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 ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bryan, R. B.
Right arrow Articles by Rockwell, D. L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Bryan, R. B.
Right arrow Articles by Rockwell, D. L.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Bryan, R. B.
Right arrow Articles by Rockwell, D. L.
Soil Science Society of America Journal 63:977-987 (1999)
© 1999 Soil Science Society of America

DIVISION S-6-SOIL & WATER MANAGEMENT & CONSERVATION

Automated Microstandpipe System for Soil Erosion Research

Rorke B. Bryana, Richard M. Hawkeb and David L. Rockwellc

a Soil Erosion Lab., Univ. of Toronto, 1265 Military Trail, Scarborough, ON, M1C 1A4, Canada
b School of Earth Sciences, Victoria Univ. of Wellington, P.O. Box 600, Wellington, New Zealand
c Teal Group Corp., 3900, University Dr., Fairfax, VA 22030 USA

r.bryan{at}utoronto.ca


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusions
 REFERENCES
 
The influence of soil water on erosion is well known, but the full effect of variations in water conditions on erosional processes has only recently been recognized. Micro time domain reflectometer (TDR) probes and microtensiometers have provided some precise temporal and spatial data necessary for soil erosion models, but data on local water table dynamics are also needed to explain rill incision and network development processes. An automated multiprobe microstandpipe system we developed provides continuous water table data at spatial and temporal scales comparable to those from microtensiometers and TDR microprobes (temporal precision: 1–3 min). The new system uses small probes etched with open-ended conductors to provide incremental information on water table height with ±0.25 cm resolution. It has been used in a range of soil erosion experiments, one of which is used to demonstrate the system by examining drainage of interrill slopes in response to rill incision. This experiment was carried out in a 10 m by 0.8 m by 0.3 m laboratory flume under simulated rainfall at 43.4 mm h-1 on a 5° slope, using a composite mixture of Arenic Hapludalf sandy mixed mesic and Aquic Hapludalf clayey mixed mesic soils at 4:1 ratio. The microstandpipe system showed sensitive response to a saturated wedge that progressively extended upflume after initiation at the terminal weir. Despite separation between instruments, agreement between the microstandpipe system, microtensiometers and micro TDR probes was good. Rigorous statistical analysis was not possible, but data suggest that temporal agreement of ±5% is realistic. Despite instrumental precision, the expected interrill drainage response to rill incision was not apparent.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Materials and methods
 Results and discussion
 Conclusions
 REFERENCES
 
RECENT SOIL erosion research has been characterized by a trend away from large-scale erosion plot monitoring and toward detailed measurement of constituent processes at progressively smaller spatial and temporal scales. Attempts to use such process measurements in developing reliable physically based erosion models have shown the limitations of available process data (e.g. Nearing et al., 1989; Elliott and Laflen, 1993; Thornes et al., 1996; Morgan et al., 1998). Most soil erosion research has been carried out in a conceptual framework in which soil entrainment and transportation processes are dominated by the critical tractive force exerted by runoff, usually modified or enhanced by rainsplash energy. Many studies have examined shallow-flow hydraulic conditions in sheetwash, rainflow, and rill erosion, with significant progress in measurement methods (Rauws and Govers, 1988; Bryan, 1990; Guy et al., 1990; Parsons and Abrahams, 1992; Nearing et al., 1997; Parsons et al., 1998).

Although soil water content has long been recognized as a significant influence on erosional response, the need for data of comparable precision on soil water dynamics has received much less attention. Until recently, simple gravimetric or volumetric measurements of average soil water content were generally regarded as adequate for soil erosion research, and more sophisticated determination of soil water energetic conditions was usually used only in surface sealing research (Sharma et al., 1981; Römkens et al., 1985, Römkens et al., 1990; Gimenez et al., 1992; Reichert, 1993; Fox et al., 1997). Some early exceptions were studies of the effects of changing matric tension on rainsplash detachment (Cruse and Larson, 1977; Al-Durrah and Bradford, 1982; Schultz et al., 1985).

