Vienna Goes VLBI Global Observing System (VGOS Squared)
In geodetic Very Long Baseline Interferometry (VLBI), we use globally distributed VLBI radio telescopes to observe quasars billions of light years away. The difference in arrival times of the signals at the stations is the primary observable used in geodesy to estimate baseline vectors between radio telescopes, positions of quasars, and Earth orientation parameters. The observables are determined by time-tagging the signals with the use of atomic clocks, e-transferring the data, and cross-correlation at special computing facilities, called correlators. With these unique capabilities, VLBI plays a key and outstanding role in the determination of terrestrial and celestial reference frames as well as Earth orientation parameters. For example, VLBI is the only technique for the observation of Universal Time (UT1) which is related to the Earth rotation angle and thus of fundamental importance for any kind of positioning and navigation with the Global Navigation Satellite Systems (GNSS). VLBI is also critical for the accuracy of the scale of the terrestrial reference frame, which is realised by positions and velocities of globally distributed sites. The stability of the scale at an accuracy level of 0.1 mm/year is a prerequisite for the observation of small geodynamic quantities, such as sea level rise at about 3 mm/year.
The VLBI community is currently working on a tremendous improvement of the VLBI technique, called VLBI Global Observing System (VGOS), and based on new fast slewing radio telescopes and increased observation bandwidth resulting in astrometric and geodetic quantities of unprecedented accuracies. From a TU Wien (Vienna) point of view, we see the following two tasks as the main fields where we can not only contribute, but really bring VLBI and in particular VGOS activities to the next level. First, there is potential in improving the scheduling of VLBI sessions, i.e., in specifying which radio telescopes should observe which quasars at what time. In particular, we will investigate the application of tree-based schedules ("looking further ahead when scheduling") in combination with graph theory as well as innovative functions to describe the sky coverage at the stations. Second, correlation is and certainly will be the bottleneck in VGOS with a dramatically increased requirement in terms of bandwidth, number of processing cores, and storage. We will use dedicated storage and computing cores for correlation activities on the Vienna Scientific Cluster (VSC), a collaboration of several Austrian universities that provides supercomputer resources and corresponding services to their users. We will investigate the correlation and fringe-fitting of VGOS data along with the automation of the processes from correlation to analysis, thus serving as a role model for other universities to deal with correlation in future.