Resolving the thickness of peat deposits with contact-less electromagnetic methods: A case study in the Venice coastland

Peat soils are typical deposits characterizing wetlands and reclaimed farmlands. They are important carbon reservoirs and when degraded (e.g., erosive processes, fires, draining and plowing) massive carbon dioxide volumes are released. This leads to increase greenhouse effect and induce serious land subsidence. Thus, mapping the volume of peat deposits is crucial in order to estimate the carbon mass and the potential release of carbon dioxide and consequent loss in soil elevation. Despite the importance of such estimations, forecasting and quantifying the peat thickness is still a challenge. Direct sediment coring provides local information that is difficult to extend to large territories. Indirect geophysical methods are unable to resolve lithological contrasts in the presence of saltwater contamination in coastal areas. In this work, we show the results obtained using two contact-less electromagnetic methods for the characterization of peat deposits in a peatland site of the Venice coastland, Italy. Specifically, a multi-frequency portable instrument (FDEM) and an airborne time-domain electromagnetic one (AEM), known for their very high and relatively low vertical resolution respectively, were used to collect data over a former wetland then reclaimed for agricultural purposes. Additional electrical resistivity tomography (ERT) data are used together with sediment core data to assess the effectiveness and accuracy of the contactless methods. Results show that both FDEM and AEM are very effective in detecting the presence of the peat layer, despite its low thickness (<2 m) and the high electro-conductive subsoil because of saltwater contamination. However, the AEM method overestimated the peat thickness while the FDEM could accurately resolve the peat thickness even where the layer was thinner than 1 m. When compared to the electrical features extracted from the ERT, discrepancies are on average lower than 30%; when compared to the borehole data, discrepancies are on average slightly higher than 6%. (c) 2020 Elsevier B.V. All rights reserved.