THE Hα LUMINOSITY FUNCTION AND STAR FORMATION RATE VOLUME DENSITY AT z=0.8 FROM THE NEWFIRM Hα SURVEY

We present new measurements of the H alpha luminosity function (LF) and star formation rate (SFR) volume density for galaxies at z similar to 0.8. Our analysis is based on 1.18 mu m narrowband data from the NEWFIRM H alpha (NewH alpha) Survey, a comprehensive program designed to capture deep samples of intermediate redshift emission-line galaxies using narrowband imaging in the near-infrared. The combination of depth (approximate to 1.9 x 10(-17) erg s(-1) cm(-2) in H alpha at 3 sigma) and areal coverage (0.82 deg(2)) of the 1.18 mu m observations complements other recent H alpha studies at similar redshifts, and enables us to minimize the impact of cosmic variance and place robust constraints on the shape of the LF. The present sample contains 818 NB118 excess objects, 394 of which are selected as H alpha emitters. Optical spectroscopy has been obtained for 62% of the NB118 excess objects. Empirical optical broadband color classification is used to sort the remainder of the sample. A comparison of the LFs constructed for the four individual fields covered by the observations reveals significant cosmic variance, emphasizing that multiple, widely separated observations are required for such analyses. The dust-corrected LF is well described by a Schechter function with L-* = 10(43.00 +/- 0.52) erg s(-1), Phi(*) = 10(-3.20 +/- 0.54) Mpc(-3), and alpha = -1.6 +/- 0.19. We compare our H alpha LF and SFR density to those at z less than or similar to 1, and find a rise in the SFR density alpha (1 + z)(3.4), which we attribute to significant L-* evolution. Our H alpha SFR density of 10(-1.00 +/- 0.18) M-circle dot yr(-1) Mpc(-3) is consistent with UV and [O-II] measurements at z similar to 1. We discuss how these results compare to other H alpha surveys at z similar to 0.8, and find that the different methods used to determine survey completeness can lead to inconsistent results. This suggests that future surveys probing fainter luminosities are needed, and more rigorous methods of estimating the completeness should be adopted as standard procedure (for example, with simulations which try to simultaneously reproduce the observed H alpha LF and equivalent width distributions).