Tentative rotation in a galaxy at z$\sim$14 with ALMA

TL;DR

This study re-analyses ALMA observations of the [OIII] 88 μm line in JADES-GS-z14-0 (z ≈ 14.18) to test for a velocity gradient and derive a dynamical mass. Using pixel-wise Gaussian fits, aperture-based spectra, and spectro-astrometry in image and uv spaces, the authors find tentative evidence for rotation. KinMS modelling of a thin exponential disc yields a dynamical mass of log10(M_dyn/M_sun)=9.4^{+0.8}_{-0.4}, a rotation speed of V_rot ≈ 164 km s^-1, and a 3σ upper limit on dispersion σ_v < 40 km s^-1, implying V_rot/σ_v > 2.5; however, M_dyn and inclination remain degenerate. Comparisons with SERRA Amaryllis mock observations indicate that a gradient can be detected even at low resolution, but higher-resolution ALMA/JWST-IFS data are essential to break degeneracies and precisely determine the kinematics and mass budget, with implications for early disc formation.

Abstract

We re-analysed ALMA observations of the [OIII]$λ$88$μ$m emission line in JADES-GS-z14.0, so far the most distant spectroscopically confirmed galaxy at z=14.18. Our analysis shows a tentative detection of a velocity gradient of [OIII]$λ$88$μ$m using three independent tests: 1) construction of moment maps; 2) extraction of integrated spectra from a grid of apertures; and 3) spectro-astrometry in both the image and uv planes. We performed kinematical fitting using the KinMS code and estimated a dynamical mass of log$_{10}$(M$_{\rm dyn}$/$\rm M_\odot$)= 9.4$^{+0.8}_{-0.4}$, with the bulk of the uncertainties due to the degeneracy between dynamical mass and inclination. We measure an upper limit on the velocity dispersion ($σ_{v}$) of $<40~$ km/s~which results in an estimate of V$_{\rm rot}/σ>$ 2.5. This result, if confirmed with higher-resolution observations, would imply that kinematically cold discs are already in place at $z\sim14$. Comparison with mock observations from the SERRA cosmological simulations confirms that even low-resolution observations are capable of detecting a velocity gradient in $z>10$ galaxies as compact as JADES-GS-z14.0. This work shows that deeper ALMA or JWST/NIRSpec IFS observations with high spatial resolution will be able to estimate an accurate dynamical mass for JADES-GS-z14.0, providing an upper limit to the stellar mass of this over-luminous galaxy.

Figures (7)

• Figure 1: Tentative detection of the velocity gradient in JADES-GS-z14-0. Left top and middle panels: FWHM and velocity maps of the [O][iii][][88$\mu$m] emission line from Gaussian fitting. We used a 3$\sigma$ cutoff to create these maps. The red squares indicate the apertures used to extract spectra on the right. Bottom left: NIRCam F200W image tracing the rest-frame UV emission. The red and blue contours show moment-0 maps extracted from [$-50$, 0] km s$^{-1}$ and [0, 50] km s$^{-1}$. The cyan-hatched ellipse shows the ALMA beam size. We observe 0.3$\pm$0.06 ȯffset between the red and blue centroids. Right panels: [O][iii][][88$\mu$m] extracted from the square apertures indicated on velocity and FWHM maps. The data and the best-fit model are shown as blue and red lines, while the black dotted line indicates the flux uncertainties. The extracted spectra confirm the derived velocity maps.
• Figure 2: Spectro-astrometric analysis in the uv-plane. We show the posterior distribution of the spatial offset of the red and blue sides of the emission line from the centre of JADES-GS-z14-0 ([0, 50] km s$^{-1}$ and [$-50$, 0] km s$^{-1}$). We see a spatial offset between the red and blue parts of the emission line, confirming the presence of the velocity gradient.
• Figure 3: Moment maps created from the observed data (top row), maps from the best fit kinematical model of KinMS (centre row), and the normalised residuals (bottom row). The left and right columns show moment-1 and moment-2 maps.
• Figure 4: V$_{\rm rot}$/$\sigma_v$ versus redshift for observations and simulations. As we are unable to constrain the velocity dispersion of JADES-GS-z14-0 and in the low-resolution mock observations of Amaryllis from SERRA simulations, we only quote a lower limit on V$_{\rm rot}$/$\sigma_{v}$ for these. We compare JADES-GS-z14-0 to a compilation from the literature: JWST/NIRSpec, Illustris TNG Pillepich:2019, SERRA Kohandel24, ground-based IFU Wisnioski15forster_schreiber_sinszc-sinf_2018, MOSFIRE price_mosdef_2019, GN-z11 Xu24_gnz11kin, ALMA z$\sim$6 Parlanti23_alm, NIRSpec/MSA degraaff24, ALMA z$\sim$4 Rizzo21, REBELS-25 Rowland24, MACS1149-JD1 Tokuoka22 and MACS0416-Y3 Li+23.
• Figure 5: Mock observations of JADES-GS-z14-0 analogue in the SERRA simulations - "Amaryllis". The top row shows flux, velocity and FWHM maps of mock low-resolution ALMA observations (0.9), while the bottom row shows mock high-resolution observations (0.15). The simulations show that a rotating galaxy shows a small velocity gradient ($<40$ km s$^{-1}$) in the low-resolution data, similar to our observations. The simulations show that a rotating (thin) disc galaxy with a size and mass as JADES-GS-z14-0, and observed under similar conditions, should show a similar velocity gradient to the one obtained. The increased FWHM in the center of high-resolution observations can be due to the beam smearing of the rotation.