Modelling Landscape Evolution - Environmental Modelling: Finding Simplicity in Complexity - page 309

large-scale catchment-fan systems at relay zones, in con-
trast to existing conceptual basin-fill models, which thus
may require modification. Petit et al . (2009) have recently
combined a planar elastic dislocation model for faulting
(similar to that used by Densmore et al ., 1998) with a
surface-process model including diffusion and an under-
capacity algorithm for fluvial processes to study under
what tectonic and climatic conditions triangular faceted
spurs develop on fault scarps, and what controls their
height and morphology. Their model does not include
landsliding; as a result, hillslope diffusivities need to be
an order of magnitude higher than in the Densmore
et al . (1998) model, for the same model resolution of
100 m, in order for realistic topography to develop. Nev-
ertheless, they show how the competition between fluvial
and hillslope processes controls fault-scarp morphology,
with faceted spurs only developing for a narrow range of
parameter values, and how fault-slip rate controls facet
slope and height.
The drainage patterns on active folds may provide
important information on blind thrust activity and asso-
ciated seismic hazard (e.g., Burbank and Anderson,
2001). Tomkin and Braun (1999) were the first to use
a landscape-evolution model to study drainage develop-
ment associated with active folding. They modelled the
propagating fold by a simple triangular uplift function
and compared model predictions to drainage patterns and
wind-gap occurrences observed on folds in New Zealand.
In particular, they showed that regular spacing of streams
crossing the fold may result from steady tectonic and
climatic forcing and developed a linear analysis of stream
diversion to predict this spacing as a function of tec-
tonic uplift rate and fluvial incision efficiency. Champel
et al . (2002) used the same landscape-evolution model
but including a bedrock-landsliding algorithm, as well as
a more realistic kinematic displacement field for fault-
propagation and fault-bend folding, to study drainage
development on frontal Himalayan folds in Nepal. They
showed that, in addition to the controls on drainage spac-
ing identified by Tomkin and Braun (1999), the dip of
the detachment underlying the propagating fold plays a
major role by controlling surface tilting and thus drainage
reorganization behind the fold. They also showed that
landsliding (modelled in a manner similar to Densmore
et al ., 1998) is required to attain realistic topography on
the active fold, given the measured uplift rates, and that
drainage spacing could reflect in part the linking of ini-
tially individual fault segments. Finally, Miller et al . (2007)
concentrated on the topographic asymmetry rather than
the drainage patterns associated with fault-bend folds and

showed how this asymmetry varies as a function of fluvial
incision efficiency. They calibrated their model to frontal
Himalayan anticlines in Nepal and obtained reasonable
fits to the topography without including landsliding, but
requiring an extremely high slope-diffusion coefficient of
∼

10 m 2 y − 1 .

19.6.2 Fault andfoldbelts

At spatial scales larger than individual fault and fold
blocks, landscape-evolution models enable the controls
on drainage evolution and associated sediment routing
systems toward sedimentary basins to be studied, as well
as the response times inherent in the sedimentary record.
Tucker and Slingerland (1996) studied the evolution of
sediment flux from an active fold and thrust belt, scaling
their model so as to represent part of the Zagros Simple
Fold Belt in Iran. Their model highlighted the effects
of sediment ponding in intermontane basins, drainage
diversion and capture, and variable rock resistance is
setting geomorphic response times and spatio-temporal
variations in sediment flux. In an exemplary application
of landscape-evolution models, Cowie et al . (2006) used
the uplift field predicted by a dynamic fault-interaction
model to predict the landscape, drainage, and sediment-
flux response to extensional fault propagation, interaction
and linkage (Figure 19.5). Their model is remarkable
because it uses a dynamic model of stress focusing or
relaxation around fault tips (rather than a simple kine-
matic model) to predict surface displacements; in spite of
this high degree of freedom in the model, the predicted
topographic and drainage patterns explain many fea-
tures encountered in extensional provinces. In particular,
Cowie et al . (2006) show how fault interaction controls
the location of the main drainage divide and thus the size
of the footwall catchments that develop along an evolving
basin-bounding normal fault. As an aside, they also show
that 'close-to-transport limited' rivers are more likely to
maintain their course across the uplifted footwall blocks
of the faults than detachment-limited rivers and that
breakdown of width scaling with discharge contributes to
maintain river courses across the fault blocks, as observed
in reality (Whittaker et al ., 2007).

19.6.3 Riftedmargins

Amongst the first applications of landscape-evolution
models was the study of the conditions required to
grow and maintain escarpments at high-elevation
rifted margins (Kooi and Beaumont, 1994; Tucker and
Slingerland, 1994). Several aspects explain the initial