Infiltration Into a Stony Soil: Modeling of the Process Using

HYDRUS Codes
Hana Hlaváčiková, Viliam Novák, and Marek Rodný

Institute of Hydrology, Slovak Academy of Sciences, Račianska 75, 83102 Bratislava, Slovakia
[email protected]

Abstract

Stony soils are composed of fractions (rocks and fine particles) with different hydrophysical
characteristics. Consequently, effective hydrophysical characteristics are needed for simulations of water
flow in stony soils. In the following contribution, the effective hydraulic conductivity of a stony soil was
estimated using the HYDRUS-2D numerical experiment. Next, infiltration into soils with different
stoniness was performed using the derived effective characteristics and the HYDRUS-1D model. Ponded
and rainfall infiltrations were modeled (with the rainfall intensity lower than the saturated hydraulic
conductivity of a stony soil) with the aim of evaluating the influence of different stoniness on the process
of water flow in stony soils.

1. State of the Art

Soils containing a significant fraction of rocks (stones, skeleton), generally denoted as stony
soils, are located mainly in forest and mountainous areas. According to Poesen and Lavee
(1994), Western Europe contains about 30% of such soils; in the Mediterranean region stony
soils cover around 60% of the territory. According to Šály (1976), the major portion of Slovak
forest soils (up to 80%) contain stones, while the stone content increases with depth.
Furthermore, about 47% of Slovak agricultural soils are referred to as stony soils (Hraško and
Bedrna, 1988). Shape, size, degree of weathering, and geological origin of the soil skeleton can
strongly influence the soil hydrophysical properties, mainly the retention capacity and hydraulic
conductivity. Therefore, it is important to characterize the properties of the stony skeleton and its
distribution in the soil profile. Obtained information could be used to elucidate the influence of
stones on soil water transport processes. It can be expected that the stone content in soils will
reduce the effective cross-sectional area through which water flows. Furthermore, an increase in
stoniness results in higher tortuosity of soil pores, which consequently leads to a lower hydraulic
conductivity of the stony soil (Ravina and Magier, 1984, Childs and Flint, 1990; Novák et al.,
2011; Bouwer and Rice, 1984; Ma et al., 2010). On the other hand, the shrinking-swelling
phenomena could lead to the presence of temporal lacunar pores (voids along soil/stone
interface), which could consequently cause the presence of preferential flow, and thus an
increase in the saturated hydraulic conductivity (Sauer a Logsdon, 2002; Shi et al., 2008; Verbist
et al., 2009; Zhou et al., 2009). Moreover, it could be hypothesized that the degree of preferential
flow in such soils will be proportionally related to the stone content and its spatial distribution.

In this study we present an original method for estimating the effective hydraulic conductivity of
stony soils (Kse) based on the HYDRUS-2D numerical simulations. Moreover, HYDRUS-1D
simulations were performed for soils with different stoniness under different infiltration

1
conditions. Simulated upper and bottom boundary fluxes for soils with different stoniness were
evaluated and discussed.

2. Soil Characteristics

2.1. Effective Saturated Hydraulic Conductivity of a Stony Soil

Effective saturated hydraulic conductivities and retention curves need to be estimated for a
quantitative description of soil water movement in stony soils. The representative elementary
volume (REV) used for the determination of the above-mentioned soil hydraulic properties is
dependent on the size of rock fragments. For example, for a soil containing stones with an
average diameter of 10 cm, a REV of 1 m3 is needed. Since it is extremely technically difficult to
perform physical experiments with this size of soil sample, solutions based on virtualization of
classical laboratory experiments are taking place. In this study, a classical Darcy´s experiment
was simulated using the HYDRUS-2D model (Šimůnek et al., 2008b). Impermeable stones of a
spheroidal shape (approximated as circles in a 2D cross-sectional area) were equally distributed
in the virtual soil sample with a volume of 1 m3. Furthermore, steady-state water flow was
simulated using the unit hydraulic gradient (1 cm of water pressure head was applied as the
upper and bottom boundary conditions). The rate of water flow through the REV therefore
equals the effective saturated hydraulic conductivity Kse (Novák et al., 2011). Results of the
numerical experiments were used to adjust the formula of Ravina and Magier (1984) to the form
presented in Novák et al. (2011):

K se = (1− a.Rv )K fe (1)

where Kse is the effective saturated hydraulic conductivity [cm h-1], Kfe is the saturated hydraulic
conductivity of a fine-textured soil fraction [cm h-1], a is a parameter that incorporates the
hydraulic resistance of the rock fragments to water flow and depends on soil texture and
dimensions of rock fragments [-] (e.g., a = 1.1 was estimated for sandy loam and spherical rock
fragments of a 10-cm diameter), and Rv stands for a relative volume fraction of stones [cm3cm-3]
(stoniness).

2.2. Effective Water Retention Curve of a Stony Soil

The relationship between water contents and pressure heads for stony soils can be estimated
using the method presented by Bouwer and Rice (1984). In this method, the bulk volumetric
water content of a stony soil is expressed from the volumetric water content of a fine soil fraction
alone (eq. 2). Bouwer and Rice (1984) have neglected the retention capacity of the rock
fragments.

θ se = (1 − Rv )θ fe (2)

where θse is the bulk volumetric water content of a stony soil [cm3cm-3], θfe is the volumetric
water content of a fine soil fraction alone [cm3cm-3], and Rv is the relative volume fraction of
stones [cm3cm-3].
2
 3. Infiltration into Soils with Different Stoniness

The impact of different stoniness on the upper and bottom boundary water fluxes during
infiltration through the stony soil was evaluated using HYDRUS-1D (Šimůnek et al., 2008a).
Two qualitatively different upper boundary conditions were used to evaluate the stony soil
behavior during infiltration. A Dirichlet boundary condition with a 2 cm layer of water ponded
on the soil surface was used to simulate ponded infiltration. The second case involved simulating
infiltration from precipitation with an 80 mm/h rainfall intensity and three hours duration (an
extremely high rainfall event). However, the rainfall intensity was lower than the saturated
hydraulic conductivity of a stony soil with 50% stoniness. Under this condition, surface runoff
did not occur. Furthermore, the effective hydraulic properties were determined for the sandy
loam soil from the study site Červenec (1500 m a.s.l.) located in the Western High Tatra Mts.
Infiltration was modeled using the standard (single domain) description of hydraulic properties
(van Genuchten, 1980; Mualem, 1976).

Infiltration into four virtual soil profiles with a depth of 95 cm and different stoniness (Rv = 0,
10, 30, and 50%) were modeled. During the simulation, evaporation from the soil surface and the
water retention of the rock fragments were neglected. For all modeled soil profiles, the initial
conditions were set as vertically uniform (h = -1000 cm). The bottom boundary was defined as a
free drainage boundary. The spatial distribution of rock fragments in the soil profile was
assumed to be uniform in order to be compatible with ideas presented in the HYDRUS-2D
numerical experiment.

4. Results and Discussion

The parameters of effective retention curves and effective saturated hydraulic conductivities
reflecting different stoniness are presented in Table 1. As can be seen in Figure 1, under ponded
conditions the infiltration rates are negatively correlated with the amount of stones present in the
soil (stoniness). As is demonstrated in Figures 2 and 3, cumulative infiltration rates and fluxes
through the bottom boundary under saturated conditions are lower for stony soils in comparison
with soil without stones (0% stoniness). This is due to the fact that rock fragments reduce the
effective cross-sectional area through which water flows. Hydraulic conductivities of the matrix
soil fraction in forest stony soils are usually high. Therefore, it can be expected, even in the case
of heavy rains, that surface runoff will rarely occur in the mild temperate climatic zone. This fact
was already confirmed by a study of Holko et al. (2011) for the same catchment.

Table 1. Effective soil hydraulic parameters determined for the Červenec study site.

Rv 0% Rv 10% Rv 30% Rv 50%

3
Figure 1. Modeled infiltration rates and bottom boundary fluxes for soils with different stoniness
under ponded conditions.

Figure 2. Modeled cumulative infiltration amounts for soils with different stoniness.

Figure 3. Cumulative outflow through the bottom boundary of a soil layer 95 cm thick, during
infiltration shown in Figure 2. Outflow rates for soils with different stoniness (Rv) are shown
using color lines.

4
The situation is completely different if infiltration under unsaturated conditions (the rainfall
intensity is lower than the saturated hydraulic conductivity) is modeled (compare Figs. 4 and 5).
As a result, an increase in the penetration depth of the infiltration front (Figure 5), and also in the
bottom boundary flux (Figs. 3 and 6), is higher for stony soils compared to the soil without
stones. Moreover, the penetration depth and bottom outflow are positively correlated with the
stone content (Figs. 4 and 5). These phenomena could be explained by reduced water retention
capacity of a stony soil. Since stones present in the soil limit the volume of fine soil matrix, less
infiltrating water is needed to increase the water content in the matrix of the stony soil and its
related unsaturated hydraulic conductivity. Consequently, a faster movement of pollutants into
deeper soil layers could be expected for soils with high stone content.

Figure 4. Soil water content profiles during ponded infiltration with the initial soil matrix water
pressure head h= -1000 cm and θi = 0.38. a) a homogeneous soil without stones, (Rv = 0.), b) a
stony soil with stoniness of 50% (Rv = 0.5), a total water content, and c) a stony soil with
stoniness of 50% (Rv = 0.5), a matrix water content.

Figure 5. Soil water content profiles during rainfall infiltration. Rain intensity was 80 mm/h, rain
duration 3 h, the initial soil matrix water pressure head h= -1000 cm and water content θi = 0.38.
a) a homogeneous soil without stones, (Rv = 0.), b) a stony soil with stoniness of 50% (Rv = 0.5), a
total water content, and c) a stony soil with stoniness of 50% (Rv = 0.5), a matrix water content.

5
 Figure 6. Cumulative rain infiltration (full line, rain duration 3 h) into a stony soil and cumulative
outflows (color lines) from the bottom boundary of a soil with different stoniness (Rv).

