| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
School Resource, Environmental, & Heritage Sciences University of Canberra Canberra, ACT 2601, Australia
kinnell{at}scides.canberra.edu.au
Abbreviations: C, sediment concentration • DDL, dynamic depositional layer • FD, flow detachment • FT, flow transport • H, degree of protection (0 to 1) • R, water discharge • S, sediment discharge rate • RD, raindrop detachment • RIFT, raindrop induced flow transport • ST, splash transport
In Huang et al. (2001), the comment was made that my detachment-transport concept (Kinnell, 2001) is deficient when the deposition process becomes dominant under high sediment feed in the experiments of Zheng et al. (2000). This is not so.
To illustrate my detachment-transport concept as it applies to the Zheng et al. experiments, consider, for example, that the input of sediment from the feeder box occurs at a rate of 10 kg h-1 for 7 min and the sediment is made up of equal parts of particles that travel between 21 and 28, and 28 and 41 mm s-1 down the test box. The duration of the infeed is consistent with that used by Zheng et al. (2000) and the infeed rate is consistent with that for 50% cover in the drained 5% slope—50 mm h-1 rain experiment. For simplicity, consider that the input of water to the test box is maintained when the sediment input ceases so that these particle travel speeds are maintained once the sediment feed is removed, the test box has a nonerodible surface, a 5% slope, and is being subjected to 50 mm h-1 rainfall. If 15 kg h-1 is the discharged rate that occurs when the transport capacity of the system is test box is reached, then all the sediment input from the feeder box will be discharged from the downstream end of the test box and the discharge rate will vary according to the line marked output in Fig. 1 here. As can be seen from Fig. 1, here the steady sediment discharge rate produced under these circumstances is equal to the steady sediment infeed rate and continues for a period of time after the infeed is removed.
|
Now consider the situation when the infeed to the noneroding 5% slope is 17 kg h-1, as occured with 0% cover in the drained 5% slope–50 mm h-1 rain experiment. In this case, the infeed rate exceeds the 15 kg h-1 that can be discharged from test box. Under these conditions, two layers of particles occur on the surface of the test box. One layer is the DDL and particles in this layer are disturbed by drop impacts. The other layer is a layer of particles that is not disturbed by drop impact because it lies below the DDL and is protected by the DDL. This layer is a deposit in the sense that it is not being temporarily stored in direct association with a transport mechanism. However, once the infeed is removed, material that was part of the DDL is discharged, reducing the thickness of the DDL so that material in the lower layer is disturbed and thus becomes part of the DDL. As a consequence, the sediment discharge from the feeder box is maintained after the infeed stops for longer than in the case when the infeed was 10 kg h-1. The cessation of the discharge of sediment from the feeder box will be delayed further by a reduction of flow velocity such as occurs in the Zheng et al. experiments. Consequently, depending on the time gap between the cessation of the infeed and the collection of the "without feeder input" designated samples, infeed material may be contained in samples thought not to contain infeed material.
In the Zheng et al. (2000) experiments, the test box had an erodible rather than nonerodible surface. As a consequence, sediment discharged from the test box resulted from raindrop detachment in both the feeder box and the test box in some circumstances, and only the test box in others, and the sediment discharge rate across the downstream boundary of an eroding surface would have been influenced by the DDL.
When, as in the situation where only RIFT occurs, the flow does not have the capacity to entrain the detached particles, and there are no detached particles on the surface, the sediment passing over the boundary at any given time comes from raindrop impacts detaching particles within a limited distance of the downstream boundary. The zone where this detachment occurs is called the active zone (Kinnell, 1991). Any material detached upstream of this zone requires a number of drop impacts to cause it to move across the downstream boundary. The detachment of material upstream of the active zone contributes to the development of the DDL in the active zone and the situation can develop where the material that passes across the downstream boundary results from drop impacts lifting only previously detached particles sitting in the active zone. Consequently, the erodibility of the surface (k) can be seen to vary between two extremes; one (kSM) when there are no predetached particles in the active zone, the other (kDDL) where the DDL in the active zone gives full protection to the underlying surface. The intermediate values are given by
 | [1] |
When the feeder box was not connected to the top of the test box, water flow and sediment movement start at the upstream end of the test box. Initially, since the test box had been subjected to prior rain when the surface was in a horizontal position, a relatively uniform layer of loose particles existed on the surface to give a relatively uniform preformed DDL. As a consequence of this, RD would have occurred uniformly over the surface at the onset of the rainfall but, as time goes by, the thickness and protective effect of the DDL increases in the downslope direction (Kinnell, 1994). Consequently, RD would have been high at the upstream end of the test box and become reduced towards the downstream end because H tends towards 1 in the downstream direction. Depending on flow of velocity, the ease by which particles are detached from the surface of the soil matrix and particle characteristics such as size and density, H can reach 1 within quite a short distance of the upstream boundary (Kinnell, 1994). Under these circumstances, RD at the upstream end is the source of detached material discharged at the downstream end.
