Interacting binaries typically have separations in the milliarcsecond regime, and hence it has been challenging to resolve them at any wavelength. However, recent advances in optical interferometry have improved our ability to discern the components in these systems and have now enabled the direct determination of physical parameters. We used the Navy Prototype Optical Interferometer to produce for the first time images resolving all three components in the well-known Algol triple system. Specifically, we have separated the tertiary component from the binary and simultaneously resolved the eclipsing binary pair, which represents the nearest and brightest eclipsing binary in the sky. We present revised orbital elements for the triple system, and we have rectified the 180 degrees ambiguity in the position angle of Algol C. Our directly determined magnitude differences and masses for this triple star system are consistent with earlier light curve modeling results.

The Observing Strategies Sub-group of IVS's Working Group 3 has been tasked with producing a vision for the following aspects of geodetic VLBI: antenna-network structure and observing strategies; source strength/structure/distribution; frequency bands, RFI; and field system and scheduling. These are high level considerations that have far reaching impact since they significantly influence performance potential and also constrain requirements for a number of other \VG3 sub-groups. The paper will present the status of the sub-group's work on these topics.

We observed the 10 quasars with the highest core radio luminosity in the Palomar-Green Bright Quasar Survey to look for both short- (i.e., intra- and interday) and long- (i.e., months to years) term optical variability. In most models, the central quasi-stellar object (QSO) radio source is thought to have high apparent radio luminosity because a jet of relativistic material is beamed close to our line of sight. We expect more variability in the radio luminous sample than in a sample of radio-quiet QSOs if a shock in a beamed jet is the cause of variability; seven of 10 quasars showed short-term variability, and all 10 quasars have long-term variability. Comparison with other samples indicates our radio-loud QSOs are more likely to have short-term variations than radio-quiet QSOs. This result supports the model of QSO variability being dominated by shock waves in the relativistic plasma jet.

A VLBA map of CTD 93 at 2 cm shows the morphology to be better described as a core-jet configuration than as a symmetric compact double structure.

Images of resolved radio sources in the Palomar Bright Quasar Survey are presented with an angular resolution of 0.5 and 18 arcsec. The observed structure of some well resolved radio quiet quasars and AGN's show large scale linear structures or unresolved central cores similar to radio loud objects in the BQS sample as well as the more luminous radio selected quasars. We suggest that at least some of these less luminous radio quiet objects may contain compact central engines characteristic of radio loud quasars and radio galaxies.

Previous VLBI observations of the peculiar superluminal quasar 4C 39.25 at wavelengths lambda = 1.3, 2.8, and 3.6 cm have revealed the presence of a superluminal component (b) moving between the western component (c) and the eastern component (a). Here we combine data published previously and reanalyzed by us with new VLBI observations between 1986 and 1989 at these wavelengths. Components a and c have remained fixed relative to each other, while component b has slowed down and brightened as it approaches component a. Our most recent lambda = 1.3 cm VLBI observations have revealed the presence of a fourth, weak component (d) (to the west of c), which could be the core of the radio source, undetected in previous observations presumably due to an inverted spectrum and the limited dynamic ranges of the older maps.These observational results support a model in which 4C 39.25 contains a bent relativistic jet which is misaligned relative to the observer near the core region, leading to a relatively low core brightness. Near the stationary components c and a, the jet curves its trajectory toward the observer, so as to become more closely aligned to the line of sight, while it is misaligned in the region between these two hotspots. We interpret component b as a shock wave propagating down the jet. This shock-in-curved-jet hypothesis is supported by the nonuniform proper motion of component b, the total and polarized flux density evolution of both the individual components and the source as a whole, the spectral evolution, and the relative position angles between the source components.

We present radio maps and flux density and polarization measurements of the quasar 4C 39.25 made with the Very Large Array at wavelengths of 20, 6, 2, and 1.3 cm, with Very Long Baseline Interferometry at 18, 3.6, and 2.8 cm, and with the University of Michigan 26 m radio telescope at 6.2, 3.8, and 2.1 cm. The milliarcsecond structure consists of a bent jet containing both superluminally moving and stationary structure. The jet extends to the east, with significant broadening apparent, on the approximately 0".01 scale of the 18 cm VLBI image. Symmetric arcsecond scale structure seen on the 20 cm VLA map is resolved at shorter wavelengths into an asymmetric, lumpy jet extending 2" to both the east and west of the compact core, embedded in a more diffuse, slightly elongated emission region. At the epoch of the maps, the polarization of the subarcsecond structure is of order a few percent at wavelengths longer than 6 cm, with E-vector position angle parallel to the compact jet at 6 cm and at an oblique angle to the jet at 2 and 1.3 cm.The 3.6 cm VLBI maps show that the previously reported apparent superluminal motion of one of the components continues. The most compact features of the arcsecond jet lie on the same side of the core as the direction of the superluminal motion. The flux density of the moving component has increased monotonically since 1984, while the overall spectrum of the source has maintained the same shape.We develop a twisted, relativistic jet model to explain (i) the absence of a distinct submilliarcsecond core, (ii) the combination of superluminal and stationary compact structure, and (iii) the morphology of the arcsecond-scale structure. We favor a scenario in which the jet bends almost directly toward the line of sight at the positions of the stationary milliarcsecond-scale components and is otherwise slightly misdirected from the line of sight. The required curvature of the jet is +/- 4-degrees from the direction that maximizes the observed proper motion of moving knots in the jet. According to this model, the superluminal feature corresponds to a shock propagating down the jet. We show that the evolution of the spectrum and polarization of the source are qualitatively reproduced by this model.

Observations from 600 Mark III VLBI experiments from 1979 to 1988, resulting in 237 681 acceptable pairs of group delay and phase delay rate observations, have been used to derive positions of 182 extragalactic radio sources with typical formal standard errors less than 1 mas. The sources are distributed fairly evenly above δ=-30^deg^, and 70 sources have δ<0^deg^. Analysis with different troposphere models, as well as internal and externalicomparisons, indicates that a coordinate frame defined by this set of radio sources should be reliable at the 1 mas level. The right ascension zero point of this reference frame has been aligned with the FK5 by using the optical positions of 28 extragalactic radio sources whose positions are on the FK5 system. Because of known defects in our knowledge of astronomical constants, daily nutation offsets in longitude and obliquity were determined relative to an arbitrary reference day in the set of experiments. The declinations of the sources are nominally absolute with respect to the pole of this reference day: 1980 October 17.

Following the Loma Prieta earthquake, two mobile Very Long Baseline Interferometry (VLBI) systems operated by the NASA Crustal Dynamics Project and the NOAA National Geodetic Survey were deployed at three previously established VLBI sites in the earthquake area: Fort Ord (near Monterey), the Presidio (in San Francisco) and Point Reyes. From repeated VLBI occupations of these sites since 1983, the pre‐earthquake rates of deformation have been determined with respect to a North American reference frame with 1σ formal standard errors of ∼1 mm/yr. The VLBI measurements immediately following the earthquake showed that the Fort Ord site was displaced 49 ± 4 mm at an azimuth of 11 ± 4° and that the Presidio site was displaced 12 ± 5 mm at an azimuth of 148 ± 13°. No anomalous change was detected at Point Reyes with 1σ uncertainty of 4 mm. The estimated displacements at Ford Ord and the Presidio are consistent with the static displacements predicted on the basis of a coseismic slip model in which slip on the southern segment is shallower than slip on the more northern segment of the fault rupture. We also give the Cartesian positions at epoch 1990.0 of a set of VLBI fiducial stations and the three mobile sites in the vicinity of the earthquake.