Geoscience Reference
In-Depth Information
flow. Transport relates to the actual movement of a sedi-
ment particle and may be by direct movement through the
air following the ejections of particles by raindrop impact
or carried in rebounding water droplets from impacting
raindrops. Alternatively, it may occur as a form of bed-
load, with particles rolling or sliding along the surface, by
saltation within a water flow or by suspension in the wa-
ter flow. Deposition will then be controlled by the mode
of transport. Splashed particles return following parabolic
trajectories to the surface and may rebound depending on
conditions at the point of return. Bedload-type transport
leads to deposition in relation to the local interaction of
flow and sediment conditions, while deposition from sus-
pension is as a consequence of settling of the sediment
through the water column. On any slope, erosion and de-
position are likely to relate to a continuum of several
of these processes, with specific thresholds controlled by
storm, flow and surface dynamics. This continuum is best
described in terms of variations in splash, unconcentrated
and concentrated erosion.
energy is usually characterised as a function of rainfall
energy (Morgan
et al.
, 1998) or the square of rainfall
energy (e.g. Ghadiri and Payne, 1977; Meyer, 1981), al-
though some authors have questioned the use of intensity
as the underlying control and have suggested raindrop
momentum as a better way of characterising the process
(Salles and Poesen, 2000). Increasing slope angle is gen-
erally considered as enhancing the effect of detachment.
For example, the results of Quansah (1981) suggest a
power-law dependency on slope angle, with the coeffi-
cient dependent on the particle size. However, Torri and
Poesen (1992) suggest that slope effects at practical scales
of measurement are more problematic, and certainly on
rough surfaces the local microtopography may counter-
balance average slope effects. Particle size and slope con-
trols on detachment are further affected by the interac-
tion with wind intensity and direction (Wainwright, 1992;
Erpul, Gabriels and Norton, 2005), so that the direction
of travel of a storm event in relation to slope and slope
aspect may be a significant control on the spatial pattern
of the process.
Soil cohesion is a further control on detachment, typ-
ically with an exponential decrease in the energy re-
quired for detachment as cohesion decreases (Al Durrah
and Bradford, 1982; Nearing and Bradford, 1985; Brad-
ford, Ferris and Remley, 1987b). These decreases will
occur dynamically through an event as the surface be-
comes saturated. Surface sealing (Bradford, Ferris and
Remley, 1987a), aggregate stability (Farres, 1987) and or-
ganic matter content (Tisdale and Oades, 1982) all have
the opposite effect, but again can change dynamically dur-
ing a storm, as noted above. Soil compaction leads to a
significant reduction in detachment (Drewry, Paton and
Monaghan, 2004).
Splash erosion occurs as the combination of raindrop
detachment and transport, either by direct ejection of par-
ticles or by entrainment in droplets that rebound from the
surface. The exact combination of these transport mech-
anisms is poorly understood but relates to the effect of
elastic versus plastic deformation of the surface on impact
(Huang, Bradford and Cushman, 1982, 1983). Splashed
sediment follows parabolic trajectories away from source
and can travel distances of a few cm to tens of cm (Savat
and Poesen, 1981; Moeyersons, 1983; Poesen and Torri,
1988; Van Dijk, Meesters and Bruijnzeel, 2002; Legout
et al.
, 2005). Splash is essentially a diffusive process, in
that it can move sediment in all directions. Movement can
occur in both and upslope and downslope directions, with
a progressively smaller proportion of upslope movement
as the slope angle increases (Savat, 1981). This diffu-
sion of sediment by splash has been interpreted as the
11.5.1
Splash
While all splash requires raindrop detachment, not all
raindrop detachment leads to splash erosion. This distinc-
tion is critical and leads to fundamental problems with
the characterisation of erosion processes, in that while
splash is measurable, detachment is not (Kinnell, 1993;
Wainwright
et al.
, 2009). Poesen and Savat (1981) sug-
gest that the minimum energies required for detachment
fall in the same range as the maximum splash rate, which
assumes that detachment and splash are directly corre-
lated. Van Dijk, Meesters and Bruijnzeel (2002) suggest
that this assumption is reasonable as long as the splash
distance is known, which is consistent with the results of
Savat and Posen (1981) using radioactive tracers. Rain-
drop detachment is controlled by particle size and density.
The dominant particle size detached by typical raindrops,
the diameter of which is generally
5 mm, is fine sand
(Ellison, 1944; Poesen and Savat, 1981; Parsons, Abra-
hams and Simanton, 1992). Finer particles require more
energy, due largely to cohesion developed by electrostatic
forces between particles; coarser particles require greater
energies. Practical thresholds of maximum movement are
of the order of 10-20 mm, but will depend on the density
of the particle and the exact configuration of the particle on
the surface (Wainwright, 1992). This maximum threshold
means that fine sediment can be preferentially removed
from the surface, leaving stones sitting atop columns of
sediment, known as splash pillars. Control of raindrop
<
