The primary reporting of the contributions VLBI could make to global change studies was through the NOAA Chapman conference [ Carter, 1991]. Douglas [1991], Sloss [1991] and Douglas [1992] discuss the contributions that VLBI measurements, with sufficient accuracy, could make to the establishing a global height reference frame and conclude that a verification and possible modification of existing post glacial rebound models would be one method for allowing contemporary solid Earth motions to be deconvolved from existing tide gauge estimates of the global sea level rise. There have been recent studies comparing VLBI measurements at northern latitude sites with the expected signals from post glacial rebound. Tushingham [1991] discusses the effects to be expected and Mitrovica et al. [1993], James and Lambert [1993], and Mitrovica et al. [1994] all report on the comparison of VLBI measurements with the signals both in the horizontal and vertical directions expected from post glacial rebound. All of these studies indicate that the precision in height rates of change or of baseline length rates of the change is not yet sufficient to make any definitive statements about the current accuracy of post glacial rebound models. Although there is strong evidence building that the VLBI site at Algonquin Canada is moving at a rate of a few mm/yr under the influence of post glacial rebound relative the VLBI sites in eastern North America at lower latitudes [ Ryan et al., 1993].
The viability of VLBI to make definitive height measurements has been
addressed by vanDam and Herring [1994] and
MacMillan and Gipson [1994] by searching for the elastic
deformation of the Earth's surface due to changes in atmospheric pressure
loading (RMS amplitude of 
5 mm) which should be seen in VLBI height
measurements. Both studies indicate that only about 60% of the expected
atmospheric loading signal is present in the VLBI measurements indicating
either errors in the VLBI measurements that are anticorrelated with the
loading signal or noise in the estimates of the loading signals.
Anticorrelated signals might be expected through atmospheric delay modeling
errors because of possible gradients in the atmospheric delay associated with
the passage of pressure fronts which themselves cause most of the changes in
the loading signal. Some combination of both reasons is likely. In both
studies, it was concluded that atmospheric loading is currently a small
percentage (
20%) of the noise on VLBI height measurements. For the most
recent VLBI data with improved vertical precision this percentage could rise
to as much as 50%.
In another study of the ability of VLBI to maintain a long term reference system, Herring [1992 b] reports of submillimeter repeatability of the relative positions of a pair of radio telescopes in Massachusetts over a 20 year interval as a measure of the long stability of the radio telescopes. The conclusion from this study was that despite the size of VLBI radio telescopes, long term stability of the reference point appear adequate for making submillimeter per year velocity measurements over long periods of time. In another study of accuracy, Ray et al. [1991] reports on the comparison of VLBI and satellite laser ranging results and confirmed agreement between the two systems at about the 10 mm level.