JEQ Grow Your Career With ASA
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 Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Magesan, G. N.
Right arrow Articles by Lee, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Magesan, G. N.
Right arrow Articles by Lee, R.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Magesan, G. N.
Right arrow Articles by Lee, R.
Related Collections
Right arrow Time Domain Reflectometry, TDR
Right arrow Solute Transport Models
Right arrow Soil Physics
Right arrow Vadose Zone Processes and Chemical Transport
Published in J. Environ. Qual. 32:2325-2333 (2003).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS

Vadose Zone Processes and Chemical Transport

Solute Movement through an Allophanic Soil

G. N. Magesan*,a, I. Vogelerb, B. E. Clothierb, S. R. Greenb and R. Leec

a Forest Research, Private Bag 3020, Rotorua, New Zealand
b HortResearch, Private Bag 11030, Palmerston North, New Zealand
c Landcare Research, Private Bag 3127, Hamilton, New Zealand

* Corresponding author (gujja.magesan{at}forestresearch.co.nz).

Received for publication March 25, 2002.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 THEORY
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 APPENDIX
 REFERENCES
 
Allophanic soils are widespread around the world, but little research has been done on their transport properties. This study reveals the effect of two soil water potential heads and two water-flow regimes of continuous and intermittent flow on solute transport through undisturbed soil columns of Horotiu silt loam (Typic Hapludand), an allophanic soil. Two different methods—breakthrough curves (BTCs) and time domain reflectometry (TDR)—were employed to determine the extent of preferential solute transport in the topsoil. The TDR data were also used to look at the depth dependence of the transport properties. The convection–dispersion equation (CDE) with the appropriate boundary conditions adequately described the movement of both Br and Cl under the various flow conditions. Although no preferential flow was found under the imposed unsaturated flow conditions, the flow of water and transport of solute became more uniform with depth. The results show that both Br and Cl are retarded in this allophanic soil. Retardation values range from 1.5 to 1.9, and, as the TDR data showed, increase from the depth of 5.0 to 10.0 cm. Intermittent leaching results showed that there was no effect on solute concentrations in the leachate following no-flow periods. This suggests that water and solute transport in this soil were either relatively uniform or that transverse mixing during flow was already fast enough to eliminate concentration gradients between regions of different "mobility."

Abbreviations: BTC, breakthrough curve • CDE, convection–dispersion equation • TDR, time domain reflectometry


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
THEORY
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 APPENDIX
 REFERENCES
 
STUDIES ON SOLUTE MOVEMENT through soil have increased recently due to public concern about water quality. Solute transport experiments conducted under controlled laboratory conditions are useful to better understand some specific solute behaviors (Magesan et al., 1995). For laboratory results to have some relevance to the field, however, it is generally best to use intact soil columns and unsaturated flow conditions.

Preferential flow is of critical importance in solute transport because surface-applied solutes can reach the saturated zone more rapidly than in soils with homogeneous wetting (Seyfried and Rao, 1987). In agricultural areas preferential flow can lead to a loss of nutrients and an increased likelihood of ground water contamination. While much work has been done on soils in which anions such as bromide and chloride act conservatively, essentially no work has been done on allophanic soils (soils formed from the volcanic ash). Volcanic soils often carry variable surface charge and can therefore adsorb anions. Because sorption critically controls the depth and pattern of leaching, it is important to quantify the degree of adsorption under different flow conditions in leaching experiments.

Another factor influencing the transport of solutes is the depth dependency of the transport processes. So far, however, little is known about this depth dependency. This lack of information is mainly due to the difficulty in measuring transport properties at various depths in the soil profile. Various studies have shown that TDR can be used to study solute transport in soils (Wraith et al., 1993; Ward et al., 1994; Risler et al., 1996). Time domain reflectometry probes can be installed at different depths into the soil column. Thus, TDR has the potential to measure the depth dependence of transport properties. Only a few studies have been done using horizontally installed TDR probes to look at changes in dispersivity ({lambda}) with depth (Vanclooster et al., 1995; Vanderborght et al., 2000). These studies were performed to evaluate if transport is stochastic–convective ({lambda} increases with depth) or convective–dispersive ({lambda} is depth invariant). Jury and Roth (1990) give a detailed discussion on the differences between stochastic–convective and convective–dispersive transport.

