Rasmus Jes Petersen

Peatlands are crucial in the global carbon and nutrient cycles for water regulation and biodiversity, highlighting their importance, especially during times of global change. Grasping the spatial extent, depth, and internal structure of peatlands is essential for effective restoration and sustainable land use. Geophysical methods provide valuable information on these characteristics at various scales, with varying spatial resolutions and accuracies. This study integrates multiple ground-penetrating radar (GPR) techniques and surface nuclear magnetic resonance (SNMR) to characterize a drained peatland in the Gnarrenburg area of Germany. Constant-offset GPR mapping was used to image peat geometry and internal stratigraphy, while common-midpoint and guided-wave soundings provided velocity–depth models for estimating volumetric water content. SNMR sounding data were inverted to obtain relaxation-time distributions and water-content profiles, and GPR-derived reflectors were incorporated as structural constraints to improve inversion sharpness of internal layering of the peat. Soil cores and ring samples served as ground truth. Our results demonstrate that guided-wave GPR yields precise interval-velocity and water-content profiles, resolving small-scale variations and transitions at the peat–mineral interface. Constrained SNMR inversions successfully identify distinct relaxation-time signatures associated with varying peat decomposition states, particularly distinguishing less and highly decomposed layers. Thus, the combination of GPR and SNMR provides a comprehensive, minimally invasive framework for assessing peatland stratigraphy, water content distribution, and peat degradation. Given their relevance in global change, these encouraging outcomes present an opportunity to observe changes over time, making them especially suitable for costeffective monitoring purposes due to their minimally invasive characteristics. Ultimately, these methods serve as valuable tools for planning and guiding peatland restoration activities.

Study regionVejrumbro, a shallow drained riparian peatland in the Nørreå River valley, central Jutland, Denmark.Study focusRiparian peatlands exhibit strong spatial and temporal heterogeneity. Consequently, controls on water table levels are not constant. We investigated how dominant controls shift through time and space using a high-resolution (5 m x 5 m) physically based model and a distance-based generalized sensitivity analysis.

Geological interpretation of borehole Nuclear Magnetic Resonance (NMR) data in unconsolidated geological materials remains manual, subjective, and time-consuming. Here, we evaluate three supervised machine learning classifiers, including Logistic Regression (LR), Random Forest (RF), and Support Vector Machine (SVM), to develop an automated geological / lithological interpretation tool specifically for borehole NMR data under fully saturated conditions. The dataset used for model development comprises 2681 measurements from nine sites across Denmark and Germany representing sand, clay, peat, and gyttja. Classifiers were trained using total water content, the base-10 logarithm of mean T2 relaxation time, and normalized water fractions. Beyond standard test set evaluation, three cross-validation strategies (Monte Carlo random splits, GroupKFold by borehole ID, and GroupKFold by site ID) were employed to assess performance on unseen boreholes and entirely new sites. RF and SVM showed very similar performance in Monte Carlo and borehole-holdout cross-validations, while both substantially outperformed LR. For the site-holdout cross-validation, performance decreased significantly for all models, and standard deviations increased approximately threefold compared with borehole-holdout, reflecting substantial variability among sites. In terms of model performance based on an independent test set, RF achieved the highest balanced accuracy (0.86) and demonstrated superior performance on minority classes, while SVM achieved the highest Matthews Correlation Coefficient (0.84); both non-linear models substantially outperformed LR (balanced accuracy = 0.72, Matthews Correlation Coefficient = 0.78), which struggled with underrepresented lithologies. Application to an unseen borehole confirmed that RF reproduced the observed geological stratigraphy well. Deployment on an independent direct-push NMR dataset without geological logs produced geologically plausible probabilistic interpretations, demonstrating the practical value of such approaches in data-sparse regions. Overall, our results demonstrate that machine learning provides a robust, objective, and reproducible framework for automated geological interpretation of borehole NMR data, with probabilistic outputs effectively quantifying uncertainty in heterogeneous Quaternary settings.

