Soil carbon stores can be greatly influenced by land cover and management, including clear-cutting, soil sealing, and agricultural intensification, especially tillage. In addition, climate change, especially temperature and precipitation patterns, can modify the dynamics of soil carbon, through an effect on soil moisture regime and microbial activity. Droughts, floods, and other extreme weather events have become more frequent and severe due to global warming, which might further affect soil carbon levels.
Also, bulk Density (BD) influences the accuracy of carbon stock calculations and is therefore an important factor in soil carbon monitoring. BD is strongly, but not solely, affected by soil compaction, tillage, and the application of organic amendments. Erroneously measured or calculated BD can thus imply errors in soil carbon stock estimation.
Nutrient availability regulates C decomposition and stabilization processes, thereby tightly linking nutrient cycles with soil C dynamics. Tools such as isotopic tracers, lysimeter studies, and digital monitoring platforms provide new insights into nutrient fluxes and their interactions with soil carbon pools.
This session addresses the dynamics of SOC, SIC, and nutrients in agroecosystems, while investigating innovative monitoring and modelling strategies for optimizing soil carbon and nutrient management, such as machine learning, process-based models, and remote sensing, to improve our knowledge of soil carbon and nutrient dynamics and to support decision-making in natural and agroecosystems. Contributions that integrate monitoring and modelling of nutrient–carbon interactions and that highlight their implications for sustainable soil management are particularly welcome.
The session is supported by the PRIMA project SHARInG-MeD (AT, SS), Horizon projects SUS-SOIL (AT, SS), Horizon project NBSOIL (EPS), PCI2025-163238 – AGRECO4CAST (GG), PID2023-146177OB-C22 – ReLandWater (GG) and PR.AVA23.INV202301.035 – ECOMED (GG).
Orals: Fri, 8 May, 14:00–17:45 | Room 0.11/12
Multi-compartment first-order kinetics models are commonly employed to represent soil organic carbon (SOC) dynamics. In such framework, SOC is partitioned into distinct theoretical pools, each characterized by its own first-order constant that determines its intrinsic potential decomposition rate. Models with only two interacting dynamic soil carbon compartments—such as RothC or ICBM—are commonly utilized in national inventories and carbon farming initiatives due to their simplicity and ease of initialization and parameter identifiability. Alternatively, streamlined soil carbon models can treat the potential fractional turnover rate of the soil layer (ρ)—the reciprocal of its turnover time—as a state variable, further minimizing the number of parameters required. Multicompartmental models can be represented via a matrix approach as a linear dynamical system, such as dC/dt = Bu + AKC, where C is the carbon stock vector, u represents external inputs, B the partition vector of input material, the matrix A defines the partitioning of carbon decomposed in each pool which is lost as CO2 or transferred to other pools. K defines each pool’s potential fractional turnover rate. Such formulation explicitly encodes both carbon exchanges between compartments, and so SOC stabilization, as well as its losses to the atmosphere. Considering the two-pool 2×2 model matrices with parameters A = [-1, a12, a21, -1], K = [k1, 0, 0, k2], and their product AK = [-k1, a12k2, a21k1, -k2], the derivative of the ratio between the compartment stocks, r = C1/C2, produces a quadratic Riccati-type differential equation dr/dt = a12k2 - (k1 - k2)r - a21k1r2 which can then be further algebraically manipulated to yield a quadratic equation of the variation of fractional turnover rate, i.e., dρ/dt = aρ2 + bρ + c. This continuous formulation is particularly relevant because it allows carbon decomposability to be represented as a state variable, rather than as a fixed property associated with discrete theoretical compartments. Consequently, SOC dynamics can be captured using only the measurable SOC stock coupled to an evolving potential fractional turnover rate (decomposability), enhancing model identifiability and initialization since total soil carbon is directly measurable in the field.
How to cite: Valladão, B., Gonçalves, D., and Barioni, L.: Analytic derivation of a carbon turnover time stabilization model from a standard two-pool model, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13116, https://doi.org/10.5194/egusphere-egu26-13116, 2026.
Soil Organic Carbon (SOC) is a foundation of soil health and global climate resilience, yet its
prediction remains difficult because of intricate physical, chemical, and biological processes. In this
study, we explore a Scientific Machine Learning (SciML) framework built on Universal Differential
Equations (UDEs) to forecast SOC dynamics across soil depth and time. UDEs blend mechanistic
physics, such as advection–diffusion transport, with neural networks that learn nonlinear microbial
production and respiration. Using synthetic datasets, we systematically evaluated six experimental
cases, progressing from clean, noise-free benchmarks to stress tests with high (35%) multiplicative,
spatially correlated noise. Our results highlight both the potential and limitations of the approach. In
noise-free and moderate-noise settings, the UDE accurately reconstructed SOC dynamics. In clean
terminal profile at 50 years (Case 4) achieved near-perfect fidelity, with MSE = 1.6 × 10−5, and
R2 = 0.9999. Case 5, with 7% noise, remained robust (MSE = 3.4×10−6, R2 = 0.99998), capturing
depth wise SOC trends while tolerating realistic measurement uncertainty. In contrast, Case 3 (35%
noise at t = 0) showed clear evidence of overfitting: the model reproduced noisy inputs with high
accuracy but lost generalization against the clean truth (R2 = 0.94). Case 6 (35% noise at t = 50)
collapsed toward overly smooth mean profiles, failing to capture depth wise variability and yielding
negative R2, underscoring the limits of standard training under severe uncertainty. Qualitatively, the
UDE framework consistently preserved broad SOC patterns, avoided overfitting in moderate noise,
and maintained physics-based plausibility even when data were corrupted. These findings suggest
that UDEs are well-suited for scalable, noise-tolerant SOC forecasting, though advancing toward field
deployment will require noise-aware loss functions, probabilistic modelling, and tighter integration
of microbial dynamics.
How to cite: Satyanarayana Raju, G. V. V.: A Study of Universal ODE Approaches to Predicting Soil Organic Carbon, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-1353, https://doi.org/10.5194/egusphere-egu26-1353, 2026.
The importance of agricultural soils as a potential carbon sink has been extensively discussed as a key step towards climate neutrality and sustainable land use. Typical measures that could enhance soil organic carbon (SOC) stocks include cover cropping, perennial crops, or establishing hedgerows or agroforestry systems. Targeted advisory services and financial incentives, including subsidies or carbon credit trading systems, could encourage farmers to establish these measures. However, accurately assessing their contribution to climate protection via enhanced SOC stocks remains challenging. Taking soil samples every few years can fulfil this purpose, but they are costly, labour-intensive, require a careful sampling regime and usually a period of at least ten years to detect significant differences in SOC stocks. As a result, various estimation methods are discussed as an alternative. However, there is currently no consensus as to which approach best balances accuracy, feasibility and practicality. This study evaluates five different methods and models of varying complexity to estimate SOC stock changes, using data from 46 German permanent soil monitoring sites. Included in the assessment is the VDLUFA humus balance method as well as the process-based model RothC, which is run with two variants regionally averaged and site-specific. Our results confirmed previous conclusions, that simple carbon balance methods perform poorly if no site-specific pedoclimatic information is considered in the methodology. By comparison, the RothC model achieved significantly better predictive performance, especially if executed with site-specific information. A hybrid approach integrating properties of the RothC model with the simplicity of the VDLUFA method achieved a comparable predictive performance while reducing methodological complexity. Our findings provide insights into the trade-offs between model complexity and prediction accuracy, offering recommendations on their applicability for climate policy and agricultural decision-making.
How to cite: Aiteew, K., Dechow, R., and Don, A.: From simple to complex: Evaluating methods for estimating soil organic carbon changes in croplands, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-7375, https://doi.org/10.5194/egusphere-egu26-7375, 2026.
Understanding the spatial and temporal variations of soil CO₂ efflux from heterotrophic respiration (HR) and straw decomposition (SDR) is essential for constraining cropland carbon budgets, yet long‑term assessments that distinguish these sources remain limited. Here, we developed process‑based models for HR and SDR using multi‑site field observations. The models were then applied across the North China Plain under the winter wheat–summer maize rotation system, driven by high‑resolution forcing data to generate 1 km, 8‑day simulations for 2003–2024. Validation with independent data showed high consistency between simulations and observations (HR: R² = 0.97; SDR: R² > 0.93). Average CO₂ efflux from HR and SDR was estimated at 505.5 ± 34.5 and 278.0 ± 84.5 g C m⁻² yr⁻¹, respectively. HR exhibited a pronounced latitudinal gradient, driven primarily by SOC content, whereas SDR showed a more gradual decline toward higher latitudes, associated with straw return amounts. Both fluxes peaked in summer, but HR displayed a substantially larger seasonal amplitude. Interannual HR increased significantly across the region (4.6 g C m⁻² yr⁻¹), largely in response to rising temperatures. SDR showed no significant trend before 2010 but increased sharply thereafter (9.3 g C m⁻² yr⁻¹), driven mainly by rising straw inputs following regional straw incorporation policies. The contribution of SDR to total soil CO₂ efflux increased from 27% in 2003–2010 to 39% in 2011–2024, highlighting the need to better account for residue‑derived CO₂ when evaluating cropland carbon processes. Collectively, the findings clarify how climate and management interact to regulate cropland CO₂ emissions and strengthen the process basis for agricultural carbon modeling.
