Atmospheric Pressure Loading - Atmospheric Effects in Space Geodesy - page 149

Ta b l e 1 C 1 coefficients at some space geodetic stations in mm/hPa

Station GPS VLBI Model

(van Dam et al. 1994 ) (MacMillan and Gipson 1994 ) (van Dam et al. 1994 )

Fairbank − 0.59 ± 0.14 − 0.35 ± 0.07 − 0.44 ± 0.006

Onsala − 0.00 ± 0.11 − 0.16 ± 0.13 − 0.29 ± 0.005

Wettzell − 0.30 ± 0.17 − 0.53 ± 0.80 − 0.43 ± 0.008

Goldstone − 0.80 ± 0.47 − 0.30 ± 0.13 − 0.40 ± 0.019

Kokee 0.35 ± 0.77 − 0.49 ± 0.35 − 0.11 ± 0.024

specify the scale length r o . Their regression coefficient C 1 (mm/hPa) changes from
approximately 5500 km.

The work of Rabbel and Zschau ( 1985 ) had been extended by determining the C 1
coefficient from the vertical displacements as deduced by VLBI (van Dam and Wahr
1987 ; MacMillan and Gipson 1994 ; Petrov and Boy 2004 ) and GNSS (van Dam
and Herring 1994 ;vanDametal. 1994 ; Kaniuth and Vetter 2006 ; Dach et al. 2011 )
observations. The C 1 coefficient determined by van Dam and Herring ( 1994 ) and
MacMillan and Gipson ( 1994 ) are in the range of

− 0 . 1mm/hPa at r o = 160 km to − 0 . 9mm/hPa at r o =

6mm/hPa for inland
sites, which corresponds to scale lengths r o of 1000-2000km (synoptic scale) in the
simple Gaussian pressure model of Rabbel and Zschau ( 1985 ). Therefore, most of the
variance of APL displacements is determined by synoptic scale pressure variations.
This is reasonable since the largest surface pressure variations are synoptic. In most
regions of the Earth it should be a good approximation to model the loading effects
at a site using only the site pressure.
Table 1 shows the C 1 coefficients at some fundamental stations derived by lin-
ear regression between GPS or VLBI vertical positions and local pressure as well
as regression between modeled vertical displacement (derived using the method in
Sect. 2.1 ) and local pressure. The coefficients determined by VLBI observations
more closely match the coefficients predicted by the model than the GPS results.
This may indicate that the loading signal is correlated with another signal in the
GPS data processing. GPS or VLBI vertical position estimates and modeled vertical
displacements produce different coefficients for Kokee Park, which may be due to
inverted barometer effects as this station is located on Kauai, a rather small island in
the Pacific Ocean. As Rabbel and Zschau ( 1985 ) noted the simple loading functions
as given by Eq. ( 12 ) can only be applied to anomalous pressure on the continental
surface far from any coastlines. On the ocean floor, passing cyclones cause a more
complicated effective pressure distribution due to reaction of the water masses. In
general, this reaction is dynamical and is affected by water depth, geometry of the
coastlines, velocity of the cyclone in a highly complex way. Without any dynamical
effects the ocean would react to air pressure changes like an inverse barometer and
would compensate an air pressure low by raising the water level so that there is no
pressure change on the ocean floor.

− 0 . 4to − 0 .