HR: 0800h
AN: G51B-0836 [Abstracts]
TI: Analysis of GPS Coordinate Time Series for Tide Gauge Benchmark Monitoring
AU: * Jia, M
EM: minghai.jia@ga.gov.au
AB:
Long term sea level change from tide gauges depends heavily on tide gauge benchmarks' stability. Therefore, monitoring of
vertical motion of tide gauge benchmarks is crucially important to obtain reliable sea level change. Continuous GPS is a
widely used technique for monitoring likely tide gauge benchmark vertical motion. However, analysis of continuous GPS
coordinate time series still is a major challenge due to influence of systematic errors such as GPS satellite orbital errors
and atmosphere loading displacement.
Up to 10 years of daily GPS coordinate time series of over 60 continuous GPS sites in the Australian region are analysed to
obtain reliable vertical velocities of tide gauge benchmarks using in-house software and a site by site analysis strategy.
Three functional models, including a linear fit model, a linear plus annual signal model, and a linear plus annual and
semi-annual signal model, are tested respectively. Results show that if a data time span is over 5 years the estimated
velocity almost does not change respective to the model used. However, the uncertainty of the estimated velocity heavily
depends on the length of a data time span. When only a white stochastic model noise is used the uncertainty of the estimated
velocity is always too optimistic. A stochastic model from a combination of white noise plus flicker noise can provide more
reliable uncertainty information of the estimated velocity than a stochastic model from white noise only.
In addition, the terrestrial reference frame can influence the vertical velocity estimation seriously. Therefore, careful
selection of reference frame sites by statistical method is also very important to obtain reliable vertical velocities of
tide gauge benchmarks.
DE: 1222 Ocean monitoring with geodetic techniques (1225, 1641, 3010, 4532, 4556, 4560, 6959)
SC: Geodesy [G]
MN: Fall Meeting 2005