Seminar in Extragalactic Astronomy

The redshifted 21-cm signal coming from the HI of the Epoch of Reionization (EoR; z~6-14) will be one of the most interesting beacon of observational cosmology once detected. Low frequency radio arrays have been and are being built all over the world to detect this signal, for example, LOFAR, MWA, PAPER, PaST, HERA, and the planned SKA. Like most cosmological observations, the greatest difficulty in its detection is removing the foreground radiations. After removing the point-like extragalactic foregrounds, we are left with the diffuse Galactic foreground which is polarized because it is created through the synchrotron mechanism. The total intensity of the diffuse foreground can be removed to a large extent utilizing its spectral smoothness. But its polarized counterpart is not smooth along frequency. We would not care about this polarization, because the EoR signal will be detected in total intensity, unless it could contaminate the total intensity. The errors of the telescope can do exactly this – a small portion of diffuse polarized foreground can leak into the total intensity mainly because of the primary beam of an antenna.

In this thesis we have presented a pipeline that can simulate realistic EoR observations taking into account the direction independent (DI) and dependent (DD) systematic effects of LOFAR. We have tested various DI and calibration errors, but finally focused on one DD effect, the primary beam. We have predicted the level of leakage caused by the LOFAR primary beam, and compared it with a simulated EoR signal. We have also tested various leakage correction strategies.

We have found that the leakage, if uncorrected, is comparable in amplitude to the expected EoR signal toward some directions, and the primary beam of LOFAR used for this prediction has an error of around 10%. The fractional leakage is almost constant over the instrumental k-space, i. e. the 2D cylindrical power spectrum, and it is higher in the fields that remain farther away from zenith. We have found a potential 'EoR window' above the chromatic side-lobe wedge of the instrumental k-space where the EoR signal dominates leakage, and this window takes up almost the whole k-space if 70% leakage is removed. The fractional leakage does not vary significantly as the field of view is increased, because the increment of leakage with distance from the phase centre is balanced by the decrement of power. These results show that leakage will not be one of the primary limiting factors in detecting the EoR signal with LOFAR, but it still has the potential of becoming a major issue for the more sensitive future telescopes like SKA.

close

abstract

abstract

abstract

abstract