Exploring the consequences of chromatic data excision in 21-cm Epoch of Reionization power spectrum observations

TL;DR

This paper investigates how chromatic radio frequency interference (RFI) flagging impacts 21-cm Epoch of Reionization power spectrum measurements, revealing that coarse flags induce excess power similar to residual RFI, and proposes a mitigation strategy of flagging entire bands.

Contribution

It demonstrates that coarse RFI flags cause significant contamination in power spectrum measurements and introduces a simple, effective mitigation method of flagging entire frequency bands when RFI is detected.

Findings

Coarse RFI flags produce excess power in the EoR window.

Flagging entire bands effectively mitigates chromatic RFI effects.

Current strategies require improvements for reliable 21-cm EoR detection.

Abstract

We explore how chromatic RFI flags affect 21-cm power spectrum measurements. We particularly study flags that are coarser than the analysis resolution. We find that such RFI flags produce excess power in the EoR window in much the same way as residual RFI. We use Fast Holographic Deconvolution (FHD) simulations to explain this as a result of chromatic disruptions in the interferometric sampling function of the array. We also use these simulations in conjunction with Error Propagated Power Spectrum with InterLeaved Observed Noise ($\varepsilon$ppsilon) to show that without modifying current flagging strategies or implementing extremely accurate and complete foreground subtraction, 21-cm EoR experiments will fail to make a significant detection. As a mitigation strategy, we find that circumventing the chromatic structure altogether by flagging the entire analysis band when RFI is detected is simple to implement and highly successful. This demands a detection strategy with a low false positive rate in order to prevent excessive data loss.