Mobile very long baseline interferometry (VLBI) and Global Positioning System (GPS) geodetic measurements have many error sources in common. Calibration of the effects of water vapor on signal transmission through the atmosphere, however, remains the primary limitation to the accuracy of vertical crustal motion measurements made by either technique. We have evaluated the two primary methods of water vapor calibration currently in use for mobile VLBI baseline measurements: radiometric measurements of the sky brightness near the 22-GHz emission line of free water molecules and surface meteorological measurements used as input to an atmospheric model.

Based upon a limited set of nine baselines, it is shown that VLBI data calibrated with water vapor radiometer measurements can, in some cases, provide a significantly better fit to the theoretical delay model than the same data calibrated with surface meteorological measurements. A detailed error model for the vertical baseline component indicates that the current mobile VLBI technology should allow accuracies of order 5 cm with water vapor radiometer calibration and 10 cm when surface meteorological calibration is used. A statistical analysis of the results of repeated measurements of the 336-km baseline from Big Pine, California, to Pasadena, California, shows the scatter in the vertical baseline component to be consistent with this model. A similar model for a hypothetical GPS baseline mesurement is presented. A covariance analysis based upon this model shows current GPS technology to be capable of accuracies in the vertical baseline component comparable to present mobile VLBI measurements.

Expected improvements in system components for both technologies should allow reduction of the uncertainty in the vertical component to less than 3 cm by 1989.