Error Analysis of Weekly Station Coordinates in the DORIS Network
Williams, Simon D.P.; Willis, Pascal. 2006 Error Analysis of Weekly Station Coordinates in the DORIS Network. Journal of Geodesy, 80 (8-11). 525-539. 10.1007/s00190-006-0056-6Full text not available from this repository.
Twelve years of DORIS data from 31 selected sites of the IGN/JPL (Institut Géographique National/Jet Propulsion Laboratory) solution IGNWD05 have been analysed using maximum likelihood estimation (MLE) in an attempt to understand the nature of the noise in the weekly station coordinate time-series. Six alternative noise models in a total of 12 different combinations were used as possible descriptions of the noise. The six noise models can be divided into two natural groups, temporally uncorrelated (white) noise and temporally correlated (coloured) noise. The noise can be described as a combination of variable white noise and one of flicker, first-order Gauss–Markov or power-law noise. The data set as a whole is best described as a combination of variable white noise plus flicker noise. The variable white noise, which is white noise with variable amplitude that is a function of the weekly formal errors multiplied by an estimated scale factor, shows a dependence on site latitude and the number of DORIS-equipped satellites used in the solution. The latitude dependence is largest in the east component due to the near polar orbit of the SPOT satellites. The amplitude of the flicker noise is similar in all three components and equal to about 20 mm/year1/4. There appears to be no latitude dependence of the flicker noise amplitude. The uncertainty in rates (site velocities) after 12 years is just under 1 mm/year. These uncertainties are around 3–4 times larger than if only variable white noise had been assumed, i.e., no temporally correlated noise. A rate uncertainty of 1 mm/year after 12 years in the vertical is similar to that achieved using Global Positioning System (GPS) data but it takes DORIS twice as long to reach 1 mm/year than GPS in the horizontal. The analysis has also helped to identify sites with either anomalous noise characteristics or large noise amplitudes, and tested the validity of previously proposed discontinuities. In addition, several new offsets were found in the time-series that should be used or at least flagged in future work.
|Item Type:||Publication - Article|
|Digital Object Identifier (DOI):||10.1007/s00190-006-0056-6|
|NORA Subject Terms:||Marine Sciences|
|Date made live:||31 Aug 2012 13:27|
Actions (login required)