Geophysical Modeling of Typical Cavity Shapes to Calculate Detection Probability and Inform Survey Design

Abstract

Feasibility analysis of near-surface cavity detection is presented using modeling of the gravity, gravity gradient, magnetic, magnetic gradient, and ground penetrating radar techniques. The geophysical signal is modeled over typical cavity shapes in three-dimensional subsurface environments with varying geologies and survey parameters. The cavity detection probability is calculated for each technique in the outlined environments and these values are used to aid technique choice, assess the feasibility of cavity detection, assess the limits of detection for each technique, and optimise survey design before entering the field. Tests in a range of conditions show that technique choice is conditional to site characteristics and site parameters, and highlight the need for modeling in the desk study stage of site investigation and survey design. Detection probability results show that standard survey direction practice in magnetometry is not always optimal, and demonstrate the importance of site specific noise level consideration. Comparisons with case study measurements demonstrate that the modelling and subsequent detection probability calculation chose appropriate techniques and survey parameters, but also highlights the limitations of the method.