Effective search for a buried layer: An approach to experimental design in geophysics

Abstract

A practical method of solution is developed of some design and analysis problems of geophysical surveys. Starting with quantitative measures of parameter resolution in general models, we derive expressions for parameter error bounds which are related to parameter resolution and measurement noise. The methods are applied to three problems involving buried layers: a deep permafrost layer which underlies a conductor, a buried electromagnetic waveguide, and a layer beneath the Moho. We assume that magnetotelluric data and several groups of DC resistivity measurements are available. For a realistic level of data inaccuracy it is shown that both DC and MT data are essential to resolve resistivity in the waveguide. DC data may be helpful in the deep permafrost case. They add no information in the Moho model. This approach can be used to select methods for a wider variety of geophysical problems, including those of exploration and engineering. It also suggests that existing data, as for example those acquired in the waveguide search, might be reevaluated to good effect.