gaussian components are then classified into bulk or non-circular motions as well as kinematically cold or warm components. The fitting based on the Bayesian MCMC technique is insensitive to initial estimates of the parameters, and suffers less from finding the global minimum in models given enough sampling points and a wide range of priors for the parameters. It is found to provide reliable profile decomposition and classification of the decomposed components in a fully automated way, together with robust error estimation of the parameters as shown by performance tests using artificial data cubes. We apply the newly developed algorithm to the H I data cubes of sample galaxies from the Local Volume HI galaxy Survey (LVHIS). We also compare the kinematically cold and warm components, and bulk velocity fields with previous analyses made in a classical method.

KW - Galaxies: kinematics and dynamics

KW - Galaxies: structure

KW - Methods: data analysis

UR - http://www.scopus.com/inward/record.url?scp=85067054063&partnerID=8YFLogxK

U2 - 10.1093/mnras/stz710

DO - 10.1093/mnras/stz710

M3 - Article

VL - 485

SP - 5021

EP - 5034

JO - Monthly Notices of the Royal Astronomical Society

JF - Monthly Notices of the Royal Astronomical Society

SN - 0035-8711

IS - 4

ER -