The adequate quantification of soil organic carbon (SOC) turnover is a pressing need for improving soil health and understanding climate dynamics. It is controlled by the complex interplay of microbial activity, availability of carbon (C) and nitrogen (N) sources, and the dynamic restructuring of the soil's architecture. Accurate modeling of SOC dynamics requires the representation of these processes at small spatial scales to help understanding the mechanisms that drive these processes.
Among them are the enzymatic degradation of particulate organic matter, the cycling of microbial necromass, but also short-term influences as root exudation. As such, microbial growth and turnover, C respiration and N cycling depend on the C/N ratios of the different organic carbon sources.
We show the feasibility to include such a variety of processes in a microscale model, along with the possibility to simulate soil structure dynamics including the stabilization of soil particles, POM or microbial necromass via organo‐mineral associations. The computational framework is a cellular automaton model that allows to create virtual soils on the basis of µCT or video analysis data of aggregates. Parameters are chosen consistently from rhizosphere experiments without parameter fitting to explore the influence of soil structural heterogeneity and connectivity, N limitation, or necromass formation on SOC storage.
Our results highlight that evolving soil architecture and pore connectivity control substrate accessibility, creating micro‐scale hot and cold spots for microbes. N availability consistently co-limits microbial growth, while a favorable C/N ratio of root exudates substantially reduces respiration and increases CUE over extended periods. Necromass emerges as long‐term SOC pool, as N from short‐term root exudation pulses promotes biomass growth and is converted into slowly degradable necromass, which can be physically protected through occlusion. The findings align with lab experiments and additionally allow us to elucidate the spatial and temporal dynamics of the drivers of carbon turnover. We also present an option to couple such microscale simulations to macroscale transport model for, e.g., CO2 across soil profiles.
How to cite: Prechtel, A., Rötzer, M., and Ray, N.: All in? Soil organic carbon and nitrogen turnover modeling including structure dynamics, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11791, https://doi.org/10.5194/egusphere-egu26-11791, 2026.
Soil organic matter turnover and microbial metabolism are fundamentally driven by the acquisition and utilization of carbon, energy, nitrogen and further nutrients. Understanding how microbial processes respond to different energy and substrate conditions is therefore essential for revealing the mechanisms controlling soil carbon turnover and storage. This study focuses on the microscale dynamics of microbes interacting with different substrates, as well as the associated evolution of metabolic energy. Using a Cellular Automaton framework, a process-based model is developed to couple microbial activity with carbon, nutrients, energy as well as structural dynamics. The model includes local interactions of microbial consumption of organic carbon, nutrient uptake, degradation, and growth, while simultaneously representing the internal energy dynamics of the system. Based on this model, we investigate how different substrate conditions—characterized by varying energy content, stoichiometric properties, and spatial distributions—and connectivity impact energy dynamics, microbial community formation, and necromass accumulation.
How to cite: Peng, C. and Ray, N.: How Substrate Properties and Spatial Connectivity Shape Microbial Energy Dynamics and SOM Turnover, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-1852, https://doi.org/10.5194/egusphere-egu26-1852, 2026.
Integrated simulation of crop growth and soil water dynamics is essential for improving the understanding of agro-hydrological processes and advancing agricultural water resource management. In this study, a coupled agro-hydrological modeling framework was developed by integrating the crop growth model WOFOST with the soil water flow model HYDRUS-1D to explicitly represent interactions among crop development, root water uptake, and soil moisture dynamics. The framework was applied to a maize cropping system located in an arid and semi-arid region characterized by shallow groundwater, where strong soil–crop–atmosphere interactions and groundwater influences pose significant challenges to conventional modeling approaches. Model parameters were calibrated and validated using field observations collected during the 2017–2018 growing seasons, incorporating site-specific climate data, cultivar parameters, and detailed agricultural management information. To address uncertainties arising from parameter variability and model structural limitations, data assimilation techniques were further embedded into the coupled framework. Observations of soil water content (SWC), leaf area index (LAI), and evapotranspiration (ET) were assimilated using the Ensemble Kalman Filter (EnKF) and four-dimensional variational data assimilation (4D-Var), enabling dynamic correction of both soil hydrological states and crop growth variables. The results demonstrate that the coupled WOFOST–HYDRUS-1D system reliably captures crop–soil–groundwater interactions under shallow groundwater conditions. Data assimilation substantially improves simulation accuracy by reducing soil moisture bias, constraining crop growth trajectories, enhancing ET estimation, and lowering predictive uncertainty throughout the growing season. The proposed framework provides a robust and potentially transferable tool for agro-hydrological simulation in water-scarce regions and supports improved irrigation management and decision-making in precision agriculture.
How to cite: Qin, X., Zhang, C., and Huo, Z.: Improving Agro-Hydrological Process Simulations in Cropping Systems by Coupling WOFOST and HYDRUS-1D with Data Assimilation, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8632, https://doi.org/10.5194/egusphere-egu26-8632, 2026.
Accurate spatio-temporal information on the soil water balance is critical for an efficient and sustainable irrigation. Recent irrigation scheduling approaches are often limited to a representation of (i) the local or point scale soil water balance by in-situ measurements, (ii) solely surface soil water contents at a coarse spatial resolution by microwave remote sensing technologies, or (iii) only selected components of the soil water balance by simple crop evapotranspiration models. To reconcile the need for accurate estimates of different components of the soil water balance with feasible effort, this study proposes the application of physically-based one-dimensional soil water balance models in a spatially-distributed manner.