How to cite: Zhao, H., Wang, X., Xu, M., and Smith, P.: Spatial and temporal variations of soil CO₂ efflux from heterotrophic respiration and straw decomposition in the North China Plain over 2003-2024: process-based model, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-3164, https://doi.org/10.5194/egusphere-egu26-3164, 2026.
Drylands are expanding globally under climate change, intensifying pressures on soil organic carbon (SOC) and nutrient cycling in agricultural landscapes. In Brazil, semi-arid agroecosystems already experience recurrent droughts, high temperatures, and structural soil constraints, making carbon–nutrient dynamics highly vulnerable to warming and drying trends. Process-based ecosystem models are essential tools for evaluating the long-term sustainability of land-use strategies in these fragile environments. This study applied the DayCent model to quantify SOC and nitrogen (N) trajectories from 2024 to 2100 under the current climate and two IPCC scenarios (SSP2-4.5 and SSP5-8.5) across three representative dryland regions: Betânia do Piauí, Petrolina, and Sobral, each encompassing contrasting soil textures, land-use histories, and intensification agricultural gradients. Model calibration used field-measured SOC and N stocks (0–30 cm), soil properties, and detailed management records from native vegetation, conventional systems, grazed pastures, crop–livestock integration (CLI), and crop–livestock–forestry integration (CLFI). DayCent showed strong performance (SOC: R² = 0.97, RMSE = 2.09 Mg C ha⁻¹; N: R² = 0.73, RMSE = 0.55 Mg N ha⁻¹), indicating robust capacity to reproduce observed carbon–nitrogen stocks in these semi-arid systems. Simulations revealed that conversion of native vegetation, especially when associated with fire or low-input management, reduced SOC stocks by 5–20%. In contrast, agricultural intensification enhanced SOC in all regions, though responses varied by site and soil texture. In Betânia, integrated crop-livestock systems with annual fertilization combined with no-till farming stored approximately 37% more SOC stocks (75 Mg C ha⁻¹) compared to conventional tillage with fertilization every 5 years. In Petrolina, reduced grazing pressure and N fertilization increased SOC stocks relative to current grazing systems, while in Sobral, no-tillage consistently reduced SOC losses compared to conventional tillage, particularly in intercropping systems. Across all sites, climate change simulations showed pervasive SOC declines under SSP2-4.5 and SSP5-8.5, with the most pronounced losses under the high-emission scenario. Reductions were strongest in sandy soils and in systems with frequent soil disturbance. Although management intensification (fertilization, reduced grazing, and no-tillage) consistently mitigated SOC losses, no strategy fully compensated for the negative impacts of increased aridity and reduced precipitation. Integrated agricultural systems were the most resilient, partially buffering climate-induced SOC stock declines through greater biomass inputs. Overall, the results demonstrate that sustainable intensification can enhance SOC under present conditions, future climate change will reduce SOC stocks across all systems, and integrated and conservation-based strategies remain essential for slowing carbon depletion in Brazilian drylands. These findings highlight the need for climate-smart soil management policies focused on minimizing soil disturbance, enhancing nutrient availability, and increasing organic inputs to maintain carbon-nutrient resilience under intensifying aridity.
How to cite: Greschuk, L., Cherubin, M. R., and Ogle, S.: Sustainable agricultural intensification mitigates but cannot prevent soil carbon losses under climate change: A DayCent model approach., EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-659, https://doi.org/10.5194/egusphere-egu26-659, 2026.
European policy aims to enhance soil organic carbon stocks in order to improve soil fertility and resilience and for mitigation of climate change. A sound impact assessment of measures requires robust baseline estimates of soil organic carbon (SOC) trends in a changing climate.
For the German greenhouse gas emission projections, we modelled organic carbon in mineral topsoils (0-30 cm) of German croplands on a 100m*100m grid for the time-period between 1990-2075. Our interdisciplinary high-resolution modelling approach took into account the projected impacts of recent agricultural policy on agricultural management as assessed by agro-economic models. Crop rotations for each field across Germany were created using a Bayesian approach combining agricultural statistics (past and projected), remote sensing data and information from the Integrated Administration and Control System (IACS). Climate change was included based on local weather projections derived by the German Weather Service from the RCP4.5 and RCP8.5 climate scenarios. Soil carbon and texture data were obtained from maps based on the German Agricultural Soil Inventory and the European LUCAS survey. Using these management, climate and soil data, we modelled the SOC dynamics with the SOC decomposition model RothC, which has been validated with data from German agricultural long-term observation sites.
Our simulations indicate a trend of declining SOC stocks. We attribute this to impacts of climate change and management changes towards lower livestock densities, which cause decreasing organic fertilizer application rates. Increased cover cropping area partly counterbalances this trend.
The projected SOC trends can be used as a business-as-usual or reference scenario to quantify the climate effect of carbon-enhancing agricultural measures.
How to cite: Seitz, D., Dechow, R., Gocht, A., Laggner, A., Rieger, J., Vos, C., and Fuß, R.: A baseline projection of soil organic carbon stocks in German mineral croplands, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19690, https://doi.org/10.5194/egusphere-egu26-19690, 2026.
Soil salinization undermines the soil structure and microbial carbon cycling efficiency. However, the pathways by which salt stress reorganizes the microbial food web to decrease soil organic carbon (SOC) and degrade soil quality remain poorly understood. We analyzed natural and agricultural sites of low- and high-salinity soils in coastal China according to SOC, soil quality index (SQI), microbial carbon use efficiency (CUE), microbial necromass carbon (MNC), enzyme activities, and microbial community composition. Compared to adjacent low-salinity soils, high-salinity soils exhibited lower SQI and SOC, CUE, and MNC (by 16.0–21.1%, 16.7–22.0%, and 34.8–40%, respectively) but double the maintenance respiration, indicating a shift from growth to survival metabolism. The SQI in highly saline soils was positively correlated with SOC, CUE, and MNC, but negatively associated with microbial C and P limitation, highlighting the pivotal role of microbially mediated C turnover in soil quality under salt stress. Salinity favored halotolerant Proteobacteria, Crenarchaeota, and protists, displacing key bacterial and fungal decomposers. Unexpectedly, network complexity increased (nodes by 50–80% and edges by 3–11‑fold) with heightened positive cohesion, reflecting close cooperative interactions that nonetheless intensified resource competition and accelerated SOC mineralization. Structural equation modeling revealed a cascade of effects, whereby salinity disrupted soil aggregation and nutrient balance, which increased network connectivity and reduced microbial metabolism efficiency, driving SOC loss and SQI decline. Saline soil management should therefore combine aggregate stabilization, inoculation with osmolyte‑producing microbes, and modular, resilient, food web architectures to sustain SOC sequestration and soil health.
How to cite: Wang, X., Zhu, Z., Ge, T., Li, X., and Chen, J.: Salinity-induced the decrease of soil organic carbon controlled by micro-food web networks complexity in saline soil, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-614, https://doi.org/10.5194/egusphere-egu26-614, 2026.
Balanced soil nutrient budgets are important for sustainable agriculture. Soil nutrient pools are interconnected within and between elements and balanced by natural processes. In intensive agriculture the excess of nutrients disrupts the balance, leading to increased leaching, volatilisation or other losses to the environment. Stabilisation of ions and molecules stores nutrients in the soil, which can make them unavailable, while mobilisation moves nutrients into available state, potentially causing leaching. This spectrum of availability is driven by different factors for each nutrient is connected to to other cycles in the soil system.
