Integrating cone penetration testing into the 1D inversion of multi-receiver EMI data to reconstruct a complex stratigraphic landscape
Saey, T.; Verhegge, J.; De Smedt, P.; Smetryns, M.; Note, N.; Van De Vijver, E.; Laloo, P.; Van Meirvenne, M.; Delefortrie, S. (2016). Integrating cone penetration testing into the 1D inversion of multi-receiver EMI data to reconstruct a complex stratigraphic landscape. Catena 147: 356-371. https://dx.doi.org/10.1016/j.catena.2016.07.023 In: Catena. Elsevier: Cremlingen. ISSN 0341-8162; e-ISSN 1872-6887, more |  |
| Keyword | | | Author keywords | Electromagnetic induction; Depth modeling; Soil mapping; 1D inversion;Stratigraphy; Paleo-topography |
| Abstract | Conventional apparent electrical conductivity (σa) inversion approaches are generally oriented towards a smooth soil profile with gradual transitions between large stratigraphic entities. In a sedimentary environment, layered models with abrupt boundaries between the different soil units represent the soil profile accurately. Therefore, a 1D σa depth modeling procedure was developed where vertical constraints were subjected by cone penetration testing (CPT). We employed the σa measurements of a multi-receiver EMI instrument to map the stratigraphy of a 53 ha estuarine polder area up to a depth of approximately 6.0 to 6.5 m. From a limited number of CPT observations, a 4-layered soil stratigraphy was deduced with a variable topsoil containing small-scale features above a layer of either sandy or clayey estuarine deposits, a peat layer of approximately 1 m thickness and a Quaternary coarse sand substrate. The detailed vertical observations from the CPT logs allowed to estimate the inversion parameters to perform a step-wise 1D inversion of the EMI σa data, as to direct the inversion towards a realistic outcome. The depth of the interfaces between either the estuarine sediments and the peat layer and between the peat and the sand were estimated with an average error of approximately 0.5 m, hereby assuming that the thickness each layer is at least 1 m. Moreover, the variable conductivity of the top 2.0 m revealed a wide range of small and/or subtle features. To conclude, integrating the laterally dense σa measurements with a limited amount of detailed vertical observations from CPT logs allowed for a pseudo 3D representation of the soil stratigraphy within a complex sedimentary environment. This information could serve as a proxy for a range of applications which necessitate detailed information about the subsoil. |
|