The HYDRUS-1D software environment is applied at 70 m spatial resolution across a 1,600 ha study farm in Mecklenburg-Western Pomerania, Germany, with heterogeneous soil textures and different crops. Depth-specific (0 cm to 60 cm, in 10 cm increments) soil water balance simulations were conducted from 1st April to 30th September 2021 and 2022 to estimate the soil water content, plant available water content, infiltration, crop evapotranspiration, root water uptake, and deep percolation, at daily intervals. Simulated soil water contents were validated against in-situ measurements and two microwave remote sensing surface soil water content datasets (“Soil Moisture Active Passive”, SMAP; Sentinel-1, S1-SWC). Spatially distributed irrigation demands and irrigation timings at daily intervals, crop-specific irrigation efficiencies and potential farm-scale water savings are estimated using the simulated soil water balance to explore the contribution of this simulation framework for precision irrigation.
The average simulation performance metrices were Root Mean Square Error (RMSE) = 0.020 m3 m-3, Mean Absolute Error (MAE) = 0.017 m3 m-3, coefficient of determination (R²) = 0.676, and bias = -0.008 m3 m-3, showing a good accuracy of spatially-distributed HYDRUS-1D simulations. The agreement with remotely-sensed data was moderate to weak (RMSEmean = 0.059 (0.150) m3 m-3, MAEmean = 0.049 (0.123) m3 m-3, R2mean = 0.208 (0.141), mean bias = 0.021 (0.108) m3 m-3 for SMAP (S1-SWC)). Average crop specific irrigation efficiencies were 65.0% (potato), 47.3% (wheat), 40.5% (rye), and 58.2% (sugar beet). Potential water savings amounted to 87,006.9 m³ (11.2 % of the applied irrigation water; 2021) and 71,396.6 m³ (10.4 %; 2022).
The proposed simulation framework offers an easy-to-adopt and physically-based foundation for the estimation of crop-specific irrigation demands and irrigation timings at high spatial resolution. Further accuracy improvements by using depth-specific remote-sensing derived soil water contents (“Soil Water Index”) for model calibration are under ongoing investigation.
How to cite: Wenzel, J. L., Conrad, C., Mahmood, T., Volk, M., and Pöhlitz, J.: Supporting precision irrigation scheduling in the heterogeneous landscape of North-Eastern Germany by spatio-temporally distributed HYDRUS-1D soil water balance simulations and remote sensing data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4765, https://doi.org/10.5194/egusphere-egu26-4765, 2026.
Anticipated future changes in precipitation patterns are expected to affect deep percolation (DP) through the vadose zone and groundwater recharge (GWR) of semi-arid regions such as Malta. Moreover, a diverse range of agricultural practices, from rainfed to irrigated agriculture, complicates the relationship between storm characteristics (magnitude, duration, intensity, antecedent dry spells) and DP. Variations in agricultural practices are often responsible for variations in wetness conditions within the vadose zone, which ultimately impact DP and GWR potential. To better establish this relationship, four years of deep vadose zone water content measurements obtained using a unique vadose zone monitoring system network across various agricultural land uses in Malta were utilised. Furthermore, the rainfall data over these four years were characterised into storms using minimum inter-event times (MITs) ranging from 12 to 168 hours. DP events in the vadose zone were identified from the VMS by detecting the first >1% absolute increase in volumetric water content at the deepest responding sensor within three days after the onset of each storm event. The optimal MIT selection was based on Cohen’s d effect sizes, which quantify how strongly each storm characteristic distinguishes DP-triggering storms from non-DP storms. MIT of 24 hours generally produces the strongest statistical link between storm characteristics and episodic DP across land uses. DP events typically occur during storms delivering approximately 30 mm of rainfall, lasting 1.5–2 days, and exhibiting peak intensities of 8 mm/h. Thus, it is the combination of event magnitude and rainfall duration that consistently distinguishes storms capable of generating DP. For rainfed agricultural land, which comprises more than half of Malta's agricultural landscape, DP is strongly controlled by storm magnitude and intensity, while also showing dependence on dry spell duration. However, in irrigated land uses, DP becomes less dependent on these storm characteristics due to elevated antecedent moisture from irrigation when compared to rainfed systems. Storm duration remains a moderately important factor in determining DP. Analysis of future rainfall projections for the SSP5-8.5 climatic scenario indicate reductions in storm magnitude and duration, together with longer dry gaps between storms. Ultimately, a decline in episodic DP frequency is expected in the rainfed agricultural land of rocky terrains in semi-arid climates such as Malta.
How to cite: Laudi, L., Dahan, O., Sapiano, M., Schembri, M., and Turkeltaub, T.: Storm intensity and duration impact on deep infiltration in rocky terrains, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13067, https://doi.org/10.5194/egusphere-egu26-13067, 2026.
