The Combined Effect of Periodic Signals and Noise on the Dilution of Precision of GNSS Station Velocity Uncertainties

Station velocity uncertainties determined from a series of Global Navigation Satellite System
(GNSS) position estimates depend on both the deterministic and stochastic models applied to the
time series. While the deterministic model generally includes parameters for a linear and several
periodic terms, the stochastic model is a representation of the noise character of the time series in
form of a power-law process. For both of these models the optimal model may vary from one time
series to another while the models also depend, to some degree, on each other. In the past various
power-law processes have been shown to fit the time series and the sources for the apparent
temporally-correlated noise were attributed to, for example, mismodelling of satellites orbits,
antenna phase centre variations, troposphere, Earth Orientation Parameters, mass loading effects
and monument instabilities.