We propose a concept of "soil safe operating space" for agriculture - a state of soil where mobilisation and stabilisation are balanced, making just enough nutrients available for plant uptake to grow optimally, so leaching is reduced. Finding that state and developing practices to reach it would improve the sustainability of agriculture greatly. To aid in this, first the interactions of nutrient dynamics and its drivers need to be clear. Therefore we synthesised connections between cycles of macronutrients C, N, P, K, Ca, Mg, S and micronutrients Cu and Zn based on a literature review of 175 articles in the field of soil chemistry. These connections also established the driving factors of the mechanisms and corresponding feedbacks. We analysed the results in the context of the classical management practice of target pH 6.5 as per the old nutrient availability diagrams. They showed that not all nutrients benefit from it, such as sulphate being leached, ammonium oxidated and phosphate adsorbed. This suggests that, depending on the agricultural targets, pH may be adjusted to achieve certain goals. The cross-cycle interactions also played a great role, with the concentrations and forms of one nutrient heavily affecting others via precipitation, fixation or adsorption. Phosphorus has the most mechanisms of interactions and factors reducing its availability, soil organic matter affects storage of most other nutrients and cycles of other macronutrients have nuanced reactions to soil factors like mineralogy, surface charge and solution concentrations. Facilitating some processes over others via controlling pH and moisture, organic matter addition or other amelioration practices would have more defined effects than setting a target pH while ignoring other factors.
This study provides a unique framework that helps modelling of soil chemistry because of its broad scope, connecting the various individual studies into an integrated system of cycles. Examples of such developments are presented in the model FarmSAFE of the project: 101060455 — NutriBudget.
How to cite: Strekalovskaya, A., Lana, M., Hendriks, C., Westerberg, L.-O., and Brown, I.: Disentangling the interactions of soil nutrient cycles for sustainable agriculture, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19672, https://doi.org/10.5194/egusphere-egu26-19672, 2026.
Soil organic carbon (SOC) sequestration plays a critical role in mitigating climate change and enhancing soil health; however, the effects of land management and climate change remain poorly quantified in agricultural landscapes. This study simulates SOC sequestration under different land uses and climate scenarios in a 28,311 ha Lower Austrian watershed. The impacts of land management change and climate scenarios on SOC stocks in the topsoil (0–15 cm) were modelled using the Rothamsted Carbon (RothC) model up to 2050. Four climate scenarios were considered: ‘Historical’, ‘SSP1-1.9’, ‘SSP2-4.5’, and ‘SSP5-8.5’. The results indicate that the implementation of two low-barrier but effective soil conservation practices - (i) grassed waterways and (ii) conservation tillage practices (mulch tillage and no-tillage) - could store more than 15,000 t C across the watershed by 2050. Grassed waterways and no-tillage could sequester up to 0.07 t C ha-1 yr-1, while mulch tillage could sequester up to 0.04 t C ha-1 yr-1 under the ‘Historical’ climate scenario. Among the future climate scenarios, only SSP5-8.5 notably reduced SOC sequestration, lowering rates to 0.06 t C ha-1 yr-1 for grassed waterways and no-tillage, and to 0.03 t C ha-1 yr-1 for mulch tillage. Overall, the study provides a scalable framework for assessing SOC sequestration under future land-management and climate scenarios, with direct relevance for carbon farming and certification schemes under current EU regulations such as the Carbon Removals and Carbon Farming Regulation (CRCF).
How to cite: Toth, M., Stumpp, C., Vasquez, C., Klik, A., Strauss, P., Liebhard, G., and Strohmeier, S.: Modelling Soil Organic Carbon Stocks in Agriculture: Conservation Tillage Practices and Grassed Waterways under various Climate Scenarios in Lower Austria, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13412, https://doi.org/10.5194/egusphere-egu26-13412, 2026.
Food production contributes significantly to environmental degradation, accounting for an estimated 78% of global ocean and freshwater eutrophication (Poore & Nemecek, 2018), being the leading driver of biodiversity loss, and representing a major driver of soil health loss (EUSO, 2024). Organic farming is often proposed as a strategy to mitigate these impacts by enhancing biodiversity, reducing nutrient losses at large-scale adoption, and improving multiple soil quality parameters (Seufert & Ramankutty, 2017). Consequently, policy initiatives such as the European Green Deal’s Farm to Fork strategy aimed to expand organic farming across Europe (European Commission, 2020). However, the sustainability benefits of organic agriculture depend strongly on local conditions. For example, transitioning to organic management can risk decreasing soil organic carbon (SOC) stocks (Gaudaré et al., 2023), reducing yields, and potentially increasing greenhouse gas emissions per unit of product due to lower productivity (Meier et al., 2015). These outcomes depend on region-specific factors such as soil properties, climatic conditions, management practices, and nutrient availability.
This study evaluates how a transition to organic agriculture influences SOC and nutrient (N, P) dynamics across the EU by comparing a business-as-usual (BAU) scenario with a scenario in which 25% of agricultural land is managed organically by 2030. We employed the spatially explicit, process-based biogeochemical model DayCent at a 1 km2 scale across the EU, which has been calibrated and tested for European conditions (Muntwyler et al., 2023), to simulate SOC turnover, nutrient cycling, and crop yields across diverse soil and climate gradients. The model integrates detailed representations of mineralization, stabilization, plant uptake, and nutrient losses, thus capturing key processes. To evaluate broader environmental consequences, model outputs were combined with a life cycle assessment (LCA) framework using regionalized characterization factors that quantify N- and P-related impacts on freshwater fish biodiversity (Zhou et al., 2024).
Achieving the 25% organic target showed potential to improve degraded soils (defined by nutrient surplus/excess), reduce reliance on mineral fertilizers, and maintain or lessen current eutrophication impacts on freshwater fish. The spatially explicit modelling framework enabled identification of hotspot regions where transitions to organic agriculture yield environmental benefits with minimal productivity losses. However, these benefits were accompanied by reduced average yields of grain and tuber crops, partly driven by increased fodder crop production in organic rotations. A complementary cover crop scenario highlighted the benefits of increased N fixation, improved yields, and mitigation of SOC decline, but also led to higher impacts on freshwater biodiversity due to increased N losses.
These results underscore the importance of jointly considering interconnected N, P, and C cycles, yield responses, and potential feedbacks when evaluating management transitions. The approach provides valuable insights into the synergies and trade-offs between agricultural practices and environmental consequences at high spatial resolution, supporting evidence-based decisions for sustainable land management and policy.
How to cite: Muntwyler, A., Lugato, E., Panagos, P., Scherrer, L., Müller, A., and Pfister, S.: Modelling SOC and Nutrient Dynamics under Organic Farming Expansion across the EU, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-23134, https://doi.org/10.5194/egusphere-egu26-23134, 2026.
Soils store both organic (SOC) and inorganic carbon (SIC), yet biogenic processes driving SIC formation remain poorly quantified. One such process is the oxalate-carbonate pathway (OCP) where plant-derived calcium oxalate is transformed by oxalotrophic microorganisms into SIC, which can sequester atmospheric CO₂ in non-calcareous soils. Yet, the OCP has rarely been investigated in connection to trees with significant agroforestry potential. To further investigate the OCP and its connection to species with agroforestry potential, we investigated three East African fig species (Ficus glumosa, F. natalensis, and F. wakefieldii) in semi-arid Samburu County, Kenya.
Across contrasting parent materials devoid of primary carbonates, soils adjacent to fig trees exhibited significantly higher pH, exchangeable Ca, SOC, and SIC content compared to control soils, indicating the trees maintained hotspots of distinct biogeochemical conditions. Fig biomass samples contained substantial calcium oxalate contents (4.9±0.5 % dry weight), predominantly as prismatic whewellite crystals (CaC2O4.H2O). Calcium carbonate coatings were observed on trunks and roots of all three species, which reacted strongly to hydrochloric acid. Synchrotron-based μ-X-ray Fluorescence coupled with μ-X-ray absorption near-edge structure spectroscopy (Ca K-edge) revealed that CaCO₃ had precipitated deeply into woody tissues, providing direct evidence for aboveground OCP. Amplicon-based sequencing showed diverse and abundant microbial communities on the aboveground biomass, litter, roots, and adjacent soils. In addition, a co-occurrence analysis of fungal and bacterial communities showed specific fungal genera and fungal oxalate-producers are tightly linked to known baterial oxalotrophs, indicating that bacterial-fungal interactions could be essential for oxalotrophy. Combined these results demonstrate an active OCP both above and belowground in connection to the food-providing fig trees (Ficus spp.) of semi-arid East Africa.
Our findings identify East African fig trees as previously unrecognised drivers of biogenic SIC sequestration. Integrating specific fig species into agroforestry systems could therefore represent a novel nature-based solution that couples food production with SOC and long-term SIC storage in dryland landscapes.
How to cite: Rowley, M. C., Cailleau, G., Olaka, L. A., Bone, S. E., Pena, J., Wiesenberg, G. L. B., Rozin, A., McMackin, C., Tikhomirov, D., Schiedung, M., Hoppe, G. A., Vaughan, L., Lisabeth, H., Nico, P., Rieder, C., Dasgupta, S., and Bindschedler, S.: Biomineralisation of inorganic carbon by agroforestry species in East Africa: The oxalate carbonate pathway of fig trees in Samburu County, Kenya, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-6791, https://doi.org/10.5194/egusphere-egu26-6791, 2026.
