First Star Signature in Infrared Background Anisotropies

Abstract

Recent cosmic microwave background anisotropy results from the Wilkinson Microwave Anisotropy Probe suggest that the universe was reionized at a redshift around 20 with an optical depth for Thomson scattering of 0.17 ± 0.04. Such an early reionization could arise through the ionizing radiation emitted by metal-free Population III stars at redshifts of 10 and higher. We discuss infrared background (IRB) surface brightness spatial fluctuations from such a generation of early star formation. We show that the spatial clustering of these stars at tens of arcminute scales generates a contribution to the angular power spectrum of the IRB anisotropies at the same angular scales. This excess can be potentially detected when resolved foreground galaxies out to a redshift of a few are removed from the clustering analysis. We do not expect faint galaxies at redshifts of ~3, with magnitudes less than 20 in the K band, to be a source of strong confusion, since the fractional contribution to the IRB from these galaxies is at a level less than a few percent, while the expected contribution from first stars can be 50% or more. Additionally, assuming a Population III stellar spectrum, we suggest that the clustering excess related to the first generation of stars can be separated from brightness fluctuations resulting from other foreground sources and galaxies using multifrequency observations in the wavelength range of ~1-5 μm. In addition to identifying the IR clustering associated with low-redshift galaxy population, the multifrequency data are essential to account for certain foreground contaminants such as zodiacal light, which, if varying spatially over degree scales, can be a significant source of confusion for the proposed study. Using various instruments, we study the extent to which spatial fluctuations of the IRB can be studied in the near future.