In studies of extragalactic radio sources with multiple compact components the determination of which components, if any, are stationary and which moving is of importance. In order to learn about the radio properties of the individual components it is also relevant to be able to register maps made at several wavelengths. Both tasks are usually not possible with VLBI because of the irrecoverable corruption of the fringe phase introduced by the propagation medium and the instrumentation. However, when two or more compact radio sources are separated by only a small angle from each other difference techniques can be used to help tackle both questions.

Development of a catalogue of approximately 100 radio sources having positional accuracy of ~0.005 arcseconds is one goal of the DSN radio reference frame program. This objective is set by the navigation requirements of the Galileo Project for reconstruction of the probe entry angle into the atmosphere of Jupiter. These radio positions are determined by VLBI observations using the antennas of the NASA Deep Space Network on two intercontinental baselines – California-Spain and California-Australia. Since 1978, measurements have been made simultaneously at 2.3 and 8.4 Ghz. Sixteen to twenty-four hours of data on each of the two baselines are usually obtained within a few days of each other, and since mid-1982 these pairs of observations have been made at approximately six week intervals.

From its inception, it was realized that VLBI had the potential for very precise determinations of baselines between antennas. Since the first early, crude results (Cohen and Shaffer 1971), the data collection systems and analysis techniques have been continuously refined, and VLBI is now producing baseline measurements to accuracies of a few centimeters over trans- and inter-continental distances. Since the Mark III system and the basic techniques of delay and delay rate baseline determination have been previously described (Rogers et al. 1983; Shapiro 1976), and there have been several recent meetings that dealt exclusively with VLBI geodesy techniques (NASA 1980; NOAA 1982), I will concentrate in this review on recent developments in equipment and procedures, followed by presentation of new results. This review is based primarily on the efforts and results of the so-called “East Coast VLBI Group” comprised of staff and support contractors at NASA Goddard Space Flight Center (NASA/GSFC), the NEROC Haystack Observatory, MIT, and the Harvard-Smithsonian Center for Astrophysics. Additional input comes mostly from staff of the National Geodetic Survey (NGS), Onsala Space Observatory, and the Jet Propulsion Laboratory.

The need for centimetric accuracies set by the application of the VLBI technique to geodesy implies a considerable computational effort, because of the intrinsic complexity of the model and because there is an increasingly large number of calibrations and corrections which can be accounted for only via software. The program VLBI 3 (Robertson, 1975) has been developed for the geodetic and astrometric analysis of VLBI data. It includes an accurate theoretical model of the observables and is supported by a number of routines for parameter fitting and input/output operations with data and results. The original VAX version of VLBI 3, due to N. Bartel and M.I. Ratner, runs in batch mode and requires routines which are in general unavailable in standard VAX systems. We have prepared (Caporali and Sylos Labini, 1982) a modified VAX version of VLBI 3. This version runs on our standard VAX/VMS computers and contains a number of changes in the FORTRAN source which allow to the user a real time interaction with the program. In addition, having a Tektronix graphic station at our disposal, we decided to replace the existing plot package - which used the line printer - with a “ad hoc” graphic program which permits interactive display of the results of each run of VLBI 3. Our work was mostly concerned with the input/output sections of VLBI 3. The theorical model of the VLBI observables has, for the moment, been left unchanged. We have, however, noticed that the theorical model could be updated and made more precise, e.g. in the computation of the nutation terms and of the aberration. This updating and a more extensive geodetic and astrometric analysis of VLBI data are planned to be done next.

The VLB Array (VLBA) is a synthesis radio telescope which has been designed to extend the resolution of the VLA in order to allow sub-milliarcsec studies of compact galactic and extragalactic radio sources over a wide range of wavelengths and spectral resolution.

On 1983 May 31, the National Research Council of Canada gave formal approval to a new national facility, the ‘Canadian Long Baseline Array’. The instrument will be made up of eight radio telescopes spaced across southern Canada, with a ninth antenna, primarily for geophysics, in the north. Though the project is still some way from being funded, the NRC approval was an important and encouraging step in a process that started five years ago.

Three radio telescopes operating at 103 MHz are being installed at Ahmedabad, Rajkot and Surat separated by about 200 km from each other for observing interplanetary scintillations (IPS) of compact radio sources for study of solar wind plasma dynamics as well as radio source size measurements for cosmological studies. Of these, two radio telescopes at Ahmedabad and Rajkot have been commissioned and started synchronous daily observations of IPS of a few compact radio sources with relative time accuracy of about ± 1 millisec. The third telescope at Surat is expected to go in operation by the end of 1983. As soon as all the three telescopes go in for simultaneous operation, it is proposed to (1) augment the telescope sensitivity so as to detect sources with flux density ~ 1 Jansky (2) incorporate better time and frequency standards at each station which can be synchronised to better than μs relative time accuracy (3) develop suitable receivers and data acquisition system for generating interference fringes using a general purpose computer and (4) take advantage of the availability of three telescopes to incorporate ‘closure phase and amplitude’ techniques which eliminate undesirable atmospheric and ionospheric phase distortions.

The Very-Long-Baseline Interferometry technique is not limited by the size of the Earth. Near-Earth-orbiting space vehicles can carry radio telescopes that can serve as VLBI stations using existing technology. By proper use of ground VLB arrays, a single orbiting VLBI satellite can yield radio maps with 2-dimensional coverage and high dynamic range at all declinations. A single orbiter can be used out to orbits that yield an effective aperture greater than two Earth diameters. Interstellar scintillations are a limiting factor only in the micro-arc-second range.

A proposal has been made to use the Tracking and Data Relay Satellite System (TDRSS) as an orbiting element for a very long baseline interferometry (VLBI) demonstration. The TDRSS is a satellite system designed to coherently track and relay data between other satellites and the central ground station. This system could also be used to coherently observe a celestial radio source. A ground-based frequency standard would be used to coherently drive the spacecraft receiver local oscillator and transmitter. The data will be telemetered to a ground station, where it will be recorded on a Mark III terminal.

Preston et al (1976) and Burke (1982, these proceedings) have long extolled the virtues of launching a radio telescope into space to increase VLBI baseline lengths and thus angular resolution, and to provide a much enhanced image formation capability. The scientific motivation for this has been covered in a number of memoranda referenced by Burke in these proceedings, and by Anderson et al (1982). Efforts to mobilise western astronomical support for space VLBI met with success in late 1982 at a meeting of US and European radio astronomers in Toulouse, France, at which a decision was taken to propose a joint mission to ESA and NASA. Shortly thereafter, a formal proposal was made to ESA (Anderson et al 1982) for a free flying satellite in an elliptical orbit out to 15000 km from the Earth, designed to observe in concert with the major ground-based VLBI networks and arrays. The mission, dubbed QUASAT, was received favourably in both ESA and NASA, with the result that formal Assessment Studies are scheduled to begin in both agencies in October 1983.