Stacking of SKA data: comparing uv-plane and image-plane stacking

TL;DR

This paper introduces a novel uv-data stacking algorithm for radio interferometry, compares it with traditional image-plane stacking through simulations, and discusses its advantages for SKA data analysis.

Contribution

It presents a new uv-data stacking method, compares its performance with image stacking, and explores its implications for SKA data processing and analysis.

Findings

Uv-stacking offers better calibration artifact analysis.

Simulations show uv-stacking can outperform image-stacking.

Uv-stacking is advantageous for source size and property studies.

Abstract

Stacking as a tool for studying objects that are not individually detected is becoming popular even for radio interferometric data, and will be widely used in the SKA era. Stacking is typically done using imaged data rather than directly using the visibilities (the uv-data). We have investigated and developed a novel algorithm to do stacking using the uv-data. We have performed exten- sive simulations comparing to image-stacking, and summarize the results of these simulations. Furthermore, we disuss the implications in light of the vast data volume produced by the SKA. Having access to the uv-stacked data provides a great advantage, as it allows the possibility to properly analyse the result with respect to calibration artifacts as well as source properties such as size. For SKA the main challenge lies in archiving the uv-data. For purposes of robust stacking analysis, it would be strongly desirable to either keep the calibrated uv-data at least in an aver- age form, or implement a stacking queue where stacking positions could be provided prior to the observations and the uv-stacking is done almost in real time.