Polarisation performance of offset phase antennas: A study for FARSIDE

Abstract Several radio telescopes have been planned or proposed to be deployed on the Lunar farside in the coming years. These will observe the unexplored ultra-long wavelengths of the electromagnetic spectrum from the lunar farside's unique radio-quiet and ionosphere-free environment. One such lunar radio array is the NASA-funded concept – the Farside Array for Radio Science Investigations of the Dark Ages and Exoplanets (FARSIDE). FARSIDE will operate over 100 kHz to 40 MHz with 128 spatially non-co-located orthogonal pairs of antenna nodes distributed over a $12\times12$ km area in a four-arm spiral configuration. Being on the lunar farside, this radio interferometer will be deployed by tele-operated rovers. The rover deployment mode could lead to a phase offset between each of the two orthogonally polarised antenna elements in the array, which are typically co-located. In this paper, we quantify the effects of such antenna phase offsets on the polarisation response and imaging performance of the lunar radio array. Modelling and analysing the FARSIDE dipole beams with and without offset, we find the latter leads to additional leakages into Stokes U and V corresponding to Muller matrix terms of $M_{2(0,1,2,3)}$ and $M_{3(0,1,2,3)}$ . Using a custom simulation pipeline to incorporate all four Stokes beams of spatially co-located and non-co-located dipoles, we produce visibilities and simulated images for the GLEAM (GaLactic and Extragalactic All-sky MWA) sky model through the FARSIDE array. We find that for a pure Stokes I input sky, the output image maximum Stokes $V/I$ flux ratio for the offset case has increased to $2.5\%$ versus $0.05\%$ for the co-located case. The additional Stokes V needs to be corrected since the detection of electron cyclotron maser emissions from exoplanets requires high-fidelity Stokes V measurements.