Diffuse Galactic emission at low frequencies is a major contaminant for studies of redshifted $21$ cm line studies. Removal of these foregrounds is essential for exploiting the signal from neutral hydrogen at high redshifts. Analysis of foregrounds and its characteristics is thus of utmost importance. It is customary to test efficacy of foreground removal techniques using simulated foregrounds. Most simulations assume that the distribution of the foreground signal is a Gaussian random field. In this work we test this assumption by computing the binned bispectrum for the all-sky $408$ MHz map. This is done by applying different brightness temperature ($T$) thresholds in order to assess whether the cooler parts of the sky have different characteristics. We find that regions with a low brightness temperature $T < 25$ K indeed have smaller departures from a Gaussian distribution. Therefore, these regions of the sky are ideal for future H{sc i} intensity mapping surveys.
Download