A Comparison of Star Formation Rates by Different Tracers in Nearby Galaxies

TL;DR

The paper tackles how different star formation rate (SFR) tracers compare in a large, high-quality SDSS sample ($ ext{z}<0.3$) and how their estimates relate to AGN strength via the Eddington ratio. It systematically cross-matches optical SFR indicators with infrared, ANN-predicted IR, and SED-based SFR catalogs, across SF, composite, Seyfert, and LINER hosts, and accounts for AGN contamination using the Rainbow $f_{AGN}$ fractions. The results reveal modest agreement among tracers for star-forming galaxies but substantial discrepancies for non-SF hosts, especially LINERs, and show a robust, mass-dependent correlation between SFR (and sSFR) and the Eddington ratio. The findings imply a linked gas supply governing both star formation and AGN activity, with implications for interpreting AGN feedback and selection biases when using different SFR indicators.

Abstract

We utilize a large sample of $\sim$113,000 galaxies ($z < 0.3$) from the Sloan Digital Sky Survey with high-quality data to compare star formation rates (SFRs) across multiple diagnostic methods and examine their connection to Active Galactic Nuclei (AGNs) strength, indicated by Eddington ratio. Our sample encompassed star-forming (SF), composite, Seyfert, and LINER galaxies. Our analysis utilizes various SFRs indicators, including observed infrared flux ($\rm SFR_{FIR}$) from AKARI/Herschel ($\sim$4,100 sources), the MPA-JHU catalog ($\rm SFR_{Dn4000}$), the ANN catalog ($\rm SFR_{ANN}$), the GSWLC catalog ($\rm SFR_{SED}$ and $\rm SFR_{MIR}$), as well as \OII\ and \Ha\ emission lines ($\rm SFR_{[OII]}$ and $\rm SFR_{Hα}$). Within SF galaxies, SFRs measurements from different tracers exhibited differences, with offsets and scatter below 0.26 dex and 0.29 dex, respectively. Moreover, non-SF galaxies (composite, Seyfert, and LINER) displayed discrepancies among SFR tracers, particularly for LINER galaxies, with offsets below 0.86 dex and a scatter of 0.57 dex. Additionally, our findings revealed robust correlations between SFRs and specific SFRs (sSFRs) with Eddington ratios. Eddington ratio exhibited gradual transitions in the (s)SFRs-stellar mass diagrams. Galaxies with high Eddington ratios displayed high star formation activity, similar to blue SF galaxies. Furthermore, we observed decreasing sSFR trends from SF galaxies to composite, Seyfert, and LINER galaxies. Our results may provide insight into our understanding of (s)SFRs traced by different approaches and their connection to AGN activities.

Figures (15)

• Figure 1: BPT diagram of the selected sample, $\sim$113,000 galaxies (z $<$ 0.3) in this work. The classification of galaxies into SF, composite, Seyfert, and LINER categories is based on Kauffmann+03 (dashed line), Kewley+06 (solid line), and Schawinski+07 (dotted line), respectively.
• Figure 2: Top panels: Stellar masses of the selected sample as a function of redshift. Bottom panels: Fraction of the u-band light captured by the SDSS fiber, defined as the difference in magnitude, $\rm f_{SDSS} = U_{Fiber} - U_{Total}$, shown as a function of redshift. The black squares show the median stellar mass values of each redshift bin $\rm \Delta z = 0.05$. Each panel corresponds to a different galaxy classification: SF, composite, Seyfert, and LINER, respectively.
• Figure 3: AGN fraction estimations obtained using the MCMC Rainbow Python package. The color scale represents the estimated fractions based on large potential referent points of pure SF galaxies (magenta) and AGNs (red).
• Figure 4: Comparison between [O3] luminosity (top panel) and $\rm SFR_{[OII]}$ (bottom panel) with and without correction for AGN fraction ($\rm f_{AGN}$). We only show sources where $\rm 0 < f_{AGN} < 1$.
• Figure 5: Ratios between SFRs determined by other tracers ($\rm SFR_{ANN}$, $\rm SFR_{Dn4000}$, $\rm SFR_{[OII]}$, $\rm SFR_{H\alpha}$, $\rm SFR_{SED}$ and $\rm SFR_{MIR}$) and $\rm SFR_{FIR}$. The circles and stars represent $\rm SFR_{FIR}$ determined by IR fluxes observed by Akari and Herschel, respectively. Each panel illustrates the classifications among galaxies, such as SF, composite, Seyfert, and LINER, respectively. The color scales represent the $\rm D_{n}(4000)$ values.