A direct comparison of EMI data and borehole data on a 1000Â ha data set

- We collect and process 1000 ha EMI data.
- We carry out a full non-linear inversion of EMI data.
- We map depth-to-clay.
- We map lateral delineation of shallow (< 6 m) hydrogeological architecture.
- We compare EMI data and borehole data.

Increasingly, the frequency domain electromagnetic induction (EMI) method is used to study soil properties and hydrological properties on different scales. Yet, EMI data are most often used as they are or with the low induction number approximation. Moreover, reported studies have been fairly small (< 100 ha) and to our best knowledge quantitative comparisons with boreholes have not been reported. In the present study we collect, process and invert a 1000 ha EMI data set, and we directly compare the EMI data with borehole data. EMI data were collected with a line spacing of 20 m, resulting in approximately 110,000 processed soundings, which were inverted with a full non-linear algorithm. The EMI results were then evaluated against lithological information from 125 boreholes. The results show that with an EMI instrument we can map the shallow (< 6 m) architecture on small-catchment scale (1000s ha) in a complicated glacial sedimentary environment. The mapping of the geological units was evaluated by a quantitative analysis in which we developed a general methodology for directly comparing EMI results and borehole data. After linking geological units to measured resistivities, we found that 103 out of 125 boreholes were reproduced by the EMI results. We attributed the few non-reproducible boreholes to be caused by local scale geological units, abrupt lateral variations in either geology or resistivity, and inaccurate borehole coordinates. We conclude that EMI can provide the architectural input with the necessary detail and quality to characterize areas of up to 1000s ha.