Environmental Engineering Reference
In-Depth Information
Publisher's Note:
Image not available
in the electronic edition
Figure 19.7
Evolution of an asymmetric 100-km wide mountain belt undergoing both tectonic uplift (at a rate of 1 mm y
−
1
)and
horizontal advection (at a rate of 20 mm y
−
1
at the left side of the model, diminishing to 0 at the right side). Left panel is after 2 My of
evolution, centre after 6 My, right after 10 My, when the model attains overall topographic steady state. An adapted version of the
Cascade
code was used for this model, including hillslope diffusion (Equation 20.3) and a detachment-limited stream-power model
(Equation 20.7) for fluvial incision. Parameter values are:
κ
=
100 m
2
y
−
1
;
K
=
×
10
−
7
m
0
.
5
y
−
1
;
m
=
.
=
1 (Modified with
permission from Willett, S.D., Slingerland, R. and Hovius, N. (2001) Uplift, shortening and steady state topography in active
mountain belts.
American Journal of Science
, 301, 455-85).
3
0
5;
n
topography? Can we extract a diagnostic signal from the
geologic record as to what drives topographic change?'
Landscape-evolution models can play an important
role in answering these questions, providing some of
the inherent uncertainties associated with their use
are elucidated. These uncertainties mostly concern
the algorithms described in Section 19.3. As discussed
there and in Section 19.4, there is uncertainty about
how to best capture different geomorphic processes
numerically and even about what are the key processes
that must be included in the models. There are two
possible approaches to this limitation. The most common
approach today is to build an
ad hoc
model that is
thought to be best suited for a particular problem
setting with its characteristic spatial and temporal scale.
Although such an approach may be justified, it does
complicate comparison of model results obtained using
different approaches and is sometimes motivated by what
particular model was available rather than what would
be the most pertinent model description for the problem
at hand. A more ambitious (or more naıve) view is that
a single model formulation should be able to address all
problems of landscape evolution, independent of spatial
and temporal scale, with only minimal and justifiable
modifications to model algorithms and with parameter
values that vary in a predictive manner between settings.
Whether such a 'unified model' approach is realistic or
even desirable is not clear at present.
Another issue concerns the controls on landscape
evolution: whereas the tectonic (or more precisely, rock-
uplift rate) control on landscape form is now reasonably
well understood (e.g. Montgomery and Brandon, 2002
and references therein), the same cannot be said for
climatic and lithologic controls. The problem here is
that we are still searching for the pertinent parameters.
As concerns lithology, rock resistance as inferred from
Schmidt-hammer tests is usually measured as a proxy of
erodibility (e.g. Snyder
et al
., 2003b; Duvall
et al
., 2004)
but how this translates into parameter values for erosion
algorithms remains unclear. Measurement of the tensile
strength has also been proposed and may represent a bet-
ter proxy (Sklar and Dietrich, 2001). Our understanding
of the potential climatic controls on erosion and relief
development is even less developed. First, only precipi-
tation has been considered to some extent and climate
controls in models are usually limited to the potential
role of precipitation. Second, the variability, rather than
the mean precipitation rate, appears to be the critical
variable in setting erosion rates (Molnar, 2001; Dadson
et al
., 2003; Lague
et al
., 2005), suggesting an important
role for thresholds in determining erosional process rates
(Tucker, 2004). Mean precipitation and its variability
can be predicted to a certain extent by global circulation
models of climate and climate change. However, data that
would permit climatic change to be linked to variations
in erosion rate is as yet cruelly lacking.
Search WWH ::
Custom Search