HS8.3.2

We present improved European hydraulic pedotransfer functions (PTFs) which now use the machine learning algorithm random forest and include prediction uncertainties. The new PTFs (euptfv2) are an update of the previously published euptfv1 (Tóth et al., 2015). With the derived hydraulic PTFs soil hydraulic properties and van Genuchten-Mualem model parameters can be predicted from easily available soil properties. The updated PTFs perform significantly better than euptfv1 and are applicable for 32 predictor variables combinations. The uncertainties reflect uncertainties from the considered input data, predictors and the applied algorithm. The euptfv2 includes transfer functions to compute soil water content at saturation (0 cm matric potential head), field capacity (both -100 and -330 cm matric potential head) and wilting point (-15,000 cm matric potential head), plant available water content computed with field capacity at -100 and -330 cm matric potential head, saturated hydraulic conductivity, and Mualem-van Genuchten parameters of the moisture retention and hydraulic conductivity curves. The influence of predictor variables on predicted soil hydraulic properties is explored and suggestions to best predictor variables given.

The algorithms have been implemented in a web interface (https://ptfinterface.rissac.hu) and an R package (https://doi.org/10.5281/ZENODO.3759442) to facilitate the use of the PTFs, where the PTFs’ selection is automated based on soil properties available for the predictions and required soil hydraulic property.

The new PTFs will be applied to derive soil hydraulic properties for field- and catchment- scale hydrological modelling in European case studies of the OPTAIN project (https://www.optain.eu/). Functional evaluation of the PTFs is performed under the iAqueduct research project.

This research has been supported by the Hungarian National Research, Development and Innovation Office (grant no. KH124765), the János Bolyai Research Scholarship of the Hungarian Academy of Sciences (grant no. BO/00088/18/4), and the German Research Foundation (grant no. SFB 1253/12017). OPTAIN is funded by the European Union’s Horizon 2020 Program for research and innovation under Grant Agreement No. 862756.

How to cite: Szabó, B., Weynants, M., and Weber, T.: euptfv2: updated hydraulic pedotransfer functions for Europe, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-904, https://doi.org/10.5194/egusphere-egu21-904, 2021.

Most models of soil water retention represent the wettest range simplistically, reflecting a high priority on facilitating computation without recognition of the active physical processes. Commonly the wet range is misleadingly represented by a straight line of zero slope, or by default using the same formulation as for the middle range, even though the mechanisms of water retention are different for the wet and middle portions of the range. Though adequate for some purposes, such treatment causes problems for applications that are sensitive to wet-range processes. It prevents accurate prediction of critical but challenging wet-range phenomena such as domain exchange between preferential flow paths and soil matrix. It limits the choices available for quantifying flow problems, for example a blowing-up of derivatives on approach to saturation prohibits the use of diffusivity-based formulations.

A new model addresses these issues for the important case where the medium is soil matrix material exclusive of macropores, thus having a well-defined air-entry value, and the moisture dynamics are the typical wet and dry cycling that achieves maximum wetness at field saturation, with a presence of trapped air at zero matric potential. The range between the air-entry value and field saturation is dominated by trapped air expansion in response to pressure change, as well as a process that increases the sensitivity to changing matric pressure. This enhanced sensitivity may be related in part to a collapse of liquid bridges between air pockets as they expand. For this wet range, the new model incorporates the Boyles’ law inverse-proportionality of trapped air volume and pressure, amplified by an empirical factor to account for the additional processes. To cover the full range of possible moisture, this wet-range formula is supplemented by two others. The middle range of capillary advance/retreat and Haines jumps is represented by a new adaptation of the lognormal distribution function. The adsorption-dominated dry range is represented by a logarithmic relation used in earlier models. Joined together with a continuous first-derivative constraint, the overall formulation recognizes the dominant processes within three segments of the full range. Optimization of five parameters can fit the model to a full data set.

