Geology Reference
In-Depth Information
The return period analysis from the water-
shed run (Table 16.3) predicted that the 10-year
return period sediment yield was 7505 Mg. The
area was determined by GeoWEPP to be 140 ha,
leading to an erosion rate for the 10-year return
period of 54 Mg ha
−1
for the 10-yr return period
sediment yield event. A review of the GeoWEPP
'Events' file showed that this event occurred on
9 April in year 15, a day when there was no pre-
cipitation, so it was a runoff event from snow-
melt only. The stochastic climate file was then
truncated to contain only years 14 and 15, and
GeoWEPP was run for the same watershed with
the 'Flowpath' option for those two years. The
results of the flowpath run with the average
annual erosion rate for two years, one of which
contained the 10-yr event, are shown in Fig. 16.7.
The erosion rate for the darkest pixels exceeded
200 Mg ha
−1
, and on the lightest pixels it was less
than 12.5 Mg ha
−1
.
The GeoWEPP predictions were greater than
the observed values or other predictions because
the slope lengths were greater, averaging 200 m.
The flowpath method showed that the areas with
greater predicted sediment yields were the areas
immediately adjacent to the streams, while the
ridge tops had lower predicted erosion rates. It
also showed that the more westerly-facing slopes
were at a higher risk of erosion than the east-fac-
ing slopes (Fig. 16.7). If additional information
about the spatial distribution of the severity of
wildfire were known, this could also be incorpo-
rated into GeoWEPP by altering the soil proper-
ties and/or ground cover on each hillslope polygon
to match the conditions determined by remote
sensing or ground survey as described by Elliot
et al
. (2006).
Table 16.4
Summary of risk-based predictions for a
10-year event. All data are daily values except
Disturbed WEPP.
Precipitation
(mm)
Runoff
(mm)
Sediment
yield (Mg ha
−1
)
Interface
WEPP Windows
34.5
52.2
2.2
Disturbed WEPP
(annual)
13.01
ERMiT
40.9
19.0
2.5
GeoWEPP
Watershed
33.4
99.1
53.6
weather whereas only 50 years were used for the
other examples. The ERMiT tool and WEPP
Windows predicted the lowest sediment delivery
rates. This is probably due to the fact that ERMiT
considers a number of different surface and soil
conditions even for high severity, whereas all of
the other tools considered a single high-severity
condition. The WEPP Windows prediction may
be lower than Disturbed WEPP because WEPP
Windows modelled cover as perennial, whereas
the Disturbed WEPP interface was developed
when this feature was not available, and thus
may have limited vegetation cover early in the
spring when significant rain-on-snow events
occur. Also, the estimate for Disturbed WEPP
was for an entire year that included several
events, whereas ERMiT and WEPP Windows
predictions were for single events. The GeoWEPP
flowpath method predicted a much higher
erosion rate, probably due to the much longer
slope lengths.
The output files portray another modelling
challenge: many of the large runoff events were
a combination of rainfall and snowmelt. This
is especially evident when runoff exceeded
precipitation for the 20-year runoff event for
ERMiT (Fig. 16.4), and the 10-year event for the
watershed example when the entire event was
snowmelt and there was no precipitation. Because
of the importance of snowmelt processes in this
climate, traditional precipitation-based risk tools
may not work as well as models that account for
snowmelt processes.
16.6 Discussion
Four different predictive tools, all based on WEPP
technology, have been presented. The results of
each method are summarized in Table 16.4, for a
ten-year return period erosion event. The ERMiT
tool estimated a higher precipitation value, prob-
ably because it uses 100 years of stochastic
