high-z galaxies

The first few hundred million years, one spectrum at a time

Before JWST, the first few hundred million years of cosmic history were essentially a blank page. They are not blank any more — but what has been written there is stranger than anyone expected. Galaxies at z > 10 are more numerous and more UV-luminous than pre-launch models allowed, and the explanations on offer range from bursty star formation, to top-heavy initial mass functions, to almost dust-free lines of sight.

My work here is mostly spectroscopic: taking the candidates that imaging surveys throw up and either confirming them or killing them, then using NIRSpec to measure what those galaxies are actually made of — their stellar populations, their ionized gas, and how quickly they turn gas into stars. I work on this through CAPERS in the blank fields and GLIMPSE, the deepest JWST imaging of a lensing cluster, at the faint end.

CAPERS: Two UV-bright Galaxies at z > 10

Kokorev et al. (2025), ApJL

The first results out of CAPERS, using just 14% of the planned data: two spectroscopic confirmations at z = 10.562 and z = 11.013 in PRIMER-UDS. One of them shows strong Hγ (EW0 ~ 70 Å), which is a direct fingerprint of a recent burst. Putting these together with everything else confirmed at z > 10, most bright galaxies at these redshifts appear to be caught mid-burst, with a median SFR10/SFR100 ~ 2 — a real, if partial, answer to why the bright end of the UV luminosity function is so overpopulated.

Burstiness, SFR10/SFR100, against the age of the Universe. The CAPERS galaxies (red) sit alongside every other spectroscopically confirmed bright galaxy at z > 10; the population mean sits above unity, i.e. most of them are caught on the way up.

A Glimpse of the New Redshift Frontier through AS1063

Kokorev et al. (2025), ApJL

Two galaxy candidates at 15.7 < z < 16.4, found in the ultra-deep GLIMPSE imaging of Abell S1063. Lensing by the cluster pushes us down to MUV = −17, roughly two magnitudes fainter than blank-field surveys reach at these redshifts, and both objects sit right at that limit with very blue slopes (β ≃ −2.87). Their number density is in clear tension with pre-JWST predictions, extending the galaxy over-abundance from z ~ 10 out to z ~ 17, and their star formation histories make them plausible progenitors of the UV-bright systems JWST now routinely finds at z = 10–14.

The GLIMPSE field on Abell S1063 — the deepest JWST imaging of a lensing cluster to date, and where both z ~ 16 candidates were found.