Soil water plays a complex role in soil erosion beyond its clearly recognized role in infiltration and runoff generation. Recent studies (Govers and Loch, 1993; Bryan, 1996) have shown that small changes in antecedent soil moisture can greatly affect soil resistance during rainstorms, sometimes altering soil loss under controlled conditions by an order of magnitude. In some cases, such changes can change dominant processes from surface rilling to subsurface micropiping (Bryan, 1996). Such observations have drawn attention to the complex interaction between surface and subsurface erosion processes. Except in areas of sodic, swelling clays that favor tunnel erosion or piping (Bryan et al., 1978; Gerits et al., 1987; Torri and Bryan, 1997) studies of surface erosion have been overwhelmingly dominant. But in the past decade several studies have examined subsurface erosion processes in less extreme soil conditions (Howard and McLane, 1988; Stolte et al., 1990; Gomez and Mullen, 1992; Huang and Laflen, 1996; Gabbard et al., 1998; Bryan et al., 1998). These studies have shown the importance of more precise information about soil water conditions at small temporal and spatial scales. Requirements vary with individual study objectives, but include, for example, detailed spatial and temporal data to test the correlation between subsurface water concentration and surface rill networks and data on the localized effect of matric conditions on soil cohesion and entrainment resistance.

During the past decade major advances have been made in instrumentation for soil water measurement. For several years now, scientists conducting rill initiation and surface sealing experiments at the Soil Erosion Laboratory at the Univ. of Toronto have been able to routinely make several different measurements of soil water conditions at small temporal and spatial scales. These include microtensiometers and micro TDR probes (Fox, 1995; Fox et al., 1997; Hawke, 1997) designed and built in the laboratory, which closely parallel instruments developed elsewhere (e.g. Roth et al., 1991; Topp et al., 1992; Heimovaara, 1993; Heimovaara et al., 1993). These instruments have considerably enhanced understanding of links between soil water conditions and erosional response. But experiments on the interaction of seepage and runoff tractive forces in rill initiation (Bryan et al., 1998) have demonstrated the need for additional instrumentation. Howard and McLane (1988) and Gomez and Mullen (1992) showed that water table or perched water table development can produce seepage forces adequate to trigger rill incision. As seepage forces are controlled by the hydraulic gradient, information on the evolving water table geometry is essential to identify the existence of critical seepage forces. In laboratory flume experiments, water tables can be quite precisely controlled, but ultimately relationships must be tested in field situations where such control is seldom possible, or appropriate.

This paper describes an automated microstandpipe system, recently developed at the Toronto Soil Erosion Lab, to provide parallel information about soil water table dynamics at temporal and spatial scales similar to those possible with microtensiometer and TDR units.

Even where seepage forces are inadequate for entrainment, complete or partial water table development can significantly reduce the resistance of soil to entrainment by tractive forces in surface flow (Bryan and Rockwell, 1998). Information on water table dynamics is therefore necessary for accurate interpretation of soil erosion patterns. In soil erosion research on agricultural fields or undisturbed soils, localized or extensive perched water table development is often encountered, sometimes very close to the surface, and can become a dominant influence on runoff and erosional response. Perched water table development is also frequently encountered in laboratory flume experiments where uniform, homogeneous packing of large soil samples is difficult. The data provided by microstandpipes do not eliminate this problem, but allow it to be monitored and its effects identified.


   Materials and methods
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
Results and discussion
 Conclusions
 REFERENCES
 
The standard measurement components of the system are microstandpipe probes linked to a data logger. In all experiments conducted so far we used either Campbell CR 10 or Campbell CR7 data loggers (Campbell Scientific, Inc., Chatham, ON). The only practical limitation for the system was the number of circuits available on the data logger. The complete instrument consists of two portions joined together. The upper portion, which stands above the soil is the electronics board. The lower section, which is an etched board 128 mm long and 9 mm wide, is the section that is buried in the soil, and measures incremental changes in water table height. The probe was etched with 12 open-ended horizontal conductors at 5-mm increments along each side of the probe. Vertical conductors connected each horizontal conductor to measurement circuits on the electronics board. A simplified probe diagram is shown in Fig. 1 , and one side of the complete instrument is shown in Fig. 2 . The vertical conductors were solder masked to provide electrical insulation, as conductivity measurements were made only between horizontal probe conductors and the ground via the conductivity of water surrounding the probe. The electronic components of the circuit board were shrink-wrapped for protection against moisture intrusion.



