SDSS galaxies with double-peaked emission lines: double starbursts or AGNs?

TL;DR

This study analyzes SDSS galaxies with double-peaked emission lines to distinguish between dual starbursts and active galactic nuclei (AGNs), using spectral decomposition and emission line diagnostics.

Contribution

It introduces a method to decompose emission lines into two components and classifies galaxies into starburst or AGN categories based on BPT diagrams and abundance estimates.

Findings

55 spectra successfully decomposed into two emission components.

18 galaxies identified as starburst-like, 37 as AGNs.

Global oxygen abundance typically between component measurements.

Abstract

With the aim of investigating galaxies with two strong simultaneous starbursts, we have extracted a sample of galaxies with double-peaked emission lines in their global spectra from the SDSS spectral database. We then fitted the emission lines Halpha, Hbeta, [OIII]5007, [NII]6584, [SII]6717 and [SII]6731 of 129 spectra by two Gaussians to separate the radiation of the two (blue and red) components. A more or less reliable decomposition of the all those emission lines have been found for 55 spectra. Using a standard BPT classification diagram, we have been able to divide the galaxies from our sample into two subsamples: Sample A consisting of 18 galaxies where both components belong to the photoionised class of objects, and Sample B containing 37 galaxies which show non-thermal ionisation (AGNs). We have examined the properties of the blue and red components, and found that the differences between radial velocities of components lie within 200 - 400 km/s for galaxies of both subsamples. The equivalent number of ionising stars is in the range 10^4 - 10^5 O7V stars for each component in the galaxies of Sample A. We have estimated the oxygen and nitrogen abundances as well as the electron temperatures for each component using the recent NS-calibration and from global spectra for galaxies from Sample A using both the NS and ON-calibration. We have found that the global oxygen abundance is typically in between the measured abundances of individual components for our sample of galaxies, and that both calibrations provide consistent global abundances. Finally, we suggest the classical O/H -- N/O diagram is used to test the reliability of the dividing lines between starburst-like objects and AGNs in the so-called BPT diagram.