| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Institute of Soil and Water Conservation, 26 Xinong Rd. Yangling, Shaanxi, 712100, People's Republic of China
b USDA-ARS National Soil Erosion Research Lab., 1196 SOIL Bldg., Purdue Univ., West Lafayette, IN 47907-1196 USA
chihua{at}purdue.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMaterials and methodsResults and discussionConclusionsREFERENCES |
|---|
Abbreviations: WEPP, Water Erosion Prediction Project
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionConclusionsREFERENCES |
|---|
An early conceptual model of the effect of sediment concentration in the eroding water on erosion processes was proposed more than 50 yr ago by Ellison (1947) and Ellison and Ellison (1947a, b). This model considers sediment in the water as an abrading agent and the detachment capacity of the flowing water increases as the sediment content is increased. On the other hand, the flow also has a limited sediment carrying capacity. Therefore, clear water has a maximum transporting capacity, minimum detaching capacity, and causes very little erosion. On the other hand, when the water is full of sediment, it has a maximum detaching capacity, minimum transporting capacity, and again very little erosion. Maximum erosion occurs when the flow contains just enough abrasive sediment to detach as much soil as the flow will carry.
Ellison also laid the foundation for an erosion process model by dividing soil erosion processes into detachment and transport by raindrop impact (Dr and Tr) and detachment and transport by surface flow (Df and Tf). On the basis of Ellison's work, Meyer and Wischmeier (1969) further postulated that sediment delivery (qs) was limited by either the detachment rate (Dr + Df) or the transport capacity (Tr +Tf), depending on which had a lesser value. In 1972, Foster and Meyer proposed a coupling concept to link flow detachment to sediment transport capacity, Tc (Foster and Meyer, 1972). The detachment-transport coupling concept and the separation of rainfall-dominated interrill and flow-dominated rill erosion processes (Meyer et al., 1975) became the principal erosion concept of the process-based Water Erosion Prediction Project (WEPP) model (Nearing et al., 1989).
The run-on water effect in the WEPP model is accounted for by the detachment-transport coupling concept. This model states that the detachment or deposition rate, E, is proportional to the difference between transport capacity, Tc and sediment load, qs, or:
 |
is a rate control constant. The sediment detachment-transport coupling is also widely known as the sediment feedback relationship in a different form:
 |
is the detachment capacity. The sediment feedback relationship implies that runoff with low sediment load will have a high detachment rate. As the sediment load approaches the transport capacity, erosion rate becomes negligible. Deposition occurs when the sediment load exceeds the transport capacity. In this manuscript, the sediment detachment-transport coupling concept is interpreted beyond the rill channel for erosion processes in general (Huang and Bradford, 1993). Despite differences in Ellison's original hypothesis and subsequent proposition by Foster and Meyer on how sediments in the runoff water would affect erosion processes, little work has been done to test the validity of these concepts, especially on the detachment and transport interaction. Nevertheless, field studies conducted on the steep hill slopes in China demonstrated the significance of run-on water and sediment effects on downslope sediment production.
Chen (1992, 1993) divided the loessial hill slopes of the Loess Plateau of China into up-slope, midslope, and downslope sections and studied the effect of runoff from upper slopes on erosion and sediment transport processes on a downslope section. Chen pointed out that an increase in runoff sediment concentration from upper slopes resulted in a decrease of erosion downslope. Zheng and coworkers (Zheng, 1997; Zheng et al., 2000) established different sizes of runoff plots on the basis of the vertical distribution of sheet, rill, and shallow gully erosion zones on a loessial hill slope. Their results showed that increased runoff from up-slope areas resulted in an increased erosion downslope. But, with an increased up-slope sediment concentration, erosion from the downslope segment was reduced.
Since slopes are steep on the loessial hill slopes of China, ranging from 3 to 5° at the top to 32 to 35° at the bottom, the applicability of these results to more gently sloping landscapes is not clear. In addition, conditions in the field are very complex and it is difficult to identify quantitatively how runoff and sediment from up slope affects the downslope sediment transport regime under different rainfall, runoff, slope steepness, and surface hydrologic conditions.
