Toggle Main Menu Toggle Search

Open Access padlockePrints

Parameter Bias When Processing GPS Data Containing Unmodelled Vertical Signals: Sensitivity Analysis

Lookup NU author(s): Professor Matt King


Full text for this publication is not currently held within this repository. Alternative links are provided below where available.


We have identified biases in sub-daily estimates of position obtained from GPS data using batch least squares due to the presence of unmodelled vertical signals in the GPS data. The estimation of sub-daily position is commonly used for ocean loading and ice movement studies, for example. The biases are most evident in the horizontal components of the coordinates and in the case of the unmodelled vertical signal being periodic (i.e., tidal), the biases are also periodic with magnitudes of about 40-50% of the amplitude of the vertical periodic signal. The shape of the periodic horizontal signals is a function of the first derivative and the square of the unmodeled vertical signal. The artifacts are predominantly evident in the east-west component and can be removed by fixing the carrier phase ambiguities to integer values. These findings have important consequences for GPS investigations into the horizontal components of ocean tide loading, for example, or for historic campaign-style measurements where ocean loading was not modelled and ambiguities were not fixed to integer values. We present an investigation into this problem by way of a sensitivity analysis. The sensitivity analysis is compatible with the results obtained using field data. In particular, we discuss the latitudinal variation of the artifacts, the effect of different ambiguity resolution success rates and the `averaging effect' as session lengths are increased. Finally, we investigate the influence of unmodeled vertical linear and periodic signals in 24 hour batch least squares solutions.

Publication metadata

Author(s): King MA, Coleman R

Publication type: Conference Proceedings (inc. Abstract)

Publication status: Published

Conference Name: Eos Transactions: Fall Meeting

Year of Conference: 2002

Pages: Abstract G52A-0972

Publisher: American Geophysical Union