Monument-antenna effects on GPS coordinate time series with application to vertical rates in Antarctica

We examine the electromagnetic coupling of aGPS antennamonument pair in terms of its simulated affecton long GPS coordinate time series. We focus on the Earthand Polar Observing System (POLENET) monument designwidely deployed in Antarctica and Greenland in projectsinterested particularly in vertical...

Full description

Bibliographic Details
Published in:Journal of Geodesy
Main Authors: King, MA, Bevis, M, Wilson, T, Johns, B, Blume, F
Format: Article in Journal/Newspaper
Language:English
Published: Springer-Verlag 2012
Subjects:
Online Access:https://doi.org/10.1007/s00190-011-0491-x
http://ecite.utas.edu.au/89659
Description
Summary:We examine the electromagnetic coupling of aGPS antennamonument pair in terms of its simulated affecton long GPS coordinate time series. We focus on the Earthand Polar Observing System (POLENET) monument designwidely deployed in Antarctica and Greenland in projectsinterested particularly in vertical velocities. We base ourtests on an absolute robot calibration that included the top∼0.15 m of the monument and use simulations to assessits effect on site coordinate time series at eight representativePOLENET sites in Antarctica over the period 2000.02011.0. We show that the neglect of this calibration wouldintroduce mean coordinate bias, and most importantly forvelocity estimation, coordinate noise which is highly sensitiveto observation geometry and hence site location andobservation period. Considering only sub-periods longer than2.5 years, we show vertical site velocities may be biased byup to 0.4 mm/year, and biases up to 0.2 mm/year may persistfor observation spans of 8 years. Changing between uniformand elevation-dependent observation weighting altersthe time series but does not remove the velocity biases, nordoes ambiguity fixing. The effect on the horizontal coordinatesis negligible. The ambiguities fixed series spectrashownoise between flicker and randomwalk with near-whitenoise at the highest frequencies, with mean spectral indices(frequencies <20 cycles per year) of approximately −1.3(uniform weighting) and −1.4 (elevation-dependent weighting).While the results are likely highly monument specific,they highlight the importance of accounting for monumenteffects when analysing vertical coordinate time series andvelocities for the highest precision and accuracy geophysicalstudies.