Death by Impact: Evidence for Merger-Driven Quenching in a Collisional Ring Galaxy at Cosmic Noon

TL;DR

The study exploits a high-redshift collisional ring galaxy at $z=1.61$ to directly link a major merger event with instantaneous quenching and starburst activity within the same system. By combining JWST and HST imaging with Keck/MOSFIRE spectroscopy, the authors reconstruct a precise collision timeline ($t_0 \sim 50$–$150$ Myr ago) from ring dynamics and match it to region-specific SFHs: a vigorous starburst in the Host and rapid quenching in the Bullet. They argue merger-driven turbulence and instabilities likely suppress star formation in the Bullet, while AGN feedback from the Host — the Dragon Effect — may externally aid this quenching. This work provides a rare, direct observational link between interaction geometry, star formation responses, and feedback processes at cosmic noon, with implications for how mergers regulate massive galaxies in the early universe.

Abstract

The role of interactions and mergers in the rapid quenching of massive galaxies in the early Universe remains uncertain, largely due to the difficulty of directly linking mergers to quenching. Collisional ring galaxies provide a unique opportunity, as their morphology allows precise dating of the interaction, which can then be compared to quenching timescales inferred from star formation histories. We study a gravitationally bound system at $z=1.61$ in the UDS field, composed of a Host galaxy ($M_\star = 10^{11.4} M_\odot$) with a collisional ring and an X-ray AGN, and the Bullet galaxy ($M_\star = 10^{11.2} M_\odot$), located at a projected distance of $\sim 8$ kpc. Combining JWST and HST imaging with Keck/MOSFIRE spectroscopy, we find compelling evidence for an ongoing starburst in the Host concurrent with rapid quenching in the Bullet. The ring, $\sim 20$ kpc in diameter, is expanding at $127^{+72}_{-29}$ km s$^{-1}$, implying the galaxies first collided 47--96 Myr ago. This timeline is consistent with the Host's current starburst and the Bullet's sudden quenching, strongly suggesting both phenomena were triggered by the interaction. Crucially, the Bullet shows no evidence of a preceding starburst, ruling out rapid gas consumption as the primary quenching channel. Instead, we suggest that merger-driven processes -- such as enhanced turbulence and disk instabilities -- may have suppressed star formation. An additional possibility, which we term the ``Dragon Effect,'' is that AGN-driven outflows from the Host disrupted the Bullet's low-density molecular gas, thereby preventing efficient star formation and accelerating quenching.

Figures (9)

• Figure 1: JWST/NIRCam images of the system. The upper panel shows the F200W-band image with labels indicating prominent components for which we have kinematic measurements. The lower panel presents a composite color image using the F090W, F115W, F150W, and F200W NIRCam bands. The tidal tail, highlighted in the inset, is more evident in the F444W filter.
• Figure 2: Star formation histories of different regions in the merging system, derived from JWST/NIRCam and HST photometric data and modeled using prospector. The inferred dust attenuation and stellar mass (in M$_\odot$) for each region are indicated within the corresponding panel. The regions are color-coded and overlaid on a NIRCam F200W cutout (left image), with the corresponding RGB image (right image) providing a visual reference for their locations within the system. Each region is connected with an arrow to its SFH panel, which is placed in a color-coded box matching the region’s color to aid visual interpretation. Note that regions are identified using segmentation map as described in Sec. \ref{['sec:photometry']}. For visualization purposes, we exclude the SFHs of two regions that are heavily contaminated by neighboring sources, while, their $\mathrm{M_\star}$ are included in the total values reported in Table \ref{['tab:Mass_RA_Dec']}.
• Figure 3: Keck/MOSFIRE Y and H band spectra extracted from three spatial regions along the slit: the Bullet and Upper Ring (top), the Host galaxy and Inner Ring (middle), and the Lower Ring (bottom), as shown in the image on the left. Black squares indicate the observed data with associated uncertainties. Emission lines in the H band are modeled using lmfit, while spectral features in the Y band are fit using pPXF. The dashed lines mark the observed wavelengths of H$\alpha$ and Ca ii H at the redshift of the Host galaxy, serving as a visual reference for identifying relative blueshifts or redshifts in other regions.
• Figure 4: The best-fit SED models of the Host and Bullet galaxies, derived using Prospector. The fits incorporate photometric data from HST and JWST/NIRCam (top panels), along with Y-band spectra from MOSFIRE (bottom panels), as described in Sec. \ref{['sec:SFH']}. The wavelength range covered by the spectra is magnified for both objects and overlaid with the corresponding best-fit spectral models. Grey shaded regions in the zoomed-in panels indicate masked intervals containing $\mathrm{H{\delta}}$ emission lines.
• Figure 5: Star formation histories of the Bullet (top panel) and the Host (bottom panel) galaxies derived using JWST/NIRCam and HST photometry, along with Y band MOSFIRE spectroscopy, modeled with Prospector. In each panel, the SFH inferred from photometry alone is shown in gray, while the combined photometry plus Y band spectroscopy result is shown in dark red. Shaded regions indicate the 1$\sigma$ confidence intervals around the maximum a posteriori estimates in each age bin.