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 1 Simplified diagram of microstandpipe showing the main design elements

 


View larger version (33K):
[in this window]
[in a new window]
  		Fig. 2 Microstandpipe probe showing one side of circuit board

 
Probe electronics comprised a "fuzzy logic" digital front end as a conductivity sensor with an analog summation device, inverter, and voltage follower as the output stage. The complete circuit diagram is too complex to reproduce at publication scale, but a simplified circuit diagram identifying the major features is shown in Fig. 3 (copies of the complete circuit diagram can be obtained from the authors). Each horizontal conductor on the probe is connected to the input of a CMOS MC14584BCP Schmitt trigger inverter (see Fig. 2), designed to compare the impedance of the water:probe conductor interface with a fixed resistance of 2.2 M{Omega}. If the interface impedance is significantly below 2.2 M{Omega}, inverter output changes by 0 to +5 v. Each inverter was linearly summed via 10-K resistors with two op-amps of an LM324N integrated circuit (Fig. 2) connected as gain stages of x -0.02, inverted by a third of gain x -1, and buffered by a fourth connected as a voltage follower. With the gains selected, the voltage follower output increases by 100 mv for each horizontal conductor immersed in water.



View larger version (12K):
[in this window]
[in a new window]
  		Fig. 3 Simplified diagram showing the main components of the microstandpipe circuitry. Detailed circuit diagrams will be made available on request

 
The circuit design used has some advantages over previous designs (e.g., MacVicar and Walter, 1984) as the probe output voltage calibration was entirely independent of the absolute value of the conductivity of the water probe interface circuit between 0 and {approx}2.2 M{Omega} impedance. This is important as it ensures that output is not affected by changes in water conductivity during experiments.

The depth probe was installed in a 130 mm long by 9.5 mm (i.d.) aluminum tube, which was perforated along its complete length with alternating 4.6-mm diam. holes to permit free water access from surrounding soil. A metal disc sealed to the top of the tube with silicone prevented entry of water from the surface. The tube was covered with coarse filter cloth to prevent blockage by soil, and was installed in an hole augured vertically into the soil. Considerable care in installing the tube, particularly with regard to packing soil around the tube, is essential for accurate results. Detailed procedures were discussed by Rockwell (1995). The electronics sector protruded above the surface, and was linked by a thin three-wire lead to the datalogger outside the flume.

Each microstandpipe allowed spot measurement of the height of the water surface within the aluminum tube to ±2.5 mm. The only theoretical limitation to the spatial density of information is the number of microstandpipes that can be conveniently linked to a datalogger. In experiments so far 10 probes have been simultaneously linked with a Campbell CR10WP datalogger. Readings for each microstandpipe were taken at intervals of 1 min or greater. The unit has been used only for laboratory experiments, but is also suitable for field experiments.

Experimental Design
The microstandpipe system has been used in laboratory experiments on rill initiation, rill network development, and surface crusting, producing consistently useful results (Hawke, 1997; Bryan and Rockwell, 1998). The results of one experiment are described here to demonstrate use of the instrument in conjunction with microtensiometers and micro TDR probes. The substantive objective of the experiment was to examine the effect of a headward-migrating rill knickpoint on drainage conditions on surrounding interrill slopes. The experiment was developed following observations by Bryan and Poesen (1989) that knickpoint scour in rill channels on sealing soils can significantly change local water conditions, both beneath the rill channel and on surrounding interrills. Within the rill channel, scour in the knickpoint plunge pool penetrated the channel seal, causing local increase in infiltration, and reduced channel discharge. As knickpoints migrated upslope, visible drainage occurred on surrounding interrills, with clearly reduced discharge input to rill channels.