Recent laboratory studies showed that the surface hydrologic condition, specifically near-surface hydraulic gradient, had a profound effect in the dominant erosion process and total sediment delivery (Huang and Laflen, 1996; Gabbard et al., 1998; Huang, 1998). Rainfall simulation experiments were conducted on a clay loam soil with a 5-m-long soil box that can be set to different seepage and drainage gradients. Under seepage or exfiltration conditions, the surface became more erodible and rilling was the dominant erosion process. The sediment regime became transport limiting. Under drainage (or infiltration) conditions, surface rilling did not occur and erosion was caused by an interrill-type surface scour. Sediment delivery under drainage conditions was under a detachment limiting regime. This detachment limiting regime was supported by a data set that showed similar sediment delivery under 10 and 15% slopes because of an increased drainage despite the increased transport capacity of the flow from a steeper slope (Huang, 1998). These studies show the importance of considering surface hydrologic conditions in quantifying soil erosion processes.
Recently, the single 5-m-long soil box was expanded to a dual box system to further study erosion process and sediment regime for a hill slope segment. A 1.8 m sediment source (or feeder) box was constructed to provide sediment input to the 5-m test box. The dual box system provides a good way of studying effects of runoff and sediment from up-slope areas on downslope erosion processes. Huang et al. (1999) used the dual-box system to evaluate erosion processes in the test box to test erosion model concepts. On the basis of the relative values of sediment delivery from the feeder and test boxes, they proposed five scenarios ranging from deposition-dominated to transport-dominated sediment regimes. They further showed that for a 5% slope under artesian seepage or a 10% slope under free drainage condition, the runoff from the feeder box caused additional sediment delivery in the test box, indicating a transport-dominated sediment regime. At 5% slope under drainage condition, deposition occurred at low rainfall intensities. Increases in either slope, rainfall intensity, or soil erodibility shifted the dominant erosion process from detachment limiting to transport limiting.
Motivated by previous field observations in China and the capability of a dual-box system to simulate hill slope erosion processes, we designed a laboratory study to quantify the up-slope run-on water and sediment effects on downslope erosion processes. The run-on effects were tested for different surface conditions, i.e., rainfall intensity, slope gradient, and seepage–drainage, in the test box. During the experiment, the sediment concentration from the feeder box and vertical drainage conditions were varied, while maintaining a constant level of runoff, to create a range of up-slope boundary conditions for the test box. The subsequent response of the test box can be used to evaluate the erosion process and transport capacity model concept. Results of this study will further the understanding of soil erosion processes and provide data for the development of a more accurate process-based erosion model.
|
Materials and methods |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionConclusionsREFERENCES |
|---|
Experimental Setup
The study was conducted on a dual-box system consisting of a 5-m-long test box and a 1.8-m-long feeder box (Fig. 1)
. Both boxes were 1.2 m wide and 0.3 m deep. These two boxes can be connected by a connecting piece to feed the runoff from the feeder box to the upper end of the test box. When these two boxes disconnected, runoff samples can be collected separately from each box. The connection and disconnection can be done quickly without stopping the rain.
|
The feeder box could be set to any slope steepness by hoisting up its up-slope end using a pulley-chain system. This soil box had 27 6-mm holes at the bottom for seepage or drainage control. In this study, the feeder box was free drained.
For both soil boxes, the depth of soil was approximately 25 cm with a 2-cm layer of sand at the bottom to promote bed saturation from the watering system. These two boxes were placed under two sets of oscillating nozzle, programmable rainfall simulation troughs (Foster et al., 1979). There were five troughs spaced 1.07 m apart for the test box and three troughs spaced 0.84 m apart for the feeder box. Each rain trough had two Veejet nozzles (part No. 80100, Spraying System Co., Wheaton, IL) spaced 1.07 m apart. Vertical distance between the nozzle and soil surface was approximately 2.8 m for the test box and 2.4 m for the feeder box. During the rainfall simulation, the nozzle pressure was kept at 41.4 kPa (6 psi). These two sets of rainfall simulators could be controlled independently and set to a selected rainfall intensity, ranging from 25 to 200 mm h-1, by programming the oscillating frequency of the nozzles.
