Geology Reference
In-Depth Information
The theory of Hairsine and Rose described
such behaviour as due to the formation, as soil
erosion proceeds, of a layer of sediment consist-
ing of previously eroded material, and formed by
net deposition. The original soil matrix in general
is characterized as having some strength, while
the deposited layer is much weaker. This depos-
ited layer can grow in depth during erosion, pro-
viding partial or perhaps complete protection
from erosion of the underlying soil matrix.
Complete protection corresponds to the essen-
tially unstable upper transport limit − unstable
since continued erosion of the deposited layer
will eventually reduce the completeness of its
coverage. At the potentially more stable source
limit the sediment concentration is determined
by the rate of erosion of sediment sourced from
the soil matrix. Such observations made during
experiments in the GUTSR facility on fluctua-
tions in sediment concentration over time during
erosion provided insight into what may occur in
field plot experimentation, where commonly
only measurements of soil loss for the entire ero-
sion event are available.
Sensitivity analysis using the Type A GUEST
program showed the importance, amongst other
factors, of land slope and a sediment characteris-
tic called 'depositability', defined as the mean set-
tling velocity of sediment components in water − a
measurable characteristic (Misra & Rose, 1996;
Lisle
et al
., 1995). Data collected in the GUTSR
facility with simulated rainfall and overland flow
on soil beds prepared with differing soil strengths
demonstrated the ability of GUEST to interpret
the effect of varied soil strengths on sediment
concentration (Misra & Rose, 1995, 1996).
The theory of Hairsine and Rose (1992a,b) for
flow-driven erosion introduced an important
fundamental parameter,
J
, related to soil
strength, and defined as the energy required, per
unit mass of soil eroded, to remove soil from
the matrix by overland flow, a process termed
'entrainment'. As noted above, where the rate
of entrainment limits the concentration of
eroded sediment, the sediment concentration
achieved is described as 'source-limited'. This
situation is distinct from that at the 'transport
limit' when the soil matrix is completely cov-
ered by sediment eroded previously in the same
event (as, for example, following rill collapse),
and when soil strength can be assumed to play
no role in determining sediment concentration.
Hairsine and Rose (1992a,b) developed analy-
tical expressions for the transport limit and
also the source limit (involving
J
). Figure 11.1
illustrates how, as the soil strength (and so
J
)
increases, sediment concentration
c
decreases
for any particular value of stream power (
).
As described earlier, careful flume experimen-
tation showed that even during a flow-driven ero-
sion event in which flow conditions are constant,
sediment concentration can fluctuate through
time, generally not remaining at either of the two
limits described earlier (e.g. Rose, 1993). In order
to describe the actual erodibility in this normal
complex flow-driven situation, a Type B version
of the family of GUEST programs was developed
in which an empirical parameter
b
was intro-
duced, defined so that at the upper or transport
limit of flow-driven erosion,
b
had the value of
unity. Thus finding a value of
b
<1 following anal-
ysis of experimental data indicates that during
the erosion event, soil strength did play some role
in reducing sediment concentration to a value
below that of the transport limit. Should sedi-
ment concentration remain at the source limit
during an erosion event, then there is a direct
relationship between
b
and
J
. This is more likely
to be the case where the eroding surface is reason-
ably stable, without very active rilling, which can
involve head cutting and rill wall collapse.
However,
b
quantifies soil erodibility whether or
not persistence of sediment concentration at the
source limit occurs. The use of
b
has the advan-
tage of not requiring evaluation of
J
, which cur-
rently depends on measurement of sediment
concentration as a function of time - data not
commonly available in field experiments.
The Type B program involving
b
is more versa-
tile than Type A since it is less demanding in data
than that required for the determination of
J
.
However the theory developed for
J
, which
describes the fundamental manner in which soil
strength can reduce sediment concentration in
Ω
