Reconstructing the nature of the first cosmic sources from the anisotropic 21-cm signal

The redshifted 21-cm background is expected to be a powerful probe of the early Universe, carrying both cosmological and astrophysical information from a wide range of redshifts. In particular, the power spectrum of fluctuations in the 21-cm brightness temperature is anisotropic due to the line-of-sight velocity gradient, which in principle allows for a simple extraction of this information in the limit of linear fluctuations. However, recent numerical studies suggest that the 21-cm signal is actually rather complex, and its analysis likely depends on detailed model fitting. We present the first realistic simulation of the anisotropic 21-cm power spectrum over a wide period of early cosmic history. We show that on observable scales, the anisotropy is large and thus measurable at most redshifts, and its form tracks the evolution of 21-cm fluctuations as they are produced early on by Lyman-a radiation from stars, then switch to X-ray radiation from early heating sources, and finally to ionizing radiation from stars. In particular, we predict a redshift window during cosmic heating (at z ~ 15), when the anisotropy is small, during which the shape of the 21-cm power spectrum on large scales is determined directly by the average radial distribution of the flux from X-ray sources. This makes possible a model-independent reconstruction of the X-ray spectrum of the earliest sources of cosmic heating.