J/MNRAS/464/3796 HI gas mass fraction estimations (Teimoorinia+, 2017)
Pattern recognition in the ALFALFA.70 and Sloan Digital Sky Surveys:
a catalogue of ∼500000 H I gas fraction estimates based on artificial
neural networks.
Teimoorinia H., Ellison S.L., Patton D.R.
<Mon. Not. R. Astron. Soc., 464, 3796-3811 (2017)>
=2017MNRAS.464.3796T 2017MNRAS.464.3796T (SIMBAD/NED BibCode)
ADC_Keywords: Galaxy catalogs ; H I data
Keywords: methods: data analysis - methods: statistical - surveys -
galaxies: evolution - galaxies: fundamental parameters
Abstract:
The application of artificial neural networks (ANNs) for the
estimation of HI gas mass fraction (MHI/M*) is investigated, based
on a sample of 13 674 galaxies in the Sloan Digital Sky Survey (SDSS)
with HI detections or upper limits from the Arecibo Legacy Fast
Arecibo L-band Feed Array (ALFALFA). We show that, for an example set
of fixed input parameters (g-r colour and i-band surface
brightness), a multidimensional quadratic model yields MHI/M*
scaling relations with a smaller scatter (0.22dex) than traditional
linear fits (0.32dex), demonstrating that non-linear methods can lead
to an improved performance over traditional approaches. A more
extensive ANN analysis is performed using 15 galaxy parameters that
capture variation in stellar mass, internal structure, environment and
star formation. Of the 15 parameters investigated, we find that g-r
colour, followed by stellar mass surface density, bulge fraction and
specific star formation rate have the best connection with MHI/M*.
By combining two control parameters, that indicate how well a given
galaxy in SDSS is represented by the ALFALFA training set (PR) and the
scatter in the training procedure (σfit), we develop a
strategy for quantifying which SDSS galaxies our ANN can be adequately
applied to, and the associated errors in the MHI/M* estimation. In
contrast to previous works, our MHI/M* estimation has no systematic
trend with galactic parameters such as M*, g-r and star
formation rate. We present a catalogue of MHI/M* estimates for more
than half a million galaxies in the SDSS, of which ∼150000 galaxies
have a secure selection parameter with average scatter in the MHI/M*
estimation of 0.22dex.
Description:
We present a novel method to estimate HI gas mass fraction and the
associated uncertainties based on the patterns found in our data sets,
using machine learning methods. The ALFALFAsurvey is used as our main
training sample, and we check our model estimations with a range of
validation sets, comprised of the GASS (Catinella et al., 2013,
Cat. J/MNRAS/436/34) and Cornell (Giovanelli et al., 2007, Cat.
J/AJ/133/2569) surveys and a small sample of PM galaxies (Ellison et
al., 2015MNRAS.448..221E 2015MNRAS.448..221E).
File Summary:
--------------------------------------------------------------------------------
FileName Lrecl Records Explanations
--------------------------------------------------------------------------------
ReadMe 80 . This file
table2.dat 109 561585 Catalogue of ANN estimated MHI/M* along with all
the control parameters described in this paper
--------------------------------------------------------------------------------
See also:
J/AJ/133/2569 : Arecibo legacy fast ALFA survey III. (Giovanelli+, 2007)
J/MNRAS/436/34 : GALEX Arecibo SDSS survey final data release (Catinella+ 2013)
Byte-by-byte Description of file: table2.dat
--------------------------------------------------------------------------------
Bytes Format Units Label Explanations
--------------------------------------------------------------------------------
1- 18 A18 --- SDSS SDSS identification number
20- 31 F12.8 deg RAdeg SDSS right ascension (J2000)
33- 44 F12.8 deg DEdeg SDSS declination (J2000)
46- 53 F8.6 --- z SDSS redshift
55- 60 F6.3 [Msun] logM* Stellar mass
62- 69 F8.5 [-] logMHI/M* Ratio of HI-to-stellar mass
71- 77 F7.5 --- Cfgas Confidence of the MHI/M* estimation
(Cfgas) (1)
79- 85 F7.5 --- PR [0/1] Pattern recognition detection metric
87- 93 F7.5 --- sigmafitN [0/1] Inverse normalized uncertainty on the
ANN estimation
95-101 F7.5 --- sigmafit [0/4.5] Uncertainty on the ANN estimation
103-109 F7.5 --- sigmafgas [0/1] Uncertainty on the MHI/M* estimation
--------------------------------------------------------------------------------
Note (1): higher values indicate more robust estimations.
--------------------------------------------------------------------------------
History:
From electronic version of the journal
(End) Patricia Vannier [CDS] 05-Jun-2018