We present new results on the evolution of the cosmic star formation rate as a function of stellar mass in the Subaru/XMM-Newton Deep Survey–Ultra Deep Survey field. We make use of narrow-band-selected emission line galaxies in four redshift slices between z = 1.46 and 0.63, and compute stellar masses by fitting a series of templates to recreate each galaxy’s star formation history. We determine mass-binned luminosity functions in each redshift slice, and derive the star formation rate density (ρSFR) as a function of mass using the [O III] or [O II] emission lines. We calculate dust extinction and metallicity as a function of stellar mass, and investigate the effect of these corrections on the shape of the overall ρSFR(M).
ρSFR as a function of mass for redshift slices at z = 0.63, 0.83, 1.19 and 1.46. ρSFR is estimated via Kennicutt (1998) and 1 mag extinction at Hα (top two panels), Kewley et al. (2004) and dust extinction as a function of stellar mass (central two panels), and the fully corrected ρSFR including dust an metallicity corrections (lower two panels). The left-hand column of figures present results for an integration to the limit of the data in each redshift slice, and the right-hand column of figures presents an integration to 1 M⊙yr−1.
We find that both these corrections are crucial for determining the shape of the ρSFR(M), and its evolution with redshift. The fully corrected ρSFR(M) is a relatively flat distribution, with the normalization moving towards lower values of ρSFR with increasing cosmic time/decreasing redshift, and requiring star formation to be truncated across all masses studied here. The peak of ρSFR(M) is found in the 1010.5 < M⊙ < 1011.0mass bin at z = 1.46. In the lower redshift slices, the location of the peak is less certain; however, low-mass galaxies in the range 107.0 < M⊙ < 108.0 play an important part in the overall ρSFR(M) out to at least z ∼ 1.2.
Further details can be found in Drake, A.B. et al. 2015, MNRAS, 454, 2015