PhD defense : Rafael C. Arango Toro

“Star Formation Histories of Galaxies from Deep Multi-Color Images.”

Abstract:

The processes governing the assembly of stellar mass and the regulation of star formation in galaxies remain fundamental open questions in understanding the emergence of the observed diversity of different populations. This thesis leverages the COSMOS-Web survey, the largest JWST Cycle 1 program, to reconstruct non-parametric star formation histories (SFHs) for galaxies up to z ~ 12, using multi-wavelength data from JWST/NIRCam, MIRI, HST, and complementary ground-based observations.

Within this framework, I employ CIGALE to model complex, time-varying SFHs without assuming predefined functional forms, thereby capturing rapid variations associated with starburst and quenching phases. From these reconstructions, I trace galaxy migration across the stellar mass–SFR plane for ~300,000 galaxies up to z < 4, introducing migration vectors that quantify both the direction and the speed of their evolution. The analysis reveals a remarkably stable star-forming main sequence, punctuated by short-lived starburst excursions and mass- and epoch-dependent quenching pathways.

A major outcome of this work is the COSMOS2025 catalog, which provides stellar masses, SFRs, non-parametric SFHs, and dust properties for over 700,000 galaxies, a public legacy dataset for the community. The methodology has been validated using the HORIZON-AGN cosmological simulation to assess accuracy and potential biases.

By combining JWST observations, advanced SED modeling, and simulations, this thesis quantitatively traces how galaxies form, grow, and quench across cosmic history.
—
Jury:

Véronique BUAT – Chair, LAM, AMU
Giulia RODIGHIERO – Reviewer, University of Padova
Emanuele DADDI – Reviewer, AIM Paris-Saclay
Sandro TACCHELLA – Examiner, Kavli Institute for Cosmology, Cambridge
Hervé AUSSEL – Examiner, AIM Paris-Saclay
Olivier ILBERT – Thesis Supervisor, LAM, AMU
Laure CIESLA – Co-Supervisor, LAM, CNRS