Chalk, a high-porosity carbonate rock, is often intersected by fractures, allowing an increase in permeability by orders of magnitude and solute bypass of the matrix, which induces rapid water flow and contaminant migration. However, depending on the level of saturation at the fracture–matrix interface, mass exchange may occur. Consequently, the matrix can store a significant fraction of infiltrating water and solutes, thereby controlling hydrological dynamics. Despite its importance, our understanding of the sensitivity and variability of exchange rates to the initial level of saturation remains limited. Therefore, this study implemented a unique experimental setup to quantify the effects of initial saturation variation on the transport using Rhenium (Re) as a conservative tracer. The system encloses a chalk core drilled from the Eocene-age Avdat Group in the northwestern Negev Desert, containing a 1 mm artificial vertical fracture along its longitudinal axis to mimic preferential flow pathways observed in fractured chalk formations. Three initial saturation levels were considered: nearly saturated conditions (≈ 95%), and unsaturated conditions (40% and 60%). Controlled Re tracer injection, followed by artificial rainwater infiltration, was performed, and outlet concentrations were collected under controlled boundary conditions and analyzed using inductively coupled plasma mass spectrometry (ICP–MS).
The Re breakthrough curve (BTC) results, under unsaturated conditions, show a higher peak and lower dispersion compared to those under nearly saturated conditions. These results were further validated by a dual-porosity model (DPM) that was solved using the Hydrus 1D code. The Latin hypercube sampling method was used to generate multiple combinations of hydraulic parameters and longitudinal dispersivity for the DPM. Any simulation that produced an NSE larger than 0.9 was identified as a behavioral simulation. The relationship between solute transfer and initial saturation conditions exhibits pronounced nonlinear behavior. At relatively wet initial conditions (low pressure head —h—), solute transfer remains very limited, indicating weak fracture–matrix exchange. As the system becomes progressively drier, solute transfer increases sharply over a relatively narrow range of pressure heads, reflecting enhanced exchange between mobile and immobile water regions. Beyond this transition zone, a further decrease in initial pressure head results in only minor changes in solute transfer due to intrinsic storage limitations.
How to cite: Jmili, H., Turkeltaub, T., Klein-BenDavid, O., De Falco, N., and Weisbrod, N.: The Effect of Initial Saturation on Solute Transport and Fracture–Matrix Exchange Rates in Chalk, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-7301, https://doi.org/10.5194/egusphere-egu26-7301, 2026.
The formation of waterlogged areas results from the combined effects of external water accumulation and internal water retention. Accurate identification of such areas is a prerequisite for implementing waterlogged farmland remediation. Existing remote sensing-based identification methods suffer from insufficient coupling of systemic factors and limited recognition accuracy. This study proposes a multi-source data coupling approach for precise waterlogged area identification. The method first utilises high-precision Digital Elevation Models (DEM) to extract topographic depressions. Subsequently, it constructs a Soil Waterlogging Potential Index (SWPI) based on soil texture to identify waterlogging-prone areas. Furthermore, it employs the Soil Water Content Index (SWCI) derived from long-term remote sensing data to identify potential waterlogged areas. Finally, spatial overlay techniques are employed to achieve precise identification of waterlogged areas. Experiments conducted on waterlogged areas within China's Northeast black soil region demonstrate the method's feasibility and accuracy through comparative analysis with traditional remote sensing approaches. This research aims to provide technical support for the conservation and utilisation of black soil farmland in Northeast China.
How to cite: Zhou, Y., Liu, R., Gao, Y., Dong, S., and Liu, Y.: Research and Application of a Precise Identification Method for Waterlogged Farmland through Multi-Source Data Integration, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-16305, https://doi.org/10.5194/egusphere-egu26-16305, 2026.
Hydraulic tomography (HT) has been proven as a robust approach to map subsurface heterogeneity through the joint inverse modeling of multiple pumping test data. However, smooth or even erroneous tomograms occur in data sparse areas. In this study, we present a novel procedure of integrating travel time inversion (TTI) results into geostatistical inversion (GI). By treating specific storage (Ss) to be homogeneous, the estimated diffusivity (D) tomogram from TTI is transferred into heterogeneous hydraulic conductivity (K) field. The derived spatial K distribution coupled with effective Ss estimate are utilized as initial guesses for GI. Comparative cases of HT analyses are designed for a numerical case study to highlight the performance of the novel procedure integrated with the TTI result, in which the geostatistical inversion is initialized with: (a) effective homogeneous K and Ss; (2) zonation model results built by different geological information for dividing zones; (3) heterogeneous initial guess of K from the novel procedure. Based on the comparison of lnK and lnSs fields, validation of drawdowns, and examination of travel times indices, results indicate that the novel procedure integrating TTI into HT analysis is demonstrated as an effective approach, which has good performance similar to when a zonation model is integrated with accurate geological information.
How to cite: Qiu, H., Luo, N., Illman, W., Zhuang, C., Huang, Y., and Hu, R.: A novel procedure for geostatistical inversion improved bycombining high-resolution initial guess from travel time inversion, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18291, https://doi.org/10.5194/egusphere-egu26-18291, 2026.
The availability of autonomous sensor networks providing information about soil status offers significant potential for optimizing water management in agricultural systems. Realizing this potential requires robust, in-situ, real-time, and non-invasive measurements of soil water content, salinity, and structure. These sensors are sensitive to many soil characteristics, requiring specific calibration or approximations based on soil types.
Among existing monitoring techniques, electrical impedance spectroscopy provides a direct means of transducing soil physical properties into measurable electrical parameters. Many existing dielectric sensing approaches perform well under specific conditions, particularly at the low and high frequency extremes of the electromagnetic spectrum and in coarse-textured soils. However, a large portion of the intermediate frequency range (10 kHz to 10 MHz) remains comparatively underexploited, despite offering rich information content linked to soil physical and structural properties.
