HR: 0830h
AN: G21C-0283    [PDF]
TI: Repeated High-Precision Gravity and GPS Measurement Techniques
AU: * Gettings, P
EM: gettings@mines.utah.edu
AF: Dep't of Geology & Geophysics University of Utah, 135 S 1460 E, Salt Lake City, UT 84112 United States
AU: Harris, R N
AF: Dep't of Geology & Geophysics University of Utah, 135 S 1460 E, Salt Lake City, UT 84112 United States
AU: Allis, R
AF: Utah Geological Survey, 1594 North Temple, Salt Lake City, UT 84114 United States
AU: Chapman, D S
AF: Dep't of Geology & Geophysics University of Utah, 135 S 1460 E, Salt Lake City, UT 84112 United States
AB: Repeated high-precision gravity and GPS measurements are becoming a common tool for tracking changes in subsurface reservoirs. Despite this, there is little literature which discusses measurement techniques and the expected errors. Our research has focused on improving measurement techniques to be applied to ground water and geothermal steam reservoirs, including quantifying the minimum error levels with modern equipment. We applied these methods in two studies: ground water monitoring of the southern Salt Lake valley, Utah, USA, and steam monitoring of The Geysers geothermal field, California, USA. Gravity measurements using modern relative high-precision meters, such as Scintrex CG-3Ms or L\&R E series, can now be routinely made to an accuracy of 5 $\mu$Gal. Such accuracy requires the use of time series analysis at each station, and non-linear instrument drift functions. Modern computerized meters are capable of internally storing a time series of measurements for each station; older meters can often be fitted to log such data to a field computer. This time series, typically of 10-15 minute duration in our work, can then be analyzed in several ways to produce stable estimates of the gravity reading. In particular, our research has emphasized using a weighted arithmetic average (for long occupations), or a Thiele extrapolation scheme (for shorter station occupations). Instrument drift is removed through a superposition of a linear long-term drift function, and an empirical staircase function formed from differences between repeated station occupations. To achieve high-accuracy GPS measurements while maximizing the number of field stations in a survey, rapid-static measurements are necessary. We have tested the effect of occupation time and processing schemes on the absolute accuracy of the resulting GPS position. Using a post-processing differential method with a fixed (but not necessarily continuous) base station within 15 km, positioning error of $<$4 cm vertical is achievable with 30 minute occupations and broadcast orbital parameters. There is a definite correlation between baseline length and positioning accuracy. Using precision orbital parameters removes this correlation. Occupations of 60 minutes or more also improve accuracy; combining 1 hour occupations with precise orbital data yields a vertical error of $<$3 cm for all stations within 20-30 km of the fixed reference. In both case studies, gravity measurements tracked known mass changes in the reservoir. In the southern Salt Lake valley study, mass changes measured along the line of gravity stations qualitatively tracked water well level changes, but very complicated local hydrology precluded more detailed quantitative comparisons. At the Geysers, the station grid shows coherent spatial signals, with gravity changes that are within measurement error of theoretical predictions.
UR: http://thermal.gg.utah.edu/~gettings
DE: 0920 Gravity methods
DE: 1208 Crustal movements--intraplate (8110)
DE: 1219 Local gravity anomalies and crustal structure
DE: 1294 Instruments and techniques
SC: Geodesy [G]
MN: 2003 Fall Meeting