The rational explanation for interrill drainage appears to be a decline in water table height under interrills, linked to reduced crest height of flow in the rill. Local reduction in water table height close to the rill would also increase hydraulic gradients under interlines, resulting in increased seepage pressure in rill channel walls and, potentially, to channel wall failure. Obviously a time lag, depending on saturated hydraulic conductivity and hydraulic gradients, would be expected between knickpoint incision and consequent effects at any location. The object of the present experiment was to determine if the instrumentation available would permit identification of the postulated effects of knickpoint migration.

The experiment was carried out in a 10-m-long by 0.8-m-wide by 0.3-m-deep laboratory flume set at an angle of 5°. Details of the flume have been described in previous papers (Bryan, 1990; Slattery and Bryan, 1992). A composite soil sample was prepared from the A horizons of two soils from southern Ontario, Canada, the Pontypool loamy sand (sandy mixed mesic Arenic Hapludalf) and the Peel clay (clayey mixed mesic Aquic Hapludalf). The Pontypool developed over coarse, cross-bedded glacifluvial deposits of the Oak Ridges kame moraine, while the Peel clay developed over lacustrine clays deposited in short-lived ponds that immediately followed Wisconsinan ice retreat.

The soil samples were mixed in 80:20 ratio, corresponding to the mixture found by Bryan and Poesen (1989) to be particularly sensitive to surface sealing processes. Basic soil properties are shown in Table 1 . Air-dry soil was sieved through an 8-mm sieve and placed in the flume to an average depth of 9 cm on an aluminum mesh base covered by landscaping weed control fabric. The sample was compacted with a bevelled roller to an average bulk density of 1.23 Mg. m-3 (SD 0.034), and shaped to ensure flow concentration in the centre of the flume, with interrill side slopes at 2.5°. An intensively instrumented section was positioned between 2.25 and 3.75 m above the terminal weir. Microstandpipes were placed vertically and both microtensiometers and micro TDR probes placed horizontally. Microtensiometers were 11 mm long and 2.2 mm (o.d.), while the micro TDR probes had three 1.6-mm diam., 100-mm-long stainless steel probes, 16 mm apart. Instruments were physically positioned as close as possible to one another, to provide the best possible synchronicity of data. However, some spacing was necessary to prevent interference between instruments, and to avoid cables running across the central rill. Details of the instrumental layout in the experimental section are shown in Fig. 4 . Microstandpipe readings were taken at 1-min intervals, and microtensiometer and micro TDR probe readings at 3-min intervals. The precise relative positions of the instruments, shown in a typical cross section (Fig. 5) , are important in interpreting experimental data. As explained above, the objective of this design was to monitor changes in soil water conditions as the rill channel incised and migrated up-flume through the instrumented section.


View this table:
[in this window]
[in a new window]
  		Table 1 Characteristics of the composite sample of two southern Ontario, Canada, soils used in the runoff experiments. The sample was composed of a 4:1 mixture of Pontypool loamy sand (sandy mixed mesic Arenic Hapludalf) and Peel clay (clayey mixed mesic Aquic Hapludalf)

 


View larger version (13K):
[in this window]
[in a new window]
  		Fig. 4 Plan view of the intensely instrument of the 10-m laboratory flume used in the rill incision experiment. Note that clusters of microtensiometers and TDR probes are superimposed at the same location rather than spaced precisely as shown

 


View larger version (13K):
[in this window]
[in a new window]
  		Fig. 5 Cross section through typical portion of the instrumented flume section showing the spatial relationship between different instruments and the incised rill

 
The experiment was carried out under simulated rainfall of average intensity 43.4 mm h-1, produced by Lechler 459.008.30BK nozzles (Spraco, Inc., Nashua, NH) set at a mean height of 5.5 m above the flume surface, with a pressure-control valve for each nozzle set to 24.13 kPa. Rainfall distribution was monitored by 12 manual raingauges set at 1-m intervals on each side of the flume. The experiment was carried out on initially air-dry soil with mean volumetric antecedent moisture content of 40 g kg-1 (SD 14).