In this study, we combine analytical modeling and experimental dielectric spectroscopy to investigate soil electrical behavior across this intermediate frequency domain. Broadband complex dielectric spectra were measured on soils spanning a range of textures, salinities, water contents and porosities. These measurements are interpreted using effective medium approximations (EMAs), including geometric mixing laws and differential effective medium (DEM) formulations, explicitly accounting for soil geometry, grain shape, and phase connectivity.
The intermediate frequency regime represents a transition zone where ionic conduction and dielectric polarization coexist, giving rise to complex spectral signatures. In this band, Maxwell–Wagner interfacial polarization, strongly controlled by soil structure and connectivity, overlaps with the rotational relaxation of bound, reflecting how water is retained within the soil matrix. Together, these mechanisms encode information on soil texture, porosity, salinity, and structure, but require appropriate theoretical frameworks to be meaningfully interpreted.
Our preliminary results demonstrate that DEM-based formulations provide a consistent and physically meaningful description of measured soil dielectric spectra across the intermediate frequency range. The agreement between modeled and experimental spectra confirms the adequacy of the analytical approach and highlights its predictive value for inferring soil texture, salinity, and water content from broadband impedance measurements. These findings reposition the intermediate frequency band from a source of interpretative complexity to a powerful indicator of soil structure for next-generation agricultural sensing. Future work will focus on extending this framework toward automated in-situ experiments, leveraging laboratory-derived datasets to support robust inversion and next-generation sensor deployment.
How to cite: Nohra, M., Vandôme, P., Belaud, G., and Blayac, S.: Linking Soil Properties and Dielectric Response in the Intermediate Frequency Domain, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20819, https://doi.org/10.5194/egusphere-egu26-20819, 2026.
Soil hydraulics is a dominant constituent of ecosystemal site conditions. How the water is redistributed and stored determines the fertility of soils, the fate of pollutants, and the capacity of carbon storage of these systems. Although studied excessively in labs and monitored in big facilities like lysimeters and field instrumentations, the modelling of the water redistribution lacks a dominant feature: non equilibrated fast flows, where water enters an unsaturated soil and gets conducted fast through a network of larger pores. This has consequences at the large scale: water can either bypass the rooting zone and be lost to plant production, or it can be stored in lower soil horizons and be preserved from soil evaporation and be available for plant transpiration. The systemic soil model BODIUM uses a new approach for modelling soil water, which is capable of reproducing these effects. By implicitely considering redistribution of water locally at the pore scale, the macroscopic behaviour follows the observed non-equilibrium dynamics and better matches field measurements. The work shows the water redistribution and its influence on the vapor exchange at the land surface for the regional water balance, both in modelling and in lysimeter measurement.
How to cite: Weller, U., König, S., and Vogel, H.-J.: Large effects of small scales: modelling non equilibrium in soil hydraulics, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19286, https://doi.org/10.5194/egusphere-egu26-19286, 2026.
Soil hydraulic properties, including saturated hydraulic conductivity and water retention parameters, play a central role in regulating infiltration, redistribution, and storage of water in the root zone, making them fundamental for understanding soil moisture dynamics and plant water availability. Their spatial variability must be characterized to enable accurate field‑scale predictions of soil water dynamics using process‑based models. The Richards equation, which serves as a core framework for modeling water movement in unsaturated soils, poses major difficulties for conventional numerical approaches because of its pronounced nonlinearity, intricate boundary conditions, and high computational demands. Physics‑informed neural networks (PINNs) have emerged as a promising tool that integrates governing physical laws into deep learning frameworks and provides a mesh‑free approach for inverse estimation of hydraulic parameters from limited and noisy datasets. While PINNs have proven effective for homogeneous soils, layered profiles remain challenging due to unknown interface depths and parameter heterogeneity. This study develops a novel PINN‑based framework with progressive physics training to estimate saturated and residual soil moisture contents and the α parameter of the van Genuchten model within layered soils by predicting volumetric water content variations from Time Domain Reflectometry (TDR) sensor data. The framework optimizes data fitting and physics regularization to predict soil moisture dynamics across multiple soil depths. Model performance is evaluated using multiple criteria, including Root Mean Square Error (RMSE), Kling–Gupta Efficiency (KGE), and the coefficient of determination (R²), at sensor‑aligned nodes. Incorporating hydraulic continuity constraints into the loss function enhances parameter identifiability and mitigates equifinality. The proposed approach advances vadose zone modeling by embedding hydrological principles within neural networks, thereby improving computational efficiency while preserving physical consistency. By coupling PINNs with field‑scale TDR observations, this framework bridges the gap between theoretical inverse modeling and practical soil monitoring.
How to cite: Saju, C. and Singh, S.: Physics-Guided Modeling of Water Flow in the Vadose Zone, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-993, https://doi.org/10.5194/egusphere-egu26-993, 2026.
Spring discharge modelling is often constrained by limited data availability. To address this challenge, we propose a hybrid framework that combines TimeGAN-based data augmentation with LSTM and GRU models for spring discharge forecasting and apply it to the Niangziguan Spring in northern China. First, TimeGAN is trained on the limited historical record to learn the underlying statistical properties and temporal dynamics and is then used to generate high-quality synthetic sequences. To evaluate the usefulness of the generated data, we generate synthetic sequences of the same length as the original training set and train LSTM and GRU models separately using (i) the observed data and (ii) the synthetic data and then compare their performance on the test set. Models trained on observed versus synthetic data show comparable test performance, indicating that the synthetic sequences reproduce the temporal dynamics and statistical properties that are critical for the prediction task and are functionally equivalent to the observed data for model training.
