Resolved spectroscopy of gravitationally lensed galaxies: global dynamics and star-forming clumps on ∼100 pc scales at 1 < z < 4

Abstract

We present adaptive optics-assisted integral field spectroscopy around the Hα or Hβ lines of 12 gravitationally lensed galaxies obtained with VLT/SINFONI, Keck/OSIRIS and Gemini/NIFS. We combine these data with previous observations and investigate the dynamics and star formation properties of 17 lensed galaxies at 1 < z < 4. Thanks to gravitational magnification of 1.4 − 90× by foreground clusters, effective spatial resolutions of 40−700 pc are achieved. The magnification also allows us to probe lower star formation rates and stellar masses than unlensed samples; our target galaxies feature dust-corrected SFRs derived from Hα or Hβ emission of ∼ 0.8 − 40M ⊙ yr −1 , and stellar masses M * ∼ 4 × 10 8 − 6 × 10 10 M ⊙ . All of the galaxies have velocity gradients, with 59% consistent with being rotating discs and a likely merger fraction of 29%, with the remaining 12% classed as 'undetermined.' We extract 50 star-forming clumps with sizes in the range 60 pc -1 kpc from the Hα (or Hβ) maps, and find that their surface brightnesses, Σ clump and their characteristic luminosities, L 0 , evolve to higher luminosities with redshift. We show that this evolution can be described by fragmentation on larger scales in gas-rich discs, and is likely to be driven by evolving gas fractions.

Paving the way for the JWST: witnessing globular cluster formation at z > 3

“…A still open question concerns the nature of the objects belonging to this luminosity domain: are we dealing with dwarf galaxies (e.g., Finlator et al 2017), Hii galaxies (e.g., Terlevich et al 2016), super star-clusters, or extremely compact star clusters or clumps? (i.e., with sizes of the order of a few tens of parsec, Bouwens et al 2016b;Kawamata et al 2015;Livermore et al 2015;Vanzella et al 2016a;Ellis et al 2001). In order to answer this question, we need to derive a few basic physical quantities such as the stellar mass, the star-formation rate, as well as the size of these systems.…”

A magnified view of galaxy formation

“…In contrast, circular rotation velocities vary significantly. Rotation velocities are on average higher for galaxies of larger size (Figure 3) and stellar mass [42]. Nonetheless random motions are significant with V rot /σ < 10 even in the largest galaxies at z 2, while random motions are dominant in many of the compact and lower-mass galaxies with V rot /σ < 1.…”

“…Nonetheless random motions are significant with V rot /σ < 10 even in the largest galaxies at z 2, while random motions are dominant in many of the compact and lower-mass galaxies with V rot /σ < 1. [42]. Left: circular rotation velocity as a function of half-light radius.…”

“…In the most massive galaxies, these clumps can reach ∼1 kpc in size such that they are marginally resolved [45,36,46]. Targeted studies of lensed galaxies, including Hα narrowband imaging with Hubble, have resolved clumps spanning over a decade in size and nearly three decades in luminosity [11,47,42] as shown in…”

“…While giant H II regions of similar size can be found in nearby galaxies, they are more frequent and typically have far higher star formation rate densities at higher redshifts. [42]. Lensed objects are color-coded according to their redshift.…”