Observations are invaluable information for testing hypotheses, increasing knowledge and advancing science. Even if this might be universally accepted, it is however quite surprising, by looking at the scientific literature, how diverse is the terminology and the interpretation used for the same monitoring strategy, ranging from the assessment of the actual measurements to the sampling designs. While this might simply mirror the diversity of the underlying assumptions developed in the different scientific communities, it raises the question if this undermines a solid scientific foundation. Among others, in this contribution we focus the discussion on the representativeness of the observations in the soil-plant-atmosphere system. We show how this has been recognized as a relevant concept since long time. We then present some formulations that have been proposed but are still not regularly adopted. By using soil water content observations as an example, we discuss the effect of explicitly considering representativeness, and provide a way forward for adopting it as basic paradigm to advance future science.

How to cite: Baroni, G., Emamalizadeh, S., Gruber, A., Evans, J. G., and Oswald, S. E.: On the representativeness of the observations: a philosophical detail or a basic paradigm?, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-9453, https://doi.org/10.5194/egusphere-egu25-9453, 2025.

The rise of affordable, autonomous in-situ sensors and IoT technology has enabled real-time monitoring of soil moisture, opening new opportunities for sustainable irrigation management. We present SWIM² (Sensor Wielded Inverse Modelling of a Soil Water Irrigation Model), a data-driven digital twin designed for field-scale soil moisture predictions, irrigation advice, and quantification of crop water stress, crop water use efficiency, and irrigation efficiency in vegetable production. SWIM² integrates real-time soil moisture sensor data with a soil water balance model using the DREAM(zs) Bayesian inverse modeling approach to estimate 12 key parameters, including soil and crop growth characteristics, along with their uncertainty distributions. This probabilistic framework with integration of weather forecasts allows for dynamic soil moisture predictions with uncertainty estimates, enabling farmers to make informed decisions about irrigation scheduling. By providing an estimate of the water required to mitigate drought risk, SWIM² supports efficient water use while maintaining crop health.

We validated and implemented SWIM² in commercial fields and irrigation trials across Flanders to evaluate its performance and demonstrate its utility in predicting soil moisture, scheduling irrigation, and quantifying water use efficiency under diverse conditions. In 2022, SWIM² was used in real-time to guide irrigation decisions during a celery trial, and in 2023, it was applied to celery, chicory, leek, and sweet potato trials. During these irrigation trials, the model was calibrated two times a week, after which soil moisture was predicted, and irrigation scheduling was based on the predicted probability of water stress over the following four days. Retrospectively, model calibration based on the 100% irrigation treatment enabled a comparison of various irrigation treatments, providing insights into crop water stress and irrigation surpluses.

Our results show that SWIM² accurately predicts soil moisture and improves irrigation scheduling, while also providing insights into resource optimization, contributing to sustainable agricultural practices. Due to the probabilistic nature of the framework, the irrigation strategy can be tailored to suit a conservative or risk-tolerant approach, depending on the farmer's preferences and water availability. By bridging advanced modeling with practical applications, SWIM² empowers farmers to make data-driven decisions for resilient and efficient crop management.

How to cite: Hendrickx, M., Vanderborght, J., Janssens, P., and Diels, J.: SWIM²: A data-driven digital twin for field-scale soil moisture predictions, irrigation advice, and quantification of water use efficiency in vegetable production in Flanders, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-4273, https://doi.org/10.5194/egusphere-egu25-4273, 2025.