Next, TimeGAN is used to expand the training set to between one and six times its original size. t-distributed stochastic neighbour embedding (t-SNE) is used to visualise the distributional consistency between observed and synthetic samples. Qualitative assessment shows that similarity in local structure and distribution patterns increases as the amount of generated data increases: synthetic data quality improves markedly when the synthetic dataset reaches three to four times the size of the original dataset, whereas further increases (four times or more) yield no evident additional improvement. Overall, the synthetic data increase sample diversity while remaining consistent with the original time-series distribution, thereby strengthening model learning when incorporated into the training set.
To quantitatively assess the effect of augmentation, we compare the hybrid models with the baseline LSTM and GRU models using training sets with observed-to-synthetic data ratios ranging from 1:1 to 1:4. Results show that both hybrid models consistently outperform their respective baselines across all evaluation metrics (MAE, MAPE, RMSE, and NSE) during training, validation, and testing, demonstrating the effectiveness of TimeGAN-based data augmentation. Notably, performance does not improve linearly with increasing volumes of synthetic data; an optimal observed-to-synthetic ratio of 1:3 is identified. At this ratio, the test NSE reaches 0.91 for the TimeGAN–LSTM model and 0.94 for the TimeGAN–GRU model. Increasing the ratio to 1:4 results in a slight performance decline (e.g. the test NSE decreases from 0.91 to 0.90 for TimeGAN–LSTM and from 0.94 to 0.93 for TimeGAN–GRU), which is likely attributable to minor distributional deviations introduced by excessive synthetic data. These findings highlight the need to determine an appropriate augmentation ratio in generative data augmentation.
Across all metrics, and particularly at the optimal ratio, TimeGAN–GRU outperforms TimeGAN–LSTM. This advantage is attributed to the GRU’s streamlined architecture, fewer parameters, and stronger adaptability to the “denoised” synthetic sequences generated by TimeGAN, thereby improving prediction accuracy and robustness under data-scarce conditions. Overall, this study demonstrates the effectiveness of TimeGAN in alleviating hydrological data scarcity and provides a practical and quantifiable approach for hydrological time-series prediction in small-sample settings.
How to cite: Hao, Y. and An, L.: A TimeGAN-Augmented LSTM/GRU Framework for Spring Discharge Forecasting Under Limited Data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-3628, https://doi.org/10.5194/egusphere-egu26-3628, 2026.
Soil moisture is a fundamental state variable governing land–atmosphere interactions and hydrological responses to extreme climate events. Although satellite remote sensing has substantially improved the spatial coverage of surface soil moisture observations, most existing products remain confined to the near-surface layer, limiting their applicability to subsurface hydrological processes. The absence of depth-resolved soil moisture information remains a key challenge for representing infiltration, drainage, and root-zone dynamics. This study examines the potential for advancing soil moisture characterization toward three-dimensional (3D) spatial representations by exploiting the complementary information content of multi-source observations. Spatially continuous surface soil moisture fields provide valuable insights into horizontal variability, whereas ground-based measurements offer essential constraints on vertical soil moisture structure. By investigating soil moisture variability across depth and space under varying hydrometeorological conditions, this work highlights the role of subsurface information in improving the interpretation of surface soil moisture patterns. Rather than presenting finalized estimates, this study adopts an exploratory perspective to emphasize the conceptual importance of incorporating subsurface soil moisture into spatial analyses. The findings aim to contribute to ongoing efforts to improve soil moisture representation for hydrological modeling and to inform future applications in flood and drought assessment using 3D soil moisture frameworks.
Keywords: Soil Moisture, Subsurface process, Hydrological extremes
Acknowledgment
This research was supported by the BK21 FOUR (Fostering Outstanding Universities for Research) funded by the Ministry of Education (MOE, Korea) and National Research Foundation of Korea (NRF). This work is financially supported by Korea Ministry of Land, Infrastructure and Transport (MOLIT) as 「Innovative Talent Education Program for Smart City」. This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (RS-2022-NR070339). This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (RS-2024-00416443). This work was supported by Korea Environment Industry & Technology Institute(KEITI) through Water Management Program for Drought Project, funded by Korea Ministry of Climate, Energy and Environment(MCEE)(RS-2023-00230286). This work was supported by Korea Environment Industry & Technology Institute (KEITI) through Research and Development on the Technology for Securing the Water Resources Stability in Response to Future Change Project, funded by Korea Ministry of Climate, Energy and Environment(MCEE)(RS-2024-00332300).
How to cite: Kim, D., Lee, S., Jeong, J., Cho, S., and Choi, M.: Exploring Three-dimensional Soil Moisture Variability using Multi-Source Observations, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10508, https://doi.org/10.5194/egusphere-egu26-10508, 2026.
