SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bachmann, J.
Right arrow Articles by van der Ploeg, R. R.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Bachmann, J.
Right arrow Articles by van der Ploeg, R. R.
Agricola
Right arrow Articles by Bachmann, J.
Right arrow Articles by van der Ploeg, R. R.
Soil Science Society of America Journal 65:1599-1607 (2001)
© 2001 Soil Science Society of America

DIVISION S-1 - SOIL PHYSICS

Isothermal and Nonisothermal Evaporation from Four Sandy Soils of Different Water Repellency

J. Bachmanna, R. Horton*,b and R. R. van der Ploega

a Institute of Soil Science, Univ. of Hannover, Herrenhaeuser Str. 2, 30419 Hannover, Germany
b Dep. of Agronomy, Iowa State Univ., Ames, Iowa 50011-1020

* Corresponding author (rhorton{at}iastate.edu)


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 THEORY
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soil water evaporation is an important component of the surface energy balance. Evaporation also affects plant available water content of soil. Soil wettability is known to affect water-holding capacity and water flow, but the impact of soil water repellency on evaporation has not been thoroughly investigated. The objective of the study was to measure and analyze isothermal and nonisothermal evaporation rates for four sandy soils showing different degrees of water repellency. Eight vertical soil columns (two of each soil) were exposed to isothermal conditions (20 ± 1°C), and another eight soil columns were exposed to nonisothermal conditions. During the nonisothermal experiment, the top boundary temperature was held constant at 21°C, and the bottom boundary temperature was constant at 55°C, thus maintaining a constant thermal gradient. After 195 d of evaporation, each column was sectioned to determine the soil water content distribution along its axis. A numerical model based on the Philip-de Vries theory was used to predict soil water flow. Deviations between predicted and measured values increased with increasing contact angle. Isothermal cumulative evaporation was 25% lower for water repellent soil than for the most wettable soil. Nonisothermal cumulative evaporation was 75% larger than isothermal evaporation for the wettable soil but only 14% larger than isothermal evaporation of the most water repellent soil. Evaporation and residual water contents, especially in the surface layer, were found to be strongly affected by soil wettability.


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 THEORY
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
WATER REPELLENCY has been observed in sand, loam, clay, peat, and volcanic ash soils all over the world (Wallis and Horne, 1992; Ellies and Hartge, 1994; Ritsema et al., 1997; Jamarillo et al., 2000). More than 75% of the cropland and grassland topsoils in The Netherlands are considered to be slightly to severely water repellent (Dekker and Ritsema, 1994). Recent research suggests that under certain conditions all soils may display water repellency to some degree. Generally, a soil might become water repellent after drying to some critical soil water content (or pressure head) level. Below this level, the soil behaves as a water repellent soil with respect to water flow; however, if moist enough, the soil behaves again as a wettable soil. The change between these two states is generally thought to be caused by certain molecular changes of all kinds of organic substances in soil. Several studies have shown that water repellency is associated with organic coatings of mineral particles (Wallis and Horne, 1992). Repellency is caused by a range of hydrophobic organic substances such as fungal hyphae, humic acids, or partly decomposed plant material. Dry, water-repellent topsoils resist or retard water infiltration into the soil matrix (Clothier et al., 2000). As a consequence, repellency enhances ponding at the soil surface. On a watershed scale, repellency potentially enhances runoff from hillslopes (Shakesby et al., 2000), a feature that also can have practical applications. Water harvesting is a technique in which the soil surface is treated chemically to induce runoff by decreasing infiltration. Runoff water is collected in a basin from which evaporation is relatively small because the basin surface area in general is much smaller than the chemically treated land area (Fink and Frasier, 1975). On a smaller scale within agricultural fields, runoff can proceed from ridges into furrows. Compared with uniformly wetted surfaces, a pattern of dry ridges and wet furrows can potentially reduce the average evaporation rate of a field (Yang et al., 1996).

Earlier studies concerning the impact of water repellency on evaporation have reported decreased evaporation because of treatment of the soil surface with a hydrophobic litter extract (DeBano, 1975). Because water repellency in natural soils is commonly not restricted to a thin layer at the soil surface, it is important to know how evaporation is affected in soils having a thick water repellent topsoil.