The soil studied in this research was Horotiu silt loam with the soil taxonomic classification Typic Hapludand (Buol et al., 1997). Extensive belts of Andisols are found around the "Ring of Fire" in the Pacific including the Andes of South America, the Central American chain, Alaska, Japan, Philippines, Indonesia, New Zealand, and the Pacific Northwest of the United States (Buol et al., 1997). Leamy et al. (1980) showed a map of the world with the global distribution of Andisols, which are present on every continent except Antartica. Because Andisols are widespread around the world and little research has been done on their transport properties, experiments performed on such soils are critical in helping to understand how the ground water under these soils can be protected from pollution.

The CDE is a widely used solute-transport model, and in general, it adequately describes solute transport through a uniform soil. Although the CDE has often been found inadequate in structured soils under saturated flow conditions (Seyfried and Rao, 1987), some researchers (e.g., Jardine et al., 1993; Magesan et al., 1995) have provided experimental justification for the validity of the CDE under unsaturated flow conditions. Vogeler et al. (1998) extended this concept to intermittent leaching conditions for the Ramiha soil (Andic Dystrochrept).

Magesan et al. (1999), working with 1-m-long Horotiu soil lysimeters, reported that no preferential flow occurred in the Horotiu soil when it was wetted to field capacity. However, McLeod et al. (1998) found evidence of preferential flow in the surface horizons of this soil. To resolve this issue, a study with laboratory-leaching experiments was performed with intact Horotiu silt loam soil columns brought from the field. We looked at the effects of different soil water potentials and water-flow regimes (continuous and intermittent flow) on preferential solute transport. We also sought to quantify the degree of anion adsorption under flow conditions.

Solute movement was studied by collecting the effluent exiting the base of the column and by TDR probes installed at various depths into the soil column. Both TDR and effluent results were compared with the CDE.


   THEORY
TOP
 ABSTRACT
 INTRODUCTION
 THEORY
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 APPENDIX
 REFERENCES
 
Modeling
For a steady state, one-dimensional transport of a reactive solute through a uniform soil, the CDE can be written as:

[1]
where R is the dimensionless retardation factor, D is the hydrodynamic dispersion coefficient (m2 s-1), {theta} is the volumetric water content (m3 m-3), C is either the flux or resident concentration (mol m-3), z is depth (m), and v is the pore water velocity (mm s-1), given by q/{theta}, where q is the water flux density (m s-1). The dispersion coefficient D is often assumed to be linearly related to the pore water velocity, D = {lambda}v, where {lambda} is the dispersivity (m). Steady state water flow at Darcy flux density q, and a uniform and constant volumetric water content ({theta}) are assumed, since equal and constant pressure heads were imposed at the top and bottom of the soil column. The intact soil columns were preleached to remove the solutes present in soil, so the soil was assumed to be initially free of the ion of interest. When time, t, equals zero, the flux concentration changes to Cf = Co in the infiltrating solution. Thus, the appropriate initial and boundary conditions relevant to this study are:

[2]

[3]

For the lower boundary it is assumed that the soil column was part of an effectively semi-infinite system, as suggested by van Genuchten and Wierenga (1986). For these initial and boundary conditions, and for the normalized concentrations at the lower boundary of the column, the analytical solution for the flux concentration of Eq. [1] in terms of cumulative infiltration is (Skaggs and Leij, 2002; where I = qt and D = {lambda}v):

[4]
where L is the column length (m).

The solution for the resident concentration (Cr) at a depth z with the above boundary and initial conditions is given by (Skaggs and Leij, 2002; where D = {lambda}v):

[5]

Equation [4] was fitted to the data of normalized solute concentrations in the effluent, and Eq. [5] to the TDR data. Least squares optimization was used with {lambda} and R as fitting parameters. The numerical approximations used for the error function and associated functions were those suggested by Jury and Roth (1990)(p. 165–167).

Time Domain Reflectometry
The theory of monitoring solute transport using TDR has been described previously (Vogeler et al., 1996, 2000). Only salient features are repeated here. The estimation of the solute resident concentration in the soil is based on the measurement of the volumetric water content ({theta}) and the apparent soil electrical conductivity ({sigma}a, S m-1). The water content can be inferred from the TDR-measured dielectric constant using the universal relationship given by Topp et al. (1980). Following the thin-sample theory of Giese and Tiemann (1975), the {sigma}a can be described by (Topp et al., 1988):

[6]
where Z0 is the characteristic impedance of the probe ({Omega}), Zu is the characteristic impedance of the TDR system (50 {Omega}), V0 is the voltage of the incident step, and Vf is the final reflected voltage. The probe impedance Z0 was calculated using Topp et al. (1988):