Soil Box Preparation
Preparation of soil boxes included adding fresh soil collected from the field, breaking up clods to 3- to 4-cm sizes, and smoothing out the visual irregularities on the surface by hand and with a rake. Both boxes were prepared the same way. Once the boxes were prepared, these two boxes were set to their level position and rained on for 30 min at 50 mm h-1 intensity. This initial rain, as part of the soil box preparation procedure, was used to wet down soil surface and to consolidate the loose aggregates to form a wet surface seal. The initial rain also reduced surface variability from preparation.
After surface water from the initial 30 min 50 mm h-1 rain event had infiltrated, the soil box was set to the selected slope steepness and hydrologic condition. Under the drainage condition, the soil box was free drained. Under the seepage condition, water tube height was set to 20 cm above the soil surface. The water system was applied continuously to maintain constant water level before and during the erosion run. Under the 20-cm-high water tube setting, water seeped out of the soil and flowed across the surface. The background seepage rates were measured before and after the erosion run.
Experimental Procedure
The erosion run was conducted 24 h after the initial rain. The test box was set to the selected rainfall intensity, slope steepness and hydrologic conditions (Table 1)
. The feeder box was set to 10% slope and exposed to 150 mm h-1 rain throughout this study. The sediment content in the feeder box was varied by progressively covering portions of the surface with landscape fabric that allowed water passage but prevented direct raindrop impact.
|
After each run, 10 to 20 mL of saturated alum solution were added to the samples to precipitate suspended sediments. After settling overnight, the excess water was poured from buckets and the sediment was washed into 1-L bottles. The bottles were decanted of excess water and placed in ovens at 105°C for at least 24 h or until the samples were dried. Dry weight was then taken to calculate the sediment transport rate and concentration.
After the run, these two boxes were drained and dried under two fans. After the soil surface was dried, gravel and coarse materials on the surface were scraped off to prevent their accumulation on the surface. The soil surface was then turned over with a shovel to a depth of 15 cm. Fresh air-dried soil was added to the surface to replace lost soil during preparation for the next run. Each run was replicated at least twice.
Runoff and sediment rates were averaged from six samples, four before connection and two after disconnection, for both test and feeder boxes separately, and from four samples when two boxes are connected. These average runoff and sediment results are presented in Table 2 .
|
, simultaneous erosion and deposition, deposition = erosion; 
, equilibrium, no effects from feeder water and sediment; The different sediment process scenarios are based on the dynamic balance of three simultaneous processes: detachment, deposition, and transport. Changes from Scenario 1 to 5 indicate the shift in the dynamic balance from a deposition-dominated (Scenario 1) to detachment-dominated (Scenario 3) and finally a transport-dominated (Scenario 5) process regime. Scenario 2 is a transition between deposition and detachment-dominated processes, and likewise, Scenario 4 a transition between detachment and transport (Huang et al., 1999).
On the basis of those five scenarios, we can analyze the effects of up-slope run-on water and sediment addition on the sediment regime of a downslope segment under different rainfall intensities, slope steepness and surface hydrologic conditions. In addition, the value of SFT – SF is the net sediment detachment (>0) or deposition (<0) on the test box with the feeder sediment input. Since ST is the sediment production from the test box without the feeder input, the difference between SFT – ST is the total sediment entrainment caused by the addition of run-on water, or the run-on water effect. The difference between total sediment entrainment caused by the run-on water, SFT – ST, and the sediment already contained in the run-on water from the feeder box, SF, i.e., SFT – ST – SF or SFT – (SF + ST), is the run-on sediment effect on the test box. The run-on sediment effect is a measure of how the up-slope sediment input can affect the downslope erosion processes. Values of the run-on water and sediment effects and the sediment regime scenario were tabulated in Table 3 .