Abstract: Effective evaluation of cultivated land quality is crucial for sustainable agricultural management. Existing research often focuses on regional scales and lacks sufficient detailed analysis of spatial distribution and limiting factors at localized scales. This study aims to select the key indicators to evaluate cultivated land quality and analyze their influence on cultivated land quality at the county level. Taking Changtu County as the research area, principal component analysis (PCA) was employed to identify the most effective Minimum Data Set (MDS) for evaluation, including cultivated layer thickness, soil organic matter (SOM), pH, bulk density (BD), and cultivated layer texture. Additionally, an obstacle degree model was used to analyze restricting factors and their degrees of impact. Results showed that the constructed MDS could replace all indicators for cultivated land quality evaluation. Quality was higher in central areas and lower in eastern and western regions. SOM content decreases with the decrease of quality grade, significantly affecting the quality distribution, and pH in the medium category (Grades III and VI) cultivated lands were lower compared to other categories. Low organic matter content and low pH levels were the main obstacles affecting cultivated land quality, with average obstacle degrees of 43.5% and 29.3%, respectively. Low SOM content significantly affected land quality, particularly in the western region, whereas acidic soils in the eastern region influenced quality grade distribution. Thus, strategies for pH control and SOM enhancement are essential for improving cultivated land quality. This study provides valuable insights into sustainable agriculture.
How to cite: Qian, F.: Quality Evaluation and Limiting Factor Diagnosis of Cultivated Land in Changtu County, Northeastern China, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-2560, https://doi.org/10.5194/egusphere-egu26-2560, 2026.
By carefully balancing the use of nitrogen, phosphorus and potassium fertilisers with crop demands, a nutrient management plan (NMP) aims to enhance crop production while reducing environmental harm due to over fertilisation. As several tools with different complexities can assist farmers in their fertiliser decisions, the characteristics of 14 of the most widely used NMP tools across Europe, China, and New Zealand were compared. All NMP tools considered the field spatial scale, seasonal time scale, and utilised a mass balance approach. To evaluate the tools, matrices of presence/absence of 24 characteristics for their practical use, 22 nutrient cycle processes and 38 required input data were compiled, and cumulative scores were calculated. In addition, two case studies were performed to compare the outputs of NMP tools. Decision support systems such as grano.net® and TUdi tool were highly adaptable and comprehensively described nutrient processes while considering many inputs. By considering only the most important nutrient processes and requiring fewer inputs, software such as PLANET_MANNER and spreadsheets such as Fert_Office were moderately adaptable. Conversely, reference tables tools such as MEM-NAK and Bulgarian tool considered only essential nutrient processes and few inputs and demonstrated limited adaptability. Fertiliser recommendations varied considerably, mainly due to differences in calculating crop nutrient uptake. For a broader application of the NMP tools, differences in the algorithms used to estimate each process, in soil and climate conditions, and in national regulations must be considered. Furthermore, interoperability should be improved in next-generation NMPs to enable data exchange between platforms
How to cite: Chiriac, O., Bakacsi, Z., Pirko, B., López, M.-L., Pareja, E., Liebhard, G., Strauss, P., Zhu, K., Gómez, J. A., Montoliu, J., Jachymova, B., Kraza, J., Ruggeri, M., Volta, B., Guzmán, G., Li, Y., Nikolov, D., Dodd, I., Hudek, C., and Zavattaro, L.: Comparing tools for determining crop nutrient requirements, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4434, https://doi.org/10.5194/egusphere-egu26-4434, 2026.
Soil Bulk Density (BD) is a key physical property integrating soil texture, structure, and soil organic carbon (SOC), and it exerts strong control on porosity, water and solute transport, root penetration, and gas exchange. Because BD links soil mass to volume, it is also a critical conversion factor for estimating SOC stocks. Even small methodological biases in BD determination can therefore propagate into substantial uncertainties in SOC stock estimates, with direct implications for soil health assessments, carbon budgeting, and evaluations of land management and climate mitigation strategies. Despite its critical importance, BD is frequently one of the most commonly missing variables in standard soil datasets. BD can be determined using a range of direct and indirect methods, each involving trade-offs among accuracy, cost, and operational feasibility. Direct approaches (e.g. core, clod, or excavation methods) are widely applied but are labour-intensive and sensitive to operator technique, sampling depth, cylinder dimensions, and soil moisture conditions. Indirect approaches, including pedotransfer functions (PTFs), can reduce field effort and, in some cases, improve spatial coverage, but they require careful calibration and high-quality ancillary data (e.g. texture, organic carbon, climatic variables) and may introduce depth-dependent uncertainties. These issues are particularly critical in heterogeneous Mediterranean landscapes, where BD exhibits strong spatial and vertical variability.
Within this context, a legacy data-rescue activity was conducted for Sicily, the largest island of the Mediterranean Basin, integrating multiple legacy datasets and the first results from the Soils4MED survey, to compile approximately 2,000 records of soil texture, SOC, and BD. Data were collected from the soil province geodatabase of Italy, the Land Use and Coverage Area frame Survey (LUCAS) topsoil, the European Hydropedological Data Inventory (EU-HYDI), the Soils4MED project, and several peer-reviewed papers. To ensure consistency across heterogeneous sampling depths, data harmonization was performed using a mass-preserving approach based on the equal-area quadratic spline method.
To generate spatially explicit estimates, a geostatistical approach was used to create maps at 1 km and 500 m resolution of all input variables of the PTFs before the determination of BD to reduce the propagation of errors in input data. These products, together with the original point data, were compared against established European-scale BD and packing density datasets to identify potentially biased areas. Moreover, an uncertainty BD map was generated by a geostatistical stochastic simulation to provide the quality of BD assessments at the different locations of Sicily. The results highlight the added value of data integration, geostatistics, and PTFs for improving BD representation and supporting robust SOC stock assessments in Mediterranean soils, in line with the objectives of the EU Soil Mission.
Acknowledgements
The SOILS4MED project is part of the PRIMA programme supported by the European Union.
How to cite: Belvisi, G., Buttafuoco, G., Gristina, L., and Scalenghe, R.: Towards soil bulk density maps with quantified uncertainty in Mediterranean soils, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-22071, https://doi.org/10.5194/egusphere-egu26-22071, 2026.
We present soil organic carbon (SOC) data collected from 2010 to 2025 across 146 rainfed and 34 irrigated cropland plots in sandy soils in Western Bahia, a heavily cultivated area in Brazil's Cerrado. Rainfed fields are cropped once annually, while irrigated fields are cultivated twice yearly. Data from 143 plots were gathered from the 0-20 cm layer between 2010 and 2018, and data from 40 plots from 0-100 cm were collected between 2017 and 2025. All plots had native vegetation cleared prior to 1990. The analysis reveals two distinct patterns: rainfed soils appear stable, with no significant carbon changes in both the 0-20 and 0-100 cm layers. Conversely, irrigated soils increased by 2.6% per year (p=0.066) from 2010 to 2018, but declined by -5.75% per year (p<10-5) from 2017 to 2025. In 2018, irrigated fields had higher SOC levels (p=0.034) than rainfed fields, but by 2025, the difference was not significant (p=0.423). Both systems showed no significant difference from the original Cerrado ecosystem (p=0.269 and 0.455, respectively, based on 2018 data). It seems that rainfed soils have reached a long-term equilibrium similar to that of the native Cerrado. The trend in irrigated soils, however, remains puzzling, showing fluctuations from increases to decreases. A combination of high water availability and increased temperatures may have contributed to the decline in SOC in recent years.
How to cite: Heil Costa, M., Espindola, I., and Chartuni Mantovani, E.: Fifteen years of soil organic carbon field measurements in rainfed and irrigated cropland soils across Brazil's Cerrado reveal mixed trends, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8057, https://doi.org/10.5194/egusphere-egu26-8057, 2026.
Measuring soil organic carbon (SOC) stock changes across space and time is crucial for comprehending ecosystem responses to climate variability, land-use change, and agricultural management. However, long-term comparisons remain challenging due to differences in sampling designs, analytical methods, and data structures.
From an integrative approach combining soil depth harmonization, equivalent soil mass method, and statistical analyses, we quantified soil organic carbon (SOC) stock changes across contrasting ecosystems within the Integrated Carbon Observation System infrastructure (ICOS). Historical and labeled? ICOS soil data were harmonized using several complementary approaches. Soil depth limits were harmonized using cubic spline interpolation and a non-model approach based on a 1-mm depth discretization and soil organic carbon (SOC) stocks were compared using equivalent soil mass approach. The comparisons between sampling campaigns were further constrained by accounting for the spatial occurrence of sampling points within similar soil types, defined by comparable coarse fragment contents and similar soil texture. The design-based approach was applied to compare SOC stocks between sampling campaigns. Differences in SOC stocks were assessed using Welch’s t-test, which does not assume equal variances.