[7]
where s is the rod spacing (m) and d the rod diameter (m). The TDR-measured apparent soil electrical conductivities were normalized using either the initial ({sigma}i) or final ({sigma}f) measured conductivity. For the step-down in Br, {sigma}i was used, and for the step-up in Br and Cl, {sigma}f was used.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 THEORY
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 APPENDIX
 REFERENCES
 
The soil studied is a Horotiu silt loam soil (Typic Hapludand; Soil Survey Staff, 1992). This soil is common in the Waikato region, which is centered on Hamilton City (37°47' S, 175°17' E), North Island, New Zealand. This is an intensively farmed soil, and is formed from alluvium, derived from a volcanic source, associated with the ancient Waikato river system. Horotiu soil, on levees, is a deep and well-drained, versatile soil, with moderate permeability. The soil has a bulk density of about 0.8 Mg m-3, a porosity of about 0.62, and it contains about 20% sand, 65% silt, and 15% clay. The saturated and unsaturated (pressure head, ho = -40 mm) hydraulic conductivities of this soil are 110 and 75 mm h-1, respectively (Singleton, 1991). Topsoil thickness is generally around 200 mm. Chemically, this topsoil has a very high phosphate retention (>98%) due to a relatively large content of allophane (10–12%). The pH of the topsoil is 4.8, the cation exchange capacity is 25 molc kg-1, and the total carbon and nitrogen are 8.2% and 0.67%, respectively.

The top 3 cm of the soil was removed and two undisturbed soil columns (20-cm length and 10-cm diameter) were taken. The leaching apparatus used in this experiment was similar to the one described by Magesan et al. (1995). In brief, the apparatus induced gravity-driven unsaturated flow of water by ensuring that the pressure head (ho) was the same at the top and bottom of the soil column. The diagram of the experimental apparatus is given as Fig. 1 in Magesan et al. (1995). A disk permeameter sits atop each soil column to allow solution entry at some pre-set pressure head. A manifold and bubbling tower in conjunction with a vacuum system maintained this. A similar apparatus underneath maintains the same pressure head at the base, yet allows routine collection of effluent aliquots. Two sets of experiments were performed with these two soil columns to study the influence of (i) soil water potential and (ii) continuous and intermittent leaching on solute transport.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 1. Flow rates for various leaching experiments for (a) Column A and (b) Column B. (•) -100 mm Br step-up, continuous; ({circ}) -20 mm Br step-down, continuous; ({blacksquare}) -20 mm Cl step-up, continuous; ({square}) -20 mm Cl step-down, intermittent; ({blacktriangleup}) -20 mm Br step-up, intermittent. The term I is drainage.

 
In one of the columns, Column A, three horizontal TDR probes were installed at 5-, 10-, and 15-cm depths below soil surface. The parallel three-wire TDR probes were 80 mm in length, with a wire diameter of 2 mm, and a spacing of 12.5 mm. The TDR probes were connected, via coaxial cables (with polyethylene as a dielectric material, and 2 m long) to a multiplexer. The multiplexer was similar in design to that of Heimovaara and Bouten (1990). The multiplexer was connected via a 0.5-m-long coaxial cable to a cable tester (Model 1502C; Tektronix, Beaverton, OR). A computer controlled the settings of the TDR, and also recorded and analyzed the waveform, based on curve-fitting algorithms described by Baker and Allmaras (1990). Time domain reflectometry measurements of the water content and the apparent soil electrical conductivity were made every 10 min.

Indigenous solutes in the soil columns were preleached with distilled water using more than two pore volumes (about 350 mm) at ho = -100 mm. In the first part of the experiment (Table 1), the soil columns continuously received about 350 mm of 0.015 M CaBr2 solution at ho = -100 mm. This was followed by continuous application of distilled water at ho = -20 mm to leach the Br. In the second part of the experiment, 0.015 M CaCl2 solution was applied continuously at ho = -20 mm, which was followed by intermittent leaching with distilled water to remove Cl. In the third part of the experiment, bromide was applied intermittently at ho = -20 mm. We used tracers (bromide and chloride) alternatively because it would help us to distinguish the ions easily, and to trace the movement of ions clearly, from each experiment.


View this table:
[in this window]
[in a new window]
  		Table 1. Data for Columns A and B of Horotiu silt loam. Step-up indicates continuous application of solute, and step-down indicates the solute is leached with water.

