Capillary Barrier Effects Research Articles

Pavement subsurface drainage: importance of appropriate subbase materials

Past and current research has clearly established the detrimental effects of inadequate subsurface drainage within a road environment. In response, pavement engineers have developed a number of theoretical models to establish the lateral drainage requirements of pavements. The intention of the present paper is to extend current drainage design by accounting for the presence of a pervious subbase layer. A parametric study of an extended form of the time to drain problem was performed while focusing on the effect of subbase material characteristics under saturated and unsaturated conditions. The numerical results established that the addition of a suitable subbase material can shift drainage time from days to minutes by combining the effects of gravity, seepage distance, and discharge area to enhance vertical water movement. Although increasing saturated hydraulic conductivity of the subbase material was generally shown to favour vertical drainage, it also created a capillary barrier effect that hindered downward flow. The results of this analysis emphasize the importance and relevance of subbase material selection for improving the drainage quality of dense-graded base materials, which display excellent bearing capacity and remain stable under repeated loading.

Laboratory Investigation Into Factors Affecting Performance of Capillary Barrier System in Unsaturated Soil

Covers of the nuclear waste repository are of great significance to the long-term safe storage and disposal of nuclear wastes. Capillary barriers have proven to be effective to resist the downward water seeping into the underlying nuclear wastes, especially in dry climate, and have been widely used worldwide. Infiltrating water is removed from the fine layer by evaporation or transpiration or through percolation into the coarse layer, which plays a critical role in preventing the water from further infiltration in the bulk wastes. In this paper, laboratory infiltration tests were conducted with an organic glass box, filled with fine-grained quartz sand in which a layer of coarse-grained quartz sand was emplaced horizontally or at various slopes (10° and 20°), and the capillary barrier effect under various conditions (different thickness, slop of coarse-grained quartz sand layer, and sprinkling intensity) was investigated in detail. The results show that the thickness of the underlying coarse layer plays a critical role in governing the performance of the capillary barrier. The efficiency of capillary barrier increases with increasing thickness and/or slope of the coarse layer, but decreases with increasing sprinkling intensity. For a sprinkling intensity of 20 mm/day, a 30-mm-thick coarse layer even emplaced horizontally can achieve 100% water diversion. In addition, a visible tracer test was performed with an inert red dyestuff to trace the streamlines; the results indicate that even with a 7-mm-thick coarse layer, the capillary barrier can offer marked ability to prevent water from percolating into the coarse layer. The findings could be useful for improving engineered uses of capillary barriers at waste repository sites.

Two-dimensional numerical modeling of 90Sr transport in an unsaturated Chinese loess under artificial sprinkling

90Sr is a fission byproduct of uranium and plutonium, and it presents a major health problem in the environment. A field test on the transport of various radionuclides including 90Sr in an unsaturated Chinese loess was conducted under artificial rain conditions from July 1997 to August 2000. The vertical concentration distribution of 90Sr displayed an unusual profile of double concentration peaks, which were separated by a thin (0.7 cm) source layer. In order to interpret the double-peak concentration profile, the transport of 3H and 90Sr in the unsaturated Chinese loess under artificial sprinkling conditions was simulated using WATERM, a numerical code for simulating flow field, and NESOR, also a numerical code but for simulating nuclide migration. The models were able to adequately simulate the double-peak concentration profile. The observation suggested that the fine arenaceous quartz layer, though 0.7 cm thick, formed a capillary barrier together with the local loess, which prevented water from penetrating. A significant discrepancy was observed between the model-fitted distribution coefficient ( K d) of 90Sr and that determined from independent laboratory experiments, which can be attributed to a number of factors such as the capillary barrier effect, solution-to-solid ratio and soil water content. Therefore, when the model is used for predictive purposes where K d is used as an input parameter, K d must be determined under well controlled conditions by taking into account these factors as well as the heterogeneity in the field.