Tests have demonstrated excellent fits, using measured data that have many closely spaced points in the wet and middle ranges. With their basis in process, the model’s parameters have a strong physical interpretation, and potentially can be assigned values without fitting, from knowledge of fundamental relationships or individual measurements. This basis in process also may permit accommodation of hysteresis by a systematic adjustment of the relation between the wet and middle ranges, and with minimal additional data may serve to facilitate estimation of other properties such as hydraulic conductivity, diffusivity, and sorptivity.

How to cite: Nimmo, J. R.: Wet-range physical realism in a model of soil water retention, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3545, https://doi.org/10.5194/egusphere-egu21-3545, 2021.

Accurate determination of the water retention curve (WRC) of a soil is essential for the understanding and modelling of the subsurface hydrological, ecological, and biogeochemical processes. Volcanic ash soils with andic properties (Andosols) are recognized as important providers of ecological and hydrological services in mountainous regions worldwide due to their outstanding water holding capacity. Previous comparative analyses of in situ (field) and standard laboratory (hydrostatic equilibrium based) methods for the determination of the WRC of Andosols showed contrasting results. Based on an extensive analysis of laboratory, experimental, and field measured WRCs of Andosols in combination with data extracted from the published literature we show that standard laboratory methods using small soil sample volumes (≤300 cm³) mimic the WRC of these soils only partially. The results obtained by the latter resemble only a small portion of the wet range of the Andosols’ WRC (from saturation up to -5 kPa, or pF 1.7), but overestimate substantially their water content for higher matric potentials. The disagreement limits our capacity to infer correctly subsurface hydrological behavior, as illustrated through the analysis of long-term soil moisture and matric potential data from an experimental site in the tropical Andes. These findings imply that results reported in past research should be used with caution and that future research should focus on determining laboratory methods that allow obtaining a correct characterization of the WRC of Andosols.

How to cite: Mosquera, G., Marín, F., Feyen, J., Célleri, R., Breur, L., Windhorst, D., and Crespo, P.: How well do standard laboratory methods represent the field water retention curve of volcanic ash soils (Andosols)?, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-13935, https://doi.org/10.5194/egusphere-egu21-13935, 2021.

The mathematical representation of the soil hydraulic properties is of central importance for modeling water, solute and energy transport in the vadose zone. The established models of the soil water retention and hydraulic conductivity curves account for capillary water retention and capillary conductivity, but neglect water adsorption and water flow in films and in pore corners. They are therefore suited for modeling flow and transport processes in the medium to wet moisture range, but are susceptible to failure in dry soil. The model system developed by Peters (2013, 2014) and Iden and Durner (2014) (PDI in the following) is a simple parametric framework that overcomes these structural shortcomings. However, it requires an additional parameter to scale the hydraulic conductivity curve in the medium to dry moisture range where non-capillary flow is dominant. Measured conductivity data are required to estimate this scaling parameter and to compute the hydraulic conductivity over the complete moisture range. In this contribution, we first analyze the original model formulation and show that it is in close agreement with a comprehensive physically-based model for film conductivity in porous media. We then suggest a physically based method to predict the film conductivity from the water retention curve. This reduces the number of free parameters by one and gives a complete prediction of the hydraulic conductivity curve if only water retention data and the saturated conductivity are known. Application to literature data covering a broad range of textures shows a very good agreement between measured data and predictions.

How to cite: Peters, A., Hohenbrink, T. L., Iden, S. C., and Durner, W.: A simple model to predict hydraulic conductivity in medium to dry soil from the water retention curve, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8717, https://doi.org/10.5194/egusphere-egu21-8717, 2021.

This work presents the first results obtained in the context of the PROFILES project. The main objective of PROFILES concerns the identification of a possible correlation between water content dynamics and the distribution of metal pollutants in the surface layer of urban soils.

