The motivation for flexible star-formation histories from spatially resolved scales within galaxies

Abstract

The estimation of galaxy stellar masses depends on the assumed prior of the star-formation history (SFH) and spatial scale of the analysis (spatially resolved versus integrated scales). In this paper, we connect the prescription of the SFH in the Spectral Energy Distribution (SED) fitting to spatially resolved scales ($\sim\mathrm{kpc}$) to shed light on the systematics involved when estimating stellar masses. Specifically, we fit the integrated photometry of $\sim970$ massive (log (M$_{\star}$/M$_{\odot}) = 9.8-11.6$), intermediate redshift ($z=0.5-2.0$) galaxies with $\texttt{Prospector}$, assuming both exponentially declining tau model and flexible SFHs. We complement these fits with the results of spatially resolved SFH estimates obtained by pixel-by-pixel SED fitting, which assume tau models for individual pixels. These spatially resolved SFHs show a large diversity in shapes, which can largely be accounted for by the flexible SFHs with $\texttt{Prospector}$. The differences in the stellar masses from those two approaches are overall in good agreement (average difference of $\sim 0.07$ dex). Contrarily, the simpler tau model SFHs typically miss the oldest episode of star formation, leading to an underestimation of the stellar mass by $\sim 0.3$ dex. We further compare the derived global specific star-formation rate (sSFR), the mass-weighted stellar age (t$_{50}$), and the star-formation timescale ($\tau_{\mathrm{SF}}$) obtained from the different SFH approaches. We conclude that the spatially resolved scales within galaxies motivate a flexible SFH on global scales to account for the diversity of SFHs and counteract the effects of outshining of older stellar populations by younger ones.