Results show that cumulative soil organic carbon (SOC) stock changes in the ~0-60 cm layer are strongly site-dependent, with no consistent trend observed across sites sharing the same land-use type. At the Grignon station (FR-Gri), SOC stocks decreased significantly by 950 ± 40 g C m-2 over 13.2 years (2005-2019), and at the Estrées-Mons A28 station (FR-EM2), a significant SOC loss of 318 ± 145 g C m-2 was observed over 6 years (2015-2021). In contrast, no detectable SOC stocks changes were observed at the Lamasquère station (FR-Lam, 2015-2020), the Lonzée station (BE-Lon, 2007-2017), and the Klingenberg station (DE-Kli, 2008-2019), although SOC stock changes at the latter site showed a non-significant tendency towards an increase (p = 0.07). Minimum detectable difference analysis demonstrates that only SOC losses at FR-Gri can be robustly detected with the current sampling design, while changes at other sites remain below detection limits, underscoring the importance of accounting for methodological sensitivity in long-term SOC assessments.
These differences among agricultural sites may be not related to land use per se, but rather to site-specific management practices, particularly the balance between organic carbon imports and exports, nitrogen fertilization, and soil type. At the Grignon station, observed SOC losses were consistent with simulations from the AMG model. Similarly, at the Lonzée site, the differences between soil inventories fell within the range simulated by RothC, indicating agreement between measured and modeled SOC stocks and supporting near-equilibrium conditions.
Overall, SOC stock changes were highly site-specific, with SOC losses detected only at Grignon, while changes at other sites were close to or below detection limits despite harmonizing the datasets obtained with different sampling methods. These results highlight the need for uncertainty-aware interpretations and monitoring designs optimized for detecting long-term SOC changes.
How to cite: Winck, B., Saby, N., and Loubet, B. and the ETC-Soil & ICOS Station Teams: Contrasting soil carbon stock changes across European agricultural ICOS ecosystems, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12449, https://doi.org/10.5194/egusphere-egu26-12449, 2026.
The Sustainable Development Goal indicator 15.3.1 is defined as the “proportion of land that is degraded over total land area.” This indicator has been used to characterize degraded land surfaces in the Mediterranean region (Cherif et al., 2023), but its specific sensitivity to soil organic carbon (SOC), a critical component of soil health and land productivity, has not yet been assessed in this biogeoclimatic context considered lower in SOC (Romanya & Rovira, 2011).
This study investigates the relationship between SDG 15.3.1 indicator levels (improving, stable, declining) and SOC contents and derived indicators (e.g. SOC/N, SOC/Clay, SOCstock) using LUCAS topsoil datasets (0–20 cm) across Europe. We compared Mediterranean and non-Mediterranean biogeoclimates to assess the indicator’s responsiveness to soil carbon variations.
Our analysis shows that areas classified as “improving” according to the SDG 15.3.1 indicator correspond to soils with significantly higher SOC and SOC stock values than those classified as “stable” or “declining” in the Mediterranean bioclimate specifically—but not in other European regions. No significant differences were found between the “stable” and “declining” SDG 15.3.1 indicator levels.
These findings suggest that SDG 15.3.1 is particularly sensitive to SOC variations in Mediterranean environments, supporting its use as a reliable indicator of soil degradation in these regions. They also highlight the critical role of SOC as a key parameter for assessing soil degradation, especially under Mediterranean climatic conditions, and reinforce the need for region-specific soil monitoring programs that integrate SOC dynamics with vegetation carbon use efficiency (CUE) for more comprehensive land degradation assessments.
Cherif, I., Kolintziki, E., & Alexandridis, T. K. (2023). Monitoring of Land Degradation in Greece and Tunisia Using Trends . Earth with a Focus on Cereal Croplands. Remote Sensing, 15(1766). https://doi.org/https:// doi.org/10.3390/rs15071766
Le, Q.B., Shiri, Z., & Zucca, C. (2025a). Maps of carbon use efficiency (CUE), inter-annual CUE trends and analyses of relationships between CUE trend and current SDG indicators 15.3.1. Deliverable 7.2. WP7 (Enhanced regional soil condition mapping in the MR including C Stock mapping), SOIL health monitoring and information systems FOR sustainable soil management in the MEDiterranean region (SOILS4MED) project, PRIMA.
Le, Q.B., Zucca, C., & Shiri, Z. (2025b). Functional early warnings of land degradation revealed by carbon-use efficiency across the Mediterranean Eco-Region. Manucript in prepation.
Romanya, J., & Rovira, P. (2011). An appraisal of soil organic C content in Mediterranean agricultural soils. Soil Use and Management, 27, 321–332. https://doi.org/10.1111/j.1475-2743.2011.00346.x
How to cite: Martin, N., Caner, L., Le, Q. B., and Zucca, C.: Sensitivity of Soil Organic Carbon and Derived Indicators to the SDG 15.1.3 Model in Mediterranean Versus Other European Biogeoclimatic Regions., EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-23198, https://doi.org/10.5194/egusphere-egu26-23198, 2026.
Groundwater recharge is an ecosystem function provided by forests. At our intensive forest monitoring station “Britz” (Germany) we measure deep seepage, which later becomes ground water recharge, for various tree species and compositions with large scale lysimeters (each covering 100 m² of forest and a depth of 5 m) operating since the 1970s.
Those long-term observations show clear differences between species and forest compositions. With deep seepage being the difference between precipitation reduced by evapotranspiration, evergreen species show little seepage in most years. In dry years our Scots pine stand actually showed no deep seepage at all.
This changed in recent years, with extreme weather events leading to unprecedented seepage patterns. The continent‑wide, multi-year drought that began in 2018 had severe impact. For the first time in 2019, the beech plot—normally a strong contributor to deep infiltration—recorded no measurable annual deep seepage. Conversely, extreme summer precipitation events occurred each year from 2018 to 2021. While this led to flooding in many regions of Germany, in Britz it led to a considerable proportion of the total annual deep seepage under pine stands.
These novel dynamics reshape our understanding of how deep seepage is generated and highlight the sensitivity of groundwater recharge in the north‑eastern German lowlands to extreme climatic fluctuations.
How to cite: Natkhin, M., Strer, M., Schad, T., Schwärzel, K., and Sanders, T. G.: Long-term deep seepage monitoring in temperate forests: lysimeter evidence of climate‑driven hydrological shifts in northeastern Germany, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12513, https://doi.org/10.5194/egusphere-egu26-12513, 2026.
Soil organic carbon (SOC) is central to regulating the global carbon cycle and underpinning food security, yet unsustainable cultivation has resulted in a continuing SOC loss and has made it highly vulnerable to climate change. In China, the Well-Facilitated Farmland Construction (WFC) initiative has sought to enhance soil conditions by integrating farmland management units (FMUs) and adopting improved practices, including optimized irrigation, straw incorporation, and targeted fertilization strategies. Since its launch in 2013, the WFC project has been implemented across more than 50 million hectares of farmland in China. However, its spatio-temporal impacts on SOC remain poorly understood. To address this gap, we focused on three representative regions, Shunyi, Rudong, and Dangtu, to examine the impact of FMU integration. A total of 1,549 soil profiles were compiled to calibrate the process-based CENTURY model and simulate long-term variations in topsoil (0-20 cm) SOC density (SOCD) across Chinese farmlands. Results show that, following the WFC project, farmland fragmentation decreased while SOCD increased, with strong negative correlations between fragmentation degree and SOCD (-0.88 in Shunyi, -0.94 in Rudong, and -0.52 in Dangtu). These findings indicate that farmland patch integration contributes significantly to SOC sequestration. By comparing future SOC trajectories under traditional versus WFC practices, we found that WFC management offers substantial sequestration potential, increasing SOC stocks by 3.34 Pg under SSP1-2.6 and 2.86 Pg under SSP5-8.5 during 2030-2100 in Chinese farmlands. This sustained increase reflects the synergistic effects of fertilization and elevated CO₂, enhanced crop productivity through optimized irrigation, greater organic inputs from straw incorporation, and reduced microbial decomposition under balanced nitrogen fertilization. In conclusion, WFC demonstrates a scalable pathway toward more resilient and climate-smart food systems.
How to cite: Wang, Z., Zhao, R., Xue, J., Lu, R., Chen, Z., Yu, Q., Chen, W., Zhang, Q., Shi, Z., and Chen, S.: Well-facilitated Farmland Construction enhances soil organic carbon storage in China, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4693, https://doi.org/10.5194/egusphere-egu26-4693, 2026.