The research activity combined experimental, analytical and numerical approaches. Field activity was conducted in Bagnolo Mella, (Brescia, Northern Italy) where a ferroalloy industry operated for more than forty years (1974-2015). Four sites within the study area, at different radial distances and up/downwind with respect to the plant, have been considered. At these sites, the distribution of As, Cu, Fe, Mn, Ni, Pb, Zn concentrations in the surface soil was inferred by means of a portable X-Ray Fluorescence device. A tension infiltrometer allowed to estimate the local value of hydraulic conductivity at saturation, K_s. Physico-chemical properties evaluated on soil samples via laboratory analysis were found to be rather homogeneous. However, metal concentrations were remarkably different at the four sites, larger values being detected up-wind and closer to the production plant, within the top layer (≈ 20 cm) of the soil column. In particular, sequential chemical extraction processes and X-ray Absorption Near Edge Spectroscopy showed that Mn exceeded considerably typical background levels and was present in a hybrid form of magnetite (Fe, Mn)₃O₄, resistant to acid dissolution. Considering that it is difficult to form Mn-substituted magnetite in surface layers at low temperatures, its presence indicates this pollutant as a by-product of ferroalloy production transported by water along the soil column. Numerical simulations with the HYDRUS 1D software have been performed to model water dynamics along the uppermost 6 meters of soil at the investigated sites, over a time range of 4 years (2013–2016). A homogeneous domain, with a constant K_s value measured in the field for the top layer, has been compared against a heterogeneous case, in which the distribution of lithological categories has been determined via indicator kriging, based on available stratigraphic data. Surface recharge and evapotranspiration have been estimated from meteorological data (temperature, relative humidity, precipitation, global solar radiation and wind speed) available on an hourly basis. Numerical results allowed to characterize the time evolution of the zero-flux-plane (ZFP) depth, defined as the plane separating zones with upward and downward water flux in a thoroughly wetted soil, when evaporation and drainage are simultaneously occurring. Key findings are: (i) for the whole simulated period, the ZFP oscillates between the ground surface and a maximum depth of about 20 cm, consistent with the vertical range where peak concentrations of heavy metals were found; (ii) simulations in the homogeneous and heterogeneous cases provided analogous results, highlighting the importance of the characterization of the top surface layer. 

Acknowledgments: The project PROFILES was awarded the 2019 edition of the Florisa Melone Award, promoted by the Italian Hydrological Society (SII). The authors thank the SII for the support. Part of the research was carried out within the ISEIA project of the University of Brescia (grant UNBSCLE 9015).

How to cite: Siena, M. and Peli, M.: Water content and metal pollution dynamics in the surface layer of urban soils: first results of the PROFILES project, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12444, https://doi.org/10.5194/egusphere-egu21-12444, 2021.

Groundwater recharge through the vadose zone is an important yet hard to quantify variable. It is estimated from lysimeter experiments or mathematical modelling. For the simulation of groundwater recharge rates with a physically based model soil hydraulic properties (SHPs) have to be inversely estimated because SHPs from laboratory experiments can only be poorly transferred to field conditions. Still, also the inverse estimation of SHPs, is associated with experimental and modeling uncertainties that propagate into the recharge prediction. New methods are thus required improving the inverse estimation of SHPs and reducing the uncertainty in groundwater recharge prediction. Therefore, this study aims to investigate how the assimilation of different types of soil water measurements for the inverse estimation of SHPs with the HYDRUS-1D software affects the estimated uncertainty. For this purpose, observations from a monolithic lysimeter experiment (i.e. lysimeter outflow, soil pressure head and volumetric soil water content at two different depths) have been combined in the different modeling scenarios and coupled with a Bayesian analysis to inversely estimate SHPs and assess their uncertainty. Posterior predictive checks showed that the simultaneous assimilation of outflow and soil pressure head led to the smallest uncertainty in groundwater recharge prediction. This represented a reduction in uncertainty compared to assimilating lysimeter outflow alone. Additional information provided by measurements of soil water content resulted in a reduced parameter uncertainty for residual and saturated water content, however, it did not further reduce the uncertainty in recharge prediction. Overall, this study shows the applicability of a Bayesian analysis for determining uncertainties in the inverse estimation of SHPs with lysimeter data and for the quantification of the associated uncertainty in groundwater recharge prediction. Based on our results for the investigated site, we recommend simultaneous assimilation of lysimeter outflow and soil pressure head measurements to minimize uncertainty in groundwater recharge prediction. However, a more comprehensive analysis is required to make a generally valid recommendation for other soils or climates.

