Effect of soil water repellency on soil hydraulic properties estimated under dynamic conditions
Highlights
► Flow experiments were performed with ethanol and water on water repellent soils. ► Effective soil hydraulic properties obtained by inverse modeling. ► Water repellency contributes to an enhanced retention curve hysteresis.
Introduction
Numerical modeling techniques and available computer power have developed tremendously during the last decades and allow scientists in principal to simulate and accurately predict the movement of water in unsaturated soils (Simunek and Bradford, 2008). To do this successfully, one has to be aware of the dominating mechanisms which govern water retention and flow in a given situation, but it is equally mandatory to have a good estimation of the soil hydraulic properties (SHP). SHP are defined on a macroscopic scale (Durner and Flühler, 2005), but are influenced by processes which occur at the pore scale and which are generally not explicitly accounted for during the upscaling procedure. Assuming that these processes do not have any influence on the SHP can lead to enormous predictive errors. One of the micro-scale phenomena which affect the SHP is soil water repellency (WR) or in extreme cases hydrophobicity, a very popular topic during the last years (Wallach, 2010, Liu et al., 2012). WR reduces the affinity of soil to water, such that soils resist wetting for different time scales. The repercussions of soil water repellency include reduced infiltration capacity of soils, accelerated soil erosion, preferential flow and consequently faster leaching of agrochemicals (Doerr et al., 2000).
The origin of water repellency in natural soils is not exactly known. It is commonly accepted that soil water repellency is caused by organic compounds attached on the mineral grains’ surfaces. Various organic molecules are potentially responsible for hydrophobic properties of water repellent soils (Almendros et al., 1988). A key point, which further complicates the estimation of SHP of natural hydrophobic soils is that when these organic molecules come into contact with water, they might change their orientation and consequently the surface of the soil grains becomes more hydrophilic (DeBano, 1981). This means that a soil can have very strong water repellent properties, but just for a finite time. This time can vary from a few seconds to days (DeBano, 1981, Doerr et al., 2000) or even longer (Täumer et al., 2006). However, even short persistence of soil hydrophobicity can enforce for example enhanced run-off in natural catchments.
During the last years, soil water repellency has been studied intensively. The studies focused on the spreading of water repellency in natural soils (Crockford et al., 1991, Woche et al., 2005), finger flow and finger formation (Liu et al., 1994, Nieber, 1996, Bauters et al., 1998, Ritsema and Dekker, 2000, Ritsema et al., 1998), 1D and 2D water flow in field studies (van Dam et al., 1996, Bachmann et al., 2001, Deuer and Bachmann, 2007, Wallach, 2010), infiltration of water into natural hydrophobic soils (Clothier et al., 2000, Feng et al., 2001, Wallach and Graber, 2007), and finally on the effect of water content on water repellency (de Jonge et al., 1999, de Jonge et al., 2007, Doerr and Thomas, 2000, Liu et al., 2012). The latter is very important and it is directly connected with the SHP of water repellent soils. It is well known that drier soils are more repellent as wetted soils. Dekker and Ritsema (1994) assumed that this is probably the reason why surface runoff is larger after the first rainfall compared with later runoff events. They examined the dependence of WR on the water content and found that there was a critical water content above which soils became hydrophilic. These critical water contents were around 2% and 5%, respectively for the two sandy materials they used.
A few studies exist which examine the effect of WR on the SHP. Bauters et al. (1998) conducted infiltration experiments for one hydrophilic and four artificially hydrophobized materials. They also estimated the wetting and drainage water saturation curves under static conditions. They reported that in their experiments the wetting curves were affected by the degree of WR. More specific, for the water saturation vs. matric potential curves, the water entry pressures increased according to the degree of WR reaching a positive value for the extremely hydrophobic materials. For the drainage curves their results showed that only the non-repellent sand was different compared with the artificial hydrophobic mixtures. Ustohal et al. (1998) examined the wettability effect on both water retention and hydraulic conductivity curves in an experimental and theoretical study. They started with initially wet materials and obtained both drying and wetting curves. They presented a model which could predict the SHP of mixtures of soil materials with different wettablility very well. Czachor et al. (2010) determined experimentally drainage and imbibition water saturation curves of various soil types with different degrees of WR. Their results showed that in case of drainage the effect of WR is less pronounced than in case of imbibition. Lamparter et al. (2010) examined the effect of WR on the water saturation curves for initially dry materials. By adding different amounts of artificially hydrophobized (silanized) sand to a hydrophilic sand they produced mixtures with increasing water repellent properties. Afterwards, they performed capillary rise experiments using water and ethanol as the wetting liquid on relatively narrow soil columns (2 cm diameter). Due to the lower surface tension of ethanol compared with water, ethanol is considered to be a complete wetting liquid (Watson and Letey, 1970). The estimated ethanol saturation curves were the same for all three materials independent of the degree of WR. This was expected since the material properties were almost the same and the only difference was the different amounts of hydrophobic grains. On the contrary the water retention curves were significantly different. For the same pressure head value less water was taken up by the dry materials for a higher degree of WR. Taking into account the different physicochemical properties of water and ethanol, Lamparter et al. (2010) scaled the ethanol curves and showed that for a complete hydrophilic material the scaled ethanol curve was very close to the water curve.
Although there are a few studies which examine the effect of WR on SHP, they are limited to classical static water retention measurements. With the exception of the steady state experiments of Ustohal et al. (1998), none of these studies examined the influence of WR on both water retention and hydraulic conductivity curves estimated under transient conditions. The Multistep inflow/outflow (MSI/MSO) experiment is a standard method for estimating both water retention and hydraulic conductivity curves under dynamic imbibition and/or drainage conditions (Hopmans et al., 2002).
The main objective of this study is to examine the effect of WR on the SHP estimated under dynamic conditions with the MSI/MSO method. Furthermore, we try to answer to the following questions.
- (i)
Can we use ethanol as the wetting liquid in multistep inflow/outflow experiments to determine the dynamic intrinsic (independent of the degree of WR) properties of the soil materials under study?
- (ii)
If yes, what is the effect of WR on the SHP estimated using water as the wetting liquid?
In order to answer these questions we conducted MSI/MSO experiments using ethanol and water for two natural materials and two artificially hydrophobized mixtures. By using inverse modeling we estimated both water retention and hydraulic conductivity curves for liquid imbibition into dry material and liquid drainage. From the comparison of the estimated curves we examined the effect of WR on SHP.
Section snippets
Soil samples and contact angle measurements
For the determination of SHP as a function of soil water repellency, we used a natural soil and artificially created two more materials with increasingly hydrophobic properties. Disturbed natural sand soil (material M1) was collected from the subsoil at a field site near Hannover, Germany. The material was homogenized (ground and sieved) and then dried for 24 h at 50 °C. The texture was 99% sand, 0.5% silt and 0.5% clay, and the organic carbon content was 0.3%. By treating hydrophilic sand with
Results and discussion
The results of this study will be presented in the following order. First we present the contact angle values measured for all materials. Afterwards we discuss the results of the MSI/MSO experiments and the inverse simulations for each material and for each fluid. For the water experiments we present figures only for the first replication. Then the estimated SHP for each material and one replication will be presented along with a discussion about the effect of WR on the estimated SHP.
Table 3
Conclusions
We investigated the influence of WR on SHP estimated from dynamic water flow experiments. To achieve this, we conducted MSI experiments on initially dry materials to get the imbibition SHP and then MSO experiments to get the SHP under drainage conditions. All the experiments were conducted for two fluids: water and ethanol. Ethanol is supposed to fully wet the soil samples independently of the degree of WR. SHP were estimated from the experimental data by means of inverse modeling using the
Acknowledgments
We thank Birgit Walter for her careful assistance during the experimental work presented in this article. This study was financially supported by the NTH Project BU 2.2.7 “Hydraulic Processes and Properties of Partially Hydrophobic Soils”.
References (46)
- et al.
Effects of fire on humic and lipid fractions in a dystric xerochrept in Spain
Geoderma
(1988) - et al.
Development and application of a new sessile drop contact angle method to assess soil water repellency
J. Hydrol.
(2000) - et al.
The breakdown of water repellency and solute transport through a hydrophobic soil
J. Hydrol.
(2000) - et al.
Water retention of repellent and subcritical repellent soils: new insights from model and experimental investigations
J. Hydrol.
(2010) - et al.
The role of soil moisture in controlling water repellency: new evidence from forest soils in Portugal
J. Hydrol.
(2000) - et al.
Soil water repellency: its causes, characteristics and hydro-geomorphological significance
Earth Sci. Rev.
(2000) - et al.
Water repellency, near-saturated infiltration and preferential solute transport in a macroporous clay soil
Geoderma
(2008) - et al.
Formation and persistence of fingered flow fields in coarse grained soils under different moisture contents
J. Hydrol.
(1994) Modeling finger development and persistence in initially dry porous media
Geoderma
(1996)- et al.
Preferential flow in water repellent sandy soils: principles and modeling implications
J. Hydrol.
(2000)