Soil Organic Carbon (SOC) is a critical indicator of soil health, yet conventional laboratory-based estimation methods remain costly, time-consuming, and environmentally burdensome. This study evaluates a rapid, low-cost, and environmentally friendly alternative for SOC estimation using high-resolution color information acquired from a Nix Spectro 2 handheld sensor, integrated with machine learning and generative data augmentation approaches. A total of 641 soil samples collected across diverse agro-ecological regions of West Bengal, India, were analyzed using Random Forest, Gradient Boosting, XGBoost, and Artificial Neural Network models. To address data imbalance and limited sample representation at higher SOC ranges, synthetic datasets were generated using Gaussian Mixture Models (GMM), Generative Adversarial Networks (GAN), k-nearest neighbors–based augmentation, and bootstrapping techniques. Among the evaluated models, Random Forest achieved the best baseline performance (R² = 0.71), which further improved with GMM-based data augmentation (R² = 0.77). The results demonstrate the strong potential of combining handheld color sensing with generative artificial intelligence to develop more accurate, robust, and scalable SOC prediction frameworks.
How to cite: Singh, R., Chakraborty, S., and Weindorf, D. C.: From Soil Color to Carbon: A Generative AI and Nix Sensor Approach, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-2786, https://doi.org/10.5194/egusphere-egu26-2786, 2026.
Assessing soil macronutrients across diverse landscapes requires a transition from conventional, time-consuming, and labour-intensive wet chemistry analysis to rapid, low-cost, and non-destructive proximal sensing techniques. In this study, the individual performance as well as the synergistic potential of Portable X-ray Fluorescence Spectrometry (PXRF) and Visible–Near-Infrared (VisNIR; 350–2500 nm) diffuse reflectance spectroscopy were evaluated to enhance the prediction accuracy of soil available Nitrogen(N) and available Phosphorus(P). A total of 609 soil samples were collected from agricultural fields across West Bengal, India, representing a wide range of land-use patterns. Laboratory analysis of N and P served as the ground-truth data for evaluating several machine learning architectures. For individual sensor modelling, Linear Regression (LR), Random Forest (RF), Support Vector Regression (SVR), Partial Least Squares Regression (PLSR), and Extreme Gradient Boosting (XGBoost) models were independently developed for the PXRF and VisNIR datasets to assess their standalone predictive performance. In addition, Granger–Ramanathan Averaging (GRA) was implemented for multi-sensor data fusion using two strategies. The first was a linear approach in which Artificial Neural Networks (ANN) served as base learners with Ordinary Least Squares (OLS) as the meta-learner. The second was a non-linear approach in which RF replaced the linear meta-learner to capture complex data interactions. The results demonstrated that the prediction performance of the single-sensor models was poor but was improved through the GRA fusion framework. The GRA approach with OLS regression showed a slight improvement for P (R² = 0.45, RMSE = 63.20 Kg/ha, ratio of performance to interquartile distance (RPIQ) = 0.89) and N (R² = 0.17, RMSE = 68.86 Kg/ha, RPIQ = 0.96) compared with PXRF and VisNIR in isolation. However, GRA with a non-linear RF meta-learner significantly outperformed the linear strategies and markedly enhanced prediction accuracy for N (R² = 0.85, RMSE = 29.54 Kg/ha, RPIQ = 2.23) and P (R² = 0.87, RMSE = 31.18 Kg/ha, RPIQ = 1.80). These findings indicated that although multi-sensor fusion consistently outperformed single-sensor models, the relationship between sensor data and soil N and P concentrations was fundamentally non-linear. Consequently, these nutrients required the complex weighting capabilities of non-linear architectures, which traditional linear models failed to capture. This methodology offers a scalable solution for assessing soil health in tropical agroecosystems and encourages further exploration of digital mapping approaches for additional soil nutrients.
How to cite: Dam, U. and Chakraborty, S.: Optimizing Multi-Sensor Fusion Architectures: The Role of Non-Linear Meta-Learning in Predicting Soil Nitrogen and Phosphorus, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-6859, https://doi.org/10.5194/egusphere-egu26-6859, 2026.
Climate change and the intensification of anthropogenic activities affect the dynamics of carbon in the soil, resulting in losses in stock and increased emissions of carbon dioxide (CO2) into the atmosphere. The balance between the continuous input and output of carbon in the soil, as well as its sequestration from the atmosphere, contributes to the formulation of strategies to mitigate climate change and global warming. Our hypothesis is that it is possible to accurately predict soil CO2 emissions and soil organic carbon (SOC) stock using hyperspectral sensing and machine learning (ML) algorithms. The objectives of the study were: (i) to predict CO2 emission and SOC stock using hyperspectral sensor and ML algorithms; (ii) to identify algorithms and dataset inputs with the highest accuracy in predicting CO2 emission and SOC stock. Samples were collected from three biomes in the State of Mato Grosso do Sul, Brazil: Cerrado, Atlantic Forest, and Pantanal. Within each biome, four land use classes were assessed: agriculture, pasture, eucalyptus plantations, and native vegetation. Data was collected from 100 points distributed in each area within each biome. In all sample point, carbono stock was measured in three deepths (0-10cm, 10-20 cm, and 20-40 cm). In situ soil CO2 (FCO2), temperature and moisture measurements were also performed. Hyperspectral data were collected by a sensor in each sample point and then the spectral bands used by MODIS sensor (seven bands) were obtained. Data were submited to ML analysis, in which two input configurations in the dataset were tested: using all the bands provided by the hyperspectral sensor (ALL) and using only the bands used by the MODIS sensor (B). Carbon stock at the three depths, FCO2, soil temperature and moisture were used as output in datasets. ML models tested were: Artificial Neural Network (ANN), Decision Tree models REPTree and M5P, Random Forest (RF), Support Vector Machine (SVM), and a simple model used as control (ZeroR). Our findings reveal that the use of hyperspectral sensing and ML algorithms enables accurate prediction of CO2 emissions and SOC stock. The choice of ML model for accuratelly predicting soil CO2 emissions and carbon stocks is dependent on the input variables used in the datasets, in which SVM provides the highest accuracy when applied to all spectral bands, while RF shows better performance when using the MODIS bands. Therefore, the approach used here can provide large-scale estimates of soil CO2 emissions and organic carbon stock.
How to cite: Teodoro, P., Teodoro, L., Silva, N., Ratke, R., Santana, D., and Campos, C.: Optimizing soil carbon and CO2 emission prediction: a integration of machine learning algorithms and VIS/NIR data inputs, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8331, https://doi.org/10.5194/egusphere-egu26-8331, 2026.
Land use and land cover (LULC), as well as the biome in which they are located, influence soil carbon stocks. Tropical soils hold a significant portion of the world’s carbon stocks, a result of high temperatures, precipitation, and management practices that drive temporal variability in soil respiration. Based on this context, the hypothesis of this study is that land use and land cover influence soil carbon stocks. Therefore, the objective of this study was to evaluate soil carbon stocks under different LULCs in the Cerrado, Pantanal, and Atlantic Forest biomes, located in the state of Mato Grosso do Sul, Brazil. The following land uses and covers were assessed in each biome: agriculture (represented by soybean cultivation), well-managed pasture, eucalyptus plantation, and native vegetation. Carbon stocks were quantified by analyzing total soil carbon in the 0–10, 10–20, and 20–40 cm layers at 100 sampling points for each LULC and biome. Principal component analysis (PCA) was performed to identify interrelationships between carbon stocks at different depths across LULCs and biomes. A Pearson correlation network was also constructed to graphically represent correlations between LULCs × depths and biomes × depths for soil carbon stock contents. Our findings reveal variation in soil carbon stocks in relation to land use, land cover, and the corresponding biome. Cerrado biome exhibited the highest carbon stocks under eucalyptus, pasture, and agriculture at the 20–40 cm depth, while the lowest carbon stocks were observed in the Atlantic Forest biome under agricultural use at the 0–10 and 20–40 cm depths. Identifying the main LULCs that contribute to carbon sequestration in each biome is essential for establishing greenhouse gas mitigation targets, thereby contributing to the minimization of global climate change.
How to cite: Silva Campos, C. N., Pereira Ribeiro Teodoro, L., Pereira da Silva, N., Felippe Ratke, R., Santana Cordeiro, D., Eduardo Teodoro, P., and Leticia Monteiro Gomes, M.: Mitigating climate change through sustainable land management: insights from soil carbon stock analysis in Brazil, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8337, https://doi.org/10.5194/egusphere-egu26-8337, 2026.