How to cite: Schübl, M., Brunetti, G., and Stumpp, C.: On the identifiability of soil hydraulic parameters in lysimeter experiments: a Bayesian perspective, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-11111, https://doi.org/10.5194/egusphere-egu21-11111, 2021.

There is increasing need for mechanistic and predictive models of transpiration and stomatal response to drought. Global measurements of transpiration showed that the decrease in soil moisture is a primary constrain on transpiration. Additionally, a recent meta-analysis indicated that stomatal closure is explained by the loss in soil hydraulic conductivity, more than that of the xylem. Despite these evidences on the role of soil drying as a key driver of transpiration reduction, the mechanisms by which soil drying impacts transpiration, including the effect of different soil hydraulic properties, are not fully understood.

Here, we propose that stomata regulate transpiration in such a way that the relation between transpiration and the difference in water potential between soil and leaves remains linear during soil drying and increasing vapor pressure deficit (VPD). The onset of hydraulic nonlinearity sets the maximum stomatal conductance at a given soil water potential and VPD. The resulting trajectory of the stomatal conductance for varying soil water potentials and VPD depends on soil and plant hydraulics, with the soil hydraulic conductivity and root length being the most sensitive parameters.

From this hydraulic framework it follows that stomatal closure is not simply a function of soil moisture, soil water potential or leaf water potential. Instead, it depends on transpiration demand and soil-plant hydraulics in a predictable way. The proposed concept allows to predict transpiration reductions during drought with a limited number of parameters: transpiration demand, plant hydraulic conductivity, soil hydraulic conductivity and active root length. In conclusion, this framework highlights the role of the soil hydraulic conductivity as primary constrain on transpiration, and thus on stomatal conductance and photosynthesis.

How to cite: Carminati, A. and Javaux, M.: A modelling framework to predict transpiration reductions during drought based on soil hydraulics, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9407, https://doi.org/10.5194/egusphere-egu21-9407, 2021.

Predicting evaporation from drying soils under limited water supply conditions, where water transfer to the atmosphere is limited primarily by soil hydraulic conductivity, is challenging. The parameterization of soil hydraulic properties (SHP) plays a crucial role in reliable predictions of evaporation. In particular, there are expected differences between traditional functions that consider water flow only in capillaries and functions that additionally consider non-capillary processes, i.e., water storage and film flow on particle surfaces and in corners and channels of pores. The non-capillary processes in simulating evaporation from soil surfaces become more important when the soil dries.

The purpose of this study was to investigate the applicability of different soil hydraulic function types in modelling the actual evaporation under water-limited conditions. Data were obtained from a large bare-soil field lysimeter (2.5 m height; 1 m² surface area), where the lysimeter mass and outflow were measured in hourly time intervals. Precipitation and actual evaporation were calculated from the mass changes of the lysimeter, using a simplified version of the AWAT filter approach of Peters et al. (2017). Meteorological parameters to calculate the potential evaporation were taken from the nearest weather station. Potential evaporation rates were obtained by (i) using the FAO-56 version of the Penman-Monteith equation and (ii) scaling these values to match the bare soil potential evaporation.

The evaporation was simulated using two different models for soil hydraulic properties: i) van Genuchten Mualem (VGM) (only capillary storage and flow), and ii) Peters-Durner-Iden (PDI) (capillary and non-capillary storage and flow). The results show a systematic difference in evaporation prediction by applying the PDI and VGM models, with higher evaporation rates for the PDI model under dry conditions.

