Phase-referenced VLBA observations offer two critical advantages over pure self-calibration techniques: (1) weaker target sources can be observed because the array's effective coherence time is increased from minutes to hours, vastly increasing the number of accessible target sources; and (2) the target source position can be measured accurately relative to the phase reference source. This second advantage permits registration of VLBA images with those from other facilities, as well as registration of VLBA movie frames that span days to years. Such VLBA movies yield information on proper motions, annual parallaxes and/or orbital motions.
What percentage of VLBA observations employ phase referencing? A total of 2773 science observations were made with the VLBA during 2003-2008. Associated with each observation is a summary file created by Sched, the NRAO scheduling software. Information from the summary file yields the number of scans scheduled per hour during the observation. To follow the effects of the dynamic ionosphere and/or troposphere, rapid switching between the reference source and the target source is required. This leads to 20 or more scans per hour for a phase-referenced observation. In contrast, an observation that will rely on self-calibration will have less than 20 scans per hour. A break point of 20 scans per hour was tested by visually inspecting all observations during several months. This break point failed for observations at 86 GHz and for observations involving in-beam phase referencing; such cases currently represent a negligible fraction of all observations.
Using 20 or more scans per hour as the break point, it is a simple matter to use awk to find the percentage of observations employing phase referencing. The figure shows this percentage year by year. On average, 63 % of VLBA observations during 2003-2008 used phase referencing. This enormous popularity stems from the two critical advantages mentioned above, as well as an end-to-end implementation of phase referencing on the VLBA. Key aspects of this implementation were community led, most notably the development of and operational support for precision geodesy and astrometry on the VLBA by NASA and the USNO (Petrov et al. 2009, Journal of Geodesy, 83, 859; arXiv:0806.0167v2).