Geology Reference
In-Depth Information
The material in this topic goes a long way to
pointing to the probable future of soil erosion
modelling. A perusal of the applications chapter
of this topic indicates that watershed- and basin-
scale assessments have become a dominant inter-
est to modellers (see chapters 9, 13, 14, 16, 18
and 19). As we look to larger areas and up-scaling
soil erosion models (chapters 6 and 19), spatial
interfaces complete with GIS will continue to
play a big role in new model development. The
use of the Internet as a delivery mechanism for
erosion models will certainly also be important
in the future (chapters 16 and 17).
Continued data collection for supporting ero-
sion modelling will be critical. Paucity of data
remains a major limitation to the development of
reliable models. The types of data needed will
correspond to the new emphases in the model
applications. Whereas in the past, plot studies
have been the basic data of the erosion modeller,
in the future modellers will increasingly need
spatial data on erosion patterns and sources
(Walling, 2005). Fortunately, there is a wide range
of sediment sourcing tools now available to test
models that accumulate sediment across com-
plex environments. Motha
et al
. (2004) applied a
combination of minor and major element chem-
istry and sediment magnetic properties to assess
the proportion of sediment reaching a river from
different sources, including gravelled and ungrav-
elled roads and hillslope erosion from different
soil types. Rhoton
et al
. (2008) performed simi-
lar work in the Walnut Gulch Experimental
Watershed in southeastern Arizona. On a smaller
spatial scale, Polyakov
et al
. (2004) used rare earth
element oxides to measure the erosion, redistri-
bution and deposition of sediment in a small agri-
cultural watershed in Ohio. Sediment tracers that
differentiate between surface and subsurface soil
have been used widely to assess the proportion of
river sediment coming from hillslope, stream
bank and gully erosion (see review by Mabit
et al
.,
2008). In some instances these techniques have
been extended to assess the contributions of rill
and sheet to hillslope erosion (Wallbrink &
Murray, 1993). Sediment tracers that differentiate
between soil and sediment that was sourced from
land uses with specific vegetation types are under
active research (Gibbs, 2008).
Sediment that is deposited can be used as an
accumulated record of the transport history.
Where this history is well defined, it can serve as
a testing ground for predictive models run retro-
spectively. A key element in this approach is the
association of the deposited sediment with events
in the rainfall record. A wide range of evidence
has been used to establish the age of recent
(within the past 100 years) layers of sediment
deposition. These include charcoal layers from
known fires, and labelling by atmospheric events
including the atmospheric testing of nuclear
weapons (e.g.
137
Cs as reviewed by Walling and He
(1999) and Walling
et al
. (1999), and plutonium as
introduced by Everett
et al
. (2008) ). Until recently,
many of the sediment dating techniques used
in geochronology and anthropology were not
applicable to the last 100 years. The advent of
optically stimulated luminescence and other
techniques has enabled the assessment of the age
of modern deposited sediment and residence
times of sediment in fluvial systems (Gale, 2009).
Sediment dating data are now available for recent
agricultural history (Olley
et al
., 1998). The appli-
cation of single-grained, optically stimulated
luminescence has now extended this develop-
ment to small samples and ages greater than
2 years (Gale, 2009; Pietsch, 2009).
Each of these forms of sediment dating and
tracing techniques serve to provide further data
to our models of sediment transport. Specifically,
we can test the question of 'Are we getting the
right answer for the right reason?' in terms of
when the sediment was transported, from what
soil type it was sourced, what were the eroding
processes (sheet, rill or gully), and what land use
was in place at the source of the sediment.
Integrating the use of remote-sensing data into
the erosion modelling process has the potential
to offer an opportunity to verify independently
and provide initial conditions to models, but also
to change the way we conceptualize modelling of
erosion. Two key methodologies include the
determination of effective vegetative soil cover
and the direct sensing of sediment concentration
