A Robust Classification of Galaxy Spectra: Dealing with Noisy and Incomplete Data

Connolly, Andrew J.; Szalay, A. S.

doi:10.1086/300839

Cited by 90 publications

(101 citation statements)

References 15 publications

Supporting

Mentioning

101

Contrasting

Order By: Relevance

“…Panel a shows the histogram of the galaxy spectral classification parameter eClass as provided by the SDSS for DR1. This spectral classification parameter is based on a principal component analysis method (Connolly et al 1995;Connolly & Szalay 1999). Note that the spectral classification of galaxies by Connolly et al (1995) is slightly different from the one used in SS99 who removed from the analysis the wavelength regions containing emission lines.…”

Section: Overall Properties Of Our Sample Of Normal Galaxiesmentioning

confidence: 99%

What drives the Balmer extinction sequence in spiral galaxies?

Stasińska

Mateus

Sodré

et al. 2004

A&A

View full text Add to dashboard Cite

Abstract. Using spectra of normal emission line galaxies from the First Data Release of the Sloan Digital Sky Survey (SDSS) we have investigated the relations between the extinction C(Hβ) as derived from the Hα/Hβ emission line ratio and various global parameters of the galaxies. Our main findings are that: 1) C(Hβ) is linked with the galaxy spectral type and colour, decreasing from early-to late-type spirals; 2) C(Hβ) increases with increasing metallicity; 3) C(Hβ) is larger in galaxies with an older stellar population; 4) C(Hβ) is larger for more luminous galaxies; 5) The extinction of the stellar light is correlated with both the extinction of the nebular light and the intrinsic galaxy colours. We propose phenomenological interpretations of our empirical results. We have also cross-correlated our sample of SDSS galaxies with the IRAS data base. Due to the lower redshift limit of our sample and to the detection limit of IRAS, such a procedure selected only luminous infrared galaxies. We found that correlations that were shown by other authors to occur between optical and infrared properties of galaxies disappear when restricted the sample to luminous infrared galaxies. We also found that the optical properties of the luminous infrared galaxies in our SDSS sample are very similar to those of our entire sample of SDSS galaxies. This may be explained by the IRAS luminosity of the galaxies originating in the regions that formed massive stars less than 1 Myr ago, while the opacity of galaxies as derived from the Hα/Hβ emission line ratio is due to diffuse dust. We show some implications of our empirical results for the determination of global star formation rates and total stellar masses in normal galaxies.

show abstract

Section: Overall Properties Of Our Sample Of Normal Galaxiesmentioning

confidence: 99%

What drives the Balmer extinction sequence in spiral galaxies?

Stasińska

Mateus

Sodré

et al. 2004

A&A

View full text Add to dashboard Cite

show abstract

“…The treatment of incomplete data in PCA has already been discussed in the literature, e.g., Everson & Sirovich (1995) and Connolly & Szalay (1999). However, the implementation is complicated and deserves further discussion.…”

Section: Dealing With Incomplete Data In the Pcamentioning

confidence: 99%

“…It can be shown that the latter issue has a relatively straightforward, clean solution involving decorrelating the eigenspectra (which are not orthogonal when we do not use their entire wavelength coverage). This has been presented in Connolly & Szalay (1999). However, estimating the eigenspectra in the first place is much more difficult to address.…”

Section: Dealing With Incomplete Data In the Pcamentioning

confidence: 99%

See 1 more Smart Citation

The DEEP2 Galaxy Redshift Survey: Spectral Classification of Galaxies at \documentclass{aastex} \usepackage{amsbsy} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{bm} \usepackage{mathrsfs} \usepackage{pifont} \usepackage{stmaryrd} \usepackage{textcomp} \usepackage{portland,xspace} \usepackage{amsmath,amsxtra} \usepackage[OT2,OT1]{fontenc} \newcommand\cyr{ \renewcommand\rmdefault{wncyr} \renewcommand\sfdefault{wncyss} \renewcommand\encodingdefault{OT2} \normalfont \selectfont} \DeclareTextFontComman

Madgwick

Coil

Conselice

et al. 2003

ApJ

View full text Add to dashboard Cite

We present a principal component analysis (PCA)-based spectral classification, , for the first 5600 galaxies observed in the DEEP2 Redshift Survey. This parameter provides a very pronounced separation between absorption-and emission-dominated galaxy spectra-corresponding to passively evolving and actively star-forming galaxies in the survey, respectively. In addition it is shown that, despite the high resolution of the observed spectra, this parameter alone can be used to quite accurately reconstruct any given galaxy spectrum, suggesting there are not many '' degrees of freedom '' in the observed spectra of this galaxy population. It is argued that this form of classification, , will be particularly valuable in making future comparisons between high-and low-redshift galaxy surveys for which very large spectroscopic samples are now readily available, particularly when used in conjunction with high-resolution spectral synthesis models, which will be made public in the near future. We also discuss the relative advantages of this approach to distant galaxy classification compared to other methods such as colors and morphologies. Finally, we compare the classification derived here with that adopted for the 2dF Galaxy Redshift Survey and in so doing show that the two systems are very similar. This will be particularly useful in subsequent analyses when making comparisons between results from each of these surveys to study evolution in the galaxy populations and large-scale structure.

show abstract

“…The first method attempted was based on a PCA of spectroscopic objects using the method of Connolly & Szalay (1999). The spectral classification coefficient is generally calculated by using two out of five eigencoefficients that build up the expansion of the eigentemplates of a galaxy's spectrum.…”

Section: Morphology Classificationmentioning

confidence: 99%

In the neighbourhood of Tame Monsters

Villarroel

2012

A&A

View full text Add to dashboard Cite

Context. The impact of quasars on their galaxy neighbours is an important factor in the understanding of galaxy evolution models. Aims. The aim of this work is to characterize the intermediate-scale environments of quasars at low redshift (z < 0.2) with the most statistically complete sample to date using the seventh data release of the Sloan Digital Sky Survey. Methods. We have used 305 quasar-galaxy associations with spectroscopically measured redshifts within the projected distance range of 350 kpc, to calculate how surface densities of galaxies, colors, degree of ionization, dust extinction and star-formation rates change as a function of the distance to our quasar sample. We also identify the companion active galactic nuclei from our main galaxy sample and calculate surface density for different galaxy types. We have done this in three different quasar-galaxy redshift difference ranges |Δz| < 0.001, 0.006, and 0.012. Results. Our results suggest that there is a significant increase of the surface density of blue neighbours around our low-redshift quasar sample that is steeper than around non-active field galaxies of the same luminosity and redshift range. This may indicate that quasar formation is accomplished via a merging scenario. No significant changes in star formation rate, dust extinction, degree of ionization or color as a function of distance from the quasars was observed. We could not observe any direct effects from quasars on the their companion galaxies.

show abstract

A Robust Classification of Galaxy Spectra: Dealing with Noisy and Incomplete Data

Cited by 90 publications

References 15 publications

What drives the Balmer extinction sequence in spiral galaxies?

What drives the Balmer extinction sequence in spiral galaxies?

In the neighbourhood of Tame Monsters

Contact Info

Product

Resources

About