Modelling Landscape Evolution - Environmental Modelling: Finding Simplicity in Complexity - page 308

algorithm in landscape-evolution models, they do point
toward the important role played by sediment flux, non-
hydraulic scaling of channel width and possibly incision
thresholds. Inclusion of these elements in the fluvial
incision algorithms may thus become a requirement for
future landscape-evolution model development.

effects may be important when comparing predicted
long-term exhumation histories with those recorded by
thermochronometry. Braun and van der Beek (2004)
addressed this issue by coupling the Cascade landscape-
evolution model to a 3-D thermal model of the crust,
to show the strongly spatially variable pattern of
thermochronological ages expected even in a relatively
passive tectonic setting (see below).

19.5 Coupling of models

In most landscape-evolution models, topography is gen-
erated by a kinematically imposed tectonic displacement
field, or simply evolves passively from given initial con-
ditions. Although it is now commonly accepted that
important feedbacks exist between tectonic and surface
processes, because of the perturbation of the thermal
and stress field in the crust due to redistribution of
mass at the surface (e.g. Willett, 1999; Beaumont et al .,
2000; Whipple, 2009), such coupling has not as yet been
addressed extensively in landscape-evolution models. The
most direct feedback between erosion and tectonics is
through the isostatic response of the lithosphere to mass
removal by erosion, and this process has been included in
several landscape-evolution models designed to address
large-scale (

19.6 Model application: some examples

Landscape-evolution models have been applied to gain
conceptual insights (e.g. Kooi and Beaumont, 1996; Simp-
son and Schlunegger, 2003; Perron et al ., 2008) but also to
obtain a better understanding of the controls on landscape
evolution in particular tectonic or geomorphic settings,
ranging in scale from individual folds or fault blocks to
entire orogens or rifted margins. Below, I will review
some of these studies in order to provide a sampling of
what problems can be addressed with landscape-evolution
models, without pretending to present an exhaustive
review of all published model studies.

10 5 m) problems (see next section).
Feedbacks with viscous or plastic deformation of the
lithosphere, as expected in rapidly uplifting and eroding
mountain belts, are numerically much more challenging
to model. Dynamic models of crustal deformation have
therefore generally been developed in two-dimensions
(along a profile) and early coupled models have included
very simplified 1-D representations of surface processes,
using either diffusion only (e.g. Avouac and Burov,
1996; Simpson, 2006) or highly simplified fluvial incision
algorithms (e.g. Willett, 1999; Beaumont et al ., 2001).
Although these models provide insights into the potential
erosional controls on deformation in active orogens, the
predicted surface evolution cannot be compared to real-
world examples. Some workers have adopted a '2 1

∼

19.6.1 Individual foldsandfaultblocks

Models of landscape evolution on individual folds and
fault blocks have been strongly motivated by understand-
ing the topographic expression of active tectonics, with
the aim of better assessing seismic hazard. Densmore
et al . (1998) developed a model for the topographic evo-
lution of normal-fault scarps that combined a tectonic
displacement field predicted by a model of faulting in
an elastic half-space with a landscape-evolution model
including soil production, diffusion, bedrock landsliding
and fluvial incision (see Table 19.1). They applied their
model to fault blocks in the western United States and
found that bedrock landsliding was crucial to satisfyingly
simulate the observed topography. In their model, flu-
vial incision and bedrock landsliding are tightly coupled
and lead to rapid landscape response to changes in fault
activity. The same model was subsequently used by Dens-
more et al . (2003) to study the controls on catchment
development and sediment flux from relay zones between
adjacent normal faults. They found that the geomorphic
evolution of the relay zone depends on the timescale of
fault propagation relative to that of catchment growth;
the ratio between these determines whether relay-zone
drainage will be captured by streams incising the adjacent
footwall. The model did not predict the development of

/ 2 D'
approach, in which crustal deformation was modelled
in profile and surface processes in planform, coupling
being achieved by supposing the displacement field later-
ally constant and feeding laterally averaged erosion rates
back into the deformation model (e.g. Beaumont et al .,
1992; Stolar et al ., 2006). Full three-dimensional coupling
of dynamic crustal deformation and landscape-evolution
models is currently in progress (Braun and Yamato, 2010).
Erosion affects the thermal field of the crust because
exhumation of rocks advects heat toward the surface
and the surface itself changes through time. These