Hurst, K.J., 1987, The measurement of vertical crustal deformation: New York City, New York, Columbia University, Ph.D. dissertation, 197 p., illust.
The measurement of vertical crustal motion can be accomplished most simply by measuring the height of a point at two different times. There are many methods for doing this ranging from traditional spirit leveling to radio-interferometric methods using extraterrestrial radio sources. These methods do not all yield the same height for a given point at a given time. Spirit leveling yields heights referenced to the geoid whereas the methods using satellites or extragalactic radio sources yield heights referenced to the ellipsoid which is in turn referenced to the center of mass of the Earth. Previous attempts to apply the principles of hydrostatic leveling to precision geodesy have been limited by the uniformity of the fluid density attainable in field environments. This is largely due to the effects of temperature variations in the fluid tube. We have succeeded in limiting the density variations to less than 1 ppm by using water maintained near its maximum density at 3.98°C inside a counterflow heat exchanger. We used a finite element computer model of the system to extrapolate the performance of a 14 m prototype to instruments up to 1 km long in environments between $-$40$\sp\circ$ and +50$\sp\circ$C. In our attempt to use the 14 m prototype pressure-transfer level we were unable to exploit the density stability that we had achieved due to inadequacies in available pressure gauges. We have installed an array of sea level gauges in the Shumagin Islands Seismic Gap, Alaska to monitor the area for vertical deformation expected to be associated with a great subduction earthquake forecast for the region within the next 10 to 20 years. An examination of noise levels in the data suggests that the network is providing relative vertical deformation data that have a lower noise level than data available from any other source. The greatest contribution to the vertical error of GPS baselines comes from the variable delay caused by water vapor in the troposphere. An investigation of the effect of this delay on synthetic GPS data yields an improved estimate of this error when surface meteorology measurements are used to model the delay.
Theses and Dissertations
