Geology Reference
In-Depth Information
One appealing aspect of this model from the sur-
face-process perspective is that it does not demand
that high straths were formed as broad surfaces
spanning the entire valley width (Fig. 7.13).
Instead, the straths can form as relatively narrow
benches on the margins of the fill. Given the sedi-
ment loading model, is it still possible to use
straths to deduce bedrock incision rates? Yes, but
the magnitude of potential alluviation in propor-
tion to strath heights has to be estimated, such that
interpretations of bedrock incision histories
become more nuanced.
Determining the age of an abandoned strath
is typically a challenge to the field geologist.
Commonly, a veneer of fluvial gravel mantles
the strath surface and may contain organic
matter that can be radiocarbon dated (Lavé and
Avouac, 2001; Merritts
et al.
, 1994; Weldon,
1986). For straths lacking a gravel cover, calcula-
tion of the duration of exposure of the strath
surface to cosmic radiation can provide an
estimate of the time since abandonment of the
strath by a downcutting river. In such cases, it
must be assumed or demonstrated that the
strath has remained uncovered by alluvial mate-
rial, landslide debris, or persistent snow cover
since abandonment, and that the strath surface
itself has not significantly degraded. If the strath
displays well-preserved original bedforms, such
as potholes, flutes, and polished surfaces, that
were created by abrasion in a river channel,
then it can be argued that little degradation has
occurred. Even so, the strath age might only
reflect the time of a modest amount of erosion
that occurred when an overlying alluvial fill was
removed (Pratt
et al.
, 2002), because only 1-2 m
of bedrock removal is required to reset the
cosmogenic age to near zero.
Well-preserved straths with associated cosmo-
genic exposure ages along the Indus River in
northern Pakistan have been interpreted to
define late Quaternary river incision rates into
metamorphic and igneous rocks (Fig. 7.14A and
B). This region is argued to contain some of the
most rapid bedrock uplift rates in the world
(Zeitler
et al.
, 2001b). In the arid Indus environ-
ment, some straths are extremely well preserved
despite their positions more than 100-200 m
above the modern river. Along a 100-km reach of
the Indus River, estimated incision rates vary
from about 1 mm/yr to greater than 8 mm/yr
(Burbank
et al.
, 1996b). Interpretation of bed-
rock incision rates using the traditional bedrock-
incision model, rather than the sediment-loading
model (Fig. 7.13), is supported by three observa-
tions: the great height of these straths (some up
to 400 m above modern flood levels), the lack
of evidence for significant depths of fluvial
aggradation, and the sequentially younger strath
ages for progressively lower terraces in the most
rapidly incising areas. Nonetheless, the very
rapid erosion and steep hillslopes in this region
would promote removal of any aggradational fill.
Consequently, a conservative approach would
attach a 20-40% uncertainty to the incision rates.
If such high rates were sustained, they would
lead to as much as 8 km of incision every million
years. The spatial pattern of incision rates of
the Indus Rive mimics the pattern of long-term
erosion and rock-uplift rate estimates from
apatite fission-track dating (Fig. 7.14C and D),
suggesting an overall balance. Such near-
equilibrium is perhaps not unexpected: if rock
uplift and denudation were not in balance, and if
rates of more than 5 mm/yr of rock uplift were
sustained for very long, rivers that failed to incise
rapidly enough to keep pace with the uplift
would instead develop very steep profiles. A
mismatch of as little as 20% would create 1 km of
relief along the river's course in less than 1 Myr!
Rates of denudation by landslides
It has long been recognized that, in actively
deforming areas, landsliding commonly provides
an important mechanism for delivering material
from hillslopes into valley bottoms occupied by
rivers or glaciers. We have suggested previously
that, wherever overall erosion rates exceed the
local conversion rate of rock to regolith, bedrock
landslides assume an increasingly important
role (Box 7.1). Because landslides are sporadic
events, however, calculating their time-averaged
contribution to erosion can be difficult. Several
different approaches could be used to assess
erosion by landslides (Box 7.3). We know that
landslides result from interactions between
rock strength (cohesion
c
and the angle of
