[Previous] | [Session 123] | [Next]
A. M. Wolfe (CASS, University of California, San Diego)
The ratio of the N(C II*) to N(H I) column densities in damped lyman alpha systems is proportional to the 158 \mum cooling rate per H atom, lc. In a steady state this equals the heating rate of the neutral gas, since 158 \mu m emission dominates the cooling in the cold neutral medium which must be present in these objects. Assuming the heating is driven by star formation, presumably via photo-ejection of electrons from grains, we compute the average star formation rate per unit area. From our collection of C II* 1335 absorption and damped lyman alpha lines we find loglc = -26.6±0.2 ergs s-1 H-1. This is about 30 times lower than the average rate deduced for the ISM. Since the heating rate \Gammad \propto the energy density of UV starlight\timesdust-to-gas ratio, and because the metallicity in damped systems is about 30 times lower than solar then we can account for the observed heating rate provided the dust-to-gas ratio equals the metallicity and the energy density of UV starlight is comparable to that in the ISM. We then compute the global star formation rate by noting that the star formation rate per unit area is \propto the energy density of UV starlgiht and that the area times the number of damped systems per unit comoving volume is uniquely determined by the incidence of damped systems per unit redshift interval. The result at z \approx 2.5 is comparable to that determined by completely independent methods that depend on UV emission from flux limited samples of Lyman Break Galaxies. We discuss the cosmological implications of deducing the star formation rate for the bulk of the protogalactic mass distribution at high redshifts.