|
|
Results and discussion |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionConclusionsREFERENCES |
|---|
Under the seepage condition, rilling started from the very beginning of the run. In some cases, the background seepage flow imposed 24 h preceding the run caused one to two small rills, 3 to 4 cm wide and 2 to 3 cm deep, in the lower 1- to 2-m portion of the test box, although the sediment content in the seepage flow was negligible. During the run, it was obvious from the muddy nature of the runoff that sediment contents were much greater than those from the drainage condition. The rilling pattern was maintained throughout the run, and after the run, the rills were roughly 2 to 4 cm deep and 12 to 24 cm wide. In addition to rills, the after-run surface of the test box also showed an overall but discernible lowering, indicating the occurrence of severe erosion. Observations during the run can help us explain the sediment data trends in a later section.
The seepage and slope effects are similar to those observed by Gabbard et al. (1998) despite differences in soil and run procedures. They used a clay loam soil, which was much less erodible as compared with the silt loam soil used in this study. They also added inflow water as high as 38 L min-1 to plots half as wide (i.e., 0.6 m) while this study used the full box width with a highest runoff value of 1300 L h-1 or 22 L min-1.
Temporal Effects on Surface Condition and Sediment Delivery
Different processes during the experimental run, which affected the surface condition and consequently, sediment delivery, were observed in the data shown in Table 3. Sediment delivery from the test box, both with and without the feeder input, showed three different trends: steady, increasing, and decreasing, as the hour-long run progressed (Fig. 2)
. An apparent steady state sediment delivery was maintained during the run when an interrill-type surface scouring was the dominant process. Data showing the steady state trend were obtained from drainage conditions at 5% slope under both 50 and 100 mm h-1 rainfalls and at 10% slope under 50 mm h-1 rainfall. At 10% slope and 100 mm h-1 rainfall, the increase in sediment delivery with run time was associated with the shift from an initial interrill-type process to severe rilling toward the end of the run. Runs made under the seepage condition showed a decreasing sediment delivery as the runs progressed. This can be attributed to the rapid loss of easily erodible materials on the surface from severe rilling, making the surface less erodible with time. Similar reduction in soil erodibility after a sequence of rainstorms was also observed in a prior study (Huang, 1998).
|
Process Scenario in the 5-m Test Box
Runoff data shown in Table 2 indicated a reasonable mass balance between total runoff from both boxes separated (RF + RT) and the runoff from the test box with feeder input (RFT). On the other hand, the sediment process scenario presented in Table 3 showed situations ranging from Scenario 1 to 5.
Under drainage and lower slope (5% slope) conditions, erosion process in the test box changed from the deposition-dominated process (Scenario 1) to transport-dominated (Scenario 5) as sediment concentration from feeder box was decreased. When slope was increased to 10% or when the hydrologic condition was changed to seepage condition, the erosion process became transport-dominated (Scenario 5). With the greatest feeder sediment input (i.e., 0% cover) to the test box at 5% slope under drainage condition, an increase in rainfall intensity from 50 to 100 mm h-1 changed the dominant erosion process from deposition to detachment. The rainfall intensity effects on sediment process scenario were also demonstrated in a separate study in which the rainfall intensity on the test box was varied from 25 to 150 mm h-1 while the feeder sediment input was maintained constant (Huang et al., 1999).
Under seepage conditions, the soil strength was low and sediment was easily detached and transported. The water from the feeder box induced additional sediment delivery from the test box and the additional sediment delivery was greater than those under drainage conditions for the same rainfall intensity and slope steepness. Increases in slope steepness, rainfall intensity and soil erodibility, from changing drainage to seepage conditions, increased sediment delivery from the test box and caused the shift in erosion process regime from Scenario 1 to 5 (Huang et al., 1999).