From an analysis of the current literature, it is evident that the effect of repellency upon evaporation requires clarification. Therefore, a study of water repellency and soil water evaporation was performed. The main objectives of the study were to measure isothermal and nonisothermal evaporation rates for loamy sand soil materials showing a different degree of water repellency, to analyze the impact of water repellency on the residual soil water content profiles after isothermal and nonisothermal temperature conditions, and to test the ability of a numerical model to describe the observed soil moisture transport.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
THEORY
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soils
Four soil materials were used in this study. The sampling site for two of the materials was soil in a former pine (Pinus L.) plantation, under agricultural use for the past 15 yr. The wettability of the A horizon (soil texture class sand) and Soil B (loamy sand) at this site varied considerably between slight water repellency (Soil AW) and strong water repellency (Soil AH, Table 2). The third material, Soil BW, was a wettable loamy sand (subsoil) from a quaternary river deposit with a similar particle-size distribution as Soils AW and AH. The fourth material, Soil BH, was hydrophobized in the laboratory. The grain surfaces were coated with 90 ml of dichlorodimethylsilane (C2H6Cl2Si) per kilogram of soil to give a strong water repellent counterpart to the wettable soil BW. The silane was added to the air-dry soil in small amounts under constant stirring. Preliminary tests showed that the standard deviations of contact angles of subsamples were in the same range as the standard deviation of replicates of the same sample. Experience gained in previous experiments had indicated that 90 ml of silane per kilogram of soil was the correct dose. A sufficient degree of repellency was achieved, but no coating was visible under the (light) microscope. Properties of the four soils that are relevant to the present study are presented in Tables 1 and 2.


View this table:
[in this window]
[in a new window]
  		Table 2. Contact angle, van Genuchten hydraulic parameters ({theta}s, {theta}r, {alpha}, n{ddagger}, ks, and l), temperature coefficient of water retention, ßo, and temperature coefficient of unsaturated hydraulic conductivity, {gamma}.

 

View this table:
[in this window]
[in a new window]
  		Table 1. Selected physical and chemical properties of the four soil materials.

 
Column Setup
Sixteen plexiglass columns of a 70-cm length and a 5-cm o. d. were used. Each column was filled with 65 cm of soil (Fig. 1) . During the course of the experiment, each column was closed by a lid (Fig. 1). Between the soil surface and the lid there was an empty space of 5 cm in height. The colums were filled step-by-step (1 cm per step). For each layer of 1-cm thickness, the same amount (by weight) of moist soil was used. Soil water content was controlled by measuring the water content gravimetrically of five subsamples per soil before filling the columns. Shortly before filling each column, the soil was mixed for a couple of minutes inside a plastic bag. Each 1-cm layer, in turn, was filled in three steps. In this way, it was made sure that there was neither bulk density nor initial water content gradients inside a column. Bulk density, as shown in Table 1, was calculated from the total amount of soil used to pack each column. Average bulk density for the 16 columns varied between 1.52 and 1.55 Mg m-3.



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 1. Schematics of a column.

 
The initial water content in the columns was 0.077 kg kg-1 for Soils AW, AH, and BW. For Soil BH, the most water repellent material, the initial water content was 0.102 kg kg-1. The higher initial water content for Soil BH was necessary to obtain a homogeneous soil water distibution in the column during packing. A gravimetric water content of 7.7 to 10.2% was chosen on purpose, because the largest thermal effects (based on previous experiments) on the liquid and vapor water movement were expected.

Eight vertical soil columns (two of each soil) were subjected to isothermal conditions (20 ± 1°C), and the other eight columns were buried vertically in a large quartzsand-filled container (base 20 by 40 cm) that was heated at the base. To ensure a one-dimensional temperature field, the sides of the container were insulated with styrofoam. Soil column boundary temperatures during the 195-d study were held at 21°C (surface) and 55°C (base). A temperature of 55°C at the bottom of the column was chosen to have a distinct temperature gradient along the column and to have measurable temperature effects in relatively short time (195 d).