Monitoring and modelling of soil hydrological processes at large scales require measuring the spatial and temporal evolution of soil volumetric water contents, θ_v. Direct measurement of θ_v can be done by sampling and laboratory analyses. Although such methods are straightforward, they require a lot of time and effort for the collection of several samples to account for the spatial and temporal variability. Time-domain reflectometry, TDR, can be a good alternative as it is used to indirectly measure θ_v in the field by measuring the travel time of an electromagnetic pulse in a probe. However, it is an intrusive, point-scale method making it impractical at large scales and at subsurface measurements. Other geophysical methods, such as earth resistivity tomography, ERT, and electromagnetic induction, EMI, sensors represent a practical solution for their time efficiency and the ability to measure at large scales. In particular, compared to the ERT, which still requires the insertion of several electrodes at the soil surface and their connection with a cable network, the EMI has the further advantage (in terms of measurement rapidity) of not requiring insertion in soil to take measurements. Nevertheless, the toll to pay for this larger scale applicability, is that they do not directly measure θ_v and require complex electromagnetic inversion models to obtain either the electrical resistivity, ρ_b, or the bulk electrical conductivity, σ_b, spatial distributions over time respectively from the pseudo-sections coming from ERT and the series of apparent electrical conductivity, ECa, coming from EMI. In this sense, these methods require further efforts to correctly translate the ρ_b and the σ_b distributions in as many θ_v distributions. Accordingly, this study aims at comparing thermogravimetric, EMI and ERT systems to obtain the spatio-temporal evolution of θ_v at a transect scale. For this purpose, a series of measurement campaigns were carried out at a transect 24 m long and 1 m wide at a sprinkler-irrigated field in Arborea area in Sardinia, Italy. A large database of spatially, vertically and temporally distributed measurements was created from auger samples, undisturbed samples, Campbell TDR-200, ERT and CMD mini-explorer EMI sensor. TDR measurements were used to find a relationship between θ_v and σ_b. Inversion models were then used to obtain the σ_b distribution from ERT and EMI measurements, utilizing a previously determined characterization of the soil profile, e.g., layering, depth to groundwater table, texture, etc. The TDR-obtained θ_v - σ_b relationship was then utilized to estimate the θ_v distributions from EMI- and ERT-based σ_b distributions.

How to cite: Coppola, A., Vacca, A., Deidda, G. P., Hassan, S. B. M., Da Pelo, S., Lobina, F., Abdelmaqsoud, M. S. M., Souid, F., Biddau, R., Manis, N., and Comegna, A.: Comparing thermogravimetric, TDR, ERT and EMI measurements of space and time evolution of water content along a transect during an infiltration experiment , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-9988, https://doi.org/10.5194/egusphere-egu25-9988, 2025.

Soil hydraulic properties, including the water retention curve (WRC) and hydraulic conductivity function (HCF), are crucial for accurately simulating hydrological processes in soils. These properties are highly variable and nonlinear, making them challenging to parameterize, particularly at field scales. This study introduces a novel physics-informed neural network (PINN) approach with constraints of Richards equation to estimate these constitutive relationships, conditioned on field soil moisture measurements in a semi-arid study area. The PINN comprises three interconnected networks: soil moisture over space and time, WRC and HCF networks. Given the high non-linearity of the soil hydraulic functions, we adopted an alternating training strategy, with an outer loop to filter the observation dataset and train the networks for the observation variable and an inner loop to train the WRC and HCF networks through the constraints of Richards equation. This two-step alternating training approach (with different loss functions) obtains reasonable observation networks, and since then it strengthens the possibility and the efficiency to learn the constitutive relations.

How to cite: Li, N., Zheng, X., and Yue, X.: Physics-informed Neural Networks for Inferring Hydraulic Properties from Field Soil Water Content Measurements, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-1286, https://doi.org/10.5194/egusphere-egu25-1286, 2025.

This study evaluates the water retention effectiveness of different good farming pactices on a hilly catchment (Felső-Válicka, Hungary) using the SWAT+ hydrological model. The Felső-Válicka case study area (124 km²), the focus of the research, is predominantly used for agricultural production (~35% of the total area). Farmers in the region face increasingly severe droughts, coupled with significant erosion damage to arable land due to intense rainfall events caused by unfavourable precipitation distribution. SWAT+ enables us to analyse management practices that potentially improve water management in agricultural fields. We followed the OPTAIN R workflow (www.optain.eu) to support the process of input data preparation, model setup, model calibration and scenario simulations. In our setup: i) individual fields/parcels can send surface runoff and lateral flow to neighbour objects (contiguous object connectivity approach), and ii) characteristic management practices can be assigned to each field annually, enabling us to analyse the effectiveness of NSWRMs with respect to their individual site-specific allocation within the catchment. The modelled good farming practices align with those defined in the agricultural subsidy framework (CAP) in Hungary.

Among the investigated practices, reducing soil tillage depth and minimizing soil disturbance through reduced tillage practices had the most significant positive impact on water retention. In contrast, linear structured measures (such as riparian buffers along streams and hedges between agricultural parcels) had a less pronounced effect. The periodic planting of perennial crops (e.g., alfalfa on a three-year cycle) or land use change from cropland to pasture yielded mixed results regarding the selected hydrological indicators.