Time series of soil moisture is an important status variable for understanding hillslope hydrological processes at the mountain hillside because soil moisture plays a critical role in regulating water retention, generating runoff and controlling vegetation dynamics. In order to explore the ultimate interpretation based on past hydrologic information (e.g., precipitation and soil moisture history) for contemporary soil moisture status, several machine learning models had been applied using systematically collected soil moisture measurements along transects in a hillslope. The fitness of models was evaluated in terms of coefficient of determination, mean absolute error and root mean square error. Appropriate lag extent for parsimonious modeling of soil moisture was explored and determined through heuristic approaches which can be explained by historic gain and loss and uncertainty contribution. Modeling results indicate that the vertical infiltration to weather rock as primary hydrological process for most measurement points. Two distinct modeling performances in soil moisture modeling at top hill and streamside points indicate the degree of hydrologic process complexity can be identified through delineated AI modeling results. This study highlights the potential of machine learning based time series modeling for prediction of soil moisture and corresponding hydrologic process configuration in the mountain hillslope.
How to cite: Kim, S. and Kim, D.: Machine learning modeling of soil moisture time series for a hillside at Sulmachun watershed, South Korea, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-6123, https://doi.org/10.5194/egusphere-egu26-6123, 2026.
Understanding the migration behavior of thallium (TI) in subsurface environments is essential for Tl pollution prevention. With the wide production and utilization of biochar, the notable ability of biochar colloids to carry environmental contaminants may make these colloids important for Tl(I) mobility. This study systematically investigated the impact of wood-derived biochar (WB) and corn straw-derived biochar (CB) colloids on Tl(I) transport in water-saturated porous media under different pH (5, 7 and 10) and ionic strengths (ISs) (1, 5 and 50 mM NaNO3). WB colloids improved Tl(I) transport under all IS conditions at pH 7 due to the adsorption capacity of biochar and competition for adsorption sites on the sand surface. However, at IS 50 mM, CB colloids slightly impeded Tl(I) mobility due to the straining. In addition, both WB and CB colloids accelerated Tl(I) mobility under all pH conditions at IS 5 mM. At pH 10, the promotion effect was more obvious due to the deprotonation of O-containing functional groups and higher fluidity of biochar colloids. Furthermore, the two-site nonequilibrium model and two-site kinetic attachment/detachment model suitably described the breakthrough curves (BTCs) of Tl(I) and biochar colloids, respectively. The colloid-facilitated solute transport model could also describe Tl(I) transport influenced by biochar colloids reasonably well. This study provides insight into the migration and fate of Tl(I) in the presence of biochar colloids.
How to cite: Chen, W. and Liu, J.: Impact of nanobiochar on thallium(I) transport in water-saturated porous media: Effects of pH and ionic strength, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-15432, https://doi.org/10.5194/egusphere-egu26-15432, 2026.
Accurately characterizing soil hydraulic properties—specifically water retention and conductivity—is essential for modeling hydrological risks such as flooding, drought, and solute transport. However, direct measurement of these properties in heterogeneous field conditions remains a significant challenge. This study proposes a novel framework for estimating hydraulic parameters using Physics-Informed Neural Networks (PINNs), which constrain deep learning architectures with the fundamental physical laws of subsurface flow. To address the inherent noise and sparsity of field-collected data, we developed a two-stage training strategy: We first introduce a specialized neural network designed to preprocess raw sensor data and capture the complex spatio-temporal dynamics of soil moisture, an the PINN is subsequently refined to map these dynamics back to the underlying hydraulic properties. Furthermore, we enhanced the model’s robustness by integrating empirical soil-water characteristic models into the Activation function ensuring stability across the full moisture spectrum, from desiccation to saturation. Results indicate that this hybrid approach significantly improves parameter estimation accuracy compared to traditional inverse modeling and standard machine learning techniques. This methodology provides a scalable and robust tool for enhancing the predictive reliability of environmental water management models.
How to cite: Li, N.: A Two-Stage Physics-Informed Neural Network Framework for Estimating Soil Hydraulic Properties, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-15711, https://doi.org/10.5194/egusphere-egu26-15711, 2026.
Soil functions in agroecosystems such as nutrient cycling, water filtering and storage, productivity, and carbon storage are highly affected by agricultural management as well as climate change. To understand and predict the complex dynamics, mechanistic modelling is a powerful tool. BODIUM is a site-specific systemic soil model, which was developed for exactly this purpose (König et al., 2023). It integrates the main important biological, physical and chemical processes in soil and at the soil-plant interface, including a dynamic soil structure and explicit microbial activity. It allows for simulating different management practices such as crop rotation, cover crops, tillage, organic and inorganic fertilization.
The web application BODIUM4Farmers builds upon this model and provides a user-friendly interface to support effective soil management (https://bodium4farmers.de/). It was developed in co-design with farmers and agricultural advisors and was already tested by several practitioners. Users can simulate the effect of different management and weather scenarios on soil functions at specific locations within Germany, where soil and weather data are directly provided from our databases.
In this contribution, we will introduce BODIUM4Farmers with selected examples and demonstrate the potential for agricultural practice, but also for teaching and scientific purposes. Although the simulation results in the web application are presented in an aggregated way to easily compare different indicators for soil functions, the underlying process-based model produces daily data along the whole soil profile and thus allows for in-depth analysis of the scenarios.
We will further give insights into ongoing development in regard to extending the management measures including intercropping and differentiated soil tillage operations. Within the EU-project DeepHorizon, we are currently also extending BODIUM4Farmers to include databases for soil and weather for whole Europe, increasing the potential of web application even more.
