Geology Reference
In-Depth Information
2004). These cap-batter gullies are a common ero-
sional feature of mines sites. Plate 15 shows an
example of such a gully at a mine in Northern
Australia (Willgoose & Loch, 1996).
Willgoose and Loch (1996) identified a 50-year
old mine site, Scinto, that had an eroding mine
waste structure with the same type of cap-batter
geometry as the proposed Ranger structure (Hancock
et al
., 2000). The runoff and erosion at the site were
monitored for three months with a series of plots
and small catchments and SIBERIA calibrated in an
identical way to that done for Ranger. The original
landform was not known, but given that erosion
had occurred in a series of valleys at the cap-batter
transition it was decided simply to infill the valleys
in the landform used for the initial condition and
then run SIBERIA forward from that initial condi-
tion. The match between SIBERIA and the observed
landform was very good, with the volume and depth
of eroded and deposited material and the location of
the valleys being well modelled.
Bell and Willgoose (1998) carried out a three-
month field trial over a single wet season at
Ranger to simulate the evolution of a gully cre-
ated by the transition from the low slope on the
cap onto the high slope of the batter. As noted
above, these gullies are a common feature of the
LEM simulations. The study indicated: (1) that a
key feature of the erosional development was the
creation of a coarse armour layer at the base of
the gully which significantly reduced erosion;
and (2) the beginnings of the development of a
depositional fan at the base of the slope where it
transitioned onto the flatter surrounding land-
scape. The LEM simulations of the gully erosion
modelled the three dominant erosion events dur-
ing the wet season of the trial and found that the
gully was developed mostly during the largest
event (also observed in the field), and that if the
erosion rate of the original material was used, the
LEM overpredicted the gully depth. In subsequent
unpublished work we found that if the erosion
rate of the armour material developed on the
gully bottom was incorporated into a model
where the erodibility of the material was a linear
function of the cumulative erosion depth, then,
in the LEM simulations, the gully depth was cor-
rectly simulated. The conclusion of this work is
that to model correctly the initial rate of gully
development on rocky material, a model of the
evolution of the soil erodibility is necessary. It
should be noted that because of the short dura-
tion of the trial, no conclusion can be inferred on
the long-term equilibrium depth of the gullies,
which are a function of the fully-developed height
of the downstream alluvial fan.
18.3.2
Example 2: Farm scale
Gyasi-Agyei and Willgoose (1996) examined the
impact of graded contour banks on long-term ero-
sion. Contour banks are a common means of
reducing erosion in landscapes. They do this in
two ways.
●
Firstly, they trap sediment behind the contour
bank so that it is captured on the hillslope. This
does not lower erosion from the hillslope, but
does at least ensure that the eroded material does
not leave the slope.
●
Secondly, they reduce erosion by breaking up the
downhill slope length into a series of short seg-
ments. Almost all traditional erosion models indi-
cate that as slope length increases, the amount of
material eroded per unit area (in t ha
−1
) from the
slope also increases. This means that a series of
shorter slopes erodes less than a single slope of the
same total length. Contour banks break up the
slope so that less material is eroded from the slope.
Gyasi-Agyei and Willgoose used an LEM to simu-
late the evolution of the landform over 100 years
and found that the reduction in erosion is main-
tained even as the landform evolves (Fig. 18.4).
The contour banks were not along-contour but
had a longitudinal slope of 1%. Technically the
1% slope means that they were graded rather than
contour banks, but the differences for our purposes
here are small. The graded bank allowed the hill-
slope to shed water but keep water velocities
behind the contour bank to a low level. Figure 18.5
shows that reduction in erosion due to contour
banks actually improves with time, with a maxi-
mum reduction in erosion in their case study of
95% after 100 years, even though the initial imp-
rovement was only about 20% after 1 year.