Life cycle assessment of mine tailings management in Canada

The effective management of mine tailings involves the control of several environmental impacts and legal requirements. Six tailings site management and closure scenarios were developed for a copper zinc underground mine located in Quebec (Canada) and compared using life cycle assessment (LCA). Two options are considered for the mine operation: tailings can be sent to the tailings disposal area where they are submerged or they can be partly used for backfilling. For each of these two operation options, three alternatives are presented for mine closure: (a) submerged tailings, (b) partial desulphurization with a cover of desulphurized material and (c) a cover with capillary barrier effects (CCBE) made of natural soils followed by revegetation. The goals of the study were to draw the inventory of these management scenarios from the development to the post-closure phase, to assess and compare their environmental impacts and to determine the importance of the land-use impact category. The potential impacts for each scenario were evaluated using the IMPACT 2002+ LCIA method. The results of the performed LCA indicate that for mine development and operation, scenarios where tailings are partly used as backfill for underground stopes appear to lead to larger impacts in 11 of the 14 midpoint categories since the backfill plant operation consumes a greater amount of material and energy. For a site closure period of 2 years, the CCBE option creates the greatest impacts, since it requires much more effort than the other techniques. The results for the post-closure phase have been analysed separately since they have a larger uncertainty. They appear to modify the comparative assessment results. The various results presented in the paper show the importance of taking land-use impacts into account.

Model-based prediction of long-term leaching of contaminants from secondary materials in road constructions and noise protection dams

In this study, contaminant leaching from three different secondary materials (demolition waste, municipal solid waste incineration ash, and blast furnace slag) to groundwater is assessed by numerical modeling. Reactive transport simulations for a noise protection dam and a road dam (a typical German autobahn), in which secondary materials are reused as base layers, were performed to predict the breakthrough of a conservative tracer (i.e., a salt) and sorbing contaminants (e.g., PAHs like naphthalene and phenanthrene or heavy metals) at the groundwater table. The dam constructions have a composite architecture with soil covers in inclined layers and distinct contrasts in the unsaturated hydraulic properties of the used materials. Capillary barrier effects result in strong spatial variabilities of flow and transport velocities. Contaminant breakthrough curves at the groundwater table show significant tailing due to slow sorption kinetics and a wide distribution of travel times. While conservative tracer breakthrough depends primarily on subsoil hydraulic properties, equilibrium distribution coefficients and sorption kinetics represent additional controlling factors for contaminant spreading. Hence, the three secondary materials show pronounced differences in the temporal development of leached contaminant concentrations with consequences for breakthrough times and peak concentrations at the groundwater table. Significant concentration reductions due to dispersion occur only if the source concentrations decrease significantly prior to the arrival of the contaminant at the groundwater table. Biodegradation causes significant reduction of breakthrough concentrations only if flow velocities are low.

Hydrostratigraphic Characterization of Glaciofluvial Deposits Underlying an Infiltration Basin Using Ground Penetrating Radar

An understanding of the heterogeneity of quaternary gravelly deposits is required to predict flow and contaminant transfer through these formations. In such deposits, preferential flow paths can lead to contamination at depths greater than predicted under the assumption of a homogeneous medium. The difficulties in characterizing their complex structure with conventional methods represent an obstacle for this prediction. In this study, we developed an approach relying on the use of ground penetrating radar (GPR) for the detection of sedimentary depositional units. A genetic interpretation of the radar stratigraphy allowed us to construct a distribution model of lithofacies. The study was conducted on glaciofluvial deposits underlying a stormwater infiltration basin. Two main system tracts were characterized: a top stratum (50–80 cm deep) corresponding to massive gravel and open‐framework gravel, and a base stratum corresponding to trough‐fill structures with associated sandy, open‐framework, massive, and matrix‐rich gravelly lithofacies. The knowledge of the hydraulic properties linked to each lithofacies led us to propose a hydrostratigraphic model. Based on this model, we formulated a hypothesis about the hydraulic behavior of the deposit during stormwater infiltration. Open‐framework gravels can act, during complete saturation, as preferential flow paths, and capillary barrier effects may occur under variably saturated conditions. These hypotheses were tested by measuring water content variations (using time domain reflectometry probes) at three depths (0, −0.5, and −1.15 m). Experimental data show infiltration behavior that can be explained by a capillary barrier effect between the two lower probes. These results suggest that our hypothesis about hydraulic behavior is reasonable.

