The Earth's crust is deforming continuously due to plate tectonics. Deformation at plate boundaries causes volcanoes and most destructive earthquakes. Monitoring such deformation is essential to gaining an insight into the mechanisms of plate tectonics. Deformation analysis is one of the most important aspects of geodetic research. Space geodesy, with which long baselines can be measured to millimetre accuracy, plays an important role in determining crustal deformation parameters, since deformation in general, means a change in geometric configuration. The main deformation monitoring problem is to determine the spatial relationship of a set of object points relative to a number of reference points. Ideally reference and object point observations are made at regular intervals. After mathematical adjustment of each epoch's observations, which includes 'data snooping', a displacement vector is obtained by simply differencing the estimated coordinates at consecutive epochs. The use of thismethod also however, increases the noise level. In this thesis, the author proposes a deformation analysis technique which mainly uses a Kalman Filter. However, Kalman filter estimation may not be optimal if local movement occurs between observation epochs. To overcome this kind of deficiency, two sub-optimal filters have been proposed: Fading Memory Filter and Adaptive Kalman Filter for a System with Unknown Measurement Bias. These two filtering techniques have been used in this research and tested on real/simulated data based on the EASTMED project. In addition to this, data from the EUREF Permanent GPS Network, and from the UK Tide Gauge Monitoring C Project are also processed and the result presented.
Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:519309 |
Date | January 1998 |
Creators | Çelik, Cahit Tăgi |
Publisher | University of Nottingham |
Source Sets | Ethos UK |
Detected Language | English |
Type | Electronic Thesis or Dissertation |
Source | http://eprints.nottingham.ac.uk/11438/ |
Page generated in 0.001 seconds