Peatland drainage for agricultural purposes has transformed wetlands into major nutrient sources, but rewetting offers potential for nutrient retention and climate mitigation. We quantified hydrological and nutrient fluxes in a 14.5 ha former pump-drained fen (Strande Enge, Denmark) across 1 year before and 2 years after rewetting. Restoration involved removing pumping infrastructure and reconnecting upstream catchments with the aim to restore natural flow paths. Prior to rewetting, the site exhibited high hydraulic loading (5904 mm yr ⁻¹ ) dominated by groundwater inputs and acted as a strong nutrient source, with net losses of total nitrogen (TN: 123 kg ha ⁻¹ yr ⁻¹ ), total phosphorus (TP: 23 kg ha ⁻¹ yr ⁻¹ ), and total organic carbon (TOC: 1336 kg ha ⁻¹ yr ⁻¹ ). Following rewetting, hydraulic loading increased more than fourfold (24,258–27,160 mm yr ⁻¹ ), and the system shifted to a flow-through wetland dominated by surface water inputs. Despite substantially higher nutrient loads after rewetting, the fen became a large sink for TN (209–242 kg ha ⁻¹ yr ⁻¹ ; 21% retention) and retained 44–72% of soluble reactive phosphorus. However, although their losses declined, particulate phosphorus and organic matter were still exported, particularly during high-flow events. At catchment scale, rewetting reduced the yearly TN export to the downstream lake by 27–38%, while the annual TP export varied strongly after rewetting (12% reduction in year 1 and 74% increase in year 2). These findings demonstrate that rewetting of pumped fens can rapidly transform nutrient source hotspots into effective nitrogen sinks, but additional measures may be needed to control particulate losses of nutrients and organic carbon, if these prove to persist over time. Overall, our results show that rewetting can reduce nutrient losses, including TOC losses, thereby contributing to mitigation of aquatic pollution and helping preserve peatland carbon stocks.

The rewetting of low‐lying organic soils previously used for agriculture can mobilize legacy phosphorus (P), increasing the risk of eutrophication in downstream water bodies. One potential mitigation strategy is biomass harvesting, which may remove mobile P fractions from the soil. At the 44‐ha, pump‐drained former shallow lake Gammelgaard, biomass was harvested annually from 2020 to 2023, prior to the re‐establishment of the lake in July 2023. Over this period, biomass harvesting removed a total of 701 kg P. In comparison, 254 kg of total P was exported via drainage pumping, corresponding to an average annual loss of 1.5 kg P ha ^{− 1} . Following rewetting, the P mass balance revealed an inlet load of 206 kg P and an outlet loss of 238 kg P, resulting in a net P loss from the rewetted area of 0.7 kg P ha ^{− 1} . Soil cores collected before and after 3 years of biomass harvesting period showed a reduction in the redox‐sensitive P pool of 222 kg P ha ^{− 1} . This reduction exceeded the sum of harvested P and outlet P fluxes, indicating that additional processes—such as internal redistribution or transformation into more stable binding forms—likely contributed to the observed changes. Although biomass harvesting appears to significantly reduce the mobile P pool in the soil, further research is needed to clarify the extent to which it effectively mitigates P losses from rewetted wetlands.

Rewetting of agricultural peatlands is widely recognised as a key climate mitigation strategy, yet its net effect remains uncertain, particularly during the early transition years. We conducted a multi-year study using continuous eddy covariance measurements to quantify carbon dioxide (CO2) and methane (CH4) fluxes across two peatland fields undergone rewetting and an adjacent shallow-drained control field in Denmark. Rewetting raised the mean annual water table level by ~7 cm relative to the control when side-by-side comparisons were possible. Our results demonstrate that rewetting can rapidly shift the ecosystem carbon balance toward net CO2 uptake, likely due to the successful establishment of productive vegetation such as Phalaris arundinacea prior to rewetting. Early vegetation development may therefore accelerate CO2 uptake compared with slower trajectories observed in more degraded peatlands. However, this benefit can be partially offset by increased CH4 emissions, particularly during wet periods, which can rival the CO2 sink strength and reduce the overall greenhouse gas mitigation potential. Even in shallow-drained control areas, modest increases in water table during wet years were sufficient to temporarily reverse net carbon loss, highlighting the sensitivity of early outcomes to hydrological conditions. Management practices, such as autumn biomass cutting, further influenced CO2 exchange by enhancing early-season uptake, though biomass removal can partially offset these gains. Collectively, our findings underscore that the net climate effect of peatland rewetting depends strongly on interannual variability, water table dynamics, vegetation establishment, and management interventions. Long-term, ecosystem-scale monitoring is thus essential to capture the full spectrum of environmental variability and optimise restoration strategies for effective climate mitigation.