König, S., Weller, U., Betancur-Corredor, B., Lang, B., Reitz, T., Wiesmeier, M., Wollschläger, U., & Vogel, H.-J. (2023). BODIUM—A systemic approach to model the dynamics of soil functions. European Journal of Soil Science, 74(5), e13411. https://doi.org/10.1111/ejss.13411
How to cite: König, S., Rüschhoff, J., Franke, L., Weller, U., Ansorge, J., Gasser, A., Ohmann, L., Wollschläger, U., and Vogel, H.-J.: Modelling soil functions in agroecosystems: the potential of BODIUM4Farmers for science and practice, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10220, https://doi.org/10.5194/egusphere-egu26-10220, 2026.
The functioning of agricultural ecosystems is increasingly threatened by global change drivers, including climate change and land-use intensification, through the disruption of vital ecosystem processes. Process-based simulation models offer a powerful tool to disentangle the complex interactions between microbiota, plants and soils, providing a foundation for long-term projections and scenario analyses.
Within the framework of the “Global Change Experimental Facility (GCEF)”, extensive datasets on plant physiology, soil nutrients, soil microbial and faunal communities, and soil physical properties have been collected across multiple agricultural land-use types. These include conventional and organic cropping systems, intensively and extensively managed meadows, and extensively grazed sheep pastures, each exposed to both ambient and experimentally simulated future-climate conditions.
Here, we present an extended version of the process-based soil model BODIUM, now capable of simulating grassland dynamics in addition to cropping systems. This extension allows for a comparison of ecosystem processes across contrasting land-use types. The model was parameterized for various GCEF land-use systems, and simulated outputs, including plant shoot and root biomass, and soil carbon, nitrogen and water dynamics, were compared with empirical data for model validation. We analyze how climate change and land management influence soil functions and ecosystem processes, highlighting differences between arable and grassland systems. Furthermore scenario simulations under future climate projections can provide insights into the potential resilience of different land-use systems, offering a basis for informing more sustainable management practices.
How to cite: Kanagarajah, L., Reitz, T., Schädler, M., Taubert, F., Vogel, H.-J., Weller, U., and König, S.: Process-based modelling of soil functions across agricultural land-use types under climate change, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17443, https://doi.org/10.5194/egusphere-egu26-17443, 2026.
Root architecture and soil-plant interactions affect carbon storage and nutrient uptake efficiency of plants. Mechanistic modeling of the soil-plant system enables a better understanding of coupled processes and allows us to predict the response of the soil-plant system to environmental change.
The fully coupled functional-structural soil-plant model "CPlantBox" can simulate plant growth and soil water flow, solute transport and microbial dynamics. In the rhizosphere, the area influenced by the root activity, focus is put on the influence that special root exudates called mucilage have on the plant water uptake and how the microbial activity promoted by the root exudation impacts the nitrogen uptake.
Both the mucilage and the nitrogen cycling in microbes are investigated using a sensitivity analysis: For a given metric of interest (e.g. total microbial biomass after 10 days of root growth) each parameter gets assigned a measure of importance and of interactions with other parameters. This standard approach of sensitivity analysis is extended to parameter inequalities, enabling the inclusion of additional information.
Through the sensitivity analysis we will be able to identify which model parameters determine the effects of root exudates on microbial N mineralization, plant water and N uptake. Measurements from drought and nitrogen limitation experiments will be used to estimate these important model parameters.
How to cite: Kopp, E., Schnepf, A., Javaux, M., Streck, T., Pagel, H., and Giraud, M.: Modeling the impact of root exudates and microbes on water and nitrogen uptake using a fully coupled soil plant model, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-22336, https://doi.org/10.5194/egusphere-egu26-22336, 2026.
Coupled soil–atmosphere heat and moisture transport is strongly influenced by phase change and water vapor dynamics. Evaporation and condensation form a crucial link between the soil water balance and the surface energy balance by coupling hydrologic and thermal processes through latent heat exchange. Accurate representation of these processes is therefore essential for modeling moisture and energy dynamics in variably saturated soil. In this study, an established physics-based model describing liquid water flow, water vapor transport, heat transfer, and the surface energy balance was calibrated using observations from an eddy covariance monitoring station. The model explicitly incorporates the surface energy balance and computes its individual components using a combination of physically based formulations and empirical parameterizations, making it particularly suitable for direct comparison with eddy covariance observations. Soil hydraulic and thermal properties, together with key surface energy balance parameters, including surface resistance, atmospheric emissivity, and surface albedo, were estimated through inverse modeling without direct soil sampling. Model calibration was performed using an evolutionary optimization approach and resulted in good agreement between simulated and observed soil moisture, temperature, and turbulent energy fluxes. The calibrated model provides a physically consistent representation of the eddy covariance observations while maintaining a closed surface energy balance, which is commonly not achieved with observation data alone.
How to cite: Hájková, M. and Kuráž, M.: Inverse modeling of coupled soil heat and moisture transport constrained by eddy covariance observations, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18178, https://doi.org/10.5194/egusphere-egu26-18178, 2026.
