<< . .

. 3
( : 51)



. . >>

linear
Kern R.
profile in (e)
1.0
slope map
x (km)
50
0 100
S (m/m)
h(km)
0.0 0.5 >1.0
3.0 4.0
1.0 2.0

Fig 1.8 Major geomorphic features of the southern Sierra Nevada. (a) Shaded relief map of topography indicating major rivers
and locations of transects plotted in b. (b) Maximum extents of the Chagoopa and Boreal Plateaux based on elevation ranges of
1750“2250 m and 2250“3500 m a.s.l. Also shown are along-strike and along-dip topographic transects illustrating the three levels
of the range in the along-strike pro¬le (i.e. incised gorges, Chagoopa and Boreal Plateaux) and the westward tilt of the Boreal
Plateau in the along-strike transect. (c) and (d) Grayscale map of topography (c) and slope (d) of the North Fork Kern River,
illustrating the plateau surfaces (e) and their associated river knickpoints (f). For color version, see plate section. Modi¬ed from
Pelletier (2007c). Reproduced with permission of Elsevier Limited.
8 INTRODUCTION


stantial stream power drain the plateau. In the
(a) sediment-¬‚ux-driven model, however, low erosion
Boreal Plateau
rates on the plateau limit the supply of sediment
that acts as cutting tools to channels draining
the plateau edge. As a result, the sediment-¬‚ux-
driven model leads to much slower rates of knick-
point retreat than those predicted by the stream-
power model in a plateau-type landscape (Gas-
parini et al., 2006).
The sediment-¬‚ux-driven model also implies
Kern R. that the evolution of hillslopes and channels
is more intimately linked than the stream-
Boreal Plateau power model would suggest. In the stream-
(b)
power model, hillslope evolution plays no ex-
plicit role in bedrock channel evolution. In the
sediment-¬‚ux-driven model, however, increased
hillslope erosion will supply more cutting tools
to bedrock channels downstream. Downstream
channels will respond to this increased supply
knickpoint
with faster incision, further lowering the base
level for hillslopes in a positive feedback. These
ideas will be further explored in Chapter 4 where
the stream-power and sediment-¬‚ux-driven mod-
els are compared in detail.
Fig 1.9 Virtual oblique aerial photographs of portions of
As one travels through the ¬‚uvial system, the
the Kern River basin. (a) The upper portion of the basin is
transport capacity of the main channel gener-
characterized by the low-relief Boreal Plateau. The mainstem
ally increases with downstream distance as a re-
Kern River is incised 1“2 km into the Boreal Plateau. (b)
sult of increasing contributing area. The effect of
Relief between the Boreal Plateau and Kern River is
declining channel slope, however, partially off-
accommodated by a series of channel knickpoints along the
sets that effect. As a result, transport capacity
Kern River and its tributaries.
increases downstream, but not as rapidly as the
sediment supply that the channel is required to
stream-power model to obtain a sediment-¬‚ux-
transport. The bedrock-alluvial transition occurs
driven model:
where the sediment supply exceeds the transport
n
‚h ‚h
= U ’ KsQm capacity. In most drainage basins of the west-
(1.10)
‚t ‚x
s
ern US, the bedrock-alluvial transition occurs up-
where Q s is the sediment ¬‚ux and K s is a new stream of the mountain front (e.g. Figure 1.10).
coef¬cient of erodibility. Only in regions that are tectonically most active
The predictions of the stream-power and does the bedrock-alluvial transition occur at the
sediment-¬‚ux-driven models are similar for cases mountain front itself.
of uniformly uplifted, steady-state mountain
belts. In such cases, erosion is spatially uniform
1.1.4 Alluvial channels
(everywhere balancing uplift) and therefore sed-
iment ¬‚ux Q s is proportional to drainage area Alluvial channels evolve in a fundamentally dif-
A. The predictions of the two models are very ferent style than that of bedrock channels (Figure
different, however, following the uplift of an 1.7). Alluvial channel evolution is governed by a
initially low-relief plateau. In such cases, the conservation of mass relationship which states
stream-power model predicts a relatively rapid that the change in channel-bed elevation is equal
response to uplift because large rivers with sub- to the gradient in bedload sediment ¬‚ux along
1.1 A TOUR OF THE FLUVIAL SYSTEM 9


load transport. If channel width is assumed to be
uniform along the longitudinal pro¬le, the com-
bination of Eqs. (1.11) and (1.12) gives the classic
diffusion equation (Begin et al., 1981):
‚h ‚ 2h
=κ 2 (1.13)
‚t ‚x
where the diffusivity is given by κ =
(BQb )/(c 0 w 0 ), and w 0 is the uniform channel
width. The diffusive evolution of alluvial chan-
nels of uniform width is illustrated schematically
in Figure 1.7b.