How to cite: Joshi, D. C., Peters, A., C. Iden, S., Zimmermann, B., and Durner, W.: Actual evaporation from bare soils - A comparison of numerical modelling and field lysimeter data, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12566, https://doi.org/10.5194/egusphere-egu21-12566, 2021.

The incorporation of a comprehensive crop module in land surface models offers the possibility to study the effect of agricultural land use and land management changes on the terrestrial water, energy and biogeochemical cycles. It may help to improve the simulation of biogeophysical and biogeochemical processes on regional and global scales in the framework of climate and land use change. In this study, the performance of the crop module of the Community Land Model version 5 (CLM5) was evaluated at point scale with site specific field data focussing on the simulation of seasonal and inter-annual variations in crop growth, planting and harvesting cycles, and crop yields as well as water, energy and carbon fluxes. In order to better represent agricultural sites, the model was modified by (1) implementing the winter wheat subroutines after Lu et al. (2017) in CLM5; (2) implementing plant specific parameters for sugar beet, potatoes and winter wheat, thereby adding the two crop functional types (CFT) for sugar beet and potatoes to the list of actively managed crops in CLM5; (3) introducing a cover cropping subroutine that allows multiple crop types on the same column within one year. The latter modification allows the simulation of cropping during winter months before usual cash crop planting begins in spring, which is an agricultural management technique with a long history that is regaining popularity to reduce erosion and improve soil health and carbon storage and is commonly used in the regions evaluated in this study. We compared simulation results with field data and found that both the new crop specific parameterization, as well as the winter wheat subroutines, led to a significant simulation improvement in terms of energy fluxes (RMSE reduction for latent and sensible heat by up to 57 % and 59 %, respectively), leaf area index (LAI), net ecosystem exchange and crop yield (up to 87 % improvement in winter wheat yield prediction) compared with default model results. The cover cropping subroutine yielded a substantial improvement in representation of field conditions after harvest of the main cash crop (winter season) in terms of LAI magnitudes and seasonal cycle of LAI, and latent heat flux (reduction of winter time RMSE for latent heat flux by 42 %). Our modifications significantly improved model simulations and should therefore be applied in future studies with CLM5 to improve regional yield predictions and to better understand large-scale impacts of agricultural management on carbon, water and energy fluxes.

How to cite: Boas, T., Bogena, H., Grünwald, T., Heinesch, B., Ryu, D., Schmidt, M., Vereecken, H., Western, A., and Hendricks-Franssen, H.-J.: Improving the representation of cropland sites in the Community Land Model (CLM) version 5.0, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10915, https://doi.org/10.5194/egusphere-egu21-10915, 2021.

In the context of climate change, strong precipitation events tend to increase in intensity and frequency. Such extreme events are associated with the genesis of fast infiltration into the soil, which can generate a very dynamic contribution of the groundwater, as it was observed during the June 2016 flood event in the Paris basin. An issue is that these events are poorly simulated, limiting the ability to monitor and forecast such event. Even with a physically based model such as the Interaction between Soil Biosphere Atmosphere soil multilayer diffusion scheme (ISBA_DF) the hydrodynamic of the simulated groundwater recharge at the basin scale was poor. Various hypotheses can explain these gaps like 1) a lack of physical processes such as preferential flows or double porosity 2) an inadapted characterization of soil properties and soil hydraulic properties 3) a bad representation of the vegetation properties; 4) or a too coarse spatial and temporal distributions of precipitation.

