Geology Reference
In-Depth Information
1E+6
1E+4
1E+2
1E+0
1E-2
1E-4
Fig. 14.4
Comparison of the simulated bulk monthly sediment yields with sediment yields generated directly from
measurements for the Rimbaud basin in southeast France. The vertical lines represent the simulated bounds; the
horizontal lines represent the measured yields. Reproduced from Lukey
et al
. (1995).
events, and an excellent ability to represent inter-
month variability and long-term sediment yield
over a two-year period (Fig. 14.4).
Scenario simulations are reported for a number
of basins in the Mediterranean area. Typically the
model is first calibrated for the current state of
the basin. Land use scenarios are then specified as
a function of possible changes in vegetation cover,
such as partial or complete deforestation or a
change of agricultural basis. The model is then
rerun with the new parameter values and the out-
puts are compared. Climate change scenarios are
developed on the basis of predictions for the rest
of this century by general circulation models.
This is a complex procedure, requiring for exam-
ple the linking of the general circulation model
data, at spatial and temporal resolutions of typi-
cally tens to hundreds of kilometres and months
respectively, to the model grid (hundreds to thou-
sands of metres) and time (hours) resolutions (e.g.
Kilsby
et al
., 2007). The model is then rerun with
the altered rainfall and evaporation data and the
outputs are compared. Bathurst
et al
. (2002) cali-
brated SHETRAN for the upper 1532 km
2
of the
Agri basin in southern Italy for the four-year
period 1985-88, and then carried out simulations
(unpublished) for land use and climate scenarios.
For the land use scenario, all croplands (including
pasture lands) were changed to native vegetation,
the cover provided by the latter thereby increas-
ing from 28% to 71% of the basin. The model
was then run for the same four-year period. As
the native vegetation has a lower actual eva-
potranspiration rate than crops, the change caused
an increase in annual runoff (17%) and a corre-
sponding increase in sediment yield (8%) (Fig.
14.5). For the climate change simulations, three
scenarios were generated from the UK Hadley
Centre Climate Model 2 (HADCM2) for 1970-79
(representing the current conditions), 2030-39
(a relevant planning horizon) and 2090-99 (for a
full global warming effect). Relative to the first
period, annual rainfall decreases by 5% and 16%
for the second and third periods (most notably
