Geology Reference
In-Depth Information
Evolution of the Southern Alps
Model age
3100 m
1750 m
2.0 Ma
0.25 Ma
5750 m
3000 m
1.0 Ma
5.0 Ma
Maximum
peak height
Fig. 11.16
First-generation planview landscape evolution model incorporating channels and non-uniform
precipitation and rock uplift patterns, Southern Alps.
Modeled temporal landscape evolution of the Southern Alps, New Zealand. Channels are dictated to have a particular
shape draping between the mountain crest and the sea. Diffusion modulates the hillslopes, and both differential rock
uplift and rainfall patterns are incorporated. Circles represent the highest topographic point in each simulation. The
complex ridge pattern reflects the essence of that observed in the Southern Alps of New Zealand. Modified after
Koons (1989).
wider than the others in the simulation. This
position reflects both the asymmetric rock uplift
pattern and the fact that the steady-state relief
depends on the square of the interfluve width
(Koons, 1989). The simplicity of this model,
which is as simple as a two-dimensional plan-
form model can get, allows rapid exploration of
the various controls on the topography. Koons
did this exploration very effectively by varying
the spatial pattern of the uplift pattern and of
the rainfall pattern (hence diffusivity).
One of the drawbacks to Koon's (1989) model
is that the channels are not interactive: their
profiles are dictated, rather than reflecting
the real dynamics of the channel system.
Nonetheless, just as much was learned from the
Stein
et al
. (1988) models in one dimension,
much is learned from Koons' (1989) two-
dimensional approach. With two-dimensional
models, the modeling community took a strong
step toward realizing the relative importance of
one or another process, and in particular alerted
the geomorphic community to the importance
of bedrock channel incision.
In the 1990s, several attempts were made to
incorporate more interactive channels and
hillslopes into the mountain range evolution
models. This generation of models required
integration of rules for the sediment transport
down channels and for the incision of bedrock
by channels. Anderson (1994) addressed the
problem of topographic evolution in a region of
30 × 100 km near a restraining bend in a major
strike-slip fault (the San Andreas Fault). The
rock-uplift pattern was simply dictated at the
outset, as a two-dimensional Gaussian rock
uplift pattern whose scale was set by the require-
ment for conservation of crustal volume arriving
at the bend. Crust was then translated through
this pattern, such that any parcel of crust
experienced a time-varying history of uplift
dictated by what portion of the uplift pattern
it intersected. The coarse, 1-km resolution of
the model required significant abstraction of
the hillslope-scale processes. Even small-scale
channels with drainage areas of less than 1 km
2
were subsumed in the numerical pixels.
The remaining major channels were inserted in
the calculation space, were tied to the bedrock
as it slid horizontally relative to the fixed fault
bend, and were allowed to evolve as bedrock
channels with their drainage area acting as a
proxy for local stream power. No attempt was
made to incorporate orographic effects: rainfall
was taken to be uniform. Arguing that landslides
were the dominant delivery mechanism for
debris to the channels, Anderson (1994) simply
set the hillslope angles in the landscape and