Soils contain a great diversity of microorganisms. Edaphoclimatic characteristics and land use affect the biological diversity of the soil. The hypothesis of this study is that the land use influences the diversity of fungi and bacteria and is correlated with the stock and emission of carbon in the soil. The aim was to identify which land uses, among native forest, agriculture, pasture and eucalyptus, in the three biomes of the State of Mato Grosso do Sul, Brazil (Cerrado, Atlantic Forest and Pantanal), have the highest microbiological diversity and to understand this relationship with soil carbon emissions and stocks. Carbon stock was assessed by analyzing total soil carbon in the layers sampled (0-10, 10-20 and 20-40 cm) at 100 sampling points for each land use and biome, while carbon emission was assessed at the same points using an EGM 5 portable device. Soil samples were grouped into a composite sample for each use and biome for microbiological identification analysis. Bacteria and fungi were identified using the 16S rRNA sequencing method and ITS1/ITS2 PCR, respectively. Our findings reveal that the abundance and diversity of bacteria and fungi is influenced by land use and biome. Cerrado has lower carbon emissions and higher stocks, and a high incidence of beneficial microorganisms of the genera Metarhizium and Bacillus and pathogenic microorganisms of the genera Penicillium and Fusarium. Atlantic Forest biome has higher carbon emissions and lower carbon stocks, and a higher number of beneficial microorganisms of the Bacillus genus. There was a greater carbon emission and stock in the pasture, with a high number of Bacillus, low temperature and high humidity. Agriculture emitted less carbon and stored little, with the presence of Fusarium and a moderate amount of Bacillus.
How to cite: Teodoro, L., Silva, N., Ratke, R., Santana, D., Teodoro, P., and Campos, C.: Microbial diversity mirrors carbon stocks and emissions in soils under contrasting land uses in Brazil, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8353, https://doi.org/10.5194/egusphere-egu26-8353, 2026.
Soil Organic Carbon (SOC) stocks in natural and semi-natural ecosystems remain poorly quantified in intensively cultivated lowland regions, such as Vojvodina, Serbia, which is part of the Pannonian Basin. To address this gap, we developed a general framework for large-scale, multipurpose soil sampling. The study further investigates the potential of machine learning-based regression and classification approaches to predict SOC stocks in forest and grassland ecosystems using diverse land cover and soil-related indicators as key predictors.
Spatial clustering for the planning of soil sampling was conducted by combining information from different sources. To implement a systematic and stratified sampling scheme, we followed the LUCAS methodology. Natural and semi-natural forest and grassland areas were delineated using Copernicus LULC data. An exploratory analysis was conducted using climate variables from C3S Copernicus (2015–2024) and soil properties (soil order and type, and silt, sand, and clay proportions) to identify spatial clusters suitable for field sampling. Further, forests and grasslands were clustered separately using an unsupervised K-prototypes approach. Based on this approach, 62 representative locations were identified across forests and grasslands, from which a total of 186 soil samples were collected using composite sampling at three sites per location.
Land cover features were collected along 250 m transects at each location, and landscape heterogeneity was quantified using LUCAS-based diversity indicators derived from the same transects. For machine learning–based SOC stock prediction, these indicators were combined with soil descriptors, including soil texture, soil type, and geomorphology-based soil groups, as well as spatial cluster information for forest and grassland areas. SOC stock values were averaged per location, and forest and grassland samples were jointly used in the modeling to capture landscape heterogeneity.
Regression modeling aimed to predict continuous SOC stock values, while classification categorized SOC stock into low, medium, and high levels based on thresholds derived from K-means clustering applied to the observed SOC distribution. Among the regression models, Elastic Net achieved the highest performance, with an R² of 0.49 and an RMSE of 13.74 t ha⁻¹, indicating moderate predictive capability given the complexity of SOC stock dynamics and the limited sample size. In contrast, classification models demonstrated higher predictive reliability. Logistic Regression achieved the best performance, with an overall accuracy of 76.9% and a macro F1-score of 77.1%, suggesting that SOC stock can be more robustly distinguished across discrete classes than predicted as a continuous variable. Permutation importance analysis revealed that soil texture was the dominant predictor in both regression and classification models.
Overall, the findings highlight the combined importance of soil properties and landscape diversity indicators for SOC stock prediction in natural and semi-natural ecosystems. While continuous SOC stock prediction remains challenging, classification into discrete SOC stock classes provides higher accuracy and more stable performance, suggesting a more reliable framework for SOC stock assessment in heterogeneous landscapes. Independently, this study establishes the first SOC reference framework for natural and semi-natural ecosystems in Vojvodina, providing a conceptual basis for spatial analysis and mapping.
How to cite: Marković, M., Brdar, S., Kalkan, K., Knežević, M., and Nikolić Lugonja, T.: From Ground Truth to Regional Insights: Soil Organic Carbon Predictions in Heterogeneous Landscapes using ML and multipurpose sampling, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11030, https://doi.org/10.5194/egusphere-egu26-11030, 2026.
Artificial soils known as technosols are increasingly promoted as a nature-based solution for land restoration and carbon sequestration, yet their long-term capacity to store and stabilise soil organic carbon (SOC) remains poorly constrained. This study evaluates the dynamics, stability and sequestration potential of carbon in experimentally constructed soils under contrasting mineral and organic amendments in Galicia (NW Spain). Six treatments (including artificial soils with and without biochar and dunite, a dunite residue soil and an untreated control) were monitored at two depths (0–15 and 15–30 cm) over an 11-month field campaign and combined with process-based modelling to assess medium-term SOC trajectories.
A comprehensive laboratory dataset including physico-chemical properties (pH, electrical conductivity, bulk density), nutrient status, and functional carbon fractions (CWE, CHE, REM, CFA, CHA, REC) was analysed using multivariate statistics. Principal component analysis revealed that the first two components explained 65–75% of total variance, with PC1 driven by total and organic carbon and PC2 reflecting carbon quality and stabilisation, largely controlled by C/N ratio. Treatments containing biochar (particularly when combined with dunite) exhibited the highest stocks of recalcitrant carbon and the most advanced progression towards stabilised organic matter.
These experimental data were integrated into a multimodel ensemble (RothC, ICBM, Century, Yasso07, AMG and SG) implemented in R using the SoilR framework. After a 1000-year spin-up, 10-year forward simulations were run under two contrasting carbon-input scenarios (2.8 and 5.8 Mg C ha⁻¹ yr⁻¹). The ensemble showed strong sensitivity to amendment type and carbon inputs. Biochar-based technosols consistently produced the highest SOC stocks, with the biochar + dunite treatment gaining up to +5.3 Mg C ha⁻¹ over 10 years under high-input conditions. Conversely, soils without biochar exhibited either near-equilibrium behaviour or limited sequestration capacity.
Overall, the results demonstrate that combining biochar with mineral amendments creates synergistic mechanisms for long-term carbon stabilisation in artificial soils. The multimodel approach provides a robust framework for quantifying uncertainty and supports the deployment of engineered technosols as effective, scalable carbon sinks in land restoration strategies.
How to cite: Bravo Garcia, J., Anaya Romero, M., and Blanco Velázquez, F. J.: Soil organic carbon dynamic in artificial soils under different treatments, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11327, https://doi.org/10.5194/egusphere-egu26-11327, 2026.
Agri-photovoltaic (Agri-PV) systems are increasingly implemented for combined food and renewable energy production, yet their impacts on soil carbon cycling and stabilization remain insufficiently understood. This study aimed to understand how shading in a horticultural Agri-PV field alters microbial activity, soil carbon (C) stocks and stability compared with a conventional open-field management in western Germany.
The experimental fields were established in 2021 and soils (homogeneous silt loam: 5% sand, 15% clay, 80% silt) were sampled (0–30 cm) in 2025 along replicated transects across under-panel (UP) and inter-row (GAP) zones in the Agri-PV field. This Agri-PV system consisted of south-facing, fixed-tilt PV modules inclined at 20° and mounted at a maximum height of 4.30 m. An adjacent, identically managed Control open-field was sampled using the same approach. We measured gravimetric soil water content, microbial biomass C (Cmic), aggregate size distribution (8–2 mm, <2 mm), total organic carbon (TOC) stocks, and organic carbon stocks in density organic matter fractions (free-light, FLF; occluded-light, OLF; mineral-associated, MAOC), together with δ¹³C of bulk soil and fractions.
Relative to the Control, Agri-PV soils showed degradation of biological functioning and carbon pools across UP and GAP zones. Soil moisture was 22–24% lower in UP zones and 11–21% higher in GAP zones, reflecting rainfall redistribution by the panel structures. Cmic declined by 39–48% in UP zones and by 18–26% in GAP zones. TOC stocks were 16–29% lower in Agri-PV than in the Control. FLF stocks declined by 44–72% in UP zones and by 36–45% in GAP zones, reflecting reduced plant growth and biomass addition to soil in Agri-PV field. This partially explains why OLF stocks were reduced by 45–53% in GAP zones, while MAOC declined by 20–29% in UP zones and 10–12% in GAP zones in the Agri-PV compared to respective sampling positions in Control. Carbon fractions were consistently enriched in ¹³C relative to the Control (+0.18‰ to +0.33‰ in Bulk Soil, up to +1.26‰ in FLF and +0.63‰ in OLF), indicating enhanced microbial processing and reduced fresh biomass inputs.
