CIRADA: VLASS Epoch 1 Quick Look Catalogue User Guide
Yjan Gordon, Christopher O'Dea, Lawrence Rudnick,
Heinz Andernach, Adrian Vantyghem, and Michelle Boyce
July 6, 2020
Abstract
This document describes the catalogue of radio components and host
identifications based on quick look imaging from epoch 1 of the Very
Large Array Sky Survey (VLASS) produced by the Canadian Initiative for
Radio Astronomy Data Analysis. The catalogue itself consists of three
tables: a Component Table listing PyBDSF detections from the imaging;
a Host ID Table listing radio sources with relatively simple morphology
for which a host has been identified using the maximum likelihood ratio;
and a Subtile Information Table providing metadata for the quick look
images themselves. The production of these tables, how to access the
data, and the data models are described within. The VLASS quick look
data used in the production of this catalogue are of variable quality,
and we detail the known issues including, but not limited to a 10 %
systematic underestimation of the flux measurements, astrometric errors
at the 100 level, and a non-homogeneous distribution of the rms noise
over the sky. Consequently we urge users to read this document in full
prior to using the data - which is useful for generating source samples
and target lists, but should be used with extreme caution for further
applications. In general we recommend using data with Duplicate flag <
2 and Quality flag == 0 from the Component Table, and P Host > 0.8 and
Source reliability flag == 0 from the Host ID Table.
VLASS Overview . . . . . . . . . . . . . . . . . .
Catalogue Production . . . . . . . . . . . . . . . .
2.1. Component Extraction . . . . . . . . . . . .
2.2. Identifying Sources and Host Galaxies . . .
2.3. The Image Subtiles . . . . . . . . . . . . . .
3. Accessing and Making Use of the Data . . . . .
3.1. Data Access . . . . . . . . . . . . . . . . . .
3.2. Component Table Data Model . . . . . . .
3.3. Host ID Table Data Model . . . . . . . . . .
3.4. Subtile Information Table Data Model . . .
4. Summary of Known Data Quality Issues . . . .
4.1. Noise Variance Between Images . . . . . . .
4.2. Reliability of Flux Density Measurements .
4.3. Astrometry . . . . . . . . . . . . . . . . . . .
4.4. Contamination of Host IDs by Large Radio
Acknowledgements . . . . . . . . . . . . . . . . . . . .
1. VLASS Overview
The Karl G. Jansky Very Large Array Sky Survey (VLASS, Lacy et al.,
2020) is an S-band (2- 4 GHz referred to hereafter as 3 GHz) continuum
survey of the entire sky at dec >-40deg currently being undertaken by the
National Radio Astronomy Observatory (NRAO). Utilising a synthesised beam
size of 200.5, three epochs of observation are planned with a median
rms of 120muJy/beam, with the ambition of a 70muJy/beam rms in the final
three-epoch stack. The first of those three epochs, VLASS1, has now been
fully observed (within two sub-epochs, 1.1 and 1.2), and work is ongoing
at NRAO to produce high-quality single epoch images and a resultant epoch
1 basic component catalogue. Prior to the availability of high-quality
single epoch images, NRAO have produced rapidly CLEANed quick look
(QL) images with a pixel size of 100, that were released within a few
weeks of each observation set being completed. The expedited nature of the
QL image processing limits the quality of this data and hence constrains
its scientific usability. Furthermore this rapid-CLEAN process was updated
between epoch 1.1 and 1.2 resulting in significant differences in image
quality between the two sub-epochs1. The epoch 1 QL imaging is described
in full in Lacy et al. (2019), but we highlight here the key issues known
a priori: Flux calibration - There is a systematic under-measurement
of flux values in the QL images. Specifically, the peak flux densities
are ~15 % and ~ 8 % too low in epoch 1.1 and 1.2 respectively. The
respective total fluxes are underestimated by ~10 % and ~ 3 %.
Astrometry - The positional accuracy of the VLASS QL imaging is limited
to 100 , improving to 000.5 at DEC > -20d. Ghost images -
These appear offset by multiples of 30 in RA from bright sources.
