Limits on the star formation rates of z>2 damped Ly-alpha systems from H-alpha spectroscopy

Andrew J. Bunker, Stephen J. Warren, D. L. Clements, Gerard M. Williger, Paul C. Hewett

Published 1999-06-10Version 1

We present the results of a long-slit K-band spectroscopic search for H-alpha emission from eight damped Ly-alpha absorbers (DLAs) at z>2 with the goal of measuring the star-formation rates in these systems. For each system we searched for compact sources of H-alpha emission within a solid angle 11x2.5arcsec^2 (44x10 kpc^2/h^2, for q_0=0.5, H_0=100 h km/s/Mpc). No H-alpha emission was detected above 3-sigma limits in the range (6.5 - 16)x1E-20 W/m^2, equivalent to star formation rates of (5.6 - 18) M_sun/yr/h^2, for a standard IMF, assuming the lines are spectrally unresolved (<650 km/s FWHM). We compare these results against the predictions of the models of Pei & Fall (1995) of the global history of star formation, under two different simplifying hypotheses: i.) the space density of DLAs at z=2.3 is equal to the space density of spiral galaxies today (implying DLA disks were larger in the past, the `large-disk' hypothesis); ii.) the sizes of DLAs at z=2.3 were the same as the gas sizes of spiral galaxies today (implying DLA disks were more common in the past, the `hierarchical' hypothesis). Compared to the previous most sensitive spectroscopic search our sample is twice as large, our limits are a factor greater than two deeper, and the solid angle surveyed is over three times as great. Despite this our results are not in conflict with either the large-disk hypothesis, because of the limited solid angle covered by the slit, or the hierarchical hypothesis, because of the limited sensitivity.