1.1.5 Alluvial fans
Deposition occurs on alluvial fans primarily be-
cause of channel widening near the mountain
Fig 1.10 Oblique perspective image of channels of front. Abrupt widening decreases the transport
Hanaupah Canyon. Smaller channels carve directly into capacity with distance downfan. Under such con-
bedrock, while the largest channels have wider beds ¬lled
ditions, Eq. (1.11) predicts aggradation and fan
with alluvium. Reproduced with permission of DigitalGlobe.
development.
Hanaupah Canyon has one of the most spec-
tacular alluvial fans in the western US. The size
the channel pro¬le:
of the Hanaupah Canyon fan re¬‚ects the large
‚h 1 ‚(wqs )
=’ sediment ¬‚ux draining from Hanaupah Canyon,
(1.11)
‚t c0 ‚ x
which, in turn, is a result of the high relief
where h is the elevation of the channel bed, t is and semi-arid climate of this drainage basin.
time, c 0 is the volumetric concentration of bed The semi-arid climate maximizes runoff intensity
sediment, w is the channel width, qs is the sedi- while minimizing erosion-suppressing vegetation
ment discharge per unit channel width, and x is cover. Many alluvial fans in the western US, in-
the distance downstream. Equation (1.11) is sim- cluding Hanaupah Canyon fan, have a series of
ply a statement of conservation of mass, i.e. the distinct terraces (Figure 1.11) that rise like a ¬‚ight
channel bed must aggrade if the sediment-¬‚ux of stairs from the active channel, with older ter-
gradient is negative (if more sediment enters the races occurring at higher topographic positions
reach from upstream than leaves it downstream) relative to the active channel. Alluvial-fan ter-
and incise if the sediment-¬‚ux gradient is posi- races in many areas of the western US can be cor-
tive (if more sediment leaves the reach than en- related on the basis of their elevation above the
ters it). A number of different relationships exist active channel, their degree of desert pavement
for quantifying bedload sediment ¬‚ux in alluvial and varnish development, and their extent of
channels, but one common approach quanti¬es degradation. Terraces form as a result of changes
sediment discharge as a linear function of chan- in the ratio of sediment supply to transport ca-
nel gradient and a nonlinear function of the dis- pacity through time. During times when the ratio
charge per unit channel width: of sediment supply to transport capacity is high,
aggradation and channel widening operate in a
b
‚h
Q
qs = ’B positive feedback that grows the fan vertically
(1.12)
‚x
w
and radially. During times when the ratio of sedi-
where B is a mobility parameter related to grain ment supply to transport capacity is low, incision
size, Q is water discharge, and b is a constant. and channel narrowing operate in a positive feed-
The value of b is constrained by sediment rating back that causes the channel to entrench into
curves and is generally between 2 and 3 for bed- older fan deposits, leaving an abandoned terrace.
10 INTRODUCTION


