HR: 0800h
AN: G41B-0355 [Abstracts]
TI: The Effect of Processing Technique and Reference Frame Definition on Noise in CGPS Position Time
Series
AU: * Teferle, F N
EM: norman.teferle@nottingham.ac.uk
AF: Institute of Engineering Surveying and Space Geodesy,
University of Nottingham, University Park, Nottingham, NG7 2RD
United Kingdom
AU: Williams, S D
EM: sdwil@pol.ac.uk
AF: Proudman Oceanographic Laboratory, Joseph Proudman Building,
6 Brownlow Street, Liverpool, L3 5DA
United Kingdom
AU: Kierulf, H P
EM: halfdan.kierulf@statkart.no
AF: Norwegian Mapping Authority, Geodetic Institute, Kartverksveien 21,
Service Box 15, Honefoss, N-3504
Norway
AU: Bingley, R M
EM: richard.bingley@nottingham.ac.uk
AF: Institute of Engineering Surveying and Space Geodesy,
University of Nottingham, University Park, Nottingham, NG7 2RD
United Kingdom
AU: Plag, H
EM: hpplag@unr.edu
AF: Nevada Bureau of Mines and Geology, University of Nevada, Mail Stop 178, Reno, Nev 89557-0088
United States
AB:
In this presentation we investigate the effects of GPS processing techniques and strategies, and the related reference frame
realization on the stochastic properties of continuous GPS (CGPS) position time series. It was of particular interest to
establish whether and how different GPS processing strategies, e.g. double differencing (DD) and precise point positioning
(PPP), the use of different orbit and clock products, and/or the definition of the reference frame, partly dependent on the
applied strategy, affect the colored noise content of time series. We used CGPS position time series from 14 different
solutions obtained from six different analysis centers as part of the European Sea Level Service - Research Infrastructure
project (ESEAS-RI) using the GIPSY OASIS II, GAMIT and Bernese GPS softwares. All time series analyzed have at least three
years of data for the period between 2000 and 2005. Furthermore, a selected set of position time series was also analyzed
using Empirical Orthogonal Function (EOF) analysis. The noise content of the first 15 modes, representing the
solution-specific common mode time series for each of the selected solutions were then also investigated for colored noise.
Using Maximum Likelihood Estimation (MLE) a white, a white plus flicker, a white plus power-law and a white plus first-order
Gauss-Markov (FOGM) noise model were fitted to the position and EOF time series data. For both the position and EOF time
series the parameter model included a constant, a rate and harmonic terms with annual, semi-annual, 4-monthly, 3-monthly,
2-monthly and 13.66 day periods. Position jumps were modeled at logged epochs or at visible discontinuities in the time
series. The MLE showed that in most cases the best fitting noise model is a combination of white plus power-law noise. This
model is closely followed by the combination of white plus flicker and white plus FOGM noise. The noise properties of the
EOF time series follow predominantly a white plus power-law character, with the first few modes indicating a white plus
flicker noise behavior. In general we find that DD solutions contain less noise than PPP solutions and that regional
reference frame definitions further reduce the amount of noise in the time series.
DE: 1200 GEODESY AND GRAVITY
DE: 1229 Reference systems
DE: 1243 Space geodetic surveys
DE: 1294 Instruments and techniques
SC: Geodesy [G]
MN: Fall Meeting 2005