Reduction of Ambiguity in Modelling and Hydrogeological Interpretation of Airborne EM Data in Complex Coastal Environment.

Abstract:

Dunea is responsible for producing high quality drinking water using the dune area at Monster (in the Netherlands), using artificial recharge and recovery with shallow wells and infiltration lakes. The system is under threat by upconing or intruding salt water. The movement of the saltwater –freshwater interface is frequently monitored in wells and with salt watcher cables (they carry EC instruments). In 2011 a SkyTEM survey was done, to obtain information in areas lacking data. Especially relevant was a known intrusion of a thin salty groundwater wedge, lying on top of a clay layer. Separating these two units (both conductors) using AEM is extremely challenging. We present a method do deal with the biggest challenge: Reduction of ambiguity using ancillary data. The original data was recently re processed, modelled and interpreted using state of the art techniques and cognitive holistic approach. Inversions were first run without taking any ancillary information into account. The main features resolvable by SkyTEM immediately appeared. However, the local conditions require extracting more information. Top and bottom of a thin, continuous clay layer obtained from cone penetration tests were then used as a-priori (with associated error) in refined inversions. Adding this a-priori information allowed separating the contribution of the clays from those of the saline water in the model. In most places, it was not possible fitting the SkyTEM data without allowing some conductivity above the clay (i.e., the saline wedge). This proves that the AEM data do carry information about the saline wedge resting directly over the clay layer. In other cases, a-priori was used to move the depth of a deep conductive layer a few meters to attempt a better match the sparse geological info. Very importantly, the ancillary information used as a-priori required thoughtful prior scrutiny in terms of their quality, quantity, density, relevance to the AEM signal and resistivity models and, last but not least, uncertainty. Some had to be re-edited (e.g., simplified) before being adopted in inversions. But at the end, the use of a-prior as extra data in the inversion allowed reducing ambiguity and the range of parts of the hydrogeological models that are congruous with the AEM measurements, providing Dunea with more information for better management.