Soil moisture content (SMC) plays a vital role in agricultural productivity, water resource management, and ecosystem sustainability in semi-arid regions. Despite this importance, most existing machine learning models mainly rely on remote sensing data to predict the soil moisture variation in the surface soil; however, they are constrained by redundant input features and limited interpretability. To address these shortcomings, this study combines the Random Forest (RF) algorithm, Convolutional Neural Networks (CNN), and the Transformer framework to develop a hybrid RF-CNN-Transformer model. Specifically, the RF algorithm, CNN, and Transformer framework are respectively used for selecting influential features, extracting spatial patterns, and capturing long-term temporal dependencies. Applied to the Mu Us Sandy Land using data from six soil depths (5, 10, 20, 40,70, and 87 cm), the model demonstrated high prediction accuracy and training efficiency across all layers compared to baseline models, with R² values ranging from 0.8586 to 0.984 (mean R² = 0.9507). Interpretability analysis revealed a shift in the controlling mechanisms of soil moisture: shallow-layer SMC is jointly influenced by meteorological conditions and groundwater level, whereas groundwater becomes the dominant factor in deeper layers. Notably, due to the extremely dry climate, precipitation has a relatively minor impact on soil moisture dynamics across all depths. Overall, the proposed RF-CNN-Transformer model enhances both the predictive capability and interpretability of soil moisture variation, supporting precision irrigation and water resource optimization in agriculture, especially in arid and semi-arid regions.
How to cite: qiao, X., wang, N., and wu, Q.: Machine Learning with Feature Selection Reveals Key Drivers of Multi-Depth Soil Moisture Content, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20326, https://doi.org/10.5194/egusphere-egu26-20326, 2026.
Soil microbes are essential to the turnover of the soil organic matter, being involved in the intricate processes of the global carbon and nutrient cycles, hence regulating climate and pedogenesis, which in turn affects plant growth and ecosystem dynamics. Assays of soil functioning such as substrate-induced respiration give access to microbial activity and substrate uptake levels and allows elucidating the biogeochemical pathways of organic matter decomposition and mineralisation. Here we investigate the effects on the soil respiration response of different land-use histories corresponding to long-term grass, arable or fallow by revisiting previous experimental work. We use high temporal-resolution respirometery datasets from the incubation of small soil samples (0.5 g, 4 replicates) collected on experimental plots from the Rothamsted Highfield long term experiment. For each land-use history, the soil respiration rate was measured using a conductimetric respirometer for about 90 hours at 6-minute intervals. Distinct respiration responses are observed depending on whether soils experienced continuous long-term land-use, or transitions from arable to grass or vice versa. Typically grassland soils show an initial exponential-looking decay of the respiration rate followed by a wide respiration pulse with bell-shape characteristics. Fallow soils usually do not exhibit this initial decay phase, while arable soils present oscillations, intermediate between grass and fallow.
Good fits of these data were obtained from developing a parsimonious mathematical model of microbial growth consisting of a set of coupled non-linear differential equations determining the time evolution of the amounts of substrate and microbial biomass in terms of carbon mass concentration, assuming that only an active fraction of the biomass can grow, while its inactive counterpart uses part of the substrate only for its maintenance needs. Our model reveals that the soil respiration rate is governed by three successive phases. For grassland especially, the initial decay originates from the maintenance respiration of the inactive biomass. This is followed by the growth of the active biomass. The last phase results from the microbial biomass degradation once most of the substrate has been consumed. Dimensional analysis of this nonlinear system shows that the dynamics are primarily determined by a single dimensionless parameter, say ρ, which is the ratio of the rate of catabolism to the rate of anabolism. Preliminary results show that the different land uses are clearly distinguished with the hierarchy ρgrass > ρarable > ρfallow > 1. This suggests that grass soils promote a faster turnover of the microbial biomass, than arable and fallow soils.
How to cite: Putelat, T. and Whitmore, A. P.: A new parsimonious approach to modelling soil microbial respiration, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19886, https://doi.org/10.5194/egusphere-egu26-19886, 2026.
Posters virtual: Thu, 7 May, 14:00–18:00 | vPoster spot 1a
EGU26-13521 | ECS | Posters virtual | VPS15
Adding reactive transport capabilities to the 2DSOIL model with the integration of PhreeqcRMThu, 07 May, 14:06–14:09 (CEST) vPoster spot 1a
Process-based crop models are often coupled with soil models to compute the soil water and nutrient status in the root zone. The integration of a geochemical module with existing soil models can enhance their accuracy and capability to simulate additional key bio-geochemical processes. 2DSOIL is a legacy soil model integrated with several prominent process based crop models such as those for maize (MAIZSIM), cotton (GOSSYM), soybean (GLYCIM) and potato (SPUDSIM). However, this soil model lacks a dedicated geochemical component. This study addresses this limitation by integrating the prominent geochemical model, PhreeqcRM, with 2DSOIL using the operator splitting approach, resulting in an improved reactive transport model named ‘2DSOIL-PhreeqcRM’. This new model was validated with two exercises: (i) benchmarking simulated reactive transport against the standard analytical solutions; and (ii) inter-model comparison between cation-exchange simulations from 2DSOIL-PhreeqcRM versus PHREEQC’s built-in transport module. 2DSOIL-PhreeqcRM performed well in both exercises, with a mean absolute percentage error less than 4.75 % and RMSE less than 0.015 mol/l. This research establishes the accuracy and robustness of the 2DSOIL-PhreeqcRM, paving the way for its future use in simulating complex agro-bio-geochemical processes such as the nutrient transformations, precipitation and dissolution of minerals, effect of the addition of lime, ammonia and urea etc.
How to cite: Kapoor, A., Beegum, S., Fleisher, D., Timlin, D., Ray, C., and Reddy, V.: Adding reactive transport capabilities to the 2DSOIL model with the integration of PhreeqcRM , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13521, https://doi.org/10.5194/egusphere-egu26-13521, 2026.