main channels on the fan branch into dozens of
distributary channels separated by horizontal dis-
tances of only a few hundred meters or less.
The triggering mechanisms for alluvial-fan
terrace formation have long been debated, but
climatic changes most likely play a signi¬cant
role in controlling the changes in sediment sup-
ply that trigger fan aggradation and incision.
A growing database of surface and stratigraphic
age estimates suggests that Quaternary geomor-
phic surfaces and underlying deposits of the west-
ern US can be correlated regionally (Christensen
and Purcell, 1983; Bull, 1991; Reheis et al., 1996;
Bull, 1996). Several studies have documented ¬ll
events and/or surface exposure dates between
700--500 ka, 150--120 ka, 70--50 ka, and 15--10 ka
Fig 1.11 Oblique perspective image of the terraces of corresponding to the Q2a, Q2b, Q2c, and Q3 ge-
Hanaupah Canyon fan. Reproduced with permission of omorphic surfaces identi¬ed by Bull (1991) (sur-
DigitalGlobe. Younger terraces are lighter in color in this
face exposure ages correspond to youngest ages
image, representing the relatively limited time available for
in these ranges). Similarity of ages regionally has
desert pavement and varnish formation on younger surfaces.
provided preliminary support for the hypothesis
that Quaternary alluvial-fan terraces are gener-
Over time, episodes of aggradation, incision, and ated by climatic changes.
lateral reworking produce a nested sequence of Changes in sediment supply due to climatic
terraces. changes can result from several factors, but vari-
Multiple cut and ¬ll cycles on alluvial fans ations in drainage density are likely to play a
create a spatially-complex, distributary channel signi¬cant role in controlling the temporal vari-
network that presents a challenge to ¬‚oodplain ations in sediment ¬‚ux from drainage basins
managers in the western US. The Tortolita Moun- in the western US. Terrace formation during
tains fan northwest of Tucson, Arizona is a clas- the Pleistocene--Holocene transition is associated
sic example. Figure 1.12 presents four views of with a ten-fold increase in sediment supply (e.g.
this topographic complexity using a shaded re- Weldon, 1980). It is unlikely that a change in pre-
lief image, aerial photo, grayscale map of a nu- cipitation alone could account for such large in-
merical model of ¬‚ow depth during a recent ex- creases in sediment supply. During Pleistocene
treme ¬‚ood, and a sur¬cial geologic map. The sur- climates, vegetation densities were higher at
¬cial geologic map (Figure 1.12d) was constructed most elevations across the western US. Higher
by integrating soil development and other indi- vegetation density results in a lower drainage
cators of terrace age to group the terraces into density. During times of lower drainage density,
distinct age ranges (Gile et al., 1981; McFadden accommodation space is created in hollows for
et al., 1989). The surface age represents the ap- the deposition of sediment eroded from higher
proximate time since deep ¬‚ooding occurred on up on the hillslope, thereby lowering sediment
the terrace because soils would be stripped from ¬‚uxes from the basin relative to long-term geo-
a terrace subjected to deep scour and buried on a logic averages. During humid-to-arid transitions,
surface subjected to signi¬cant ¬‚uvial deposition. drainage densities increase, removing sediment
Sur¬cial geologic mapping indicates that ¬‚ood stored as colluvium in hollows during the pre-
risk (which is inversely correlated with surface vious humid interval. Sediment ¬‚uxes decrease
age) varies greatly even at scales less than 1 km. when the drainage density reaches a new maxi-
The modeled ¬‚ow depths also illustrate the spa- mum in equilibrium with the drier climate. Ac-
tial complexity of ¬‚ooding. As Figure 1.12c, the cording to this model, it is the change in climate,
1.1 A TOUR OF THE FLUVIAL SYSTEM 11



111.10° W
111.15° W
32.50° N
Cochie
(b)
1 km
(a)
Wild Burro


Ruela's




32.45° N




aerial photo
shaded relief

(c) Cochie Map units
(d) Qy1
bedrock
Qy2
Qm

Qyc
Ql




surficial geology
model flow depth

Fig 1.12 (a) Shaded-relief image, (b) aerial photo, (c) ¬‚ow
phase with climate because it is responding to
depth model prediction, and (d) sur¬cial-geologic map of the
the rate of change of climate, not its absolute
Tortolita Mountains fan northwest of Tucson. Multiple cycles
state. The duration of the climate oscillation also
of cutting and ¬lling on this fan have resulted in a complex
plays a role in landscape response. Short-term cli-
distributary channel network in which ¬‚ooding from con¬ned
matic oscillations may not allow suf¬cient time
channels upstream branches into dozens of active channels
downstream on the fan. for sediment to be stored on the landscape during
times of lower drainage density. Long-term cli-
mate changes, however, allow time for more sed-
not the absolute state of climate (wet or dry), iment to be stored during low-drainage-density
that most strongly controls sediment ¬‚uxes from humid intervals, resulting in a larger response
drainage basins. when the climate shifts to more arid conditions.
Figure 1.13 schematically illustrates the rela- In the western US, the elevation zone of
tionships between precipitation, vegetation den- greatest vegetation change is well constrained
sity, and sediment ¬‚ux in the western US during through the Last Glacial Maximum (LGM) by
¬‚uctuating climates according to the ¬‚uctuating packrat middens. Today, the modern tree line (i.e.
drainage density model. Precipitation and vegeta- the location where shrub vegetation transitions
tion changes oscillate in phase over geologic time to mature woody species) occurs at approximately
scales. Sediment ¬‚ux, however, oscillates out of 2000 m a.s.l. (with variations of a few hundred
12 INTRODUCTION



high frequency
low frequency

precip.

