Resolving the molecular gas emission of the z~2.5-2.8 starburst galaxies SPT0125-47 and SPT 2134-50

TL;DR

Dusty star-forming galaxies at $z \sim 2.5-2.8$ power the cosmic star-formation peak, and this study uses high-resolution ALMA CO$(3-2)$ observations of two strongly lensed DSFGs, SPT0125-47 and SPT2134-50, to resolve their molecular gas morphologies and kinematics. By applying parametric and nonparametric lens modeling with PyAutoLens and by splitting the line into red and blue bins to probe differential lensing, the authors find that each source is well described by a single Sérsic profile in the source plane and derive robust line magnifications of $\mu_{CO(3-2)} \approx 10.7$ for SPT0125-47 and $\approx 7.6$ for SPT2134-50. They detect a significant differential magnification across the line, and while a tentative velocity gradient is observed, there is no strong evidence for clumpy structure or ongoing mergers; the very high SFRs and short depletion times imply a dramatic evolutionary phase, possibly triggered by a recent interaction or by disks settling. The work highlights both the power and the challenges of pixelized source-plane reconstructions in lensed DSFGs and underscores the need for deeper, brighter-line observations to robustly constrain high-$z$ ISM kinematics.

Abstract

The comoving cosmic star formation rate density peaks at z~2-3, with dusty star-forming galaxies being significant contributors to this peak. These galaxies are characterized by their high star formation rates and substantial infrared luminosities. The formation mechanisms remain an open question for these galaxies, particularly with respect to how such intense levels of star formation are triggered and maintained. We aim to resolve CO(3-2) emission toward two strongly lensed galaxies, SPT0125-47 and SPT2134-50, at z~2.5-2.8 to determine their morphology and physical properties. We used high-resolution ALMA band 3 observations of CO(3-2) emission toward both sources to investigate their properties. We performed parametric and nonparametric lens modeling using the publicly available lens modeling software PyAutoLens. We divided the CO(3-2) emission line into two bins corresponding to the red and blue portions of the emission line and nonparametrically modeled the source plane emission for both bins. We performed a basic analysis of the morphology and kinematics in the source plane using nonparametric lens modeling of the red and blue bins. We found tentative evidence of a velocity gradient across both sources and no evidence of any clumpy structure, companions, or ongoing mergers. The previously calculated high star formation rates and low depletion times of both SPT0125-47 and SPT2134-50 suggest that these galaxies are undergoing a dramatic phase in their evolution. Given the lack of evidence of ongoing interactions or mergers in our source plane models, we suggest that the intense star formation was triggered by a recent interaction and/or merger. We also consider the possibility that these galaxies might be in the process of settling into disks.

Figures (8)

• Figure 1: Continuum image of SPT 0125-47 (top) and SPT 2134-50 (bottom). The contours are shown at $-3, -2, 3, 4, 5\sigma$ levels. The synthesized beam is shown in the bottom left of each image. The black annulus shows the regions used to extract the dust continuum emission.
• Figure 2: Moment-0 and moment-1 maps of the CO(3--2) emission for SPT 0125-47 (top) and SPT 2134-50 (bottom). The first column shows the unmasked moment-0 map, the second column shows the moment-0 map masked to show only values above $3\sigma$, and the third column shows the moment-1 map. The contours in the first two columns are shown at $-3, -2, 3, 4, 5, 6, 7, 8, 9, 10\sigma$ levels. The synthesized beam is shown in the bottom left of each image. The black annulus in the unmasked moment-0 map for each source shows where the spectrum was extracted from.
• Figure 3: Spectra of the CO(3--2) toward SPT 0125-47 (left) and SPT 2134-50 (right). In both cases, the spectrum is shown in the top panel, and the residuals from the Gaussian fit are shown in the bottom panel. The dashed blue line shows the two Gaussian fit to the spectra. The dashed gray line in the top panel and the shaded gray region in the lower panel indicate the per-channel rms. Additionally, the top axis in the top panel of each spectrum displays the corresponding frequency.
• Figure 4: Parametric (rows 1 and 3) and pixelized (rows 2 and 4) lens modeling of the CO(3--2) emission detected in SPT 0125-47 (top two rows) and SPT 2134-50 (bottom two rows). The first column displays the dirty image generated by PyAutoLens, with contours at $-3, 3, 4, 5, 6, 7, 8, 9, 10\sigma$ levels. Note: this is not a cleaned image, so structures may differ slightly from those in cleaned images. The second column presents the dirty model image from PyAutoLens, also with contours at $-3, 3, 4, 5, 6, 7, 8, 9, 10\sigma$ levels. The third column shows the dirty residual image produced by PyAutoLens, with contours at $-3, 3, 4, 5\sigma$ levels. The fourth column illustrates the image plane emission parametric/pixelized model of the data, produced by PyAutoLens, with the black line representing the critical line. The fifth column shows the source plane emission parametric+pixelized model of the data from PyAutoLens, with the black line indicating the caustic line. All images are centered around the ALMA phase center.
• Figure 5: Source plane intensity (left), intensity error maps (middle), and S/N maps (i.e., intensity divided by intensity error, right) for SPT 0125-47 and SPT 2134-50. The gray polygons show the Voronoi mesh used for the pixelized reconstruction and the black line shows the caustic line. Both the intensity and intensity error maps are normalized, where the error maps have been normalized to the maximum of the intensity map, meaning that the value can be seen as a percentage error per pixel. The S/N map provides a measure of the significance of the reconstructed emission.