A deep / wide 1-2 GHz snapshot survey of SDSS Stripe 82 using the Karl G. Jansky Very Large Array in a compact hybrid configuration

TL;DR

This survey used the VLA to image a large sky area at 1-2 GHz, producing a detailed catalog of radio sources with spectral information, revealing biases and limitations in snapshot radio observations.

Contribution

It presents a comprehensive 1-2 GHz snapshot survey of SDSS Stripe 82 with detailed data products, spectral index analysis, and insights into clean bias effects in VLA snapshot imaging.

Findings

Detected 11,782 radio components and 8,948 sources.

Spectral index estimates agree with previous surveys.

Bias towards flat spectral indices at low flux densities.

Abstract

We have used the Karl G. Jansky Very Large Array to image ~100 sq. deg. of SDSS Stripe 82 at 1-2 GHz. The survey consists of 1,026 snapshot observations of 2.5 minutes duration, using the hybrid CnB configuration. The survey has good sensitivity to diffuse, low surface brightness structures and extended radio emission, making it highly synergistic with existing 1.4 GHz radio observations of the region. The principal data products are continuum images, with 16 x 10 arcsecond resolution, and a catalogue containing 11,782 point and Gaussian components resulting from fits to the thresholded Stokes-I brightness distribution, forming approximately 8,948 unique radio sources. The typical effective 1{\sigma} noise level is 88 {\mu}Jy / beam. Spectral index estimates are included, as derived from the 1 GHz of instantaneous bandwidth. Astrometric and photometric accuracy are in excellent agreement with existing narrowband observations. A large-scale simulation is used to investigate clean bias, which we extend into the spectral domain. Clean bias remains an issue for snapshot surveys with the VLA, affecting our total intensity measurements at the ~1{\sigma} level. Statistical spectral index measurements are in good agreement with existing measurements derived from matching separate surveys at two frequencies. At flux densities below ~35{\sigma} the median in-band spectral index measurements begin to exhibit a bias towards flatness that is dependent on both flux density and the intrinsic spectral index. In-band spectral curvature measurements are likely to be unreliable for all but the very brightest components. Image products and catalogues are publicly available via an FTP server.