This project evaluated the effectiveness of in-field quality control (QC) procedures during cued electromagnetic induction (EMI) data collection. The primary objective of the demonstration phase of this project was to gain further insight into the field practices that lead to the most effective and efficient cued EMI surveys. For this demonstration phase, the project team worked with a field team to apply an in-field QC software module during a cued survey of the Former WMA to identify anomalies that may have been insufficiently characterized by the initial data collection. After the survey and final ground truth stages were completed, a retrospective analysis of the WMA data set was performed to identify cases where recollects based on the in-field decision led to an improvement in data quality as well as cases where the recollect was unnecessary (i.e., it did not provide any improvements in classification features).                               

It should also be noted that during most of the survey, one of the corner receivers for the MetalMapper software was malfunctioning. It is unclear to what extent this faulty receiver influenced the in-field decisions; however, it is possible that it had some effect on the accuracy of in-field target location estimates.

The in-field QC approach included the use by cued sensor operators of a real-time inversion software module that provided immediate output of features associated with each anomaly investigated by cued EMI data collection. Among the relevant features provided by the software was an estimate of the location of the buried target. If the lateral offset of this estimated location is >30 centimeters (cm) from the center of the cued sensor, the sensor operator can reposition the sensor over the estimated source location and recollect the cued data. Visual interpretation of the sensor location, the estimated target location, and other target features such as electromagnetic polarizabilities was enabled by the in-field QC software.

During the field demonstration of this project, the Parsons field team was supplied with the in-field QC software during their cued MetalMapper survey at the Former Waikoloa Maneuver Area (WMA) on the island of Hawaii. During this survey, the field team encountered 1,032 unique anomaly locations with the MetalMapper. Out of these 1,032 encounters, 231 resulted in recollects based on the estimated target location feedback provided by the in-field QC software.

A retrospective analysis was performed of these MetalMapper data to determine:

• if there were any missed recollect opportunities, i.e., cases where a recollect was not performed, but should have been performed;
• the effectiveness of the in-field QC process by quantifying any improvements in target features obtained by recollecting the data; and
• the efficiency of the in-field QC process by identifying the number of cases where the recollect was unnecessary, i.e., it did not produce better characterization of the target.

These recollect statistics were used to develop estimates of production rates for surveys conducted using the in-field QC approach and for surveys where no in-field recollect decision is made. A summary of the statistics and estimated production rates are as follows:

• out of 1,032 anomalies investigated, of which 231 resulted in a recollect, there was 1 potential missed recollect opportunity;
• out of 231 recollects, 153 recollects appeared to be a result of magnetic geology creating false source locations;
• out of the remaining 78 recollects that were due to legitimate sources (i.e., a metal object), 46 resulted in improvements in target characterization; and
• of the remaining 32 recollects that did not significantly improve target characterization, 11 cases were found where the unnecessary recollect may have been avoided with the application of additional quality metrics (i.e., in addition to the estimated target location metric).

Estimated production rates for surveying with and without the in-field QC process were 23 anomalies/hour (hr) (with in-field QC) and 26 anomalies/hr (without in-field QC).

The magnetic geology at the site presented the most significant challenge to the technology and contributed to the lower-than-expected production rate for the in-field QC approach. The 46 cases that resulted in quantifiable improvements in target features are an example of the potential benefits of applying in-field QC to cued EMI surveys. Possible ways to improve the efficiency of the technology at challenging sites, such as the WMA, could include improving background selection and removal during in-field QC of the data, or implementing multi-source solvers in the in-field inversion to account for magnetic soil effects and high-target densities.