Erratum: “On the Normalization of the Cosmic Star Formation History” (ApJ, 651, 142 [20067])

Abstract

The stellar mass density data shown in Figure 5 of our original paper are in error; a corrected version of the figure is given here. The confidence regions arising from the integration over the star formation history (SFH) are correct, and the original figure caption remains correct. The description given in § 4 regarding converting between different initial mass functions (IMFs) corresponds to a factor of A(1-R1)/(1-R2), where A is the IMF-dependent SFR conversion factor, and R1 and R2 are the recycling fractions for the IMFs being converted to and from, respectively. This is the correct factor to use in the conversion of the integrated SFH, as both the SFR conversion factor and the ratio of the mass-lockup factors are required. When converting the measured stellar mass density data from the literature, however, only the mass-lockup ratio (1-R1)/(1-R2) is necessary. The consequence is that the data points in Figure 5 are erroneously low by a factor A. For conversions from the Salpeter (1955) IMF (R2=0.72), the SalA IMF has A=0.77 and R1=0.40, and the BG IMF has A=0.50 and R1=0.56. The discussion in § 5.1 of the discrepancy between the stellar mass density measurements and the integrated SFH is largely superfluous as a consequence of this correction, although the discussion of the metal mass densities is unaffected. In particular, we note that using the BG IMF for the SFH integral (as indicated by K. Glazebrook et al. (Nature, 430, 181 [2004]) reproduces the observed stellar mass density at z=0, and up to z