Tomography of the reionization epoch with multifrequency CMB observations

We study the constraints that future multifirequency cosmic microwave background (CMB) experiments will be able to set on the metal enrichment history of the intergalactic medium at the epoch of reionization. We forecast the signal-to-noise ratio for the detection of the signal introduced into the CMB by resonant scattering off metals at the end of the cosmic dark ages. We take into account systematics associated with cross-channel calibration, errors in reconstruction of the point-spread function, and inaccurate foreground removal. We develop an algorithm to optimally extract the signal generated by metals during reionization and to accurately remove the contamination due to the thermal Sunyaev-Zel'dovich effect. Although demanding levels of foreground characterization and control of systematics are required, they are very distinct from those encountered in H I 21 cm studies and CMB polarization, and this fact encourages the study of resonant scattering off metals as an alternative way of conducting tomography of the reionization epoch. A realistic experiment, looking at clean regions of the sky, can detect changes of 3%-12% (95% confidence level) in the O III abundance (with respect to its solar value) in the redshift range z = 12-22 for reionization redshift z(re) > 10. However, for z(re) < 10 one can only set upper limits on N II abundance increments of similar to 60% solar in the redshift range z = 5.5-9 (95% c.l.). These constraints assume that cross-channel calibration is accurate to I part in 104, which constitutes the most critical technical requirement of this method but is still achievable with current technology.