The standing wave phenomenon in radio telescopes

Frequency modulation of the WSRT primary beam

¹ Kapteyn Astronomical Institute, PO Box 800, 9700 AV Groningen, The Netherlands e-mail: popping@astro.rug.nl
² Australia Telescope National Facility, CSIRO, PO Box 76, Epping NSW 1710, Australia

Received: 22 November 2007
Accepted: 7 December 2007

Abstract

Context.Inadequacies in the knowledge of the primary beam response of current interferometric arrays often form a limitation to the image fidelity, particularly when “mosaicing” over multiple telescope pointings.

Aims.We hope to overcome these limitations by constructing a frequency-resolved, full-polarization empirical model for the primary beam of the Westerbork Synthesis Radio Telescope (WSRT).

Methods.Holographic observations, sampling angular scales between about 5 arcmin and 11 degrees, were obtained of a bright compact source (3C 147). These permitted measurement of voltage response patterns for seven of the fourteen telescopes in the array and allowed calculation of the mean cross-correlated power beam. Good sampling of the main-lobe, near-in, and far-side-lobes out to a radius of more than 5 degrees was obtained.

Results.A robust empirical beam model was detemined in all polarization products (XX, XY, YX and YY) and at frequencies between 1322 and 1457 MHz with 1 MHz resolution. Substantial departures from axi-symmetry are apparent in the main-lobe as well as systematic differences between the polarization properties. Surprisingly, many beam properties are modulated at the 5 to 10% level with changing frequency. These include: (1) the main beam area, (2) the side-lobe to main-lobe power ratio, and (3) the effective telescope aperture. These semi-sinusoidsal modulations have a basic period of about 17 MHz, consistent with the natural “standing wave” period of a 8.75 m focal distance. The deduced frequency modulations of the beam pattern were verified in an independent long duration observation using compact continuum sources at very large off-axis distances.

Conclusions.Application of our frequency-resolved beam model should enable higher dynamic range and improved image fidelity for interferometric observations in complex fields, although at the expense of an increased computational load. The beam modulation with frequency can not be as easily overcome in total power observations. In that case it may prove effective to combat the underlying multi-path interference by coating all shadowed telescope surfaces with a broad-band isotropic scattering treatment.