A small cup containing 10 cm3 of saturated CaCl-2 solution was placed in the empty space (chamber) on the soil surface in the upper part of each column to establish a constant sink for water vapor (Fig. 1). A supply of precipitated salt at the bottom of each cup was considered to prevent the solution in the cup from being diluted with the increasing accumulation of evaporated water. In view of the small size of the chamber and the relatively high temperature in the chamber (21°C), the assumption that Brownian movement prevented concentration gradients in the chamber and in the cup with the CaClCaCl-2 solution was made. Each cup was weighed 10 times during the 195-d study, on Days 7, 12, 27, 38, 52, 67, 90, 115, 151, and 195 of the experiment. The cups were weighed by opening briefly the top cover lid of the column and by replacing the cups. Because of the small air-filled volume of the chamber, the assumption was made that only a minor impact on the measured total amount of evaporated water.

At the end of the experiment, the soil columns were sectioned in 30 slices of variable thickness. The segments (layers) were 1 cm thick near the column surface and 2.5 cm thick near the bottom of the column (Fig. 1). The water content of each layer was measured to determine the soil water content distribution in the column at the conclusion of the experiment. Two soil column replicates were used for each of the four soils. Standard deviation of the residual water content was calculated for each soil with the 30 pairs of water content (one pair per depth for each pair of soil columns) according to Vermeulen (1953). It was found that standard deviation was <0.001% for isothermal and <0.003 to 0.272% (by weight) for the nonisothermal columns. Mass balances were calculated from initial water content minus the evaporated water and minus the residual water content. The absolute mass balance error was <0.26% by weight for the isothermal runs and <0.79% for the nonisothermal soil columns.

Hydraulic Properties
Hydraulic properties of all soils were determined independently with separate soil column outflow experiments. The technical setup is reported by Bachmann et al. (2002). To analyze simultaneously the water retention curve and the hydraulic conductivity, water tension and water content was determined in five depths as a function of time. Water tension of the first depth was used as upper boundary condition. For the bottom, a zero flux boundary condition was applied. These data sets, obtained at least 10 different times, were used as input data for the inverse modelling program FLOWFIT (Kool et al., 1988). The bulk densities of these columns differed only slightly from the bulk densities of the columns used for the evaporation experiments. For Soil BW , the bulk density was 1.52 Mg m-3 and for Soils BH, AW, and AH it was 1.46, 1.61, and 1.48 Mg m-3, respectively. Compared with Table 1, a mean difference of the bulk density of both soil column sets was <0.055 Mg m-3. The van Genuchten parameters {theta}S, {alpha}, n, kS, and l (van Genuchten, 1980) were estimated simultaneously for 20°C by using the inverse modeling approach reported by Kool et al. (1988). Water retention and hydraulic conductivity curves (Eq. [1] and [2]) of the four soils for 20°C are presented in Fig. 2 .



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 2. Water retention curve and unsaturated hydraulic conductivity of Soil BW (contact angle <5°), Soil AW (contact angle 45°), Soil AH (contact angle 93°), and Soil BH (contact angle 110°) at 20°C. Symbols do not represent data but are shown to identify the curves.

 

[1]

[2]

In Eq. [1] and [2], the quantity {theta}r is the residual water content (m3 m-3), {theta}s is the saturated water content (m3 m-3), k is the unsaturated and ks the saturated hydraulic conductivity (m s-1), and l (-), {alpha} (m-1), n, and m are empirical fitting parameters. We used the relation m = 1 - . The effective water content, {Theta}, in Eq. [2] is defined as {Theta} = .

The temperature dependence of the water retention curves was estimated according to Grant and Salehzadeh (1996). These authors proposed an expression that can be incorporated into any analytical model for a general description of the temperature-dependent matric potential {psi}({theta},T) at a given water content, {theta}. This expression can be written as:

[3]
where T is an arbitrary temperature (K), Tr is the reference temperature (K), and ßo is a soil-specific parameter. For experimental details concerning the determination of ßo, see Bachmann et al. (2001). Using the hydraulic conductivity at 20°C, as determined with the method of Kool et al. (1988), the hydraulic conductivity, kT, was estimated at an arbitrary temperature T with Eq. [2]:

[4]

In Eq. [4], T0 is a reference temperature (in this case T0 = 293 K), and {gamma}(K-1) is a soil-specific parameter, similar to ßo in Eq. [3]. This equation is based on the temperature dependence of the viscosity of pure liquid water (Döll, 1996). The temperature factor (for the unsaturated hydraulic conductivity was evaluated with Eq. [4] from the hydraulic conductivity curves derived from outflow experiments at 5 and 38°C. Preliminary experiments showed that higher temperatures may lead to a disconnection of the liquid phase in the transition zone between ceramic plate and soil.

