GPS tropospheric modelling: new developments and insights

Date

2017

Authors

Masoumi, Salim

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

GPS is widely used to monitor temporal and spatial variations of Earth’s crust, oceans and atmosphere. Of particular interest to this research is the use of GPS for studying variations in the Earth’s lower atmosphere. While there have been significant advances in the techniques and models used in GPS analyses over the past two decades, there is still room for improvement. In particular, observations at very low elevation angles still suffer greatly from modelling errors. These low-elevation observations provide useful information about the moisture content of the atmosphere and its variability around a GPS station, and are thus valuable data for meteorological studies if properly modelled. The main focus of this thesis is on optimization of the techniques and models used in GPS analysis for more accurate estimates of the tropospheric delays. Particular attention is paid to modelling low-elevation observations and challenging weather conditions. Throughout the thesis, we investigate several different aspects of modelling techniques and how each of them affect the tropospheric estimates. By applying a previously developed empirical model [Moore, 2015], the site-specific errors are shown to have large impacts on the tropospheric delay estimates: empirical mitigation of site-specific errors leads to improved repeatabilities of heights and tropospheric zenith delays for the majority of the stations in our analysis. The empirical site-specific model also significantly reduces the sensitivity of tropospheric zenith delay estimates to the choice of elevation cut-off. Another important potential source of error, the GPS estimates of tropospheric horizontal gradients are shown to be more accurate than the model values currently available. However, the conventional two-axis planar model of gradients does not accurately represent the actual gradients of the refractivity under weather conditions with asymmetric horizontal changes of refractivity. Such abnormal conditions may occur due to topography-driven gravity waves in the troposphere, and the mismodelled tropospheric horizontal gradients induce errors in the parameter estimates, sometimes leading to skewed position time series and inaccurate tropospheric zenith delays. A new parametrization of tropospheric gradients whereby an arbitrary number of gradients are estimated as discrete directional wedges is shown via both simulations and real case studies to largely improve the accuracy of recovered tropospheric zenith delays in asymmetric gradient scenarios. The new directional model significantly improves the repeatabilities of the station height time series in asymmetric gradient situations while causing slightly degraded repeatabilities for the stations in normal symmetric gradient conditions. The constraints on the temporal variations of the tropospheric delays are also investigated. It is shown via simulations and real experiments that it is generally preferable to avoid constraints on both tropospheric zenith delays and horizontal gradients. However, since the conventional model of horizontal gradients oversimplifies the horizontal variations of the refractivity in asymmetric gradient conditions, it is important to use a more complete model of gradients like the directional gradient model introduced in this thesis in conjunction with the relaxed constraints to avoid errors caused by the simplifying assumption of symmetric gradients by the conventional model.

Description

Keywords

Global Positioning System, Tropospheric delay, Site-specific errors, Tropospheric horizontal gradients, Directional gradient model, Temporal constraints

Citation

Source

Type

Thesis (PhD)

Book Title

Entity type

Access Statement

License Rights

Restricted until