The impact of diffuse Galactic emission on direction-independent gain calibration in high-redshift 21 cm observations
Abstract
This study examines the impact of diffuse Galactic emission (DGE) on sky-based direction-independent (DI) gain calibration using realistic forward simulations of Low-Frequency Array (LOFAR) observations of the high-redshift 21 cm signal of neutral hydrogen during the Epoch of Reionization (EoR). We simulated LOFAR observations between 147 and 159 MHz using a sky model that includes a point source catalog and DGE. The simulated observations were DI-gain calibrated with the point source catalog alone, utilizing the LOFAR-EoR data analysis pipeline. A full power spectrum (PS) analysis was conducted to measure the systematic bias, relative to thermal noise, caused by DI-gain calibration using a point-source-only (PSO) sky model, when applied to simulated data that include both point sources and DGE. The results are compared to a ground truth scenario where both the simulated sky and the calibration model include only point sources. Additionally, the cross-coherence between observation pairs was computed to determine whether DI-gain calibration errors are coherent or incoherent in specific regions of PS space as a function of integration time. We find that DI-gain calibration with a PSO sky model that omits DGE introduces a systematic bias in the PS for k bins < 0.2 h\,Mpc-1. The PS errors in these bins are coherent in time and frequency; therefore, the resulting bias could be mitigated during the foreground removal step using Gaussian Process Regression, as demonstrated in previous studies. In contrast, errors for k > 0.2 h\,Mpc-1 are largely incoherent and average down as noise. We conclude that, based on our analysis prior to foreground removal, missing DGE in the sky model during DI-gain calibration is unlikely to be a dominant contributor to the excess noise observed in the current LOFAR-EoR upper limits on the 21 cm signal PS.
Turn this paper into a full lesson
ArcXiv compiles a staged curriculum from this paper: 8-12 lessons across beginner → advanced, synthesised section guides, visuals, flashcards, a quiz, exercises, and on-demand deep dives per section. Grounded in the abstract, never invented.