Prospects for redshifted 21 cm observations of quasar HII regions

The introduction of low-frequency radio arrays over the coming decade is expected to revolutionize the study of the reionization epoch. Observation of the contrast in redshifted 21 cm emission between a large H (II) region and the surrounding neutral IGM will be the simplest and most easily interpreted signature. We assess the sensitivity of three generations of planned low-frequency arrays to quasar-generated H (II) regions. We find that an instrument such as the planned Mileura Widefield Array Low Frequency Demonstrator (LFD) will be able to obtain good signal-to-noise ratios on H (II) regions around the most luminous quasars and determine some gross geometric properties, e. g., whether the H (II) region is spherical or conical. A hypothetical follow-up instrument with 10 times the collecting area of the LFD (MWA-5000) will be capable of mapping the detailed geometry of H (II) regions, while the Square Kilometer Array (SKA) will be capable of detecting very narrow spectral features, as well as the sharpness of the H (II) region boundary. The SKA will most likely be limited by irreducible noise from fluctuations in the IGM itself. The MWA-5000 ( and even the LFD under favorable circumstances) will discover serendipitous H (II) regions in wide-field observations. We estimate the number of H (II) regions that are expected to be generated by quasars based on the observed number counts of quasars at z similar to 6. Assuming a late reionization at z similar to 6, we find that there should be several tens of quasar H (II) regions larger than 4 Mpc at z similar to 6-8 per field of view. Identification of H (II) regions in forthcoming 21 cm surveys can guide a search for bright galaxies in the middle of these regions. Most of the discovered galaxies would be the massive hosts of dormant quasars that left behind fossil H (II) cavities that persisted long after the quasar emission ended, owing to the long recombination time of intergalactic hydrogen. A snapshot survey of candidate H (II) regions selected in redshifted 21 cm image cubes may prove to be the most efficient method for finding very high redshift quasars and galaxies.