   Results and discussion
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
Conclusions
 REFERENCES
 
The composite soil sample used was unusually sensitive to surface sealing. Local ponding occurred swiftly under the high-intensity rainfall, producing localized runoff at several locations after 7 min.

Flow was continuous across the entire flume. Discharge started at the weir after 13 min. Discharge (measured manually at the weir) rose rapidly and equilibrated near 5.2 L min-1 after 90 min. Flow concentrated rapidly in the central depression, almost immediately exceeding recognized threshold hydraulic values for rill initiation (Govers, 1985; Bryan, 1990; Slattery and Bryan, 1992).

Significant incision with formation of a vertical knickpoint, 1.5 cm wide and 16 cm deep, occurred 1.86 m above the terminal weir after 47.5 min. This migrated up-flume at a mean rate of 0.13 m min-1. The knickpoint widened and deepened to a final mean width of 10 cm and depth of 2.13 cm. Rill walls and the retreating knickpoint were both undercut, reflecting the greater strength of the surface seal compared with subsurface soil (Fig. 6) . Rill widening occurred primarily by rill wall scour and periodic collapse, which was enhanced by undercutting from meandering flow and by tension crack formation parallel to the rill walls (Fig. 6). Rill wall collapse did not coincide precisely with knickpoint migration but occurred after an apparently irregular time lag. Wall collapse produced irregular inputs of sediment to the rill channel and caused significant perturbations of rill flow and sediment transport.



View larger version (150K):
[in this window]
[in a new window]
  		Fig. 6 Incised rill knickpoint migrating up-flume through the instrumented section, with microstandpipes visible. Note also evidence of tension cracking and incipient collapse of rill walls

 
In the later stages of the experiment flow transport capacity was frequently exceeded, causing local channel aggradation, as shown in Fig. 5. These interacting processes produced considerable variations in knickpoint dimensions and migration rates, but migration through the instrumented test section occurred between 53 and 68 min after the start of the experiment. The rill remained as a single channel through the test section, then bifurcated in the upper part of the flume.

Microstandpipe response during the experiment (Fig. 7) was a clearly time lagged, with individual instruments responding progressively later with increased distance from the weir (except Units 6a and 7a). Other notable features were the similarity in indicated rates of water table rise and final indicated water table heights, and the occurrence of periodic stillstand. Several factors influenced response patterns, but the dominant one appears to have been development of a saturated wedge in the soil sample at the terminal weir that extended progressively up-flume. There was no direct indication of wedge initiation at the weir, but the time-lagged response for Microstandpipes 1a to 6a coincided almost perfectly with the progression of a saturated wedge up-flume from 2.25 m to 3.5 m, calculated with a 0.30 cm min-1 mean rate of water table rise (Table 2) . We can discern no apparent reason for the atypical response of Unit 7a, which preceded the calculated time by 11 min, but this could reflect soil packing irregularities near this instrument. Responses on the opposite side of the flume were also delayed: by 7 min (Unit 8b), 13 min (Unit 9b), and 19 min (Unit 10b). The main apparent reason was interference with up-flume progression of the saturated wedge on that side of the flume by the assymetrically meandering rill channel (Fig. 8) .



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 7 Measured response of microstandpipes during 120-min duration experiment under simulated rainfall at 43 mm h-1 intensity. Note time-lagged progression up-flume related to propagation of a saturated wedge from the terminal weir

 

View this table:
[in this window]
[in a new window]
  		Table 2 Summary of critical features of mirostandpipe response during rill incision experiment in Canada

 


View larger version (126K):
[in this window]
[in a new window]
  		Fig. 8 Aerial view of the instrumented flume section in the latter stages of the rill incision experiment showing effect of rill meandering on Microstandpipe 8b

 
Calculation using the mean water table rise rate of 0.30 cm min-1 (SD 0.10) predicted up-flume wedge progression from 2.25 m accurately, but this figure could not be used for wedge progression from the terminal weir. This would have required wedge development before the experiment started. Only a very approximate solution can be suggested from the data available. Microstandpipe rise above 2.25 m occurred while runoff discharge was approaching equilibrium, with a mean infiltration between 50 and 90 min of 1.23 L min-1.

