Return to search

Environmental and instrumental effects on high precision gravimetry : a case study in Britain

This thesis investigates the removal of environmental and instrumental effects from high precision gravity records and applies the findings to measurements made in Britain. Ocean loading calculations have been developed by manipulating ocean tide model grids to be of variable size. This has improved the fit to irregular coastlines and the approximation of grid cell masses to point mass loads. Ocean loading is calculated for 12 components at the 69 sites of the British Precise Gravity Network (BPGN). The new ocean loading correction amplitude for Newlyn has increased from 16 to 23 <i>μ</i>gal; the set standard deviation of a 24 hour FG5 absolute gravity data set reduces from 4.3 to 1.8 <i>μ</i>gal. Adjustment of the BPGN with the new correction reduces the mean site standard error from 3.32 ± 0.60 <i>μ</i>gal to 2.96 ± 0.46 <i>μ</i>gal. FG5-103 Bidston measurements show a seasonal change of 8 <i>μ</i>gal. Water table data after cor­rection for air pressure show a correlation with gravity, having a lag of 3 months. Using 6 months of continuous LaCoste & Romberg (LCR) Earth-tide gravimeter data, the air pressure admittance at Bidston varies seasonally from 0.35 to 0.50 <i>μ</i>gal mbar<sup>-</sup><sup>1</sup>. Frequency dependent admittances correcting for instrumental humidity effects and environmental air pressure and rainfall effects removed structure in the data. Admittances derived from LCR Earth-tide data for air pressure and rainfall do not improve the absolute data. Using synthetic data, the frequencies of damped sinusoids found from FG5 residuals are dif­ferent to those originally present. Noise found at 9.5<i> </i>Hz causes gravity changes of < 1 <i>μ</i>gal when fitted to the equation of motion with real data. However, gravity changes by 6.3 <i>μ</i>gal upon removal of its cause. Fitting frequencies from 4 -13 Hz to this data creates changes in the order of 10 <i>μ</i>gal. Existing noise correction algorithms may miss errors of this size.

Identiferoai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:652566
Date January 1999
CreatorsHopewell, Hugh
PublisherUniversity of Edinburgh
Source SetsEthos UK
Detected LanguageEnglish
TypeElectronic Thesis or Dissertation
Sourcehttp://hdl.handle.net/1842/12086

Page generated in 0.0024 seconds