Estimating the Volume of Recoverable Materials from a Former Gypsum Landfill Using Frequency-Domain Electromagnetics

Dynamic Landfill Management wants to limit the negative effects of landfills on the environment in which they are embedded, while, if possible, mining and re-introducing waste materials back into the circular economy. However, the available information about a specific site may be limited. In this case study, we investigate the potential and effectiveness of Electrical Resistivity Tomography (ERT) and Frequency-Domain Electromagnetics (FDEM) methods to estimate the volumes of recoverable materials contained in a former gypsum landfill, covered with a high-density polyethylene (HDPE) liner. The presence of an HDPE liner, acting as an insulator, made ERT unsuitable for looking underneath the liner. We show that it is feasible to fit an interpretable inversion model to the rather noisy FDEM data, combining two instruments with varying sensitivity, and that EM methods are more suitable in this context. The presence of the unknown noise level does not allow the use of the common chi-squared criterion, but the L-curve offered a viable alternative. By varying the lateral smoothness constraint, the trend of the data is fitted well. Adding the smallness constraint prevents extreme smoothing in regions with fewer data (larger depths in this case).