Louis W. Dekker

Modeling and field evidence of finger formation and finger recurrence in a water repellent sandy soil

With prolonged rainfall, infiltrating wetting fronts in water repellent soils may become unstable, leading to the formation of high-velocity flow paths, the so-called fingers. Finger formation is generally regarded as a potential cause for the rapid transport of water and contaminants through the unsaturated zone of soils. For the first time, field evidence of the process of finger formation and finger recurrence is given for a water repellent sandy soil. Theoretical analysis and model simulations indicate that finger formation results from hysteresis in the water retention function, and the character of the formation depends on the shape of the main wetting and main drainage branches of that function. Once fingers are established, hysteresis causes fingers to recur along the same pathways during following rain events. Leaching of hydrophobic substances from these fingered pathways makes the soil within the pathways more wettable than the surrounding soil. Thus, in the long-term, instability-driven fingers might become heterogeneity-driven fingers.

Physics of water repellent soils

Although it is generally well known that water repellent soils have distinct preferential flow patterns, the physics of this phenomenon is not well understood. In this paper, we show that water repellency affects the soil water contact angle and this, in turn, has a distinct effect on the constitutive relationships during imbibing. Using these constitutive relationships, unstable flow theory developed for coarse grained soils can be used to predict the shape and water content distribution for water repellent soils. A practical result of this paper is that with a basic experimental setup, we can characterize the imbibing front behavior by measuring the water entry pressure and the imbibing soil characteristic curve from the same heat treated soil. q 2000 Elsevier Science B.V. All rights reserved.

Extent and significance of water repellency in dunes along the Dutch coast

Abstract Depth, degree and spatial variability of water repellency were examined in the surface layers of dune sands along the coast of the Netherlands. Soil samples were collected at six depths of up to 50xa0cm at 865 dune sand sites in nature reserves. The potential water repellency was measured on dried samples using the water drop penetration time (WDPT) test. The vegetation at the sites consisted of marram grass, buckthorn, grey hair grass, pine, oak, other grasses and heather. The 5190 samples were dried at the laboratory, after which the potential water repellency was measured using the WDPT test. About 60–70% of the samples taken at several depths in the young dunes with a sparse vegetation of marram grass were wettable, whereas the other samples were slightly to strongly water repellent. The samples taken at a depth of 0–5xa0cm in the surface layer at the sites with different vegetations were all strongly to extremely water repellent. At all of these sites, the severity of water repellency decreased with depth. The decrease was most evident at the grey hair grass sites. No significant differences in severity of water repellency were found between the samples taken under a cover of buckthorn, pine and oak, any of the grasses and heather. The large variability over short distances in the water repellency and water content of the soil in the dune sands is shown by the intensive sampling of soil blocks at the Ouddorp, Westduinen, Schoorl and Zwanenwater sites. Drier as well as wetter soil areas were visualized in contour plots of the soil water content distributions in transects from the blocks. Large differences in wetting capacity between samples taken at several depths at the Ouddorp site were assessed by measurements of the wetting rate. In all cases, wetter samples wetted faster than their drier counterparts.

Predicted and observed finger diameters in field soils

Abstract Wetting front instability resulting in fingered flow has been found in both wettable and nonwettable soils. Understanding how and when this phenomenon occurs under field conditions is greatly limited. Laboratory research has resulted in a number of expressions for finger diameter. In this paper we test the applicability of one of these equations for three different soils in the Netherlands where detailed soil sampling of moisture content was done earlier. In addition, information needed for finger prediction, such as the main wetting and drying loops of the soil moisture characteristic curves and the unsaturated and saturated soil conductivities, were measured in the laboratory. Results show that predicted finger diameters for the two sandy soils agreed well with the observed moisture patterns, while for the loess soil the wetting front was flat as predicted. The finger diameters in dry soil were based on the main wetting loop and in the wet soils were dependent on the main drying loop.

Modeling gravity driven unstable flow in a water repellent soil

Abstract One mechanism for the initiation of unstable flow in porous media is a condition of hydrophobicity of the solid phase. Recent continuous and nondestructive measurements of water content distribution in a 200xa0cm wide by 70xa0cm deep trench of a Netherlands’ field soil containing a hydrophobic layer, reveals a complicated wetting pattern with fingered flow being quite prevalent. The soil profile consists of a humic top layer, a second layer consisting of hydrophobic sand, and a hydrophilic sandy layer at the bottom of the soil profile. In this paper we show our attempts to simulate the unstable flow pattern observed in the field using a numerical solution developed for modeling gravity-driven unstable flow. The unstable flow simulation method employs a globally mass conservative finite element solution of the Richards equation applied to the soil trench. The overall patterns of simulated saturation are similar to those of observed saturation. Statistical analysis shows that pointwise predicted saturation is reasonably close to the observed.