 
The effluent was collected in aliquots ranging in size from 0.5 to 15 mm, until about two liquid-filled pore volumes of the applied solutions had infiltrated. Bromide and chloride concentrations in the leachate were analyzed using an ion chromatograph (Model IC5000; Lachat Instruments, Milwaukee, WI), which comprised a pump, an autosampler, a conductivity detector, a temperature-controlled column heater cabinet, and an integrator. The chromatography columns used were an IC-PAK, an anion guard column, and anion chromatography column. The eluent used was a sodium gluconate–borate buffer (pH of approximately 8.5). The working standard was a mixed anion standard containing 20 g m-3 each of bromide and chloride. Run conditions were an eluent flow rate of 1.2 mL min-1, column temperature of 35°C, and an injection volume of 50 µL.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 THEORY
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 APPENDIX
 REFERENCES
 
Water Flow
The outflow rates, measured by weighing and timing the effluent aliquots, are shown in Fig. 1. When the flow is steady, the measured outflow rate is the Darcy flux density q. At ho = -100 mm, during continuous application of bromide solution (as shown by filled circles), the flow rates were very steady. The average flow rates were 9 and 6 mm h-1 for Columns A and B, respectively. In the next phase, when this solution was changed to distilled water at ho = -20 mm, the flow rate increased. This increase is most probably due to the change in potential head and hence changes in the wetted pore geometry (Magesan et al., 1995). After about 100 to 150 mm of drainage, the flow became reasonably steady. In the next phase, with the same potential head but with Cl solution, the flow rates still increased but became steady after about 100 mm of drainage. This increase could be due to the change in applied solution. But in general, the flow rates varied somewhat during each change in applied solution or water, and as expected, especially during the intermittent application. The average flow rates ranged from about 9 to 28 mm h-1 for Column A, and from 6 to 15 mm h-1 for Column B (Table 1). Consistently, Column A had higher flow rates than Column B, despite the fact that the same potential was applied. This could be a result of the natural variability (i.e., variations in pore distribution) in the soil (Langner et al., 1999). However, the flow rates are not atypical of rainfall intensities and drainage rates through this soil. For example, as mentioned earlier, the saturated and unsaturated (ho = -40 mm) hydraulic conductivities of this soil are 110 and 75 mm h-1, respectively (Singleton, 1991; Magesan et al., 1999).

The gravity-induced redistribution and some minor evaporation when infiltration was stopped during the intermittent leaching regime meant that it took some time for the outflow rate to recover once infiltration resumed.

Influence of Soil Water Potentials on Solute Transport
Effluent Concentrations
Figures 2a and b give the relative concentrations of bromide (ho = -100 mm) and chloride (ho = -20 mm) in the leachate for the two columns. The breakthrough curves (BTCs) are presented in terms of cumulative infiltration, rather than time, to diminish the effect of the different flow rates (Table 1). Still, in both cases the duplicates look slightly different (Fig. 2; Runs 1 and 3 in Table 1), which is in part due to the slightly different water contents of the two columns. Breakthrough curves displaying a classical "sigmoid" shape generally indicate reasonably uniform flow through soil. Factors such as flow rate, unsaturated flow, and retardation affect the "sigmoid" shape (Biggar and Nielsen, 1967; Kirkham and Powers, 1972). Higher flow rates are generally associated with more preferential flow. Surprisingly, however, the BTCs obtained at ho = -20 mm for the higher flow rates of 28 and 15 mm h-1 (Fig. 2b) displayed a more classical "sigmoid" shape than those obtained at ho = -100 mm and lower flow rates of 9 and 6 mm h-1. After about two liquid-filled pore volumes the effluent concentration had reached >90% of the influent concentration in both cases.



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 2. Measured and predicted (using convection–dispersion equation, CDE) concentrations of (a) bromide and (b) chloride for (•) Column A and ({circ}) Column B as a function of drainage, I (mm). Both solutions were applied continuously.

 
It has been shown that the CDE can be successfully fitted to similar BTCs under unsaturated flow conditions by maintaining intact soil columns under pressure heads between -150 and -40 mm (Magesan et al., 1995). Here, Eq. [4] was fitted to the BTC data for these soil columns (Fig. 2; Runs 1 and 3 in Table 2), with the dispersivity ({lambda}) and the retardation factor (R) optimized using a least-squares optimization. The optimized values of {lambda} and R are given in Table 2. The {lambda} values obtained for the BTCs at ho = -20 mm were much lower (14 and 12 mm for Columns A and B, respectively) than those obtained for BTCs at ho = -100 mm (53 and 24 mm for Columns A and B, respectively). The R values ranged between 1.7 and 1.8 in both cases.