These findings underline the potential of optimized tillage practices as a key strategy for improving water retention in agricultural landscapes.

This work was funded by the Széchenyi Plan Plus program, supported by the RRF-2.3.1-21-2022-00008 project and by the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 862756, project OPTAIN.

How to cite: Kassai, P., Braun, P., Kolcsár, R., Farkas-Iványi, K., Mészáros, J., and Szabó, B.: Assessing the effects of good farming practices on water retention on a hilly catchment in Hungary using SWAT+, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-9348, https://doi.org/10.5194/egusphere-egu25-9348, 2025.

As aridity increases, the importance of focused groundwater recharge becomes more significant. It is well-recognized that groundwater recharge in arid areas primarily occurs during flash floods in ephemeral streams. However, previous studies suggest that diffuse recharge—spatially distributed groundwater replenishment from precipitation or irrigation—may also play a substantial role in aquifers recharge. This study aims to quantify the contribution of diffuse groundwater recharge through unsaturated fractured Chalk under the arid conditions of the Negev Desert.

Three models were evaluated to quantify and understand the processes of diffuse groundwater recharge: the Richards equation, the dual porosity model, and the dual permeability model. The models were calibrated against three unsaturated zone tritium profiles.  Climate data and tritium concentrations in the rain of 1960 to 2023 were prescribed as the atmospheric boundary conditions. The model calibration involved running multiple simulations, incorporating 800 combinations of hydraulic parameters generated by the Latin hypercube sampling method. Model selection and performance were evaluated using statistical metrics, including reduced root mean square error (RRMSE) and the Akaike information criteria. The simulation results indicated that the dual porosity model outperformed the dual permeability and Richards’ equations. This suggests that water flow and solute transport occur within the loess layers and chalk fractures through preferential pathways while highlighting the exchange of water and solutes between the flowing system and the immobile matrix. The calibrated dual porosity models allowed for studying the relationship between diffusive recharge and precipitation. Two cross-correlation analyses (CCA) were conducted: one between yearly rainfall and yearly potential recharge at a depth of 5 meters, and another between all precipitation events above the 95th quantile and the yearly potential recharge at 5 meters depth. Both rain statistical characteristics exhibited similar CCA trends, while the quantile-95 values demonstrated stronger correlation coefficients. This illustrates that heavy rainfall drives deep water infiltration, ultimately replenishing the aquifer.

How to cite: Jmili, H., Weisbrod, N., and Turkeltaub, T.: How important is diffuse recharge in the Chalk aquitard under desert conditions?, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-10381, https://doi.org/10.5194/egusphere-egu25-10381, 2025.

Beneficial Management Practices (BMPs) are designed to reduce nitrate leaching from agricultural fields and protect groundwater quality. However, temporal groundwater monitoring results from wells beneath or downgradient agricultural fields often fail to show evidence for nitrate reduction even years after BMP implementation, and the mechanisms underlying this delayed response remain poorly understood. This study conducted high-resolution characterization and monitoring to investigate nitrate transport from soil to groundwater in a 7-hectare potato rotation field in Prince Edward Island, Canada. The site features fine sandy loam soil underlain by 7–9 m of glacial till, which overlies a regional fractured “red-bed” sandstone aquifer. The water table fluctuates seasonally between 2 and 6 m below ground surface (bgs). Multi-depth groundwater monitoring was conducted over 5 years from 2011 to 2016. Historically, the field was uniformly managed under a grain-forage-potato rotation. For this study, it was divided into four management zones (A–D). Zone D was removed from crop production to eliminate agricultural nitrogen inputs, while Zones A–C continued the crop rotation. This ensured that results from Zone D were not influenced by active cropping. Additionally, the up-gradient areas of Zones C and D were forested, minimizing lateral nitrate input from outside the study area. Multilevel wells were installed along a transect in Zone D to measure nitrate concentrations at various aquifer depths bi-weekly, while water levels were monitored daily using transducers. Rock core collection with detailed core sub-sampling for nitrate distribution was conducted in 2012 to track legacy nitrate in the subsurface. Soil sampling was conducted in each zone during spring and fall. Daily tile drainage samples for nitrate analysis were collected in Zone B using an ISCO sampler. Initial soil and tile drainage sampling detected exceptionally high residual nitrate levels following the 2011 potato harvest. Using this nitrate pulse as a marker, rock coring identified it at ~3 m bgs in December 2012, while piezometer sampling detected it at the water table in spring 2014. Despite seasonal recharge, these results indicate that nitrate required approximately 2.5 years to travel through the 6-m-thick vadose zone to the aquifer. Seasonal recharge processes pushed older nitrate stored in the vadose zone downward via hydraulic pressure, creating a piston-like movement. This caused a rapid water table response but a delayed nitrate concentration response in the aquifer, highlighting that uniform rather than preferential flow dominated nitrate transport through the glacial till vadose zone. By 2016, the nitrate plume in Zone D had disappeared. The short presence of a nitrate plume in the groundwater zone suggested that aquifer matrix diffusion had a minor influence on nitrate transport at this site. Instead, the delayed response of groundwater nitrate levels to surface remediation was attributed to processes occurring in the vadose zone. This study underscores the critical role of vadose zone dynamics in governing the time lag between implementing BMPs and observing groundwater quality improvements. High-resolution monitoring of soil, drainage, and aquifer systems is essential for understanding these processes and accurately predicting the outcomes of agricultural nitrate mitigation efforts.

How to cite: Jiang, Y., Nyiraneza, J., Chapman, S., Malatesta, A., and Parker, B.: Deciphering the Role of Vadose Zone Processes in Delayed Groundwater Nitrate Reductions, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2879, https://doi.org/10.5194/egusphere-egu25-2879, 2025.

Nitrate pollution of groundwater is often attributed to excess fertilisation in agriculture. Low quality water, enriched by nitrate and other pollutants percolates down from the root zone through the unsaturated zone to the water table. Accordingly, the vadose zone holds the footprint of all possible groundwater pollution events occurring at the land surface before the pollution imprint arrives in the groundwater. Here we present a study were detailed long-term monitoring of the unsaturated zone reveals the groundwater pollution potential of various representative agricultural setups over the island of Malta. Malta is a semi-arid island in the Mediterranean Sea where groundwater, which is the only natural freshwater resource, suffers from nitrate pollution due to the intensive agricultural landscape. A national monitoring network, comprising of 16 Vadose zone Monitoring Systems (VMS) were installed under the different agricultural setups which represent Malta’s main agricultural practices. The VMS enables continuous monitoring of variations in the unsaturated zone water content, as indication to percolation processes, and frequent sampling of the sediment pore water for chemical analysis and characterisation of pollutant transport across the unsaturated zone.  

Results show that the mean nitrate concentrations in the vadose zone underlying fields of potato, forage, mixed outdoor vegetables, greenhouses, vineyards, and orchards were 923 mg/L, 673 mg/L, 416 mg/L, 416 mg/L, 252 mg/L and 33.6 mg/L, respectively. Spatial distribution of the different agricultural setups shows that forage and potato fields are among the most common agricultural setups, with higher occurrence in the central and eastern areas of the island (>50% of the agricultural land area). This agricultural land use distribution spatially links to the high average nitrate concentrations in groundwater under those fields (ranging from 75 to 200 mg/L). On the other hand, lower proportions of potato and wheat fields are cultivated in the north-western areas (25-50% of the total agricultural land area). In the north-western areas, a clay layer situated in the unsaturated zone creates a shallow perched aquifer with a rock matric thickness ranging from 20 to 50 m, which impedes water fluxes from the agricultural fields to the main groundwater system below. The mean nitrate concentrations in the shallow perched aquifer are relatively high ranging from 200 to 350 mg/L due to the aquifer’s low water storage. On the other hand, nitrate concentrations in the regional aquifer underlying the perched aquifer are relatively low ranging from 25 to 100 mg/L.

In conclusion, the results show that potato and wheat fields are likely to have the greatest impact on nitrate pollution in the vadose zone and eventual groundwater nitrate contamination. Furthermore, these agricultural land uses are among the most common land uses cultivated in Malta. This implies the significant potential spatial impact of potato and wheat fields on groundwater nitrate pollution. With data being made available from this vadose zone monitoring network we can increasingly understand the pollution potential of different agricultural land uses on groundwater.

How to cite: Laudi, L., Dahan, O., Sapiano, M., Schembri, M., and Turkeltaub, T.: Using vadose zone data to determine agricultural impact on groundwater pollution, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-13344, https://doi.org/10.5194/egusphere-egu25-13344, 2025.

