Osaka Electro-Communication University

This paper was selected for presentation by the JFES program committee following the review of abstract submitted by author(s). In this paper, we discuss criterion representing reliability of results estimated by directional borehole radar. This radar was developed for 3-D imaging of some objects such as fractures and faults. The operating frequencies of the radiated electromagnetic waves are between 10 and 500 MHz. Electrically small dipole antennas in a circle are arranged in a borehole, and we can estimate directions of arrival waves (DOA) with difference of the arrival times among the antenna elements. We may scan the radar sonde in a borehole, and we can estimate reflection points in 3-D space at each depth of the radar sonde. However we sometimes find that some estimated positions are not correlated with ones estimated at neighbor depths of the sonde. In this paper, we introduce criterion calculated by difference between the theoretical model and the actually measured data. With the criterion, we interpret the estimated results of 3-D position of a target objects in both an experiment in air and a field experiment in granite. In the experiment in air, where a true position of the target is known, we found that the proposed criterion represents magnitude of the error in the estimation, and we may interpret reliability of the estimated positions with the proposed criterion. We conducted a field experiments in granite at the field test site in Osaka Electro-Communication University. In this site, we guessed that there is a fault crossing the borehole. In the field experiment, we found there were estimations with high criterion values at certain depth around the expected depth of the fault, and we could estimate reliability of the estimated position