View this table:
[in this window]
[in a new window]
  		Table 2. Model parameters obtained using the convection–dispersion equation (CDE) and the effluent data or the time domain reflectometry (TDR) data at depths of 5 (TDR1), 10 (TDR2), and 15 cm (TDR3) for Columns A and B.

 
Generally, the dispersivity is either assumed to be independent of the flow rate (and thus the imposed potential), or to increase with increasing flow rate (Vanderborght et al., 2000). We do not know the reason for the decreased dispersivity found in our study with increased flow rate. The dispersivity values found were similar in magnitude to the values implicit in the results of Seyfried and Rao (1987), Jardine et al. (1993), Magesan et al. (1995), and Vogeler et al. (1998) for unsaturated soil, which ranged from 10 to 120 mm. The dispersivity can be thought of as an indicator of how variable the local solute mobility is in a soil. Working with a well-structured Ramiha silt loam soil and a weakly structured Manawatu fine sandy loam soil, Magesan et al. (1995) suggested that the dispersivity could also be a useful length scale with which to assess soil structure. Generally, values for well-structured soils are much lower than for weakly structured soils. Magesan et al. (1995) suggested that the difference is due to the well-structured soil having a more permeable matrix, a more uniform local velocity distribution, and closer spacing between faster and slower flow pathways than the weakly structured soil.

The R values were similar in both columns (Table 2), and ranged between 1.7 and 1.8. These results suggest that some anion adsorption occurred during solute movement. Volcanic soils with variable surface charge are known to adsorb anions. The R values are in the range of values reported by Katou et al. (1996) for an Andisol in Japan. Adsorption critically controls the depth of leaching, and retards the downward movement of anions such as Br, Cl, and NO3.

Time Domain Reflectometry Measurements
The TDR measured relative {sigma}a following the application of bromide (Br up), water (Br down), and chloride (Cl up) for Column A and the three different depths are shown in Fig. 3 . At constant {theta} the {sigma}a is linearly related to the solute resident concentration. The gap in the TDR measurements was due to an unfortunate failure of our automated system. However, this failure had no influence on the results obtained. The values for the dispersivity and the retardation factor obtained from the effluent BTCs were used to predict the relative {sigma}a as measured by TDR following the various solute applications (Fig. 3). The agreement between the TDR measurements and the simulations using the CDE are good for the first TDR probe at a 5-cm depth below the soil surface. However, further down some deviations occur, which suggest that retardation increases and dispersion decreases at a depth of 5 cm opposed to 10 and 15 cm. Although no preferential flow occurred under the imposed unsaturated flow conditions, the flow appears to become more uniform with depth. This is in contrast with the findings by McLeod et al. (1998), who found preferential flow in the topsoil of the Horotiu soil. This contrasting finding might be due to the application method. Whereas we used disc permeameters set to various pressure heads, McLeod et al. used a rainfall simulator. This resulted in incipient ponding, which can create preferential flow.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 3. Measured and predicted (solid lines) relative {sigma}a for Column A, and time domain reflectometry (TDR) probe at ({square}) 5-, ({circ}) 10-, and (+) 15-cm depths. The values for the model parameters were those obtained from the effluent.

 
In contrast to our findings of decreasing dispersivity with depth, Vanclooster et al. (1995) found in their topsoil an increasing dispersivity with depth. Thus the flow appears to become more uniform with depth. This is expected for stochastic–convective transport (Jury and Roth, 1990). Deeper in the column Vanclooster et al. (1995) found that the dispersivity remained constant, suggesting convective–dispersive transport. Similar observations were made by Vanderborght et al. (2000) who found depth-invariant dispersivities in both their sandy loam and loam soil at low flow rates. However, at high flow rates the dispersivities increased with depth in the loam soil. They suggested that lateral mixing between regions of different mobilities is incomplete at high flow rates, thereby resulting in stochastic–convective flow.