Peatlands store a substantial fraction of the global soil carbon pool. Widespread drainage and land-use change have accelerated peat decomposition, while current restoration efforts aim to slow or reverse peat carbon oxidation and associated greenhouse gas emissions through rewetting. Understanding when and to what extent rewetting restores carbon sequestration and long-term peat accumulation remains a key scientific and management challenge.Over recent decades, substantial progress has been made in identifying the biogeochemical controls on peat carbon turnover in rewetted systems. Water table dynamics, hydrological connectivity, redox conditions, substrate availability, nutrient status, and vegetation composition jointly regulate microbial processes driving organic matter decomposition and carbon fluxes. Yet, key questions remain: what really drives carbon turnover in peatlands? Is it “the overriding role of the water table”, the “iron gate” of mineral interactions, or the “iron wheel”? Could there even be a single enzyme controlling global carbon storage (Wen et al., 2019)? These questions highlight how peat carbon cycling defies simple explanations and directly challenges long-standing paradigms. Classical concepts emphasizing intrinsic substrate recalcitrance, single inhibitory controls (e.g., phenolics or water table position), or strictly separated aerobic–anaerobic microbial pathways fail to capture the complexity of carbon stabilization and turnover observed in rewetted peatlands (Zak et al., 2019). Instead, emerging evidence points to a network of interacting, context-dependent processes, including microbial community turnover, mineral–organic interactions, and dynamic redox condition changes, underpinning peat carbon persistence.Yet, these mechanisms are typically studied at micro- to plot scales, while restoration success and climate feedbacks are evaluated at ecosystem to landscape scales, posing persistent challenges for upscaling. In this contribution, we synthesize current insights into carbon cycling in rewetted riparian peatlands by explicitly linking microbial and biogeochemical controls on carbon decomposition with restoration approaches aimed at managing carbon fluxes. Emphasis is placed on spatial and temporal heterogeneity in peat properties, hydrology, and microbial functioning, and on how this variability propagates uncertainty in carbon balance assessments and model predictions.By integrating process-based understanding with measurements and modeling perspectives, both recent advances and remaining knowledge gaps in predicting peat carbon cycling under restoration will be highlighted. Re-assessing prevailing paradigms and strengthening cross-scale linkages are essential for designing effective rewetting strategies. ReferencesWen, Y., Zang, H., Ma, Q., Evans, C. D., Chadwick, D. R., & Jones, D. L. (2019). Is the ‘enzyme latch’or ‘iron gate’the key to protecting soil organic carbon in peatlands?. Geoderma, 349, 107-113.Zak, D., Roth, C., Unger, V., Goldhammer, T., Fenner, N., Freeman, C., & Jurasinski, G. (2019). Unraveling the importance of polyphenols for microbial carbon mineralization in rewetted riparian peatlands. Frontiers in Environmental Science, 7, 147.

Floodplains and coastal wetlands provide essential ecosystem services such as flood protection, water quality enhancement, carbon sequestration, and the production of raw materials and food. Despite numerous studies on wetland restoration, the complexities of restoring ecosystem functions remain challenging. Complete restoration of natural functions is rare, and therefore seldom the primary target; instead, restoration efforts typically focus on reactivating key ecological processes. This study examines the restoration of ecosystem functions in floodplains and coastal wetlands, focusing on water quality improvement. It summarizes both nitrogen and phosphorus dynamics, highlighting the potential to restore processes like denitrification and sedimentation within relatively short time frames using selected case studies from Central Europe. However, restoration success is hindered by altered soil properties, nutrient legacy, and other degradation-related constraints. Ultimately, the effectiveness of restoration endeavors depends on site-specific factors such as wetland degradation, size, hydrological connectivity, plant recovery, and ongoing global changes. This underscores the importance of considering both spatial and temporal dimensions and adopting a custom-tailored restoration strategy. Highlights Recovery of nutrient retention processes can be either rapid or delayed Nutrient legacies and soil degradation hinder ecosystem restoration. Process insights and site traits guide effective restoration efforts.