Environmental Engineering Reference
In-Depth Information
no values exceeded 15 t ha
−
1
year
−
1
. However, at all four
sites, locally severe areas of erosion in excess of tolera-
ble levels were predicted. For example, at the Kent site,
which has extensive hortucultural land uses, predicted
soil-erosion rates were in excess of 1 t ha
−
1
year
−
1
for
the majority of the silty loam soils found throughout the
transect. These rates are notable given the low slope angles
of most of the site. Regardless of whether average slopes
or slope distributions are used, in excess of 7% of the land
is also predicted to erode at or above the higher tolera-
ble threshold, highlighting the risks of intensive farming
practices on medium loam soils, as identified by Evans
(1988) and Boardman (1993).
A simple relationship between soil texture and erodi-
bility that is often referred to in the literature (Reed,
1979; Speirs and Frost, 1987; Evans, 1990) seems hard to
justify on the basis of these results. Whilst it is clear that
soil texture must play an important role in determining
erosion rates, the MIRSED results demonstrate that soil
loss in excess of tolerable levels are predicted for a wide
variety of soils. Thus, it has been demonstrated that it
is the complex interaction between the combinations of
slope gradient and land use that combine with soil types to
produce predictions of 'accelerated' rates of soil erosion.
Table 15.2
Ranges (1982-4) of observed soil erosion rates
across each overflight transect.
Observed soil loss rates (t ha
−
1
year
−
1
)
Site
Median soil
Maximum soil
loss
loss
Herefordshire
0.86-1.29
2.89-13.22
Beds and Cambs
0.28-0.45
0.35-3.30
Dorset
0.81-2.03
1.97-31.08
Gwent
0.92-1.33
1.61-21.87
Kent
0.85-11.09
4.84-17.86
Norfolk east
1.12-2.06
2.04-9.45
Nottinghamshire 0.34-2.43
2.1-66.15
Shropshire
0.56-1.89
11.99-49.34
Somerset
3.85-6.02
10.99-55.64
Staffs
1.05-1.83
15.08-108.28
Sussex west
0.29-0.60
0.57-10.01
from the volumetric rill-survey data. Though results
generated using average slope somewhat underestimate
maximumsoil-erosion rates, the use of slope distributions
produces results that fall within the survey ranges for seven
of the 11 sites. Thus, the approach of describing topogra-
phy within a grid cell using slope distributions instead of
average slopes seems to provide a better representation of
the process. Of those sites where predicted rates do not
fall within the survey ranges, Herefordshire, Dorset and
Gwent all produce overpredictions of maximumobserved
rates. This overprediction may be explained by consider-
ing the steep topography at these sites, which is susceptible
tomovement of particles downslope, by interrill processes
such as sheetwash and rainsplash, which are not quantified
by the volumetric surveys but are modelled by WEPP.
Considering the ranges produced from the MIRSED
runs, with average results from the observed survey data,
all but the Kent site were captured by the envelope of val-
ues from the model. The range of results generated, using
mean slopes from each grid cell, is shown in Figure 15.4
reflecting the influences of crop type, slope gradient and
soil type over areas that range from 36-112 km
2
.The
Kent results failed to capture the full effect of the inten-
sive agriculture that exists in the locality. It is possible that
the dynamic nature of market gardening such as chang-
ing crop types on an annual and sometimes biannual
basis is not fully represented by the use of the current
land-use data, which provide input to MIRSED via the
WEPP model. Without observed data detailing erosion
rates from specific soil, slope and crop-type combinations
15.8 Comparison with published data
Survey data, summarized by Evans (1988) are used to eval-
uate the simplified maps of predicted soil-erosion rates.
These data, by their very nature, exclude the processes of
interrill erosion - rainsplash, unconcentrated-flow ero-
sion andmicro-rilling - as theyweremade volumetrically,
thus care must be taken when evaluating predictions from
a modelling scheme that should incorporate both rill and
interrill processes. In addition, previous work assumes a
bulk density of 1.4 g cm
−
3
for all soils at these sites. This
assumption was retained to provide a consistent basis for
intercomparison between sites and within the literature.
Output from theMIRSEDmodel for the 11 sites is illus-
trated in Figure 15.3, with average erosion rates generated
as a function of slope distributions from each grid cell.
Observed data for the SSEW/MAFF surveys are included
(as in Table 15.2.) as error bars, describing the range of
erosion rates from each overflight transect. Average pre-
dictions of erosion rates from the transects (Table 15.3)
generally fall within the rates observed over the three years
of the SSEW/MAFF survey between 1982 and 1984, shown
in Table 15.2. There is also some agreement with the max-
imum soil-loss predictions in terms of the ranges derived
Search WWH ::
Custom Search