Next, the TDR data were fitted to the solution of the CDE (Eq. [3]). The fitted parameters {lambda} and R are given in Table 2. In all three cases, the fitted dispersivities decrease with depth. Apart from the bromide application, the dispersivity obtained from the effluent is an average of that found from the TDR probes. The retardation factor increases for all three cases with depth, suggesting that anion adsorption increases with depth. Again the R value obtained from the effluent is an average over the entire depth. Values for both {lambda} and R are quite similar for a specific depth and the various solute applications. Changes in transport properties with depth cannot be inferred from the effluent, as the effluent averages the transport properties over the entire volume studied. Here, the TDR setup offers a means for studying transport properties with depth. However, as TDR probes measure only a small volume around the probes, several replications need to be made in structured soils with highly nonuniform flow (Mallants et al., 1994).

Influence of Continuous and Intermittent Flow Conditions on Solute Transport
The flow-regime experiment was performed at ho = -20 mm. Step-up BTC (Fig. 4 ; Runs 3 and 5 in Table 1) and step-down BTC (Fig. 5 ; Runs 2 and 4 in Table 1), and fitted curves are described for the effluent concentrations. The TDR data are not shown, as the interruption of flow in these intermittent experiments resulted in a change in {theta}, and thus {sigma}a. The BTCs obtained from continuous application of Cl were uniform and the duplicates were similar. Although they had different flow rates (28 and 15 mm h-1 for Columns A and B, respectively), the dispersivity values (14 and 12 mm) and the retardation factors (1.7 and 1.8) were similar (Table 2). However, for intermittent application of Br, the flow rates were 21 and 14 mm h-1 for Columns A and B, respectively. Although the dispersivity values were almost similar (10 and 7 mm), the retardation factors were somewhat different (1.5 and 1.9). This is reflected in the BTCs with a translational delay in Column B.



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 4. Measured and predicted concentrations of (a) chloride and (b) bromide for (•) Column A and ({circ}) Column B as a function of drainage, I (mm). Chloride was applied continuously, and bromide intermittently.

 


View larger version (16K):
[in this window]
[in a new window]
  		Fig. 5. Measured and predicted concentrations of (a) bromide and (b) chloride for (•) Column A and ({circ}) Column B as a function of drainage, I (mm). Bromide was washed out by continuous application with water, and chloride by intermittent water application.

 
As mentioned earlier, there were two breaks during the Br intermittent leaching experiment. The first break of nearly 20 h occurred after about 90 mm of drainage. The relative concentration until then was <0.03, and was about to increase, as a "sigmoid" curve was set to rise. The relative concentration of Br increased from 0.03 to 0.17 in the first sample after the break. It then increased to 0.89 just before the second break of about 19 h. The second break occurred around 180 mm of drainage for Column A. The relative concentration increased from 0.84 to 1.0 after the second break. The decrease in concentration in the BTC from 0.89 to 0.84 may be due, in this case, to some molecular diffusion from fast- to slower-moving water. Another possibility is a change in the flow-path geometry following the interruption in the flow (Roth and Hammel, 1996). However, this drop was only short-lived. It appears that a drop would be noticed only when the concentration in the mobile region is high. Such an effect was not noticed after the first break when the solute concentration in the mobile region was very low.

The Column A results look similar to those reported by Vogeler et al. (1998) who found that a one-week pause in leaching had no effect on Cl concentration, because inflow and outflow concentrations were equal when it occurred. But the effect of the overnight pause during leaching at the high flux density appeared as a sudden drop in relative concentration when leaching resumed. Hu and Brusseau (1995) have also observed similar effluent concentration decreases following an interruption in the flow through a column containing porous spheres. However, in Column B, no drop in concentration was noticed after either break, which occurred at 80 and 220 mm of drainage.

As stated in the Theory section, steady state water flow was assumed for modeling solute transport under the different flow regimes, which is appropriate if the concentrations of solutes in the effluent are presented as a function of drainage (Russo et al., 1989). The modeling results from our study showed that the measured and predicted curves agreed well except for a few data points after the second break (Fig. 4). This is similar to the results of Meyer-Windel et al. (1999) who found similar transport behavior of solutes through homogenous soils for steady state and transient flow conditions, provided that the transport regime is not preferential. The fitted dispersivity and retardation coefficient values may here be more or less independent of variable water content and transient flow regime since there is no preferential transport.

Figure 5 describes the measured and predicted curves for continuous and intermittent leaching of solutes by water. In general, the duplicates are similar. The slight difference between the columns could be due to the different flow rates. Once again, for intermittent leaching, the two breaks, now on step-down, did not have any significant effect on the solute concentrations. Porro and Wierenga (1993) also indicated that transport behavior could be essentially the same under such different flow regimes.

