HR: 0800h
AN: NG31B-0874    [Abstracts]
TI: Optimal Survey Design Using Appraisal Analysis
AU: * Routh, P S
EM: routh@cgiss.boisestate.edu
AF: Dept. of Geosciences, Boise State University, 1910 University Drive, 221A Math & Geoscience Bldg., Boise, ID 83725 United States
AU: Oldenburg, D W
EM: doug@geop.ubc.ca
AF: Dept. of Earth and Ocean Sciences, University of British Columbia, 6339 Stores Road, Vancouver, BC V6T 1Z4 Canada
AB: Geophysical inversion produces images of the Earth model from finite noisy data. The scale of the model that can be resolved by inversion is affected by the sampling strategy of the data acquisition, associated data noise and a priori information that regularizes the ill-posed solution. Therefore an important questions is: How best to design surveys to obtain enhanced resolution in specific regions of the model? Thus the objective in survey design is to determine optimal survey parameters, such as position of sources/receivers and possibly frequencies in EM experiments, that would provide 'better' model resolution in a region of interest. We pose this as an inverse problem by maximising a resolution measure, the point spread function. The point spread function quantifies how an impulse in the true model is observed in the inversion result and hence the goal is to adjust the survey parameters so that the point spread function is as delta-like as possible. This is solved as a nonlinear optimisation problem with constraints on the parameters. Due to the highly nonlinear nature of the problem we examine two approaches for its solution. The first is a local, (Newton) strategy that use a primal interior point method to incorporate bounds on the parameters. We formulate it such that it can be applied to large scale problems. The second approach is a global method, simulated annealing. This proposed method is capable of enhancing resolution locally as well as globally. We examine various aspects of our method such as effects of data noise, data redundancy, multiregion survey design and compare with survey design techniques that enhances model resolution globally. We illustrate our methodology with a cross-well tomography example.
DE: 8180 Tomography
DE: 7294 Instruments and techniques
DE: 3260 Inverse theory
DE: 0915 Downhole methods
DE: 0994 Instruments and techniques
SC: Nonlinear Geophysics [NG]
MN: 2004 AGU Fall Meeting