CANDELS GALAXY PARAMETERS CATALOG

Osborne & Salim

Motivation

The estimation of robust SFRs for galaxies at high redshifts (i.e., z > 1) from SED fitting is a challenge given the degeneracy between dust, metallicity, and age. M/FIR observational data can break this degeneracy, but ARE often unavailable, leading to difficulty in determining which set of SED models is most effective at breaking the degeneracy, particularly when fitting only to UV-optical-NIR photometry.

Study

In Osborne & Salim 2024, we identified a sample of galaxies from the HST CANDELS survey that possess both abundant UV-optical-NIR photometry as well as robust detections in the MIR from Spitzer. We adopt the Spitzer-inferred total IR luminosity (LTIR), estimated by the BOSA templates, as a ground truth. We fit several sets of SED models to the UV-optical-NIR photometry to estimate a dust luminosity, which should agree with the LTIR. Using this method, we identify which models produce the best agreement between the LTIR and the dust luminosity. The best models assume a flexible dust attenuation with variable slope and UV bump, a fixed stellar metallicity, and a semi-flexible star formation history. We then apply this modeling to all the galaxies, irrespective of IR detection.

Catalog Description

The catalog includes 63,266 total galaxies from GOODS-S, UDS, COSMOS and EGS fields. Galaxies were selected to lie in the range 0.7 < z < 2.3, have reliable SED fits (reduced chi squared < 10), and have a difference in redshift between CANDELS and 3DHST of less than 0.4. No mass, sSFR, or IR photometric cuts have been applied. All parameters are from the UV-optical-NIR SED fitting using our preferred models. We include the stellar mass in log solar masses, the SFR in log solar masses per year, the FUV and V band attenuations in magnitudes, and the inferred dust luminosity in log solar luminosities. Errors for these quantities are also provided. We also include "best" redshift from CANDELS (MAST HLSP v1, as of Sep 2018), the galaxy identifier corresponding to line number of v1 data files at MAST, and the field name. Identifier is unique to field, but not overall. Chabrier IMF is assumed.