Run-On Water Effects and Dynamic Equilibrium
Table 3 showed that under drainage conditions, sediment deliveries from the 5-m test box were somewhat constant with or without feeder input except under 10% slope and 100 mm h-1 rainfall intensity. Even the variation in feeder sediment content did not affect the sediment output from the test box, SFT, as well as the run-on water effect, SFT – ST. This indicates that the sediment regime has reached a dynamic equilibrium for detachment, transport and deposition processes in the 5-m-long test box. As the run-on water contains more sediment than the amount that can be transported, deposition occurs. As the feeder sediment input is reduced, the flow causes additional detachment to fill the transport capacity.
At 10% slope and 100 mm h-1 rainfall, rills developed during the run, causing an increased total sediment delivery from the text box. Despite the increase in sediment delivery associated with the rilling process, a near-constant value of sediment entrainment from different levels of feeder sediment content indicate a dynamic response to an equilibrium regime in the 5-m test box.
Under seepage conditions, soil strength was low and sediment was easily detached and transported, and rill erosion became dominant. At the beginning of the rain, the sediment delivery was at its greatest value. As the run progressed, flow was concentrated in the rill channels and the decrease in sediment delivery indicates that soil erodibility has decreased, possible due to the combination of raindrop impact, surface armoring and removal of easily erodible fractions. In fact, the change in soil erodibility was so great making it difficult to show the characteristic run-on water and sediment effects as observed under drainage conditions.
A summary of the run-on water effect for the experimental runs is tabulated in Table 4 . Data from different levels of feeder sediment input were averaged to show effects of changing surface condition of the test box on sediment delivery from the run-on water. These data showed an increased sediment entrainment from the run-on water as the rainfall intensity and slope gradient were increased and the near-surface hydraulic gradient was changed from drainage to seepage. One interesting feature in the data that we would like to point out but cannot explain is the apparent additive nature of the run-on sediment entrainment under drainage conditions. Our data showed a similar increase in sediment entrainment from the run-on water (e.g., 12.4 vs. 11.8 kg h-1) as the slope gradient was increased form 5 to 10% under 50 and 100 mm h-1 rainfall intensities. Likewise, the increases in run-on sediment entrainment due to an increased rainfall intensity from 50 to 100 mm h-1 were also similar for the two slope gradients used in the study (i.e., 7.0 kg h-1 at 5% slope vs. 6.4 kg h-1 at 10% slope). Additional studies will be necessary to verify whether the entrainment process is indeed additive for slope and rainfall factors.
|
The experimental data demonstrated different types of dynamic equilibrium between surface conditions and erosion processes. Surface rilling tends to cause an increase in sediment delivery while severe rilling and erosion reduce soil erodibility. Changes in surface conditions and soil erodibility during the run are reflected in the sediment delivery data. This dynamic equivalent depends on specific conditions, i.e., slope steepness, rainfall and runoff intensity, and seepage and drainage gradient, during the rainfall event.
Sediment Transport Capacity Concept
Experimental data collected in this study can be used to evaluate the WEPP model concepts of sediment transport capacity (Tc) and detachment-transport coupling. In the WEPP model, the value of Tc sets the upper limit of sediment delivery and is the threshold for the detachment and deposition processes. Our experimental data under drainage conditions appeared to support the WEPP model concept. Excessive sediments from the feeder input were deposited and a decrease in feeder sediment concentration triggered a corresponding increase in downslope sediment detachment to maintain an equilibrium sediment delivery. This apparent equilibrium sediment delivery under drainage conditions can be considered a measure of the sediment transport capacity. The inverse relationship between up-slope feeder sediment input and downslope sediment detachment observed under drainage conditions were similar to field results found in freely drained plots in the Loess Plateau of China (Chen, 1992, 1993; Zheng, 1997).