Laboratory Study of Immiscible Contaminant Flow in Unsaturated Layered Sands

Little quantitative experimental data are available describing the behavior of immiscible contaminants in unsaturated heterogeneous porous media. Such data are, however, essential to the fundamental understanding of the processes governing nonaqueous phase liquid behavior and for the validation of modeling tools. The effect of macro‐heterogeneity on light nonaqueous phase liquid (LNAPL) flow and distribution in the unsaturated zone was investigated experimentally by simulating LNAPL spills in layered soil systems consisting of sands with various textures. Two multiphase flow experiments were conducted in a two‐dimensional flume (180 × 120 × 8 cm). The vertical distribution of water and LNAPL pressure were measured using hydrophilic and hydrophobic tensiometers. An image analysis technique was used to estimate the saturation distribution of the fluids in a two‐dimensional vertical plane. The experiments show that LNAPL entrapment, which contributes to long‐term soil and water contamination, depends strongly on the initial water saturation and water pressure at the layer interfaces and on the texture contrasts between the soil layers, which lead to permeability and capillary barrier effects. Thus, the knowledge of the initial water pressure and saturation distribution in unsaturated layered soil formations is critical to the correct prediction of LNAPL infiltration and drainage.

Geochemical transport modelling of drainage from experimental mine tailings cells covered by capillary barriers

The use of covers with capillary barrier effects (CCBEs) for reducing acid mine drainage (AMD) from sulphidic mine tailings is simulated using the MIN3P finite volume model for coupled groundwater flow, O2 diffusion and multi-component reactive transport. The model is applied to simulate five pilot-scale in situ test cells containing reactive tailings from the Manitou mine site, Val d’Or, Que., Canada. Four of the cells were constructed with CCBEs over the tailings, while the fifth tailings cell was left uncovered. Observed and simulated discharge from the base of each cell showed that the capillary barrier covers significantly reduced sulphide oxidation and AMD. Compared to acidic discharge from the uncovered cell, discharge from the four CCBE-covered cells had neutral pH levels and 1–7 orders of magnitude lower concentrations of SO4, Fe, Zn, Cu and Al. The simulations showed that the moisture retaining layer of the CCBEs reduced AMD by inhibiting O2 diffusion into the underlying reactive wastes. Provided the moisture-retention layer of the CCBE remains close to saturation, its thickness had a relatively minor effect. Under such near-saturated conditions, O2 availability is limited by its diffusion rate through the bulk porous medium and not by the diffusion rate through the oxidized grain shells. The model is providing important new insights for comparing design alternatives for reducing or controlling AMD.

Design of Inclined Covers with Capillary Barrier Effect by S.-E. Parent and A. Cabral

The paper was read with interest by the discussers, who have been working on the behavior of cover systems for many years, analyzing their response under a variety of conditions using experimental, numerical, and in situ monitoring tools (e.g. Aubertin et al. 1994, 1995, 1996, 1997, 2006; Bussiere 1999; Bussiere et al. 1995, 1998, 2000, 2002, 2003a, b; Zhan et al. 2001a, b, 2006). The Authors are attempting to develop an approach to design an optimal inclined ‘‘Covers with capillary barrier effect—CCBE’’ (a term coined by the discussers over 10 years ago) to limit water infiltration, based in part on the analytical solution proposed by Ross (1990). Their treatment of the topic has raised many questions, and some of the major concerns are presented below. More specifically, the following points are discussed herein: (i) the limitations of analytical solutions, (ii) the transient nature (and time-dependency) of the diversion capacity of inclined CCBEs, and (iii) the steps needed to define a more appropriate design methodology for inclined CCBEs.

Function of permeable geosynthetics in unsaturated embankments subjected to rainfall infiltration

This paper presents laboratory tests and experiments using scale models to study the hydraulic behavior of permeable geosynthetics within unsaturated embankments subjected to rainfall infiltration. Water retention curves were measured in the laboratory in order to evaluate the unsaturated hydraulic characteristics of soil and geosynthetics. Model embankments were built using two layers of permeable geosynthetics; rainfall was simulated using an irrigation pipe. Embankments were subjected to wetting and drying processes; negative and positive pore water pressure and water contents within the model were measured. Comparison between Tempe pressure cell and hanging column test results showed that geosynthetics embedded within the soil approached saturation only when the pore water pressures of the surrounding soil were close to zero or positive. This behavior was also observed in the model tests. Local failure during the wetting processes was observed while pore water pressure increased immediately above the geosynthetic layers. In the models, where strips of geotextile were used, water could not accumulate above the geotextile, and instead drained down between strips. Strips of geotextile prevented the capillary barrier effect and allowed the free drainage of water through the embankment.

Field experimental cells to evaluate the hydrogeological behaviour of oxygen barriers made of silty materials

An in situ study was initiated to evaluate the efficiency of silts as moisture-retaining material in covers with capillary barrier effects (CCBEs) that aim at controlling the production of acid mine drainage (AMD). This investigation included the construction of four experimental cells covered by different layered cover configurations. Each CCBE was instrumented to monitor its hydrogeological behaviour. The degree of saturation (Sr) in the moisture-retaining (silty) materials was usually maintained above 85%, while Sr in the two sand layers placed above and below the silts was generally lower than 30%–50% (typical hydrogeological behaviour of efficient CCBEs). Suction measurements in the different layers of the four CCBEs were consistent with the observed Sr measurements during the four years of monitoring. The suctions measured in the moisture-retaining materials were lower than their respective air-entry values (AEV), while suctions largely exceeded the AEV of the sand in the capillary break layers. The oxygen flux calculations confirm that in each of the four CCBEs the flux through the CCBEs remained low for the entire duration of the experiment, hence limiting the generation of AMD. This study also confirmed the ability of silty materials to act as moisture retaining in a CCBE. Key words: cover with capillary barrier effects, acid mine drainage, experimental cells, in situ monitoring, unsaturated flow, diffusion, oxygen flux.

Discussion of "Variations in moisture content for a soil cover over a 10 year period"

The authors have presented a very interesting paper on the in situ behaviour of a soil cover constructed on a large waste rock pile. As this type of cover is relatively new (at least in the mining industry), few investigations with such extensive field data have been reported in the literature. This type of information is quite useful to those who must design cover systems, as it brings an improved understanding of how such a complex system responds to actual field conditions. The discussers (and their collaborators) have also been involved in somewhat similar projects (e.g., Ricard et al. 1997, 1999; Bussiere et al. 2003c; Dagenais et al. 2005), and the results obtained from the corresponding monitoring and modelling programs have revealed similar trends. Many of the main findings presented by the authors are in close agreement with those gathered by the discussers. The results confirm that under actual field conditions, cover system behaviour is intrinsically dynamic, and rarely (if ever) reaches a steady state under a relatively humid climate common in many parts of Canada. This means that the monitoring program and modelling work should focus on the transient conditions that dominate the cover response (Aubertin and Bussiere 2001). Aubertin et al. 106 The authors’ results also show that a cover with capillary barrier effects (CCBE) can induce the favourable conditions required to control the production of acid rock (or mine) drainage (ARD or AMD). The authors should also be complimented for reporting on a diversity of performance indicators, and for pointing out some of the challenges and difficulties encountered when monitoring the in situ behaviour of a CCBE on a large scale, and in the interpretation of the data. In the following paragraphs, the discussers would like to draw attention to some complementary issues that may be of interest but were not addressed in detail in the authors’ paper (however, the authors have probably considered these issues elsewhere). As mentioned by the authors in their paper, and also pointed out in a recently published companion article (Fredlund and Wilson 2006), soil covers seldom behave as a one-dimensional system. Except for the relatively flat surfaces encountered on some tailing stacks, the actual flow of water in a layered cover usually involves both vertical and lateral movements within and between the different soil layers. In this regard, previous work by Aubertin et al. (1997) and Bussiere et al. (2000, 2003b) has shown, using numerical simulations and in situ observations, that the difference in elevation between the bottom and top portions of an inclined cover can induce a significant suction gradient. This in turn can create a variation in the moisture distribution in the cover along the slope. In many instances, there is a tendency for water to accumulate at lower elevations, with increasing desaturation towards the top of the cover (Aubertin et al. 1997; Bussiere et al. 2000). This is an important aspect of the behaviour of CCBEs that may not have received enough attention in the past. For inclined cover systems, the local degree of saturation along the slope depends on many factors, including the water retention curve of the materials (particularly the air entry value (AEV) of the moisture retaining layer), the thickness of the layer(s), the water budget components and sequence, the slope angle, length, and other geometric features such as benches (Bussiere et al. 2003b). It would thus be interesting if the authors could use available field data (or new data gathered in the future) to investigate how the slopes may influence the cover behaviour at the Equity Silver mine site in Houston, British Columbia. This aspect is important because, as the authors pointed out, a CCBE that aims to control the production of ARD by reducing the oxygen flux must rely on a highly saturated layer to act as an oxygen barrier. Limiting oxygen penetration into reactive mine wastes has in fact been shown to be an effective control method that is well suited to relatively humid climates, particularly on flat areas (Bussiere et al. 2003b, 2003c; Dagenais et al. 2005). The cover geometry may, however, create additional difficulties in maintaining an effective oxygen barrier along the entire sloping area, because the suction that develops at higher elevation can in some cases exceed the AEV of the material used for the moisture-retaining layer. The conditions at the site described by the authors do not seem to have created such situations (based on the properties given in the paper), but it is not clear from the presented data if (and how) the

