The 154 MHz radio sky observed by the Murchison Widefield Array: noise, confusion and first source count analyses

TL;DR

This study analyzes a 154 MHz radio sky image from the Murchison Widefield Array, assessing noise, source counts, and confusion limits, providing the most precise counts in the 30-200 mJy range and highlighting calibration needs for deeper surveys.

Contribution

It presents the most precise low-frequency source counts in the 30-200 mJy range and discusses the impact of confusion noise and calibration challenges for future deep surveys.

Findings

Source counts agree with other instruments.

Confusion noise affects the image at ~1.7 mJy/beam.

Further calibration improvements are needed for deeper observations.

Abstract

We analyse a 154 MHz image made from a 12 h observation with the Murchison Widefield Array (MWA) to determine the noise contribution and behaviour of the source counts down to 30 mJy. The MWA image has a bandwidth of 30.72 MHz, a field-of-view within the half-power contour of the primary beam of 570 deg^2, a resolution of 2.3 arcmin and contains 13,458 sources above 5 sigma. The rms noise in the centre of the image is 4-5 mJy/beam. The MWA counts are in excellent agreement with counts from other instruments and are the most precise ever derived in the flux density range 30-200 mJy due to the sky area covered. Using the deepest available source count data, we find that the MWA image is affected by sidelobe confusion noise at the ~3.5 mJy/beam level, due to incompletely-peeled and out-of-image sources, and classical confusion becomes apparent at ~1.7 mJy/beam. This work highlights that (i) further improvements in ionospheric calibration and deconvolution imaging techniques would be required to probe to the classical confusion limit and (ii) the shape of low-frequency source counts, including any flattening towards lower flux densities, must be determined from deeper ~150 MHz surveys as it cannot be directly inferred from higher frequency data.