Geology Reference
In-Depth Information
lower boundary of a hillslope represents the
junction between two different process domains,
the hillslope and the river channel. Through
bank erosion, the river may undercut the slope
and cause its boundary to retreat upslope.
Alternatively, where the hillslope extends on to a
flood plain, migration of the river channel away
from the slope may cause the hillslope to lengthen.
The need to consider changes in the location of
the boundary becomes more important with long-
term erosion models that simulate landscape
evolution. In such models it may also be neces-
sary to recognize changes at the upslope bound-
ary, for example its lowering of altitude through
erosion over thousands of years. Model users need
to assess the importance and nature of the bound-
ary conditions and select a model accordingly.
For most applications involving hillslopes or
small catchments and with short time scales,
a model with a simple approach to boundary con-
ditions will suffice.
lar stage in the cropping year needs to be adjusted
by the proportion of the rainfall erosivity (
R
-factor)
occurring during that stage. The
C
-factor value is
thus the sum of these adjusted values. A similar
approach is used in the Revised USLE (Renard
et al
., 1991) to take account of seasonal variations
in soil erodibility (
K
-factor).
With erosion models that operate on a storm
or daily basis, it is possible to change the input
parameters regularly over time to take account of
changes in, for example, infiltration, soil mois-
ture, soil erodibility and crop cover. Often calcu-
lations are made within the model for even
shorter time steps, varying between one and ten
minutes, which allow the effects of varying rain-
fall intensities within a storm to be simulated.
Some models include the effect of feedback
mechanisms whereby, for example, surface rough-
ness or soil erodibility as affected by surface
crusting or sealing, alter within a storm as a direct
effect of the rainfall. The output from such mod-
els invariably includes a storm hydrograph and a
storm sediment graph, enabling the timing and
magnitude of runoff and sediment peaks to be
determined. This output is essential for dealing
with problems where concentrations of pollut-
ants associated with sediment concentrations
exceeding some critical value are the issue, rather
than total or average sediment levels. The suc-
cess of models operating over short time periods
depends on how well the conditions at the start
of the time period can be specified. In particular,
initial soil moisture, soil erodibility, surface
roughness and the nature of the vegetation or
crop cover need to be stated. A distinction can be
made here between single storm models, like
EUROSEM, where starting conditions must be
defined by the model user, and continuous simu-
lation models, like WEPP, which calculate the
starting conditions for a given time period from
the simulations of previous time periods. These
latter models can therefore be 'run in' because,
after a certain number of simulations, their out-
put tends to stabilize almost regardless of the ini-
tial conditions chosen. Judgment is then required
on how many simulations are needed to complete
the 'running in' period.
2.6 Temporal Considerations
The USLE predicts the rate of mean annual ero-
sion, which means that it operates on factors that
express the average condition of the study area
over a period of years. This is best illustrated by
the way the
C
-factor value is calculated. The
C
-factor expresses the effect of crop management.
It is a dimensionless coefficient based on the ratio
of the mean annual soil loss under a specific man-
agement system to that from bare soil under oth-
erwise identical conditions of rainfall, soil and
slope. Since the
C
-factor is dynamic and varies
with the percentage ground cover, percentage
canopy cover, height and root density of the crop,
the values change throughout a cropping season
and also from one year to another within a crop
rotation. In addition, the effect of these crop prop-
erties at any given time depends on how erosive
the rainfall is. Erosive rain on a bare soil during
the off-season is likely to result in a high rate of
erosion, but if the off-season coincides with the
dry season, the effect of the lack of plant cover
will be minimal. Thus the
C
-value for a particu-
