Geology Reference
In-Depth Information
sediment yield and gully erosion were found to be
more sensitive to changes in rainfall than erosion
at the slope scale, due to non-linear relationships
with changes to surface runoff rates and channel
peak flow rates; however, changes to vegetation
cover had similar impacts on soil erosion at all
scales. The author also analysed the impact of
changes to soil water deficit (−20 to +20%), which
can have important consequences for water and
sediment connectivity between hillslopes and
the river network. While erosion at the slope
scale showed a relatively low sensitivity to this
parameter, gully erosion and sediment yield were
significantly more sensitive. Finally, the author
noted the relationship between low soil depth
and increased sensitivity of soil erosion to cli-
mate change, as a low water-holding capacity of
soils increases the response of water runoff gen-
eration to storm rainfall characteristics.
Nunes (2007) also applied the MEFIDIS model
for the A2 and B2 climate change scenarios, down-
scaled with the PROMES RCM; scenarios for
changes to vegetation cover and average soil water
deficit were created by applying the SWAT model
(described above). For the study areas, the scenar-
ios combine higher temperatures (+2 to +4°C)
with lower rainfall (−30 to −40%) and higher rain-
fall intensity in extreme events (10 to 20% in the
A2 scenario, for winter and spring). The SWAT
model indicates that this would lead to slightly
higher vegetation cover (from c. 5% for wheat and
vineyards to c. 30% for Mediterranean oaks and
shrubs) and lower soil water content (−40% in
winter and spring, −80% in autumn). These sce-
narios combine to cause different impacts on soil
erosion according to climate scenario and season.
Soil erosion is expected to decrease in both catch-
ments (−20 to −60%), due to the combined impacts
of higher vegetation cover, lower soil moisture
and modest increases in rainfall intensity. How-
ever, this decrease is more marked for gully ero-
sion and sediment yield due to the lower water
and sediment connectivity; at the slope scale, ero-
sion decreases are expected to be halved (−10 to
−30%). Also, this decrease is mostly under forests
and Mediterranean vegetation cover types; wheat
croplands experience little or no reduction in soil
erosion rates for the A2 scenario. These changes
are coupled with increased seasonal differences in
soil erosion; in the humid catchment, erosion is
expected to increase by 20% in winter and spring.
These results highlight how different soil erosion
processes have different responses to climate
change, and the role of soil moisture in determin-
ing changes to sediment connectivity as well as to
soil erosion, an issue which is not explored in
most studies. However, an analysis of these results
must take into account the lack of data to validate
gully erosion simulations in the study watersheds,
a problem which is widespread in erosion studies
(Boardman, 2006).
A final example of extreme event modelling
approaches was given by Michael
et al
. (2005), who
applied the EROSION 2D model (Schmidt, 1990)
to two agricultural slopes in Germany. The authors
used GCM scenarios for the B2 emission scenario,
2030-2050, downscaled to 5 min rainfall data using
a statistical approach driven by prevailing weather
types; the scenario points to an increase of 23% in
the intensity of the most extreme events, but a
38% decrease in frequency. The results point to an
increase in soil erosion between 22% and 66%, but
it should be noted that changes to vegetation cover
were not taken into account.
Overall, these studies highlight the complexity
and potential of an event-based modelling
approach for climate change impact assessment.
The models are often difficult to parameterize and
evaluate, and require both high-resolution climate
change scenarios and predictions for longer-term
impacts (in these examples, vegetation cover and
soil moisture). However, predictions can be made
for within-storm processes dependent upon sur-
face runoff concentrations (gully erosion) or peak
flow rates (sediment transport) with more detail
than that achieved by continuous models.
15.5 Conclusions, Limitations
and Research Needs
The case studies presented in the previous section
are representative of the typical modelling
approaches used to study the impacts of climate