Infiltration earlier in the experiment was much higher, with a mean value of 3.99 L min-1. Adjustment of the mean rate of water table rise to reflect relative infiltration yielded a mean rate of 0.97 cm min-1 for the first 50 min. Use of this figure would indicate saturated wedge initiation after {approx}30 min of rainfall.

While the rates of water table rise indicated by different microstandpipes were similar, the greatest similarity was during the initial stages of rise before interruption by stillstand (Table 2), with a mean rate of 0.49 cm min-1 (SD 0.09). Use of this figure would indicate wedge initiation after 10 min of rainfall, which agrees quite well with calculated time required to saturate the lowest 2 m of the flume, using measured infiltration rates, bulk density, and saturated volumetric water content (12.5 min).

Periodic stillstand interrupted water table rise at all microstandpipes, but the timing, duration, and water table height at which this occurred varied considerably (Fig. 6). Three factors were potentially involved: water impedance due to irregular packing of the soil sample; impedance due to air entrapment beneath a surface seal; and short-term drainage of the surrounding soil due to rill incision, as discussed in the experimental design.

Although the soil was placed in the flume in layers, compaction was with a surface roller when the complete sample was in place. This should have minimized compact layering within the sample, but coincidence of stillstand at a water table height of 4.5 cm in Microstandpipes 2a, 5a, and 9b may indicate impedance. Air entrapment has been linked to formation of vesicles in sealed and desert soils (Bryan, 1973; McFadden et al., 1998), and in soil column experiments Bryan and Rockwell (1998) showed the effect of such entrapment in impeding water table rise. The effect of surface sealing would be expected as the water table approached the surface. Microstandpipes 1a, 3a, 4a, and 7a all showed brief intervals of stillstand, possibly caused by sealing and air entrapment, and the final height of 6 cm for 5a and 6a was possibly limited for the same reason.

There was no evidence of water table decline caused by rill incision during the experiment, and it was not possible with any confidence to link the stillstand features shown in Fig. 7 to knickpoint migration. As noted above, the maximum depth of rill incision was just over 2 cm, and this clearly produced insufficient hydraulic head for rapid drainage of surrounding interrill slopes.

Table 2 shows a rather close agreement between the final water table height indicated by the different instruments, with a mean value of 6.9 cm (SD 0.49). The overall interrill surface height declined during the experiment by a mean value of 0.36 cm (SD 0.09) due to hydrocompaction, surface sealing, and localized splash and rainwash erosion. Because of this surface drop and the precise positions of the microstandpipes on interrill slopes (Fig. 5), final indicated water table heights coincided closely with actual interrill surface heights. Precise coincidence of the water table with the interrill surface was precluded by proximity of the microstandpipes to the incised rill (1-5 cm distance).

Responses of microtensiometers and TDR probes are shown together with the responses of the nearest microstandpipes in Fig. 9a to 9d . In each case, two or three microstandpipes were almost equidistant from the microtensiometer and TDR cluster and responses of each have therefore been plotted.






View larger version (98K):
[in this window]
[in a new window]
  		Fig. 9 (a) Response patterns of microtensiometer and TDR cluster at 2.5 m above the weir, and those of nearest microstandpipes. (b) Response patterns of microtensiometer and TDR cluster at 2.875 m above the weir and those of nearest microstandpipes. (c) Response patterns of microtensiometer and TDR cluster at 3.125 m above the weir and those of nearest microstandpipes. (d) Response patterns of microtensiometer and TDR cluster at 3.5 m above the weir and those of nearest microstandpipes

 
In comparing the records it is important to emphasize that different properties of soil water conditions are recorded by the three instruments. Water could only enter microstandpipe tubes when the pressure gradient from the surrounding soil was positive, so these instruments responded only after water table formation at the base of the soil sample. On the other hand, both microtensiometers and TDR probes respond to any change in water conditions, so traces started to rise as soon as percolating water reached the instruments. The best basis for comparison is therefore the time at which each instrument indicated saturated conditions in the adjacent soil.