5. Conclusions

In this paper, the effective saturated hydraulic conductivities reflecting different soil stoniness
were calculated by the HYDRUS-2D numerical model, which was used to simulate the classic
Darcy’s experiment. Furthermore, soil water retention curves for stony soils were estimated
under an assumption of zero retention capacity of the rock fragments. Obtained effective
hydraulic characteristics were then used to simulate the infiltration process using HYDRUS-1D.

Calculated cumulative infiltration rates during ponded infiltration were lower for stony soils
compared to a soil without stones (0% stoniness). When an infiltration from rainfall was
modeled (on the assumption that the rainfall intensity was lower than the saturated hydraulic
conductivity of a stony soil), a faster penetration of the infiltration front and higher cumulative
bottom fluxes were calculated for soils containing stones. Moisture front penetration depths and
related bottom fluxes were positively correlated with the stone content.

Acknowledgement
This contribution was partially supported by grant agency of the Slovak Academy of Sciences
VEGA, project No. 2/0032/13 and is the result of the project implementation ITMS
26240120004 Centre of excellence for integrated flood protection of land supported by the
Research & Development Operational Programme funded by the ERDF.

References
Bouwer, H., and R. C. Rice, Hydraulic Properties of Stony Vadose Zones, Ground Water, 22(6), 696-705,
1984.
Childs, S. W., A. L. Flint, Physical properties of forest soils containing rock fragments. In: Gessel, S. P.,
D. S. Lacate, G. F. Weetman, and R. F. Powers, R.F. (Eds.), Sustained productivity of forest soils,

6
 Proceedings of the 7th North American forest soils conference. University of British Columbia,
Faculty of Forestry Publication, Vancouver, Canada, 95–121, 1990.
Hraško, J., and Z. Bedrna, Applied Soil Science, Priroda Publ. House, Bratislava, pp. 474, 1998 (in
Slovak).
Holko, L., Z. Kostka, and M. Šanda, Assessment of frequency and areal extent of overland flow
generation in a forested mountain catchment, Soil & Water Res., 6(1), 43-53, 2011.
Ma, D. H., M. A. Shao, J. B. Zhang, and Q. J. Wang, Validation of an analytical method for determining
soil hydraulic properties of stony soils using experimental data, Geoderma, 159, 262-269, 2010.
Mualem, Y., A new model for predicting the hydraulic conductivity of unsaturated porous media, Wat.
Resour. Res., 12(3), 513-522, 1976.
Novák, V., K. Kňava, and J. Šimůnek, Determining the influence of stones on hydraulic conductivity of
saturated soils using numerical method, Geoderma, 161, 177-181, 2011.
Ravina, I., and J. Magier, Hydraulic conductivity and water retention of clay soils containing coarse
fragments, Soil Sci. Soc. Am. J., 48, 736–740, 1984.
Poesen, J., and H. Lavee, Rock fragments in top soils: significance and processes, Catena, 23, 1–28,
1994.
Sauer, T. J., and S. D. Logsdon, Hydraulic and physical properties of stony soils in a small watershed,
Soil Sci. Soc. Am. J. 66, 1947-1956, 2002.
Shi, Z., Y. Wang, P. Yu, L. Xu, W. Xiong, and H. Guo, Effect of rock fragments on the percolation and
evaporation of forest soil in Liupan Mountains, China, Acta Ecologica Sinica, 28(12), 6090-6098,
2008.
Šály, R., Soil - basic component of forest production, Príroda, Bratislava, p. 235, 1978 (in Slovak).
Šimůnek, J., M. Šejna, HYDRUS (2D/3D), Software Package for Simulating Two- and Three-
Dimensional Movement of Water, Heat, and Multiple Solutes in Variably-Saturated Media, User
Manual, Version 1.02, PC Progress, Prague, Czech Republic, pp. 203, 2007.
Šimůnek, J., M. Šejna, H. Saito, M. Sakai, and M. Th. van Genuchten, The HYDRUS-1D Software
Package for Simulating the One-Dimensional Movement of Water, Heat, and Multiple Solutes in
Variably-Saturated Media, Version 4.0, Hydrus Series 3, Department of Environmental Sciences,
University of California Riverside, Riverside, CA, USA, pp. 281, 2008a.
Šimůnek, J., M. Th. van Genuchten, and M. Šejna, Development and applications of the HYDRUS and
STANMOD software packages, and related codes, Vadose Zone Journal, 7(2), 587-600, 2008b.
van Genuchten, M. Th., A closed-form equation for predicting the hydraulic conductivity of unsaturated
soils, Soil Sci. Soc. Am. J., 44, 987-996, 1980.
Verbist, K., J. Baetens, W. M. Cornelis, D. Gabriels, C. Torres, and G. Soto, Hydraulic conductivity as
influenced by stoniness in degraded drylands of Chile, Soil Sci. Soc. Am. J., 73(2), doi:
10.2136/sssaj2008.0066, 471-484, 2009.
Zhou, B., M. Shao, and H. Shao, Effects of rock fragments on water movement and solute transport in a
Loess Plateau soil, C.R. Geoscience, 341, doi: 10.1016/j.crte.2009.03.009, 462-472, 2009.

7
8