Design of Inclined Covers with Capillary Barrier Effect

A design procedure is proposed to minimize water infiltration into landfills by optimizing the water diversion length of inclined covers with capillary barrier effect (CCBE). This design procedure is based on a conceptual, mathematical and numerical approach and aims at selecting materials and optimizing layer thickness. Selection among candidate materials is made based on their hydraulic conductivity functions and on a threshold infiltration rate imposed on the designer. The capillary break layer (CBL; bottom layer) is characterized by a weak capillarity, while the moisture retention layer (MRL; upper layer) is characterized by a compromise between strong capillarity and high hydraulic conductivity. The thickness of the CBL corresponds to the height where suction reaches its maximum value for a given infiltration rate. This height can be calculated using the Kisch [Géotechnique 9 (1959)] model. The optimal thickness of the MRL is determined by applying an adaptation of the Ross [Water Resources Research 26 (1990)] model. The results obtained using the proposed design procedure were compared to those obtained from numerical simulations performed using a finite element unsaturated seepage software. The procedure was applied for two cover systems; one where deinking by-products (DBP) were used as MRL and sand as CBL and another where sand was used as MRL and gravel as CBL. Using this procedure, it has been shown that an infiltration control system composed of thin layers of sand over gravel is highly efficient in terms of diversion length and that its efficiency can be enhanced by placing a hydraulic barrier – such as a layer of DBP – above the MRL.

Numerical Modelling of Flow and Capillary Barrier Effects in Unsaturated Waste Rock Piles

Flow systems in unsaturated waste rock piles were simulated using a two-dimensional numerical model (HYDRUS). The conceptual models are based on homogeneous (unstructured) waste piles, and on structured piles that include either horizontal or inclined fine-grained layers within a coarser host material, forming a capillary barrier system. The approach considers fully transient conditions and uses observed climatic data from a mine site in northern Quebec, Canada. All physical properties of the porous media, including the water retention curves, were obtained from measured data. Different geometric configurations were tested to determine their effect on moisture distribution and water flow, which ultimately control the potential for acid rock drainage (ARD). The simulations begin with a relatively dry initial condition under hydrostatic equilibrium. After an initial transient period, the simulated internal moisture distribution became periodic with a regular pattern of seasonal fluctuations. The simulations suggest that flow can be controlled in such systems using inclined fine-grained layers that retain and divert moisture due to capillary barrier effects. With horizontal layers, the local flow regimes become unstable, causing vertical preferential flow zones to develop below the barriers wherever the local water pressure first exceeds the entry pressure of the underlying coarser material. In this scenario, ARD production can remain high since a large fraction of the internal pile is being flushed. A shallow downward slope in the layers forces drainage toward the outer boundary and maintains lower saturation in the centre of the pile, thus potentially reducing the amount and mobility of ARD.