The more complex approach of the mobile–immobile model (MIM) could have described the discontinuity after flow resumed, but we did not attempt this model in our study. Some studies (e.g., Schulin et al., 1987; Vogeler et al., 1998) have suggested that MIM model simulations are only slightly better than the ones obtained using the simpler CDE. Vogeler et al. (1998), using a fine sandy loam soil, concluded that "the relatively small improvement gained compared with the CDE might not justify the use of the mobile–immobile approach, which requires estimation of more model parameters." However, the degree of improvement using MIM may depend on the soil and many other factors. So, further testing of this approach under different conditions may be needed to assess its practical use. However, it is important to minimize model parameters for reasons of practicality in field-scale modeling situations.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 THEORY
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
APPENDIX
 REFERENCES
 
For intact soil columns of Horotiu silt loam (allophanic soil), the effects of two different prescribed matric potential heads and two water flow regimes of continuous and intermittent flow were found to be of minor importance for solute transport displaying a classic convective–dispersive behavior. At flow rates ranging from 9 to 28 mm h-1, water and solute movement was quite uniform with dispersivity values ranging from 7 to 53 mm. Under these imposed conditions no preferential flow occurred in the topsoil of the Horotiu soil, at least when averaged over the entire column length of 200 mm.

We found that some adsorption of anions occurred in the allophonic Horotiu soil. This adsorption critically controls the depth of leaching of surface-applied fertilizer. Depending on the depth distribution of active roots, irrigation practices could therefore be optimized to take advantage of this.

The results from the TDR measurements demonstrated the potential of TDR for studying depth-dependent solute transport properties in structured soils. The TDR results suggest that anion adsorption increases from a depth of 5 to 10 cm and then remains about constant. Under the imposed unsaturated flow conditions no preferential flow was found for the soil. However, the dispersivity was found to decrease from a depth of 5 to 10 cm, suggesting that water and solute flow became more uniform with depth. This depth dependency of solute transport properties could not be inferred from the effluent data.

Results from the intermittent leaching experiments showed that there was no effect on solute concentrations in the leachate following no-flow periods. This suggests, again, that water and solute flow in this soil were relatively uniform.


   APPENDIX
TOP
 ABSTRACT
 INTRODUCTION
 THEORY
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 APPENDIX
REFERENCES
 
Symbols used in this manuscript: C, concentration in soil (mol m-3); Cf, flux-averaged solute concentration (mol m-3); Co, time-dependent input solution concentration (mol m-3); Cr, resident soil solution concentration (mol m-3); D, hydrodynamic dispersion coefficient (m2 s-1); d, rod diameter (m); ho, pressure head (m); I, cumulative drainage or infiltration (m); L, column length (m); q, water flux density (m s-1); R, solute retardation factor (unitless); s, rod spacing (m); t, time (s); v, pore water velocity (m s-1); Vf, final reflected voltage at very long time (V); Vo, voltage of incident step (V); z, depth (m); Zo, characteristic impedance ({Omega}); Zu, impedance of TDR system ({Omega}); {lambda}, dispersivity (m); {sigma}a, apparent soil electrical conductivity (S m-1); {sigma}f, final apparent soil electrical conductivity (S m-1); {sigma}i, initial apparent soil electrical conductivity (S m-1); {theta}, volumetric water content (m3 m-3).


   ACKNOWLEDGMENTS
 
This research was supported by the New Zealand Foundation for Research, Science, and Technology. This work was carried out at Landcare Research, Hamilton. The authors thank the anonymous referees for their useful comments and suggestions.


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




This article has been cited by other articles:


Home page
 Vadose Zone JHome page
A. Ritter, R. Munoz-Carpena, C. M. Regalado, M. Javaux, and M. Vanclooster
Using TDR and Inverse Modeling to Characterize Solute Transport in a Layered Agricultural Volcanic Soil
Vadose Zone J., May 12, 2005; 4(2): 300 - 309.
[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 Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Magesan, G. N.
Right arrow Articles by Lee, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Magesan, G. N.
Right arrow Articles by Lee, R.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Magesan, G. N.
Right arrow Articles by Lee, R.
Related Collections
Right arrow Time Domain Reflectometry, TDR
Right arrow Solute Transport Models
Right arrow Soil Physics
Right arrow Vadose Zone Processes and Chemical Transport

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
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome