Sensor-based mapping of Danish peatlands

Pristine peatlands are precious for their Carbon (C) storage ability and the vast range of ecosystem services they provide. Globally, peatlands were heavily altered over the years especially by draining the water table for meeting energy and agricultural needs. Draining the peat results in its enhanced microbial decomposition, increased dissolved C leaching and increased susceptibility to peat fires, thus turning peatlands into C-source ecosystems. Currently, the carbon dioxide (CO₂) released from degraded peatlands amounts to approximately 5% of global anthropogenic emissions. Climate change concerns have sparked an interest to reduce these emissions and different initiatives are put forward for the protection, proper management, and restoration of the peatlands. Denmark has its own national goal of reducing CO₂ emissions by 70% by 2030; of which agriculture is expected to be a significant contributor. Comprehensive characterization of peat inventory providing status on the C stocks, water table depths and emissions is required for improved land use planning as almost 4.8 million tonnes of CO₂ per annum is released from cultivated organic lands (~ 170,000 ha in total). To achieve this, measurements of peat depth (PD) for volume characterization are invaluable. The conventional mapping approach of PD using peat probes is laborious, time-consuming, and provides only localized and discrete measurements. In addition, these manual probing measurements are also prone to errors as occasionally the probes are obstructed by stones, wood and human artefacts causing underestimation and other times they might easily penetrate the soil underlying the actual peat causing overestimation. In Denmark, we are comparing and contrasting the suitability of different electromagnetic sensors, precisely, working on electromagnetic induction (EMI), ground penetrating radar (GPR), and gamma-ray radiometric (GR) principles to accurately characterize the Danish peatlands. We are testing the sensors on both ground-based and air-borne configurations to improve the feasibility, increase accessibility and save costs. A novel drone-based transient EMI sensor is being designed in this direction. So far the results suggest that the EMI and GR techniques are promising to demarcate the peatland boundaries and estimate the PDs up to a certain extent; depending on the gradient in transition between the mineral and organic soils. Ground penetrating radar provided unequivocal results in high-resistive ombrotrophic peat while failing in low-resistive minerotrophic peat due to low signal penetration. In the drone-borne configuration, GR proved superior due to its ease of use and less to no success was achieved using a GPR. Moving forward, we plan on fusing the multisensor datasets using machine learning to improve the prediction accuracy of PDs, find a means for mapping water table depths and perform advanced modelling for comprehending the effects of different management scenarios on CO₂ emissions.