Weathering profile of non-welded ignimbrite and the water infiltration behavior within it in relation to the generation of shallow landslides

The mechanism underlying rain-induced shallow landslides of non-welded ignimbrite is found to be a special type of weathering profile and the behavior of water infiltrating through that profile, according to our study of Ito ignimbrite in southern Kyushu, Japan. Rhyolitic volcanic glass, the primary component of ignimbrite, is first hydrated and dissolved, forming halloysite. Halloysite near ground surface is then transported through the ignimbrite by infiltrating water and becomes clogged in interstices to form clay bands. Suction monitoring across a weathering profile indicated that downward infiltration of water is disrupted once by a zone of less-permeable clay bands and again at the weathering front. This disruption at the front is caused by a capillary barrier effect caused by the structure where finer, weathered material overlies coarser, fresh material. This results in a well-defined weathering front, particularly beneath a slope where water flux is parallel to the front, whereas the front is transitional beneath a ridge top where the front is nearly horizontal and the water flux is normal to the front. Infiltrating water from rain increases the weight of weathered material and decreases the suction within the material, which is the final trigger of a shallow landslide of non-welded ignimbrite; long-term weathering, which proceeds on the order of years, provides slide material.

Evaluation of Evapotranspirative Covers for Waste Containment in Arid and Semiarid Regions in the Southwestern USA

Performance evaluation of evapotranspirative (ET) covers is critical for waste containment. The purpose of this study was to evaluate ET covers at sites in Texas and New Mexico representative of arid and semiarid regions in the southwestern USA using water balance monitoring during 4- and 5-yr periods and water balance simulations using short-term (1–5 yr) and long-term (25 yr) climate forcing. Estimated drainage at the Texas site was related to irrigation while measured drainage at the New Mexico site was restricted to the first 2 yr of the 5-yr monitoring period. Evapotranspirative covers work extremely well in these regions because of the dominance of summer precipitation (62–80%) that corresponds to periods of highest ET. Strong relationships between decreases in soil water storage and vegetation productivity at both sites underscore the importance of vegetation in controlling the water balance in these systems. Simulations of the Texas site indicate that drainage can occur in response to high precipitation near the end of the growing season, but such drainage can be eliminated with a capillary barrier. Inclusion of a capillary barrier increased available water storage by a factor of about 2.5 at both sites. The capillary barrier effect of drainage lysimeters can result in underestimation of drainage and overestimation of water storage relative to covers not underlain by capillary barriers. The data from this study indicate that a 1-m-thick ET cover underlain by a capillary barrier should be adequate to minimize drainage to ≤1 mm yr −1 in these arid and semiarid regions. Comprehensive monitoring integrated with modeling is required to assess total system performance to develop a predictive understanding of ET covers.

Part 3: Pavement: Subsurface Drainage and Reliability in Design and Performance: Modeling Water Flow Patterns in Flexible Pavements

Most road design models do not explicitly account for moisture transport mechanisms in roads, even though it is well known that water content plays an important part in the deterioration of roads. The Swedish National Road Administration aims to improve the current situation by supporting the development of models that eventually can provide a better and more complete description of the road and environment system. The applicability of hydrological theories and methods to the road and environment system was investigated. Particular attention was paid to flow patterns inside the road as affected by capillary barriers and generated by mechanisms of surface runoff, followed by infiltration into cracks, and the embankment. Particle tracking was used to investigate the effect of rain intensity, precipitated amount, or fracture conductivity on the flow patterns. Changes in rain intensity had a small effect on flow patterns, but velocities were higher for larger rain intensities. Both changes in precipitated amount and fracture conductivity controlled the appearance of the flow patterns, but capillary barrier effects were limited. The numerical code used proved to be mostly appropriate in describing the relevant processes.