Modeling solute transport in a water repellent soil

Abstract A bromide tracer was applied on a 2.2xa0m long and 0.4xa0m wide plot at the location of a hydrophobic soil in the southwest of The Netherlands. At the end of the experiment, the plot was excavated to a depth of 0.7xa0m using 100xa0cm 3 samples, yielding a total of 1680 samples to quantify the three-dimensional spatial distribution of water content, pH, bromide concentration and degree of water repellency. Measured water content and solute distributions indicated that unstable (fingered) flow prevails. It is considered that contaminant transport under such conditions can proceed at rates that are higher than that which would normally occur if the flow were stable. This article illustrates an attempt at modeling contaminant transport under unstable flow conditions using measurements obtained from the experimental plot. A finite element solution of the two-dimensional Richards equation forms the basis for the unstable flow simulation, while a particle tracking random walk solution of the two-dimensional convection–dispersion equation forms the basis of the transport simulation. The water flow simulation and the solute transport simulation were compared with the measured data. Initial results indicate that model predictions compared fairly well with measured water content and solute transport data.

Short communication: A new, flexible and widely applicable software package for the simulation of one-dimensional moisture flow: SoWaM

Most one-dimensional soil moisture flow simulation models have restricted applicability due to (amongst other things): i) insufficient user flexibility; ii) a lack of user friendliness; iii) dependency on scale, temporal and/or spatial, and iv) fixed boundary conditions. Therefore, we developed a simple and highly flexible software package to simulate, visualize and analyze 1-D moisture flow in soils: SoWaM (Soil Water Model). The package has a modular setup and consists of a range of tools to visualize, analyze and compare input data and results. Soil hydraulic properties for each specified soil layer can be defined by either Van Genuchten parameters or cubical splines. Since the model does not impose limits on element size or time interval, it is possible to perform simulations in very high detail, both spatially and temporally. Furthermore, four different criteria for irrigation scheduling have been implemented. The SoWaM package provides an accurate, simple and highly flexible tool to simulate soil moisture flow and to evaluate the effects of various factors on soil water movement, such as timing and amount of irrigation, soil hydraulic properties and soil layering. Results of a case study are presented to illustrate model performance.

Effects of clay amendment on adsorption and desorption of copper in water repellent soils

Copper is an important micronutrient and trace amounts are essential for crop growth. However, high concentrations of copper will produce toxic effects. Australia is increasingly developing production of crops in water repellent soils. Clay amendment, a common amelioration techniques used in Australia, has demonstrated agronomic benefits in increased crop or pasture production. The sorption and desorption of copper and the effect of clay treatment on copper behaviour in a water repellent soil collected from an experimental farm in South Australia is studied. We found that the water repellent soils amended with clay have an increased adsorption capacity of copper. Also the clay-amended soils had an increased ratio of specific sorption to total sorption of copper. The implications of this study to the sustainable agro-environmental management of water repellent soils is discussed.

Animating measured precipitation and soil moisture data

Nowadays more and more measurement sites are installed in the field to gain insight in the process of 2-dimensional moisture flow in topsoils in dependence of the weather conditions. As these measurements yield a large amount of data, visualization is essential and therefore a software package was developed consisting of several tools to process the measured data by creating animated movies of the changes in soil moisture content in time. This paper presents the software, the dataflow between the tools, a description of the tools and some examples of input and output.

Preferential flow in water repellent sandy soils: principles and modeling approaches

Leaching risks of surface -applied agrichemicals in water repellent soils can only be quantified with an acceptable degree of accuracy if knowledge of the underlying principles and an appropriate simulation model are available. The present study aimed to investigate water flow and solute transport processes in a water repellent sandy soil, and to introduce and apply new modeling approaches. Automated TDR measurements revealed that preferential pathways develop rapidly during severe rain storms, causing infiltrating water to be preferentially transported to the deeper subsoil. Furthermore, preferred pathways recurred at the same sites during all rain events. Simulations with a 2-D, numerical finite element flow and transport model indicate that preferential flow paths will only form during infiltration into dry water repellent soils, i.e. in the range below the so-called critical soil water content. Incorporation of hysteresis is essential to generate the formation and recurrence of preferential flow paths with the model. The process of preferential flow and transport has been incorporated in the well-known SWAP model also, and applied to field data of tracer transport through a water repellent sandy soil in the Netherlands. Results indicate early arrival times of bromide in the subsoil in case preferential flow is taken into account.