LMC+: Large-scale mapping of [CII] and [OIII] in the LMC molecular ridge, I. Dataset and line ratio analyses

TL;DR

We investigate how different ISM phases regulate star formation in the LMC by mapping [CII] $\lambda$158 μm and [OIII] $\lambda$88 μm across a 610 pc × 260 pc portion of the molecular ridge with SOFIA/FIFI-LS at ~2.5 pc resolution. The study integrates these maps with ALMA CO(2-1), a dust-based $L_ ext{TIR}$ map, Spitzer/MIPS 24 μm, and Hα data to dissect the heating and cooling balance across diverse environments in a low-metallicity context. We find that [CII] is widespread and traces CO-dark molecular gas, while [OIII] is strong in SF regions yet also extended in diffuse gas; the ratio $L_{[CII]}/L_{TIR}$ declines with rising $L_{TIR}$, pointing to a local origin of the [CII]-deficit rather than a global effect. The results imply a porous, clumpy ISM in the LMC where UV photons permeate beyond classical PDRs, informing how [CII] serves as a star-formation tracer in the early universe and guiding interpretations of low-metallicity ISM cooling.

Abstract

The fundamental process of star formation in galaxies involves the interplay between the fueling of star formation via molecular gas and the feedback from recently formed massive stars. This process, by which galaxies evolve, is also closely connected to the intrinsic properties of the interstellar medium (ISM). To study the role that different molecular and atomic phases of the ISM play in star formation, and to characterize their physical conditions, we zoom into our nearest neighboring galaxy, the Large Magellanic Cloud (LMC; 50 kpc). The LMC offers a view of the ISM and star formation conditions in a low metallicity environment similar to, in that regard, the epoch of the peak of star formation in the earlier universe. We present an unprecedentedly detailed analysis of a well-known star-forming regions (SFRs) at a spatial resolution of a few pc. We mapped a 610pcx260pc region in the LMC molecular ridge in [CII] and the [OIII] using the FIFI-LS instrument on the SOFIA telescope. We compare the data with the distribution of the CO (2-1) emission from ALMA, the modeled TIR luminosity as well as Spitzer/MIPS continuum and Halpha. We also provide a detailed description of the observing strategy and the data reduction. We find that [CII] and [OIII] emission is associated with the SFRs in the molecular ridge, but also extends throughout the mapped region, not obviously associated with ongoing star formation. The CO emission is clumpier than the [CII] emission and we find plentiful [CII] present where there is little CO emission, possibly holding important implications for CO-dark gas. We find a clear trend of the [CII]/TIR ratio decreasing with increasing TIR. This suggests a strong link between the [CII]-deficit and the local physical conditions instead of global properties.

Figures (9)