Modeling Water Flow Patterns in Flexible Pavements

Most road design models do not explicitly account for moisture transport mechanisms in roads, even though it is well known that water content plays an important part in the deterioration of roads. The Swedish National Road Administration aims to improve the current situation by supporting the development of models that eventually can provide a better and more complete description of the road and environment system. The applicability of hydrological theories and methods to the road and environment system was investigated. Particular attention was paid to flow patterns inside the road as affected by capillary barriers and generated by mechanisms of surface runoff, followed by infiltration into cracks, and the embankment. Particle tracking was used to investigate the effect of rain intensity, precipitated amount, or fracture conductivity on the flow patterns. Changes in rain intensity had a small effect on flow patterns, but velocities were higher for larger rain intensities. Both changes in precipitated amount and fracture conductivity controlled the appearance of the flow patterns, but capillary barrier effects were limited. The numerical code used proved to be mostly appropriate in describing the relevant processes.

A study of infiltration on three sand capillary barriers

The capillary barrier effect was investigated by conducting infiltration tests on three soil columns of fine sand over medium sand, medium sand over gravelly sand, and fine sand over gravelly sand. The barrier effect was verified in the underlying layer of coarser material, and the water-entry values of the coarser layers were confirmed to be nearly equal to the residual matric suctions of the soils. The coarser layer of gravelly sand, which had a lower water-entry value, was more effective in forming a barrier than the coarser layer of medium sand, which had a higher water-entry value. When the capillary barrier was comprised of a coarser layer of gravelly sand, there was more water stored in the finer layer at the end of the drying stage than when the capillary barrier was comprised of a coarser layer of medium sand. Non-equilibrium static conditions of pore-water pressure profiles were observed in the three soil columns, and a generalized ultimate pore-water pressure profile of a capillary barrier system was proposed. In addition, the final volumetric water contents versus matric suctions of the soils as measured from the soil columns were reasonably consistent with the soil-water characteristic curves (SWCCs) of the soils, suggesting that the drying SWCC of a soil could also be obtained from the drying process in a soil column (or a capillary open tube). The drying SWCC could be established from measurements in the soil column up to a height corresponding to two times the residual matric suction head of the soil.Key words: capillary barrier, soil column, soil-water characteristic curve, pore-water pressure, water content, matric suction.

Effect of snow structure on water flow and solute transport

AbstractWater flow through a melting snow pack modifies its structure and stability and affects the release of water and nutrients into soils and surface waters. Field and laboratory observations indicate a large spatial variability on various scales of the liquid water content and flow, a dominant system feature currently not included in numerical models. We investigated experimentally water and dye tracer movement through microstructurally different snow pack horizons and the persistence of preferential flow paths. Naturally rounded snow of varying grain size was artificially packed to obtain well known conditions by sieving it into rectangular bins. Surface melt was induced with infrared lamps. The flow paths were visualized with tracers and liquid water content was monitored with time domain reflectometry probes. Vertical cuts through the snow pack were imaged. The dye tracer patterns allowed the two flow regimes ‘matrix flow’ and ‘preferential flow’ to be distinguished. Matrix flow is apparently dominated by film and capillary flow in the unsaturated snow matrix. The capillary barrier effect at a boundary between a fine over a coarse textured layer on matrix flow in snow was confirmed. In contrast, preferential flow appears as well‐defined flow fingers that advance from 0·1 to 1 cm s−1. During a melt phase, the advancing flow fingers enlarge and are only partially time invariant. It remains to be shown whether the continuum concept, including the Darcy–Buckingham law is apt to describe the extremely non‐linear nature of water flow and the travel time of solutes in snow under conditions of melt water percolation. Probably, snow packs that include faceted crystals and large variations in bulk density, feature more pronounced capillary barriers and preferential flow triggering, but also stronger impeding of fingers by lateral dispersion. Further, triggering and persistence of preferential flow is complicated by the usually transient infiltration rate. Copyright © 2004 John Wiley & Sons, Ltd.