Measured data are shown in Table 2. Because the soils were expected to dry out during the course of the experiment, especially in the lower part of the columns for the nonisothermal conditions, the water retention relations were extended into the high-suction region. With a conventional pressure-plate apparatus, the soil water content was determined for soil water tension of 1.5 MPa. Additionally, the water content of the soil materials was determined at tension of 30 MPa by equilibrating samples with the atmosphere in a closed chamber containing a saturated CaCl-2 solution at 20°C. Both values were obtained under desorption conditions. The values were assumed to be independent of the temperature. These additional values for the water retention relation were used, together with the other values in Table 2, to derive the van Genuchten retentivity curves and the related conductivity functions. Possible deviations from the calculated hydraulic conductivity in the high-tension region (>1.5 MPa) because of water vapor transport (see Milly, 1984), were not taken into consideration.

Particle Surface Properties
The degree of repellency of the soils was assessed by a modified sessile drop method. After sieving, the air-dried soil fraction was sprinkled on a glass plate (2 by 6 cm) with double-sided adhesive tape. A flat, one-grain layer of soil particles with uniform size was formed on the plate. Contact angles were measured at room temperature (20 ± 1°C) using a microscope fitted with a goniometer scale. After placing six droplets at the same time (drop volume 1.7 mm3) of deionized water on the horizontal plate, 12 static contact angle readings were measured within 1 min by adjusting the goniometer as a tangent at the point of three-phase contact. For details of the method see Bachmann et al. (2000a)(b). Results of the contact angles of the different soil fractions are presented in Table 2. The surface area of all soils was determined with the BET method (N2) (Brunauer et al., 1938) with the use of a Ströhlein areameter.


   THEORY
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 THEORY
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Simulation of Moisture Movement
A thermal gradient can cause water movement both in the liquid and in the vapor phase (Philip and de Vries, 1957; Milly, 1982). Philip and de Vries theory for describing soil water movement under a thermal gradient was recently used by Nassar et al. (1997a) for a salinized soil, by Nassar and Horton (1997a) for a layered and compacted soil, and by Nassar et al. (1997b) for a clayey soil exposed to a daily temperature wave. The present study also used the Philip and de Vries theory in addition to a model developed by Nassar and Horton (1997b) for the simultaneous transport of heat and water. However, the heat fluxes across the boundaries of the one-dimensional flow region were not considered. Also the heat fluxes within our columns were not considered, but assumed a linear temperature distribution between the bottom (55°C) and the top (21°C) of our columns. In this case, the nonsteady-state mass balance equation for the flow of water may be written (Nassar and Horton, 1997b) as:

[5]
where {psi} is matric pressure head (m), T is temperature (K), t is time (s), {theta} is volumetric water content in the liquid phase (m3 m-3), {theta}a is volumetric air content (m3 m-3), k is unsaturated hydraulic conductivity (m s-1), D{psi}V is matric pressure head dependent diffusion coefficient for water vapor flow (m s-1), DTV is thermally induced diffusion coefficient for water vapor flow (m2 s-1 K-1), {rho}L is liquid water density (kg m-3), {rho}V is water vapor density (kg m-3), and is a unit vector, directed downward. Diffusivities D{psi}V and DTV were defined and calculated according to Milly (1984).

An implicit finite-difference algorithm was used to solve Eq. [5]. The spatial discretization of a soil column, starting from the soil surface, was as follows: for the first 4 cm of the column, a space step of 1 cm was used; for the soil depth of 4 to 10 cm, a space step of 2 cm was used; and for the soil depth of 10 to 65 cm, it was 2.5 cm. The initial time step was 0.1 s, the maximum time step was 5 x 104 s, and the minimum time step was 1 x 10-3 s.