In recent years, the coastal areas of the Mediterranean have experienced extreme climatic events, including the intrusion of the salt wedge into freshwater systems. In Italy, this phenomenon is causing an increasing salinisation of coastal agricultural soils. Among these, the Lazio coast, characterized by dune, fluvial-marsh and marine wind deposits, represents a particularly vulnerable area. This territory, historically reclaimed with landfills and drainage, sees agriculture as one of the main socioeconomic drivers. This research aims to evaluate the extent of salinisation in the coastal area around the Tevere River and its potential impact on agriculture. Monitoring was conducted during the 2024 agricultural season through monthly sampling campaigns and measuring soil and surface water parameters. A particular focus was placed on analyzing the salt wedge intrusion in water bodies and in cultivated soils to understand its progression and impacts on crops. Among the techniques used, time domain reflectometry (TDR) proved essential for collecting spatial and temporal data on salt concentration variability. Using TDR both volumetric water content (VWC%) and Electrical conductivity (EC dS/m) were measured, with EC linked soil properties such as water content, texture and organic matter. Additionally, to soil sampling, leaching tests were carried out in the laboratory to assess the release of salts from the soil into water. Preliminary chemical analyses of major ions in the leachates were used to quantify the contribution of soil salinity to water salinization. Monthly soil samples collected from 10 monitoring points showed in several cases salinity levels exceeding the minimum threshold of 2 dS/m indicated by the FAO.  This can adversely affect the growth and yield of various cultivated species. The results highlight the importance of continuous monitoring of parameters such as electrical conductivity, temperature and soil humidity to track salinization trends. Understanding these dynamics is fundamental for developing adaptation and mitigation strategies to preserve the agricultural productivity of the Lazio coast.

Keywords: Lazio Coast, Soil Salinization, Coastal Agriculture, Seawater Intrusion.

How to cite: Burbulea, R., Marchina, C., Bartoletta, S., Manoni, A., and Tarolli, P.: Analysis of Soil Salinization Induced by Seawater Intrusion on the Coastal Agriculture of Lazio Region (Italy), EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-9637, https://doi.org/10.5194/egusphere-egu25-9637, 2025.

Soil erosion damages many fertile regions world-wide. At the same time, scientists expect more frequent and intense erosive precipitation events and droughts with proceeding climate change.

The Erosion and Runoff Laboratory (EARL) of the Bavarian State Research Center for Agriculture (LfL) is an extraordinary experimental site under construction in Lower Bavaria (Germany) to study the physical, social, and economic factors driving the change in landscape water balance, as well as the increase in surface runoff and erosion under agricultural use. During the twenty-year study period, future-proof cropland systems (i.e., a combination of crop rotation and soil management regimes) will be investigated in terms of their effect on water retention, erosion protection, as well as contaminant and nutrient leaching in the hills.

Volumetric water content and matric potential are measured at three depths in 36 plots, each measuring 330 m² (55 m long and 6 m wide), and the data are validated using a central cosmic ray neutron scattering sensor. Surface and interflow runoff are monitored continuously for each precipitation event. Runoff samples collected automatically by samplers will be analyzed in the laboratory regarding the transported materials and substances. Root growth scans and regular drone flights provide data on phenological growth and soil conditions, while soil water sampling and an extensive meteorological station including distrometers and precipitation-impulse gauges ensure the boundary conditions.

We believe that this design enables comprehensive process-based modeling and a validated balance of energy, water, and material flows on the hillside scale for each individual plot. As the EARL is designed as a collaborative research facility, the created dataset will enable the scientific community to understand critical processes within agricultural soils’ vadose zone better.

Comparatively high financial resources for long-term monitoring projects in the vadose zone are very rarely available. However, the precise measurement of water fluxes in the vadose zone is difficult, the availability of non-destructive measuring instruments is limited, and the results are often compromised by individual unconsidered aspects. Consequently, the risk of an erroneous measurement design and the associated bad investment due to the naturally limited perspective of a small research group is high. Therefore, we want to present our measurement concept for the vadose zone in detail and discuss it with experts before the final installation of the measuring equipment in summer 2025.

How to cite: Mitterer, J. and Ebertseder, F.: Monitoring of water balance in soils dominated by arable farming – development of a measurement concept for the Erosion and Runoff Laboratory (EARL), EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-20353, https://doi.org/10.5194/egusphere-egu25-20353, 2025.