Relating spatially resolved optical attenuation, dust and gas in nearby galaxies

TL;DR

This work tackles how optical attenuation inferred from the Balmer decrement $A_{V,\mathrm{BD}}$ and from SED-fitting $A_{V,\mathrm{SED}}$ relate to dust distribution and gas surface density in galaxy discs at sub-kpc scales. It employs pixel-by-pixel SED fitting with CIGALE to derive $\Sigma_{\mathrm{dust}}$ and $A_{V,\mathrm{SED}}$, and computes $A_{V,\mathrm{BD}}$ per pixel from Balmer decrement measurements, enabling direct, resolved comparisons with $\Sigma_{\mathrm{HI}}$, $\Sigma_{\mathrm{H2}}$, and $\Sigma_{\mathrm{gas}}$. The results show that $\Sigma_{\mathrm{dust}}$ and $A_{V,\mathrm{BD}}$ trace molecular and total gas more closely than atomic gas, with dust-to-gas-ratio variations influencing HI shielding; notably $A_{V,\mathrm{SED}}^{\mathrm{young}}$ agrees with $A_{V,\mathrm{BD}}$ for most regions, allowing $A_{V,\mathrm{SED}}^{\mathrm{young}}$ to serve as a practical attenuation tracer in external samples lacking IFU data. The comparison with simple geometries indicates old-star attenuation is consistent with a mixed disc, while $A_{V,\mathrm{BD}}$ and $A_{V,\mathrm{SED}}^{\mathrm{young}}$ lie between foreground-screen and mixed configurations, highlighting the complex relative geometry of dust and stars in star-forming discs.

Abstract

We relate the optical attenuation inferred by the Balmer decrement, $A_{V\mathrm{,BD}}$, and by the SED-fitting, $A_{V\mathrm{,SED}}$, to the dust distribution and gas surface density throughout the disc of galaxies, down to scales smaller than 0.5 kpc. We investigate five nearby star-forming spirals with available FUV to sub-mm observations, along with atomic and molecular gas surface density maps and optical integral-field spectroscopic data. We use the CIGALE SED-fitting code to map the dust mass surface density ($Σ_\mathrm{dust}$) and $A_{V,\mathrm{SED}}$ of different stellar populations. For each pixel, we independently estimate the attenuation from the BD. We find that both $Σ_\mathrm{dust}$ and $A_{V,\mathrm{BD}}$ trace better the molecular and total gas mass surface density, rather than the atomic gas. Since regions sampled in this study have high molecular fractions, atomic gas surface densities, indicative of molecular gas shielding layers, decrease as the mean dust-to-gas ratio increases from galaxy to galaxy. The fitted attenuation towards young stars, $A^\mathrm{young}_{V,\mathrm{SED}}$, is in good agreement with $A_{V,\mathrm{BD}}$ and it can then be used to trace the attenuation in star forming galaxies where integral-field observations are not available. We estimate the ratio of $A_{V,\mathrm{BD}}$ over the total stellar $A_{V,\mathrm{SED}}$ and find it slightly larger than what has been found in previous studies. Finally, we investigate which dust distribution reproduces better the estimated $A_{V,\mathrm{BD}}$ and $A_{V,\mathrm{SED}}$. We find that the attenuation towards old stars is consistent with the expectations for a standard galactic disc, where the stellar and dust distributions are mixed, while $A_{V, \mathrm{BD}}$ and the $A^\mathrm{young}_{V, \mathrm{SED}}$ are between the values expected for a foreground dust screen and a mixed configuration.

Figures (15)

• Figure 1: Maps of observed and derived properties of NGC 4321. Top: Convolved and regridded SDSS r-band, logarithms of stellar mass surface density ($\Sigma_\mathrm{star}$), star-formation rate surface density ($\Sigma_\mathrm{SFR}$), dust mass surface density ($\Sigma_\mathrm{dust}$), as well as SED-fitting derived attenuation in the V-band ($A_{V,\mathrm{SED}}$), from left to right. Bottom: Surface density of atomic gas mass ($\Sigma_\mathrm{HI}$), surface density of molecular gas mass ($\Sigma_\mathrm{H2}$), logarithm of the total gas mass surface density ($\Sigma_\mathrm{gas}$), dust-to-gas ratio and attenuation in the V-band as derived by the Balmer decrement ($A_{V,\mathrm{BD}}$), from left to right. All maps are at a resolution of 18". Gray points correspond to pixels that are rejected because they do not have a sufficient number of data points in the FIR regime (see Sec. \ref{['sec:steps']}) or because they are unreliable (see Sec. \ref{['sec:fitting']} for more details). The $\Sigma_\mathrm{HI}$ and $\Sigma_\mathrm{H2}$ maps extend up to their corresponding 3$\sigmaup$ limit. In the $\Sigma_\mathrm{HI}$, $\Sigma_\mathrm{H2}$ and $A_{V,\mathrm{BD}}$ maps pixels that correspond to the areas that are excluded by the SED-fitting analysis are plotted with smaller dots. In the $\log_{10}\Sigma_\mathrm{gas}$ map, pixels excluded by the SED-fitting analysis and having both Hi and CO detection are depicted by smaller squares, while pixels with only Hi and not CO are plotted with dots. In the dust-to-gas ratio maps we limit the colour-coding to the $5^{th}-95^{th}$ percentile range for illustrative purposes. Contours are taken from the $\log_{10}$$\Sigma_\mathrm{dust}$ [M$_\odot$ pc$^{-2}$] maps with a lowest contour at -1.5 and linear spacing with the highest at 0.
• Figure 2: Resolved scaling relations of dust mass surface density with the atomic (left panel), molecular (middle panel) and total gas (right panel) surface density. Each galaxy is represented by a different colour. The median uncertainties for each galaxy are shown in the lower-right or lower-left corner of the panels. Solid lines are the best linear fits to the full sample, while gray shaded area indicates the fit uncertainty. The slope, intercept, and scatter of each fit are also given along with the correlation coefficients. In the right panel open circles show pixels where only Hi is detected.
• Figure 3: Dust mass surface density as a function of atomic gas mass surface density, colour-coded with atomic gas mass fraction (top-panel) and stellar mass surface density (bottom panel) for the whole sample. In both panels, crosses refer to pixels where only Hi gas is detected.
• Figure 4: Molecular gas mass fraction as a function dust-to-gas mass ratio, for each galaxy in the sample. The data points are colour coded as a function of galactocentric radius in units of $R_{25}$. Uncertainties are plotted by gray error-bars.
• Figure 5: Attenuation on the V-band, derived from the Balmer decrement, as a function of the line-of-sight surface density of the atomic (left panel), molecular (middle panel) and total gas (right panel) mass. Each galaxy is represented by a different colour. The median uncertainties for each galaxy are shown in the top- or the lower-right corner of the panels. Solid lines are the best linear fits to the full sample, while a shaded area indicates the fit uncertainty. The slope, intercept, and scatter of each fit are also given along with the correlation coefficients.