• Figure 1: Intensity of the far-infrared fine-structure lines [C$\,$ ii]$\lambda 158$$\mu$m (left) and [O$\,$ iii]$\lambda 88$$\mu$m (right) in W m$^{-2}$ sr$^{-1}$, observed by SOFIA/FIFI-LS. The intensities are the result of spectral Gaussian fits and only those fits with a S/N $\geq$ 3 are shown. Pixels with lower S/N are displayed as 0 (in black). The white contours are from L$_\mathrm{TIR}$ (300 and 600 L$_{sol}$/pc$^2$) shown in Fig. \ref{['fig:LTIR']}. The thin black contours shown in the [C$\,$ ii] map are 8$\mu$m from Spitzer/IRAC (4 and 8 MJy/Sr). The red contours in the [O$\,$ iii] map are Spitzer/MIPS 24$\mu$m continuum (5 and 150 MJy/sr). The [C$\,$ ii] map is shown in the native resolution with a beam FWHM of 15.3. The [O$\,$ iii] data in the three bright star formation regions marked by blue, green and red dashed lines) was smoothed to the same resolution (15.3) to increase the S/N per beam. Outside of the regions the data was further re-binned to 24 and fitted within a 48 beam to further increase the S/N in the diffuse regions. The 7 gray crosses, in identical positions in the [C$\,$ ii] and [O$\,$ iii] maps, mark the positions of the spectra shown in Appendix \ref{['spectra']}.
• Figure 2: RGB map of [C$\,$ ii] (r), [O$\,$ iii] (g) and $\rm ^{12}CO(2-1)$ (b) from Tarantino et al. (in preparation) and Chen et al. (in preparation), with the map of the $\rm ^{12}CO(2-1)$ outlined in white. Contours are shown for $\rm ^{12}CO(2-1)$ line flux levels of $5*10^{11}$ and $5*10^{10} W m^{-2} sr^{-1}$. Regions for analyses in Sect. \ref{['sec:dis']} are outlined, corresponding respectively to the SFRs, N159(green), N160(red), N158(blue), an extended, less active region in [O$\,$ iii] (yellow) and three CO-bright filaments (brown). They are also referred to in Fig. \ref{['fig:LTIR']}, and in the plots in Fig. \ref{['fig:ratio_diagrams']} and Fig. \ref{['fig:oiii_24']}.
• Figure 3: Map of TIR power in the molecular ridge. Regions for the analyses in Chapter \ref{['sec:dis']} are outlined, corresponding, respectively, to the SFRs, N159(green), N160(red dotted), N158(blue), an extended, less active region in [O$\,$ iii] (yellow), and three CO-bright filaments (brown dashed). They are also shown in Fig. \ref{['fig:rgb']}, and used in the plots in Figs. \ref{['fig:ratio_diagrams']}, \ref{['fig:CII_LTIR']}, and \ref{['fig:oiii_24']}.
• Figure 4: L$_\mathrm{[C\,{\sc II}]}$/L$_\mathrm{TIR}$ versus L$_\mathrm{TIR}$ in the molecular ridge in 12 × 12 pixels. Yellow, blue, red, green and brown colors refer to the regions outlined in Figs. \ref{['fig:rgb']} and \ref{['fig:LTIR']}. Gray points represent the pixels in the ridge outside of these regions. Data for 30$\;$Doradus from chevance20 is added. The gray line represents the average minimum detectable [C$\,$ ii] line flux for S/N = 3. The large circles represent the ratio values for the whole ridge (cyan) and the three bright SFRs (N158: blue; N160: red; N159:green), relative to the mean L$_\mathrm{TIR}$. The areas with filled colors represent observations from: Orion A Pabst21; M51 Pineda18; N11B lebouteiller12).
• Figure 5: Line ratio plots for the molecular ridge. Yellow, blue, red, green and brown colors refer to the regions outlined in Figs. \ref{['fig:rgb']} and \ref{['fig:LTIR']}. Gray points represent the pixels in the ridge outside of these identified regions. Left: L$_\mathrm{[O\,{\sc III}]}$/L$_\mathrm{[C\,{\sc II}]}$ versus L$_\mathrm{TIR}$ in 12 pixels for the three bright star formation regions and in 24 pixels for the rest of the molecular ridge, where pixels with S/N below 5 are shown as upper limits. 30$\;$Doradus data points from chevance20 have been added with 12 binning. The horizontal lines show the mean L$_\mathrm{[O\,{\sc III}]}$/L$_\mathrm{[C\,{\sc II}]}$ ratio for the regions. Right: L$_\mathrm{[C\,{\sc II}]}$/L$_\mathrm{TIR}$ versus L$_\mathrm{[C\,{\sc O}]}$/L$_\mathrm{TIR}$ in 12 pixels. CO (1-0) was estimated from CO (2-1) assuming a ratio of CO (2-1)/CO (1-0) of 0.6 (e.g. denBrok2021; Maeda2022). The average ratio (no fit) of L$_\mathrm{[C\,{\sc II}]}$/L$_\mathrm{CO(1-0)}$ from the molecular ridge is indicated with the cyan line. The large circles represent the ratios based on the fluxes for the whole regions, in cyan for the whole molecular ridge and in matching colors for the three star formation regions. The area filled in red represents the data range for dwarf galaxies from madden20 using a L$_\mathrm{TIR}$ to L$_\mathrm{FIR}$ ratio of 0.6.