An interference detection strategy for Apertif based on AOFlagger 3

¹ ASTRON, the Netherlands Institute for Radio Astronomy, Oude Hoogeveensedijk 4, 7991 PD Dwingeloo, The Netherlands
e-mail: offringa@astron.nl
² Kapteyn Astronomical Institute, University of Groningen, PO Box 800, 9700 AV Groningen, The Netherlands
³ Astronomisches Institut der Ruhr-Universität Bochum (AIRUB), Universitätsstrasse 150, 44780 Bochum, Germany
⁴ Department of Physics, Virginia Polytechnic Institute and State University, 50 West Campus Drive, Blacksburg, VA 24061, USA
⁵ Dept. of Astronomy, Univ. of Cape Town, Private Bag X3, Rondebosch 7701, South Africa
⁶ Instituto de Astrofísica de Andalucía (CSIC), Glorieta de la Astronomía s/n, 18008 Granada, Spain
⁷ National Centre for Radio Astrophysics, Tata Institute of Fundamental Research, Pune 411007, Maharashtra, India
⁸ Anton Pannekoek Institute, University of Amsterdam, Postbus 94249, 1090 GE Amsterdam, The Netherlands
⁹ Netherlands eScience Center, Science Park 402, 1098 XH Amsterdam, The Netherlands
¹⁰ University of Oslo Center for Information Technology, PO Box 1059, 0316 Oslo, Norway

Received: 20 September 2022
Accepted: 3 January 2023

Abstract

Context. Apertif is a multi-beam receiver system for the Westerbork Synthesis Radio Telescope that operates at 1.1–1.5 GHz, which overlaps with various radio services, resulting in contamination of astronomical signals with radio-frequency interference (RFI).

Aims. We analyse approaches to mitigate Apertif interference and design an automated detection procedure for its imaging mode. Using this approach, we present long-term RFI detection results of over 300 Apertif observations.

Methods. Our approach is based on the AOFlagger detection approach. We introduce several new features, including ways to deal with ranges of invalid data (e.g. caused by shadowing) in both the SumThreshold and scale-invariant rank operator steps; pre-calibration bandpass calibration; auto-correlation flagging; and HI flagging avoidance. These methods have been implemented in a new framework that uses the Lua language for scripting, which is new in AOFlagger version 3.

Results. Our approach removes RFI fully automatically, and it is robust and effective enough for further calibration and (continuum) imaging of these data. The analysis of 304 observations shows an average of 11.1% of lost data due to RFI with a large spread. We observe 14.6% RFI in auto-correlations. Computationally, AOFlagger achieves a throughput of 370 MB/s on a single computing node. Compared to published machine learning results, the method is one to two orders of magnitude faster.