Geophysical applications of the global positioning syste

Observation of land surface motions plays a critical role in the study of a wide range of geophysical phenomena. The Global Positioning System (GPS), designed by the US Department of Defense for military and civilian navigation and positioning, has been used to measure crustal motion. The ever expanding global network of GPS sites improves both temporal and spatial resolution of geophysical phenomena.;I assess the noise characteristics in time series of daily position estimates for 23 GPS stations with three years of data, using spectral analysis and Maximum Likelihood Estimation. A combination of white plus flicker noise can be used to model the noise characteristics of all three position components. Both white and flicker noise amplitudes are smallest in the north component and largest in the vertical component. The white noise part of the vertical component is higher for tropical stations ($pm$23$spcirc$ latitude) compared to mid-latitude stations, which may be related to high wet tropospheric path delay and/or higher variability. Velocity error in the coordinate time series may be underestimated by factors of 5-11 if a pure white noise model is assumed.;I analyze data from 20 permanent GPS stations broadly distributed through the interior of the North American plate, and use the resulting velocities to estimate an Euler vector describing motion of "stable" North America as a single rigid plate. The site velocities fit the single plate model with a mean residual of 1.3 mm/yr. The residuals do not appear to reflect post-glacial rebound, and tests for differential motion between eastern and western North America at the New Madrid seismic zone show no resolvable motion within uncertainties. The residuals more likely reflect observational error, and thus our estimate of the stability of the plate interior is likely an upper bound. In addition, I found that GPS velocity errors estimated with the white plus flicker noise model fit the observations significantly better than a pure white noise model.;Although the "Wilson cycle", in which continents rift apart to form new ocean basins which later close, is a vital feature of plate tectonics, the dynamics of rifting are poorly understood. Many of our ideas about continental rifting come from the best present example, the East African Rift, that separates East Africa (Somalia) from West Africa (Nubia). It is not clear whether the rift is failing, and will ultimately leave a fossil rift like those common in the geologic record, or whether it will succeed in splitting Africa into two plates separated by an ocean. To study this issue, we inverted space geodetic data to estimate an Euler vector describing relative motion between Nubia and Somalia. The Euler vector predicts extension directed between E-W and WNW-ESE at rates of 5-7 mm/yr, considerably higher than geologic estimates averaged over tens of millions of years. The rate of extension appears to have accelerated since 5-6 million years ago, when motion of the African plate changed.