Investigating ionospheric calibration for LOFAR 2.0 with simulated observations

¹ Hamburger Sternwarte, University of Hamburg, Gojenbergsweg 112, 21029 Hamburg, Germany
e-mail: henrik.edler@hs.uni-hamburg.de
² INAF – Istituto di Radioastronomia, Via P. Gobetti 101, 40129 Bologna, Italy

Received: 1 February 2021
Accepted: 7 May 2021

Abstract

Context. There are a number of hardware upgrades for the Low-Frequency Array (LOFAR) currently under development. These upgrades are collectively referred to as the LOFAR 2.0 upgrade. The first stage of LOFAR 2.0 will introduce a distributed clock signal and allow for simultaneous observations using all the low-band and high-band antennas of the array.

Aims. Our aim is to provide a tool for obtaining accurate simulations for LOFAR 2.0.

Methods. We present our software for simulating LOFAR and LOFAR 2.0 observations, which includes realistic models for all important systematic effects such as the first- and second-order ionospheric corruptions, time-variable primary-beam attenuation, station-based delays, and bandpass response. The ionosphere is represented as a thin layer of frozen turbulence. Furthermore, thermal noise can be added to the simulation at the expected level. We simulate a full eight-hour simultaneous low- and high-band antenna observation of a calibrator source and a target field with the LOFAR 2.0 instrument. The simulated data are calibrated using readjusted LOFAR calibration strategies. We examine novel approaches of solution-transfer and joint calibration to improve direction-dependent ionospheric calibration for LOFAR.

Results. We find that the calibration of the simulated data behaves very similarly to a real observation and reproduces certain characteristic properties of LOFAR data, such as realistic solutions and image quality. We analyze strategies for direction-dependent calibration of LOFAR 2.0 and find that the ionospheric parameters can be determined most accurately when combining the information of the high-band and low-band in a joint calibration approach. In contrast, the transfer of total electron content solutions from the high-band to the low-band shows good convergence but is highly susceptible to the presence of non-ionospheric phase errors in the data.