Title

Utility of Ionosphere and Troposphere Models for Extending the Range of High-Accuracy GPS

Date of Award

2007

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Marine Science

First Advisor

Stephan Howden

Advisor Department

Marine Science

Abstract

This dissertation studied the use of NOAA real-time ionosphere and troposphere products in extending the range of long-baseline, high-accuracy DGPS for real-time positioning. The question being addressed by this work is; can existing real-time ionosphere and troposphere models reduce the observation uncertainties to the level where they can be used to reliably resolve integer ambiguities, in real-time, over long baselines (>30km). In-house GPS processing software (USM_OTF) was developed to ingest the models and compute epoch-to-epoch, float and fixed ambiguity position solutions. Single baseline processing, ranging from 20 to 740 kin, over several days in four separate sessions (July 2004, January 2005, August 2005 and July 2006) incorporating four regions of the U.S.A. (Michigan, California, Central and the South East), were evaluated. The first session looked at the NOAA real-time troposphere model and the second session looked at the NOAA real-time ionosphere model. The third and fourth sessions looked at the use of both the NOAA real-time ionosphere and troposphere models. Results showed that the NOAA troposphere model reduced the height bias uncertainty by up to 30 cm, under high activity conditions. They also showed that the troposphere model increased the uncertainty standard deviation under these high activity conditions. The results from the first tests of the real-time NOAA ionosphere model showed that, due to satellite coverage issues, it produced worse results than other real-time models. The NOAA model suffered from lack of satellite coverage corrections, especially in areas near the limits of the model, and at the beginning and end of a satellite's flight path. A reduction in satellite numbers lead to weaker geometry and less reliable position solutions. These tests showed that it was better to provide less accurate ionosphere estimates than to leave the satellites out of the solution. These problems were addressed by NOAA prior to the final tests. The final ionosphere testing results showed that, overall, the ionosphere-free float solution with the NOAA troposphere model produced the best results. The float solution determined from the four observables (L1, L2, P1 and P2) using the real-time ionosphere model, when combined with the L1-L2 observation, produced results similar to, but slightly worse than, the ionosphere-free solution. For the short (∼20 km) baselines, the four-observable, fixed solution produced the best results, but as the range increased the ability of the ambiguity algorithm to resolve the correct integers, reliably, was degraded due to un-modeled residual ionosphere and troposphere effects. The real-time NOAA ionosphere and troposphere models, along with the methods developed for this research, greatly reduce the position uncertainty for both float and fixed solutions. However, residual effects hamper the processes ability to reliably fix ambiguities. The four-observation float solutions are comparable to the ionosphere-free solutions, but if the float solutions do not lead directly to fixed solutions, the computational and logistical overhead associated with using ionosphere models is not worth the effort, for real-time differential applications.