To answer these questions, four lysimeters of the experimental station of the French Scientific Interest Group for Industrial Wasteland (GISFI, http://gisfi.univ-lorraine.fr/) in Homécourt, France are used. Between 2009 and 2016, the total weight, the drained flow at the base of the lysimeter, and suction/moisture content/temperature were measured of each soil column every hour at 50cm, 100cm and 150cm depth. Ten intense rainfall events were selected by the daily rainfall quantile 99. These lysimeters are cylinders (2m depth*1m diameter) and contain monoliths industrialized soils with a silty-sandy texture. Lysimeters 1 and 2 have no vegetation, contrary to the lysimeter 3 with alfalfa cover, and lysimeter 4 with Noccaea caerulescens cover. In this study, ISBA_DF was used to simulate each lysimeter. The local scale and high frequency evolution of soil moisture, temperature profile, drainage rate and water storage were assessed, based on soil-water retention and conductivity curves from the Brooks and Corey model. For each lysimeter, hydrodynamic parameters were determined based on the observations. The values found for these particular soils are very different from the values expected by the literature.

With the fitted soil parameters, ISBA_DF shows a high performance in reproducing temperature profiles (R² & NSE>0,9) and has good scores for other parameters (R² & NSE>0,6). During intense precipitation, we can reproduce the drained flow at 2m depth; nevertheless, differences on the drainage period and the maximum intensity of drainage were noticed: drainage duration is generally longer for simulations with vegetation and maximum intensity seems too important for the simulations.

Finally, this multi-parameter analysis at a local scale and high frequency demonstrates the good infiltration and vegetation processes of the ISBA_DF model, even during intense rainfall. This study seems to show that there is no need to modify the physics of the model processes, but that efforts should focus on the characterization of soil properties.

How to cite: Sobaga, A., Habets, F., Decharme, B., and Enjelvin, N.: How soil hydrology reacts during strong precipitation events?, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-4128, https://doi.org/10.5194/egusphere-egu21-4128, 2021.

Understanding the hydrology of hydrologically sensitive areas (or runoff source areas) is crucial for evaluating and predicting runoff and the environmental fate of applied chemicals. However, while modeling these areas, one must deal with an overwhelmingly complex, coupled nonlinear system with feedbacks that operate at multiple spatiotemporal scales. Sufficient detailed information on the physical environment that these models represent is often not available. Consequently, the simulation's results, even after extensive calibration, are often disappointing. Fortunately, self-organization of hydrological systems' makes it possible to simplify watershed models and consider the landscape functions instead of small-scale physics. These simplified (or surrogate) models provide the same or better objective results than their complex counterparts, are much less data-intensive, and can be used for engineering applications and planning purposes.

This study aims to experimentally expose the landscape hydrological self-organization of a periodically saturated variable source area with a shallow perched water table and a humid climate. The study site is a four-hectare runoff source area near Cornell University, Ithaca, NY, US. The saturated hydraulic conductivity is greater than the rainfall intensity. The area has a single outlet through a notched weir, and the only inflow is from precipitation. We analyzed observed water table heights and field outflow and found the theory behind the self-organization of runoff processes specific to that landscape type. We determined a priori the thresholds for runoff in a surrogate model using the soil moisture retention curve. 

Weir measurements showed that outflow on the day following rainfall had decreased by orders of magnitude, indicating the soil water had returned to static equilibrium. Under the equilibrated state, established theory indicates that the matric potential decreases linearly with depth above the shallow groundwater. The matric potential (and thus the retention curve) determined the soil water distribution. Another property from the whole field perspective is that excess rainfall above saturation becomes runoff.

The reason for self-organization of the source area was that the soil moisture retention curve (which is similar for the whole source area) determined daily both the soil moisture content and the water table change using rainfall and evaporation as drivers. Since the source area behaved similarly, a simple surrogate water balance could predict the aggregated area's hydrological behavior. The nonlinear and small-scale physics associated with the field's complexity determined the rate that equilibrium is reached, which is always less than one day due to high macropore conductivity, greatly simplifying surrogate models that make daily predictions.

How to cite: Brindt, N., Pacenka, S., Richards, B. K., and Steenhuis, T. S.: Modeling groundwater table and runoff in self-organizing hydrologically sensitive areas, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8063, https://doi.org/10.5194/egusphere-egu21-8063, 2021.