On the other hand, our data also revealed features that were not considered in the original development of the Tc concept. When comparing the apparent Tc values under drainage conditions to sediment deliveries under seepage conditions, one can see that the flow indeed transported a considerably more sediment under seepage conditions than under drainage conditions for the same rainfall intensity and slope gradient. If the sediment regime in the 5-m test section has reached a dynamic equilibrium for the imposed rainfall, slope, and hydrologic conditions, then the sediment deliveries from the test box, i.e., both SFT and ST, would have been the sediment transport capacity of the flow for the specific condition. This means that the concept of the sediment transport capacity needs to be redefined to include either a soil erodibility term or terms that affect soil erodibility such as the near-surface hydraulic gradient. Our data show that different transport capacities will be associated with different dominant erosion processes depending on whether the surface is at a detachment-limiting drainage condition or a transport-limiting seepage condition. Before the concept of Tc and its role in controlling detachment and deposition are fully understood and supported by experimental data, it may be wise to use a more empirical approach to identify effects of surface condition on soil erodibility, and consequently, the controlling sediment regime.
|
Conclusions |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionConclusionsREFERENCES |
|---|
Under drainage conditions, a reduction in run-on sediment content caused a corresponding increase in downslope sediment detachment. Under seepage conditions, sediment delivery was greatly increased and the addition of run-on water always caused additional erosion. These data sets indicate that sediment delivery from the 5-m test box has reached a dynamic equilibrium. The concept of sediment transport capacity, Tc, and its role in sediment detachment and deposition appears to be a functional one under drainage conditions. Nevertheless, this concept is challenged when the soil erodibility was reduced by the imposed seepage condition.
Our results showed that the dual-box system is capable of recreating surface boundary conditions that occur on hill slopes to study erosion processes. An apparent dynamic equilibrium from the 5-m text box indicates that the dual box can be used to study different slope length effects on sediment delivery. This box system also allows us to examine sediment mass balance relationships in a hill slope segment, a first step in formulating new and more accurate erosion equations.
Although physically based erosion process models have been proposed for some time, few experiments were conducted to test the model concepts. This and other recent studies showed the dramatic effects of near-surface hydraulic gradients on soil erodibility, and consequently, the dominant erosion process and controlling sediment regime. These findings will help to improve the understanding of erosion processes and may eventually change the way soil erosion processes are modeled.
Received for publication July 23, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionConclusionsREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. R. Ahuja, L. Ma, and D. J. Timlin Trans-Disciplinary Soil Physics Research Critical to Synthesis and Modeling of Agricultural Systems Soil Sci. Soc. Am. J., February 2, 2006; 70(2): 311 - 326. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Darboux and C.-h. Huang Does Soil Surface Roughness Increase or Decrease Water and Particle Transfers? Soil Sci. Soc. Am. J., May 6, 2005; 69(3): 748 - 756. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F.-L. Zheng, C.-H. Huang, and L. D. Norton Effects of Near-Surface Hydraulic Gradients on Nitrate and Phosphorus Losses in Surface Runoff J. Environ. Qual., November 1, 2004; 33(6): 2174 - 2182. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F.-l. Zheng, S. D. Merrill, C.-h. Huang, D. L. Tanaka, F. Darboux, M. A. Liebig, and A. D. Halvorson Runoff, Soil Erosion, and Erodibility of Conservation Reserve Program Land under Crop and Hay Production Soil Sci. Soc. Am. J., July 1, 2004; 68(4): 1332 - 1341. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P.I.A. Kinnell Rejoinder to "Reply to Comment on 'Vertical Hydraulic Gradient and Run-On Water and Sediment Effects on Erosion Process and Sediment Regimes'" Soil Sci. Soc. Am. J., July 1, 2002; 66(4): 1402 - 1404. [Full Text] [PDF]
|
|
|
|
|
|
C. Huang, L. D. Norton, and F. Zheng Second reply to Kinnell's comments on "Vertical Hydraulic Gradient and Run-on Water and Sediment Effects on Erosion Process and Sediment Regimes" Soil Sci. Soc. Am. J., July 1, 2002; 66(4): 1404 - 1406. [Full Text] [PDF]
|
|
|
|
|
|
P.I.A. Kinnell, C.-h. Huang, L.D. Norton, and F.-l. Zheng Comments on ""Vertical Hydraulic Gradient and Run-On Water and Sediment Effects on Erosion Processes and Sediment Regimes"" Soil Sci. Soc. Am. J., May 1, 2001; 65(3): 953 - 956. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
