The USNO/ATNF VLBI astrometry and imaging of southern ICRF sources (UVAISIS): A progress report

Introduction

Since the 1980s VLBI observations at two frequencies, 8.4 (X-band) and 2.3 GHz (S-band) have been used to locate the positions of compact radio sources with unprecedented accuracy. The basic observable is the group delay with the use of two frequencies allowing calibration of the frequency-dependent propagation delay in the ionosphere. Observations of selected strong compact extragalactic radio sources, using this now mature technique, have been used to define and maintain a radio reference frame with sub-milli-arcsecond (mas) precision. This International Celestial Reference Frame (ICRF) was adopted by the XXIII IAU General Assembly in 1997 to replace the traditional optical fundamental reference system, the FK5 reference frame. The ICRF defines the axes of the International Celestial Reference System (ICRS) with a precision of approximately 20 micro arcseconds. The Hipparcos catalog is the realization of this frame at optical wavelengths.

Johnston et al. (1988) set out to establish a global reference frame of 400 sources in 1986. The first catalog produced (Ma et al. 1990) had 182 sources with a positional accuracy of 1 mas and all located north of -30 degrees. Subsequent observing campaigns increased the density of sources in the northern hemisphere and added sources in the southern hemisphere (Russell et al. 1994; Reynolds et al. 1994; Johnston et al 1995).

Goals

The UVAISIS project seeks to strengthen the southern hemisphere VLBI astrometric position reference frame. It consists of two distinct but related programs. The first will increase the reference source density and the accuracy of existing sources with additional astrometric VLBI observations. Additionally, it will provide more reference sources for phase referencing VLBI observations.

The second program involves VLBI imaging at 8.4 GHz (X-band) of a complete sample of compact, flat-spectrum sources south of -20 degrees declination, primarily in order to determine the contribution of the intrinsic source structure on measured source positions. The observations will also look for systematic errors in source positions and provide a tie between the northern and southern hemisphere frames through the overlapping sources whose positions have already been measured in the north.

Astrometry

Whereas, in principle, two fixed points can define the reference frame (principal plane, pole and zero point), in practice numerous fixed points are needed to define the reference frame globally as the definition of the axes of the system improves with the number of points used to define them. Also, more reference points are necessary to allow access to the reference system anywhere on the celestial sphere. Compared to the northern hemisphere, the southern hemisphere reference frame is defined by far fewer Class 1 sources i.e. sources with positional weighted rms accuracies of less than one mas.

To increase the accuracy of existing southern source positions we have now made four, 24 hour bandwidth synthesis VLBI observations. A 24 hour experiment duration is necessary to separate the parameters for nutation and polar motion, which have diurnal signatures. The telescopes involved were Parkes and Hobart (Australia), Fortaleza (Brazil), Hartebeesthoek (South Africa), Kashima (Japan) and Kokee Park (USA). The data is being recorded in the standard MkIIIA 2.3/8.4 GHz bandwidth synthesis mode and is being correlated at the USNO correlator in Washington DC.

To increase the density of southern reference sources we have made short observations of 81 candidate sources during our imaging observations (see below). Suitable candidate sources will be included in upcoming astrometric observations. Further selection and observations of candidate sources is ongoing.

Imaging

Extragalactic radio sources, such as those that define the ICRF, display a variety of structure down to mas scales. Furthermore they are all variable on scales of years to weeks and indeed, in some cases, minutes. This departure from the point source approximation introduces error in the computable variables (delay and rate). The effect of source structure on position can be as large as tens of mas (e.g. Fey and Charlot, 1997). Also, as the structure of these sources varies with time, it is important to image their structure at several epochs in order to define a time-dependent source model. Multi-epoch observations, using the VLBA, to image northern hemisphere sources at 8.4 GHz have been progressing successfully for a few years.

We have now completed one epoch of observations of 184 southern ICRF radio sources at 8.4 GHz. These observations, made in full polarimetric mode (two telescopes, Hobart and Kokee Park record only one polarization), have been distributed over nine, 24 hour observations. The telescopes involved are Parkes, ATCA, Mopra, Hobart, Ceduna (Australia), Hartbeesthoek (South Africa) and Kokee Park (USA).

An image from the imaging program, PKS 1921-293.

The data are recorded in the S2 format. Modifications to the LBA correlator, located at Marsfield, now allow routine handling of up to seven antennas in full polarization mode. All the observed data have now been successfully correlated. Calibration is being done using standard techniques and using the National Radio Astronomy Observatory (NRAO) AIPS package while imaging is being done using the Caltech Difmap package. Images obtained so far look good though some sources may need additional time, mostly to get better uv-coverage.

Apart from yielding source structure information for astrometric purposes, these are often the first VLBI images of the flat-spectrum, extragalactic sources that constitute the sample. Thus we expect this project to provide valuable new information on the physics of these objects.

Conclusion

Both the astrometric and imaging components of the UVAYSIS program are progressing well. New sources are being selected and added to increase the density of the southern astrometric reference frame. The first set of observations for imaging all the existing southern reference sources are complete and calibration and imaging of the data is in progress. Information from these images is being used to optimise the next round of observations. We expect to be able to observe all sources at least twice and perhaps three times over the five year life of the project.

References

Fey & Charlot 1997, ApJS 111, 95

Johnston et al. 1995, AJ 110, 880

Johnston et al. 1988, IAU Symp. 129, The Impact of VLBI on Astrophysics and

Geophysics ed. Reid, Moran, 317

Ma et al. 1990, AJ 99, 1284

Reynolds et al. 1994, AJ 108, 725

Russell et al. 1994, AJ 107, 379

Roopesh Ojha, on behalf of the USNO/ATNF VLBI Astrometry and Imaging team
(Roopesh.Ojha@csiro.au)