Geology Reference
In-Depth Information
interactions between climate change, runoff,
vegetation cover, and so on, which are required to
upscale model predictions in these regions. It can
therefore be argued that more slope-scale studies
are still needed in order to increase our under-
standing of the processes linking climate change
and soil erosion, thereby increasing our confi-
dence in subsequent upscaling exercises.
studies (e.g. changes to the hydrological cycle and
vegetation patterns) and therefore this issue is
likely to remain a problem in the near future.
Furthermore, it is doubtful if some sources of
uncertainty can be quantified since they can be
defined as 'deep uncertainties': processes which are
not yet fully understood (Dessai & Hulme, 2007)
and therefore not well simulated. The studies
shown in this chapter focus mainly on long-term
hillslope erosion rates and, to a lesser degree, chan-
nel processes. Gully erosion and sediment connec-
tivity issues are poorly represented; however, the
current lack of data and knowledge on these proc-
esses has limited their integration in soil erosion
models (Boardman, 2006), which may severely
hamper progress in this area. Disturbances are also
poorly represented in these studies. While continu-
ous models can represent the impacts of drought on
vegetation cover, other important impacts of severe
droughts - such as woody plant mortality, changes
in vegetation patterns, desertification - are not well
described; this issue might potentially be addressed
by associating erosion models with more detailed
vegetation models. The impacts on soil erosion of
increased wildfire frequency are also poorly repre-
sented; in this case, while links between climate
change and wildfires have been studied, the impacts
upon soil erosion are still not well understood
(Shakesby & Doerr, 2006) and require additional
research before being included in climate change
impact studies. Finally, the relationships between
climate change, soil carbon processes and erodibil-
ity have been studied mainly through space-for-
time approaches (e.g. Lavee
et al
., 1998), as described
earlier in this chapter. Soil structure processes are
poorly represented in erosion models, and some
attention to this issue is required in order to obtain
improved predictions, including more data on
climate-soil structure relationships.
15.5.2
Uncertainty in climate change
impact estimates
Knowing the uncertainty of a climate change
impact assessment is necessary to provide robust
adaptation measures, i.e. measures that can be
expected to provide acceptable results under a
large range of conditions. While sources of uncer-
tainty in soil erosion studies have been discussed
since the first modelling experiments (e.g. Favis-
Mortlock & Guerra, 1999), published studies usu-
ally only quantify uncertainty due to model
errors; even the assessment of this kind of uncer-
tainty may be hampered by the lack of soil ero-
sion data, an issue which is a broader problem in
soil erosion modelling (Boardman, 2006).
An example of other sources of uncertainty
which can significantly hinder modelling results
can be taken from the water resources sciences
(e.g. Dessai & Hulme, 2007). These include uncer-
tainties in: greenhouse gas emission scenarios, dif-
ferent GCM results for a similar emission scenario,
downscaling (particularly important for extreme
event predictions), model results for surface runoff
generation and vegetation response, and future
land-use changes; these uncertainties may propa-
gate through the modelling approach. Some of
these sources of uncertainty have been taken into
account in the studies presented in this chapter,
for example by driving erosion models with out-
puts from different GCMs or different climate
change scenarios; the remaining sources of uncer-
tainty are rarely taken into account. While progress
has been made on quantifying and reducing these
uncertainties, such as the PRUDENCE project
(Déqué
et al
., 2005) which provided downscaled
climate change estimates for Europe, soil erosion
is downstream from a number of climate change
15.5.3 Links and feedbacks between
erosion and land use/land cover
Links between soil erosion, land use/land cover
and socio-economic issues are also barely touched
by the studies presented in this chapter. Future
land-use changes are not usually considered in