veg.
cover


sed.
flux



time

Fig 1.13 Schematic diagram of the relationships between
precipitation, vegetation density, and sediment ¬‚ux for
Fig 1.14 A HiRISE Camera image of erosional gullies on
low-frequency (long-duration) and high-frequency (short
Mars and their adjacent alluvial fans.
duration) climate changes.


scientists are currently grappling with many of
meters due to slope aspect). At the LGM, the tree
the same science questions that geomorpholo-
line was approximately 500 m a.s.l. In the western
gists puzzle over on Earth. Figure 1.14 presents
US, therefore, the elevational range from 500--
one of many images of the Martian surface that
2000 m has undergone signi¬cant changes in veg-
is tantalizingly similar to ¬‚uvial landforms on
etation type and density many times within the
Earth. Figure 1.14 shows a series of erosional val-
Plio--Quaternary period (Spaulding, 1990). It is
leys in the upper right-hand side of the image,
within this elevation range that we can expect to
transitioning into entrenched depositional fans
see the most signi¬cant impact of climate change
with multiple terrace levels. Whether these ter-
on ¬‚uvial drainage basin processes. Of course, at
races represent the effects of deposition during
much higher elevations some drainage basins in
dry or wet ¬‚ow events is not currently known.
the western US were also glaciated. Glaciation has
Either way, the similarity between many of the
a huge impact on ¬‚uvial processes downstream
¬‚uvial landforms on Mars and those on Earth is
by delivering large volumes of sediment due to
striking, especially given that liquid water is un-
the erosional ef¬ciency of alpine glaciers.
stable on Mars under the temperature and pres-
The ¬‚oor of Death Valley is a saline playa that
sure conditions of today. The wealth of imagery
has been an ephemeral shallow lake many times
and elevation data from Mars provides an excel-
in the past. Pluvial lakes occupied large areas in
lent opportunity to compare and contrast Mar-
the western US during much of the Pleistocene.
tian landforms to those on Earth in order to
During pluvial lake highstands, prominent shore-
learn more about how landforms evolved on both
lines were created by wave-cut action in many
planets.
areas of Utah (Bonneville shoreline) and Nevada
(Lahontan shoreline) (Reheis, 1999). Degradation
of these shorelines over time provides natural ex-
1.2 A tour of the eolian system
periments in hillslope evolution that have been
exploited by hillslope geomorphologists. Fluctu-
ations in lake levels also store and release vast The eolian system can be divided into two types:
amounts of windblown dust in arid regions, systems dominated by the transport of silt and
thereby in¬‚uencing soil formation and hydrology clay, and systems dominated by the transport of
on nearby alluvial-fan terraces (Reheis et al., 1995). sand. Both types of systems involve particle en-
One of the most exciting developments in trainment from the ground, atmospheric trans-
surface process research is the wealth of new port, and redeposition by gravitational settling.
data emerging on the surface of Mars. Planetary The primary difference between these two types
1.2 A TOUR OF THE EOLIAN SYSTEM 13



evaporation
hydrological balance/
water table fluctuations
Franklin Playa
moisture-controlled entrainment l
g
advection/
turbulent dispersion 2
u*t u*td
= Penman“Montieth equation
r
Eagle M. r
u* u*
2 2
u*t

deposition
ponded water surface (occasional)
ground surface
h z=0
unsaturated zone
unsaturated flow
q q
Amargosa R. q q

Richards equation




Fig 1.15 Oblique perspective image of Franklin Lake Playa
and adjacent Eagle Mountain piedmont. Franklin Lake Playa
water table (moving boundary)
acts as a strong regional source of dust. Terraces on
z = zm
piedmonts such as that of Eagle Mountain act as net sinks for q = qs saturated zone
dust in arid environments. