Earth deformation signals caused by atmospheric pressure loading are detected in vertical position estimates at Global Positioning System (GPS) stations. Surface displacements due to changes in atmospheric pressure account for up to 24% of the total variance in the GPS height estimates. The detected loading signals are larger at higher latitudes where pressure variations are greatest; the largest effect is observed at Fairbanks, Alaska (latitude 65°), with a signal RMS of 5 mm. Out of 19 continuously operating GPS sites (with a mean of 281 daily solutions per site), 18 show a positive correlation between the GPS vertical estimates and the modeled loading displacements. Accounting for loading reduces the variance of the vertical station positions on 12 of the 19 sites investigated. Removing the modeled pressure loading from GPS determinations of baseline length for baselines longer than 6000 km reduces the variance on 73 of the 117 baselines investigated. The slight increase in variance for some of the sites and baselines is consistent with expected statistical fluctuations. The results from most stations are consistent with ∼65% of the modeled pressure load being found in the GPS vertical position measurements. Removing an annual signal from both the measured heights and the modeled load time series leaves this value unchanged. The source of the remaining discrepancy between the modeled and observed loading signal may be the result of (1) anisotropic effects in the Earth's loading response, (2) errors in GPS estimates of tropospheric delay, (3) errors in the surface pressure data, or (4) annual signals in the time series of loading and station heights. In addition, we find that using site dependent coefficients, determined by fitting local pressure to the modeled radial displacements, reduces the variance of the measured station heights as well as or better than using the global convolution sum.

Loading of the Earth by the temporal redistribution of global atmospheric mass is likely to displace the positions of geodetic monuments by tens of millimeters both vertically and horizontally. Estimates of these displacements are determined by convolving National Meteorological Center (NMC) global values of atmospheric surface pressure with Farrell's elastic Green's functions. An analysis of the distances between radio telescopes determined by very long baseline interferometry (VLBI) between 1984 and 1992 reveals that in many of the cases studied there is a significant contribution to baseline length change due to atmospheric pressure loading. Our analysis covers intersite distances of between 1000 and 10,000 km and is restricted to those baselines measured more than 100 times. Accounting for the load effects (after first removing a best fit slope) reduces the weighted root‐mean‐square (WRMS) scatter of the baseline length residuals on 11 of the 22 baselines investigated. The slight degradation observed in the WRMS scatter on the remaining baselines is largely consistent with the expected statistical fluctuations when a small correction is applied to a data set having a much larger random noise. The results from all baselines are consistent with ∼60% of the computed pressure contribution being present in the VLBI length determinations. Site dependent coefficients determined by fitting local pressure to the theoretical radial displacement are found to reproduce the deformation caused by the regional pressure to within 25% for most inland sites. The coefficients are less reliable at near coastal and island stations.

The fifth in the series of International Comparisons of Absolute Gravimeters (ICAG) was held at the Bureau International des Poids et Measures (BIPM) in November 1997. Fifteen absolute gravimeters participated in the comparison. The mean gravity value obtained at station A (0.9 m) at the BIPM was found to be 980 925 707.8 µGal with a standard uncertainty of 2.8 µGal. This is consistent with the results obtained in previous comparisons at this site. Conclusions based on the analysis of the present results and proposals for future activities are presented.

S U M M A R YApproximately one year's worth of altimeter-derived sea-surface heights are compared with global sea-level pressure fields to verify the open ocean inverted barometer response (-1 cm mb-'). When pressure is fit to the sea-surface height along individual altimeter tracks, the response is found to be only 60-70 per cent of the theoretical response or approximately -0.6 to -0.7cmmb-'. Fits at fixed geographic locations show a clear dependence on latitude. There is a steady decrease in the absolute value of the regression coefficient between 70" and 20", and then an abrupt increase again closer to the equator. A simple error analysis demonstrates that errors in the pressure data would reduce the along-track regression values, as is observed, and could produce a similar latitude dependence. But, the errors are unlikely to be large enough to explain the entire departure from inverted barometer. We estimate that pressure errors are apt to perturb the along-track track results by no more than about 0.1-0.2 cm mb-'.The possibility that the remaining disagreement is due to a global coherence between wind-and pressure-driven sea-surface height variability is considered. Winds driven by the pressure gradients of synoptic storms induce a sea-surface height response that is opposite in direction to that caused by the pressure cell. The wind-driven response is estimated for a stationary storm over a homogeneous barotropic ocean and for a moving storm over a two-layer baroclinic ocean by modelling the pressure cell as an idealized Gaussian distribution. The model results indicate that the wind-induced sea-surface height depends on both the radius and the translational velocity of the pressure cell. But, the winds associated with storms moving at average speeds of 10ms-' are apt to lower the theoretical pressure response in the model by only approximately 0.1 cm mb-'. The surface stress associated with those winds has the same latitudinal trend between 70" and 20" as the regression coefficients. But, the response of the ocean to that stress does not appear to exhibit the same trend. Nevertheless, the abrupt change in the regression coefficients near the equator suggests the apparent non-inverted barometer response may reflect a real change in sea-surface height related to atmospheric forcing (though the results near the equator are not as well defined as those at higher latitudes).