In the drained 5% slope–50 mm h-1 rain experiment, as with the other experiments on drained surfaces, the sediment discharge rates did not vary in any consistent way with the variations in cover in the feeder box when the boxes were connected. Thus, the values of H operating in the active zones in the experiments with drained surfaces were not influenced by variations in the sediment infeed, and because of this, it would appear that sediment discharge from the test box was controlled by the H = 1 condition when the feeder box was connected. Since, when the cover is 100%, the H = 1 condition seems to have occurred when there was no substantial infeed of sediment when the feeder box was connected, it seems that the H = 1 condition probably applied irrespective of whether the feeder box was connected or not.
The sediment discharge rate (S) is given by the product of water discharge (R) and sediment concentration (C).
 | [2] |
Considering that H = 1 condition seems to control the sediment discharge from the test box irrespective of whether the feeder box was connected or not, the sediment discharge rate resulting from RIFT depends on the erodibility of the DDL (kDDL), rainfall intensity, the drop-size characteristics of the rain, flow depth, and flow velocity. As can be seen from Eq. 2, there is an inherent direct association between the sediment S and R, and this association can mask some of the effects that result from variations in some of the factors that influence sediment discharge. For example, flow discharge is often related to flow depth to a power between 1.0 and 2.0. This results in an increase in flow velocity with the water discharge rate which tends to increase the sediment discharge rate, but it also results in an increase in flow depth which reduces the capacity of raindrop impact to lift soil particles up into the flow. Under these circumstances, the sediment discharge rate associated with RIFT tends to increase with the water discharge rate, and the effect of flow depth on the sediment discharge rate (Kinnell, 1993) will not become evident until an analysis of the effect of water discharge rate, or better still, flow depth on sediment concentration is made.
Figure 2 in Kinnell (2001) was aimed at looking at the effect of the variation of water discharge between the time when the two boxes were connected and the time when they were not. If there was an effect of water discharge on the sediment discharge rate other than through the direct effect of R on S indicated in Eq. 1, then the ratio of the sediment concentrations with and without infeed will equal a value other than 1.0. This was the case for all experiments on the drained surfaces except for 100 mm h-1 rain on the drained 5% slope and 100% feeder box cover. If there was an effect on C that is directly attributable to the variation in water discharge rate alone, then the ratio will remain constant with variations in feeder cover. That was not the case. The ratio decreased with feeder cover and this decrease resulted from increases in sediment concentration in the outflow from the test box when the feeder box was not connected. The increases in sediment concentrations in the test box follow a sequence of decreasing sediment infeed. The result depicted in Fig. 2 in Kinnell (2001) indicates that infeed material influenced the sediment concentrations obtained later when the feeder box was not connected.
As noted earlier, rainfall erosion is limited by either detachment or transport processes. In summary, the analyses presented by Zheng et al. (2000) indicate that sediment discharge in their experiments with drained surfaces occurred at the transport limit when the feeder box was connected to the test box irrespective of variations in sediment infeed. The data for the near zero sediment infeed condition indicated that sufficient detached material was provided to the sediment transport system in the test box to satisfy the capacity of the sediment transport system when 
320 L h-1 of water was fed in at the upstream end of the test box. However, the sediment concentrations produced when the feeder box was not connected indicate that infeed material contributed to the sediment discharge rate measured when the boxes were not connected some time later and Zheng et al. did not consider this fact in their analysis.
Variations in rainfall intensity influence both R and C in Eq. 1. Differences in sediment concentration resulting from variations in rainfall intensity can be observed in Fig. 2 in Kinnell (2001) and the effect of rainfall intensity in influencing the sediment discharge rate other than through the direct effect on R can be examined by comparing the sediment concentration ratios associated with the 100 and 50 mm h-1 rainfall. The ratio of the sediment concentration associated with 100 mm h-1 rainfall and the sediment concentration associated with 50 mm h-1 rainfall (zc100/c50) is 2.17 when no there is no infeed on the drained 5% slope and 1.56 when there is no infeed on the drained 10% slope. When there is an infeed, the ratios are 1.75 and 1.25, respectively. In the experiments of Walker et al. (1978) with a 3-m long noncohesive eroding surface inclined at 5%, sediment concentrations varied with rainfall intensity to a power <1 when the slope gradient was 0.5% but varied directly with rainfall intensity on the 5% slope. Consequently, the value of 2.17 for zc100/c50 when there was no infeed on the drained 5% slope appears consistent with the Walker at al. result. However, the decrease in zc100/50 by a factor of 0.71 between 5% and 10% slope gradients appears inconsistent with the Walker et al. result. Also, the effect of rainfall intensity on sediment concentration decreases by a factor of 0.80 between the no infeed and with infeed conditions on both slope gradients. The effect of rain intensity on sediment concentration is greatest on the 5% slope when there is no infeed and least on the 10% slope when the two boxes are connected together. The decrease in rainfall intensity effect between the no infeed and infeed situations may be a flow depth associated effect. The decrease between the 5 and 10% slope gradients may be associated with FT becoming increasingly important as a transport mechanism, and FD becoming increasingly important as a detachment mechanism. However, if, as is suspected, the results obtained by Zheng et al. (2000) differ from results obtained if the experiments were repeated using increasing rather than decreasing sediment infeed, then drawing major inferences from analysis of the Zheng et al. data may lead to incorrect conclusions.
Received for publication August 29, 2001.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
