Foreground modelling via Gaussian process regression: an application to HERA data

The key challenge in the observation of the redshifted 21-cm signal from cosmic reionization is its separation from the much brighter foreground emission. Such separation relies on the different spectral properties of the two components, however, the foreground intrinsic spectral smoothness is often corrupted by the instrumental response making such separation much harder. In this paper, we apply Gaussian Process Regression to model the foreground emission using $\sim 2$ hours of data from the Hydrogen Epoch of Reionization Array. We find that a simple co-variance model with three components matches the data well, the residual being noise like. These consist of an "intrinsic" and instrumentally corrupted component with a coherence-scale of 20 MHz and 2.4 MHz respectively (dominating the line of sight power spectrum over scales $k_{\parallel} \le 0.2 \, $hcMpc$^{-1}$) and a baseline dependent periodic signal with a period of $\sim$ 1 MHz (dominating over $k_{\parallel} \sim 0.4 - 0.8 \, $hcMpc$^{-1}$) which should be distinguishable from the 21-cm EoR signal whose typical coherence-scales is $\sim 0.8$ MHz.