The QL image production pipeline is designed to flag and remove these
but some faint ghosts may still remain. Moreover the removal of ghosts
was refined between epoch 1.1 and 1.2 meaning that the earlier images
are more likely to contain residual faint ghosts. The reliability of the
flux and astrometric measurements in this catalogue are further detailed,
alongside further discovered issues in Section 4. The remainder of this
document describes the catalogue of radio components and sources with
identified hosts produced from this QL imaging by the Canadian Initiative
for Radio Astronomy Data Analysis (CIRADA)2. Throughout this document and
the associated data we explicitly refer to detections from source-finding
algorithms as radio components, with the term radio source retained for
describing one or more components grouped together and associated with
a host. Furthermore, the individual lists of components, sources with
hosts, and quick look images are explicitly referred to as tables, with
the term catalogue being used to describe the entire product consisting
of the three tables. The actual production of the catalogue is outlined
in Section 2. Details on how to access the catalogue data, and the data
models are described in Section 3. In Section 4 we detail the known data
reliability issues associated with this catalogue.
NRAO have indicated an intention to reprocess the epoch 1.1 QL images with
the updated QL processing pipeline which should make future versions
of QL images from the two sub-epochs more consistent. However, these
were not available when this catalogue was produced and the differences
between these image sets applies to this data. http://cirada.ca/
2. Catalogue Production
The CIRADA VLASS QL catalogue consists of three tables (called Component
Table, Host ID Table, and Subtile Information Table) provides basic
component parameters as produced by PyBDSF3 (Mohan & Rafferty, 2015),
flagging of likely spurious detections, and identification of likely
individual radio sources (either single-component or close-double/triple)
with a cross identification obtained from unWISE (Lang, 2014; Schlafly et
al., 2019) where a host for the radio source can be found. The Component
Table is linkable to the Host ID Table via the Source name column,
and to the Subtile Info Table via the Subtile column. The catalogue is
complete in the sense that poor detections are left in but flagged with
a recommendation provided on their usability. The actual usage however
is left to the discretion of the end user.
In addition to basic radio component parameters in the Component
Table, the Host ID Table provides a list of radio sources for which
a host can be identified using the likelihood ratio method (McAlpine
et al., 2012). For identified radio sources (single-component or
close-double/triple), source parameters such as source size, source
flux (peak and total) and uncertainties are provided in addition to the
individual component parameters. Note that this catalogue only deals with
simple (single-component) to slightly complex (close-double/triples)
sources. More complex morphologies do exist but are not included in
this catalogue. The components of such sources are still listed in the
component table, but are not grouped into sources (i.e. they have no
entry in the Source name column).
Finally a table describing the QL images (hereafter referred to as
Subtiles) used in the production of this catalogue is provided. The
image information detailed in this Subtile Information Table is obtained
from the fits headers and a url to the image is included. Thus, this
catalogue consists of three distinct, linkable tables: components
(3.4M rows), sources with hosts (700k rows), and subtile information
(35k rows). These are described in detail below, and the table column
definitions are provided in Section 3.
....
References
Becker R. H., White R. L., Helfand D. J., 1995, ApJ, 450, 559
Condon J. J., Cotton W. D., Greisen E. W., Yin Q. F.,
Perley R. A., Taylor G. B., Broderick J. J., 1998, AJ, 115, 1693
Gaia Collaboration et al., 2016, A&A, 595, A1
Gaia Collaboration et al., 2018, A&A, 616, A1
Intema H. T., Jagannathan P., Mooley K. P., Frail D. A., 2017, A&A, 598, A78
Kimball A., 2017, VLASS Project Memo #7: VLASS Tiling and Sky Coverage,
https://library.nrao.edu/public/memos/vla/vlass/VLASS_007.pdf
Lacy M., et al., 2019, VLASS Project Memo #13:
Pilot and Epoch 1 Quick Look Data Release,
https://library.nrao.edu/public/memos/vla/vlass/VLASS_013.pdf
Lacy M., et al., 2020, PASP, 132, 035001
Lang D., 2014, AJ, 147, 108
McAlpine K., Smith D. J. B., Jarvis M. J., Bonfield
D. G., Fleuren S., 2012, MNRAS, 423, 132
Mohan N., Rafferty D., 2015, PyBDSF: Python Blob
Detection and Source Finder (ascl:1502.007)
Schlafly E. F., Meisner A. M., Green G. M., 2019, ApJS, 240, 30
Shimwell T. W., et al., 2017, A&A, 598, A104
Williams W. L., et al., 2019, A&A, 622, A2
de Gasperin F., Intema H. T., Frail D. A., 2018, MNRAS, 474, 5008