Within the Agri-PV system, strong spatial heterogeneity emerged due to shading. UP zones were 20–40% drier than GAP zones and contained 20–30% less Cmic. The <2 mm aggregate size percentage was 12–30% higher in UP zones than in GAP zones, indicating pronounced aggregate breakdown beneath the panels. These microenvironmental gradients drove a clear redistribution of carbon pools within the Agri-PV system: TOC, MAOC, and FLF stocks were 9–17%, 14–22%, and 13–53% higher in GAP than in UP, respectively, whereas OLF stocks accumulated preferentially in UP, where they were 36–70% higher than in GAP. No comparable spatial gradients were observed in the Control, indicating that the patterns in Agri-PV are attributable to shading.
Our results demonstrate that fixed-tilt south-facing Agri-PV systems can substantially disrupt soil C stabilization by simultaneously reducing biomass inputs and by destabilizing soil structure (in silt-rich soils), with important implications for long-term soil resilience and carbon stabilization.
How to cite: Bendijo, J. R. M., Brüggemann, N., Muller, O., Meier-Grüll, M., Siebers, N., Wissel, H., Moreno-Racero, F. J., Rodríguez, L. G., Jedmowski, C., and dos Anjos Leal, O.: Shading gradients shape soil microbial biomass carbon and carbon stability in a horticultural agri-photovoltaic field in Germany, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-15513, https://doi.org/10.5194/egusphere-egu26-15513, 2026.
Organic soils store a disproportionately large share of terrestrial soil carbon and play a key role in climate change mitigation. However, their high spatial variability and sensitivity to sampling and preparation procedures make routine monitoring of soil organic carbon (SOC) and related properties challenging. In this study, we evaluated the potential of visible–near infrared (VIS–NIR, 350–2500 nm) spectroscopy for the assessment of SOC and pH in organic soils under both field-moist and laboratory-dried conditions.
A dataset of more than 300 organic soil samples, including peat, muck, gyttja, and mineral–organic soils, was collected from reference soil profiles across Poland at two depth intervals (0–20 cm and 40–60 cm). Spectral measurements were acquired using a PSR-3500 spectroradiometer. Principal component analysis (PCA) was used to explore spectral variability among soil types, while partial least squares regression (PLS) and support vector machine (SVM) models were developed for SOC and pH prediction. Model performance was evaluated using independent validation datasets.
PCA revealed clear separation of major organic soil groups, reflecting differences in organic matter composition and degree of decomposition. SOC prediction accuracy was consistently higher for models developed on dried samples, while models based on field-moist samples showed reduced but still informative performance. Among the tested approaches, SVM generally outperformed PLS for SOC prediction, although model performance varied depending on soil type and calibration subset. Predictions of soil pH were less accurate than those for SOC but captured broad trends relevant for monitoring applications.
Overall, the results indicate that VIS–NIR spectroscopy provides a robust and non-destructive tool for SOC and pH assessment in organic soils, particularly under standardized (dried) conditions. While moisture effects remain a limitation for field-moist measurements, the approach shows strong potential for supporting soil carbon monitoring and digital soil assessment frameworks in natural and agroecosystems.
How to cite: Debaene, G. and Bartosiewicz, B.: Monitoring soil organic carbon and pH in organic soils using VIS–NIR spectroscopy under field and laboratory conditions, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17173, https://doi.org/10.5194/egusphere-egu26-17173, 2026.
Boreal soils are vital carbon reservoirs, playing a crucial role in the global carbon cycle. The northern boreal landscape is characterized by small-scale variations of forests, wetlands and fells, each with widely differing carbon stocks. These ecosystems are warming at rates significantly faster than the global average, making them a priority for climate change research. To better understand carbon dynamics, it is essential to investigate the factors influencing the magnitude and spatial distribution of carbon stocks.
This study analyzed soil organic carbon (SOC) stocks, the environmental factors driving their spatial distribution, and the reliability of various modeling methods for SOC variability. The data set includes 217 soil organic matter and carbon content samples, total soil depth, soil organic layer thickness, and remote sensing data (land cover, topography, vegetation). SOC stocks were modeled and predicted across the entire study area using an ensemble of five approaches: Generalized Linear Models, Generalized Additive Models, Generalized Boosted Models, Support Vector Machines and Random Forest, validated with leave‑one‑out cross‑validation
The results indicate that biotope, groundcover and soil wetness index are the primary factors influencing SOC variation, while secondary factors include slope, elevation and topographic position. Soil organic layer thickness ranges from 0.0 to 4.4 meters, with an average of 0.4 meters. The total estimated carbon stock for the 150 km² study area is approximately 1.86 Mt (14.3 kg/m²), with the highest stocks (205 kg/m²) found in aapa mire wetlands. The median leave-one-out cross validation result across the five methods was RMSE = 29.3, MAE = 14.0, and R² = 0.41.
The study shows that fine‑scale variation in biotopes, groundcover and terrain‑driven wetness shapes SOC patterns across northern boreal landscapes. Valley‑bottom wetlands, especially aapa mires, hold exceptionally large carbon stocks and play a central role in the boreal carbon cycle. Robust carbon stock data is essential for improving climate predictions and guiding effective mitigation efforts.
How to cite: Reijonen, A., Virkkala, A.-M., Lehtinen, J., and Luoto, M.: Spatial patterns and modelling of soil organic carbon stocks across the northern boreal landscape , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17826, https://doi.org/10.5194/egusphere-egu26-17826, 2026.
Soil organic carbon (SOC) is a key indicator of soil quality and an essential component in climate change mitigation. Its monitoring faces limitations when based solely on field data, which drives the search for complementary methodologies such as remote sensing and simulation models. The aim of this study was to assess the potential of integrating remote sensing–derived information into the estimation and modeling of SOC dynamics in the Guadiamar Green Corridor (Seville, Spain), an area undergoing restoration following the 1998 mining spill. Two methodological approaches were employed at landscape and sublandscape level: (i) the spatial prediction of SOC and clay content using a Random Forest (RF) model trained with Sentinel-2 spectral variables, and (ii) the simulation of SOC dynamics with the RothC model under seven boundary conditions (BC0–BC6), in which field-measured variables were progressively replaced by proxies obtained from remote sensing.
The Random Forest model showed moderate performance (R² ≈ 0.47 in training and validation), displaying spatial coherence between areas with higher clay content and higher SOC levels. In the case of RothC, except for BC1, all simulations reproduced a decreasing trend in SOC but did not reach the magnitude of loss observed in the field. Scenario BC2, which simulated with clay percentage data obtained through RF, showed the greatest similarity to the reference scenario (BC0), while BC5, based on remote sensing–derived potential evapotranspiration data, generated a marked underestimation of final SOC, highlighting the model’s sensitivity to this parameter. The results suggest that remote sensing is a valuable tool to complement field measurements in SOC modeling, especially in contexts with limited data availability. However, accuracy depends on the variable being substituted and on model calibration to the specific conditions of ecological restoration.
How to cite: Bravo Garcia, J., Blanco Velázquez, F. J., and Anaya Romero, M.: Assessing and modelling soil organic carbon dynamics in a mining spill restoration area, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20731, https://doi.org/10.5194/egusphere-egu26-20731, 2026.
Maintaining soil fertility in areas at risk of soil degradation is crucial, particularly given their vulnerability in the current context of climate change. Decision Support Tools (DSTs) designed specifically for farmers are essential for evaluating risks for soil health, analysing the impact of agricultural practices, and defining strategies to mitigate the negative impacts on soil health. The TUdi DSTs (available in mobile app and web format) integrate tools for farmers that address different types of soil degradation processes, such as soil biology, erosion, compaction, structure, soil organic carbon dynamics and fertilization, with the aim of restoring and enhancing soil health.
This work presents the results of the operational implementation of the TUdi APP on a commercial farm under arid conditions. The TUdi APP was used to monitor and mitigate actual and potential risks related to soil erosion and fertility loss across different crops and fields, fostering more sustainable agricultural practices.
How to cite: López Gonzalez, M.-L., Pareja-Serrano, E., Benavente-Ferraces, I., and Guzmán, G.: Implementation of the TUdi APP Decision Support Tool for soil health assessment in commercial farms, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5878, https://doi.org/10.5194/egusphere-egu26-5878, 2026.