For microtensiometers, this indication was the transition from negative to positive matric tension, and for TDR probes, an essentially constant water content. In comparing responses, it is important to note the relative heights of instruments above the base of the sample, as shown in Fig. 5. So microtensiometers at 2, 5, and 8 cm below the surface at the edge of the flume corresponded to heights of 7, 4, and 1 cm above the base of the sample at each microstandpipe.

Table 3a shows the time from the start of the experiment at which each instrument indicated saturated conditions in the adjacent soil. Comparison is complicated by the separation between instruments and progressive time lag in water table rise up-flume. There is no basis to choose which of the nearly equidistant microstandpipes best reflected conditions at adjacent instrument clusters, so mean values have been used in Table 3a.


View this table:
[in this window]
[in a new window]
  		Table 3 Time-to-saturation from beginning of experiment as indicated by various instruments recording runoff characteristics

 
Precise comparison is also complicated as the microtensiometers and TDR probes integrated soil water conditions for different soil volumes. It was also difficult to establish equilibration levels for TDR probes. In Table 3a, mean times to saturation measured for two or, where possible, three instruments, are shown. Too few instruments were available to carry out rigorous statistical analysis. However, if the standard deviations for instrument groups are considered (Table 3b), they indicate an average agreement of ±5 min, or {approx}4% of the duration of the experiment. Closer agreement can be achieved when instruments can be placed very close together, and the precise position of each is precisely controlled, as in the soil column experiments reported by Bryan and Rockwell (1998). However, it would be unrealistic to expect such close agreement in large scale erosion experiments with more heterogeneous soil and microtopography conditions.

The substantive objective of the experiment described was to examine evolving soil water conditions in interrill slopes during rill incision and migration. Despite the comparatively high density of instrumentation, it was not possible to detect significant drainage of interrill slopes as knickpoints migrated upflume, as observed by Bryan and Poesen (1989). The only apparent explanations, apart from inadequate instrumentation, are differences in soil bulk density, and rill incision depth relative to overall soil sample depth. In Bryan and Poesen's (1989) experiments, soil was not compacted and therefore was presumably of significantly lower bulk density, though this was not reported. Rill incision depth in those experiments was similar, but overall sample depth was only 3.5 cm, so the actual volume of water to be removed for complete drainage of interrill slopes was much smaller.


   Conclusions
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusions
REFERENCES
 
Several recent studies have noted the influence of a water table or perched water table on soil erosional response. Water tables can strongly influence soil erodibility and, in extreme cases, can alter the nature of the dominant erosional processes. Information about water table dynamics is therefore a useful addition in experimental analysis. The microstandpipe system described provides a relatively simple, sensitive, and inexpensive method for monitoring water table dynamics during laboratory soil erosion experiments. Comparison with near-synchronous microtensiometer and TDR data encourages confidence in the response patterns shown by the instruments. However, interpretation of results require precise information on instrument location and considerable care must be exercised in installing probes and particularly in packing surrounding soil. When used appropriately, the microstandpipe system can significantly augment available information on soil water dynamics. It has enhanced interpretation of results in a range of laboratory erosion experiments. It is sufficiently robust for field use, but the precision of information about placement necessary for accurate interpretation of response patterns means that laboratory use is more realistic.

The substantive objective of the experiment described was examination of the drainage of interrill slopes due to rill incision. Despite the precision of the instrumentation available, no significant drainage effects could be demonstrated, indicating that observed results in some previous experiments were probably artifacts of the specific experimental design used.Parsons Abrahams Luk 1990

Received for publication April 13, 1998.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results and discussion
 Conclusions
 REFERENCES
 





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 ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bryan, R. B.
Right arrow Articles by Rockwell, D. L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Bryan, R. B.
Right arrow Articles by Rockwell, D. L.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Bryan, R. B.
Right arrow Articles by Rockwell, D. L.

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