Impact of Instrument Responses on the Detectability of One-point Statistics from Redshifted 21 cm Observations [CEA]

http://arxiv.org/abs/1610.06100


We study the impact of instrumental systematics on the variance, skewness, and kurtosis of redshifted 21 cm intensity fluctuation observations from the Epoch of Reionization. We simulate realistic 21 cm observations based on the Murchison Widefield Array (MWA) Phase I reionization experiment, using the array’s point spread function (PSF) and antenna beam patterns, full-sky 21 cm models, and the FHD imaging pipeline. We measure the observed redshift evolution of pixel probability density functions (PDF) and one-point statistics from the simulated maps, comparing them to the measurements derived from simpler simulations that represent the instrument PSFs with Gaussian kernels. We find that both methods yield statistics with similar trends with greater than 80% correlation. We perform additional simulations based on the Hydrogen Epoch of Reionization Array (HERA), using Gaussian kernels as the instrument PSFs, and study the effect of frequency binning on the statistics. We find that PSF smoothing and sampling variance from measuring the statistics over limited field of view dilute intrinsic features and add fluctuations to the statistics but reveal new detectable features. Observed kurtosis will increase when a few extremely high or low temperature regions are present in the maps. Frequency binning reduces the thermal uncertainty but can also blur regions along the frequency dimension, resulting in kurtosis peaks that only appear in statistics derived from maps of certain frequency bins. We further find that the kurtosis peaks will reach their maxima when the angular resolution of the PSFs match the size scale of the extreme regions that produce the peaks. The HERA array should be capable of charting the evolution of the observed skewness and kurtosis of the 21 cm fluctuations with high sensitivity while the MWA Phase I will likely be capable of detecting the peak in variance.

Read this paper on arXiv…

P. Kittiwisit, J. Bowman, D. Jacobs, et. al.
Thu, 20 Oct 16
29/44

Comments: 14 pages, 12 figures, submitted to ApJ