Water Flux Surface Boundary Condition
The boundary condition at the soil surface was defined as a time-dependent flux of water vapor, calculated with the use of Fick's law. The vapor-pressure gradient was calculated from the simulated, time-variable soil water potential in the upper (0–1 cm) soil layer, as determined in the previous time step from the constant humidity in the small air chamber (Fig. 1), maintained by the saturated CaCl-2 solution, and from the the mean macroscopic path length of water vapor from the soil surface to the center of the surface of the CaCl-2 solution in the small cup. The diffusion coefficient of water vapor in air, Do (m2 s-1) was approximated for each time step from the mean vapor density of the air between the soil surface and the surface of the CaCl-2 solution. The diffusion coefficient, Do, was computed according to Kimball et al. (1976). Vapor density, {rho}V (kg m-3) at the soil surface was derived following Horton (1989) from the soil water potential in the upper soil layer (0–1 cm), as calculated for the previous time step.


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 THEORY
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Evaporation
Evaporation during the initial phase of evaporation (Days 1–7) showed little differences for the soils with contact angles <100° (3.61 kg m-2 for Soil BW, 3.96 kg m-2 for Soil AW, and 3.86 kg m-2 for Soil AH), but evaporation was lower by ~20% (3.19 kg m-2) for Soil BH with a contact angle >100° (see Fig. 3) . The slope of water loss squared vs. time for the entire period (195 d) suggested that unsteady second-stage evaporation took place for all soils under isothermal conditions. This stage is characterized by a relatively larger decrease of hydraulic conductivity than an increase of matric potential gradient, thus leading to decreasing evaporation rates. This particular evaporation stage is generally accompanied by a downward drying of the soil profile, starting from the soil surface (Kutilek and Nielsen, 1994). The slopes of the lines fitted by linear regression to the cumulative water loss data indicated that the evaporation rates decreased with increasing soil contact angles. For nonisothermal conditions, however, the evaporation rates were considerably larger than for isothermal conditions, e.g., Soil BW, with a contact angle <5°, showed an increase of cumulative water loss by a factor of two after 195 d. In contrast, Soil BH (with a contact angle of 110°) showed a nonisothermal water loss that was only slightly larger than that for isothermal evaporation (Fig. 3).



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 3. Measured cumulative amount of evaporated water (squared) versus time for all soils. All values are averages of the two experiments.

 
Moisture Transport to the Soil Surface
Different mechanisms may be considered to describe the evaporation process. Under the assumption that the available energy at corresponding evaporation sites in different columns is identical for all soils (which is appropriate because of the controlled temperature fields and the identical vapor sink) at least two processes determine the evaporation rate. Both processes potentially can be influenced by water repellency. These processes are the transport of liquid water to the vaporizing locations in the soil profile and the phase transition from liquid to vapor (vaporization) with diffusion of water vapor into the air-soil boundary layer.

The (known) soil water content in the upper layers (0–1 and 1–2 cm) in each column allows the partitionining of liquid water and vapor fluxes through the first two soil layers to analyze the first process mentioned. To evaluate the isothermal vapor flux, matric potential gradients and vapor density were estimated from the water content of the soil Layers 1 and 2 at Day 195 using the water retention curves. A comparison of the calculated isothermal vapor transport with the observed average moisture flux at the soil surface (Days 155–195) led to the conclusion that the water transport in the upper part of the column (0–2 cm) occurred mainly in the liquid phase. The calculated Fickian vapor diffusion was 2 to 2.5 orders of magnitude smaller than the evaporation flux observed at the surface, except for the wettable Soil BW, in which ~10% of the total water flux occurred as vapor (Table 3). Considering this final soil water content in relation to the evaporation rate of the different soils, it was apparent that for Soil BH (the most water repellent soil), the reduction of the moisture flux took place directly at the soil surface. The water content of this top layer of the repellent soil (BH) was ~0.10, whereas the corresponding water content of the wettable soil (BW) was only ~0.02. Under these conditions, the hydraulic conductivity of BH was approximately three orders of magnitude larger than for Soil BW. The average evaporation rate of BH, however, was ~25% larger than the rate for BW (Fig. 3). Probably, the average water content of Soil BH of the first soil layer (0–1 cm), or even the microscopic water content of the shallow particle layer directly at the soil surface, was beyond the critical water content, where the soil wettability changed from wettable to water repellent (Ritsema et al., 1997; de Jonge et al., 1999). This effect may disconnect liquid domains in the pore system. An inspection of the average film thickness does support this assumption. Average water film thickness was derived from the volumetric water content in the 0- to 1- and 1- to 2-cm soil layers, the bulk density, and the specific surface area of the soil. For the top layer of the soil (0–1 cm), a direct relation between the average thickness of water films on the grain surface and the observed evaporation flow was not observed (Table 3).


View this table:
[in this window]
[in a new window]
  		Table 3. Calculated values for water film thickness and water vapor flux in the surface layer (0–2 cm) based on water content data of Day 195 for isothermal conditions. (All values are averages of two experiments.)

 
The 0.2-µm thick water film of the wettable Soil BW appeared to support a larger liquid flux compared with the thicker water films of the water repellent Soils AW, AH, and BH. Again, in the uppermost first or second grain layer, the continuity of the liquid phase probably decreased with increasing contact angle. This phenomenon seems to be more effective for water flow retardation than the increasing viscosity of the water film that must be generally expected if the average water film on the grain surfaces becomes thinner. Because of the five times larger surface water content of the strongest repellent soil, BH, an insufficient moisture flux into the shallow surface zone of the columns is considered not to be the process that reduced the evaporation rate with increasing contact angle. The vaporization process itself, as mentioned above as a second important process, seems to be more limiting to evaporation than does the liquid water flux.

Vaporization and Evaporation at the Soil Surface
At least two repellency-influenced processes during evaporation have been suggested. First, Birdy and Vu (1993) observed that the evaporation (or vaporization) rate of fluid drops on a plane surface was affected by the wettability of the surface. If the contact angle was <90° (water on glass), the evaporation rates varied linearly with time, whereas at contact angles >90° (water on Teflon) the evaporation rate was decreased by a factor of ~30% and the evaporation rate decreased with the size of the drop. These results were attributed to the constant contact area of drops but decreasing contact angles on wettable surfaces, hence maintaining a relatively large liquid–gaseous interface during the entire process of evaporation. On the repellent surface, however, the contact angle remained constant, but the contact area decreased during evaporation, resulting in a smaller vaporizing drop surface area. Similar arguments may apply to the vaporization process near the surface of our columns filled with extremely water repellent soil, even if the menisci had concave curvatures, thus causing negative capillary pressure.

Second, with respect to our column experiment, the increasing path length from the vaporizing sites to the soil surface seems to be of importance. Churaev (1975) showed that in thin capillaries, vaporization from adsorbed water films above the water menisci was on the same order as vaporization from the liquid–gaseous interface of the meniscus itself. Moreover, it was shown for identical capillaries with hydrophobic walls without a corresponding water film that evaporation was reduced by 50%, as compared with completely wettable capillaries. A related observation made may serve to clarify the process that determines the evaporation directly at the soil surface.

Figure 4 shows salt crystals precipitated after a saturated salt solution had dried in different glass beakers. One beaker had a wettable glass surface and the other a hydrophobic surface. The inner surface of the nonwettable beaker had been hydrophobized with the same chemical (Dichlorodimethylsilane) used to produce our hydrophobic soil. The salt distribution after evaporation showed that with respect to the wettable beaker the salt crystals were distributed over the whole inner surface. Whereas in the case of the hydrophobic beaker, the salt crystals remained at the bottom, indicating that evaporation proceeded only directly from the salt solution surface. In case of the hydrophilic beaker, however, evaporation took place from a thin water film that entirely covered the inner glass wall of the beaker. A similar effect may have occurred within the first particle layers of our columns with Soils BW and BH. In summary, both mentioned processes may have affected evaporation, resulting in a decreased evaporation rate of ~25% for the isothermal water repellent soil columns and 50% for the nonisothermal water repellent soil columns.



View larger version (149K):
[in this window]
[in a new window]
  		Fig. 4. Salt residues after evaporation of a saturated NaCl-KCl salt solution in a wettable and a hydrophobic glass beaker.

 
Bulk organic matter itself may also have an effect on evaporation. As shown in Table 1, Soils AW and AH had an organic C content of 1.4%, and Soils BW and BH contained only 0.6%. It is probable that organic matter enhances evaporation because of its large microporosity (Tan, 1985) that may support larger liquid–gaseous interfaces. For the isothermal columns, Soil AH had a slightly larger evaporation rate than Soil AW with a smaller contact angle. For these soils, the larger surface area of Soil AH might compensate the contact angle effect. The grain surface area of the wettable Soil BW and of the hydrophobized Soil BH were similar, thus indicating that the contact angle dominated the evaporation process in these soils.

Predicting Evaporation
Model results for the transient evaporation fluxes and for the soil column water distribution at the end of the experiment were satisfying for the wettable Soil BW (Fig. 5) . For the other soils, it was found that deviations between predicted and measured total evaporation and residual water content became larger with increasing repellency. With increasing contact angle, the evaporation rates were overestimated by the model, especially for the most repellent soil with a contact angle of ~110°. It was further observed that differences between measured and simulated water content in the upper 100 mm of a column increased with increasing repellency. Hence, it appears that in general the use of a standard model approach will overestimate evaporation with increasing water repellency. Based on the observation, that evaporation is affected by wettability, we assumed that the thermal diffusion of vapor (which is both a vaporization and condensation process) is also affected by the wettability inside the porous medium. A sensitivity analysis for Soil BH showed that variation of the nonisothermal vapor transport coefficient DTV reduced effectively the bias between predicted and observed data if DTV was multiplied by a water content independent factor of 0.35 (Fig. 6) . On the other hand, a slightly better approach for the residual water content distribution was obtained for Soils AW, AH, and BW when the transport coefficient DTV was multipied by a factor of 3.5, which is in fairly good agreement with the enhancement factors reported by Döll (1996). These findings suggest that only soils with an extreme water repellency, i.e., with contact angles >100°, have a mechanism for thermal water vapor diffusion that is different from the mechanism of wettable soils.



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 5. Measured and simulated isothermal and nonisothermal cumulative evaporation (upper graphs) and measured and simulated residual volumetric water content after 195 d (lower graphs) for the wettable Soil BW. All values are averages of the two experiments.

 


View larger version (29K):
[in this window]
[in a new window]
  		Fig. 6. Measured and simulated isothermal and nonisothermal cumulative evaporation (upper graphs) and measured and simulated residual volumetric water content after 195 d (lower graphs) for the water repellent Soil BH. All values are averages of the two experiments.

 

   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 THEORY
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Our experiment showed that under isothermal conditions water repellency decreased evaporation rates by as much as 25%. The contact angle influenced evaporation rates were different during the entire period of the experiment (195 d), i.e., convergence of the evaporation rates was not observed. Predicted evaporation, based on independently measured water retention and hydraulic conductivity functions, and based further on theoretically calculated vapor transport coefficients, D{psi}V and DTV, resulted in an increasing error with increasing repellency. Compared with isothermal conditions, surface-directed thermal gradients enhanced the evaporation rate by 75% for the wettable and slightly repellent soil and by 14% for the strongly hydrophobic soil. The thermal vapor transport within a soil column became smaller with increasing repellency. The experiments indicated further that the process of evaporation was predominantly controlled by a thin surface layer, apparently <1 cm. It appears that an increasing contact angle can be used to identify soils having the capability to reduce evaporation. The extent of reduction needs more information because other factors like organic C content, porosity, pore size and particle-size distribution may have also an impact on evaporation.


   ACKNOWLEDGMENTS
 
The authors thank Patrick Becker and Rolf Ruenger, Hannover, for expert technical assistance.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 THEORY
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Journal Paper no. J-19035 of the Iowa Agric. and Home Econ. Exp. Stn., Ames, Iowa; Project No. 3287, and supported in part by Hatch Act and State of Iowa.

Received for publication September 21, 2000.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 THEORY
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Vadose Zone JHome page
J. W. Nyhan
A Seven-Year Water Balance Study of an Evapotranspiration Landfill Cover Varying in Slope for Semiarid Regions
Vadose Zone J., June 9, 2005; 4(3): 466 - 480.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
J. Bachmann, R. Horton, T. Ren, and R. R. van der Ploeg
Comparison of the thermal properties of four wettable and four water-repellent soils
Soil Sci. Soc. Am. J., November 1, 2001; 65(6): 1675 - 1679.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bachmann, J.
Right arrow Articles by van der Ploeg, R. R.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Bachmann, J.
Right arrow Articles by van der Ploeg, R. R.
Agricola
Right arrow Articles by Bachmann, J.
Right arrow Articles by van der Ploeg, R. R.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome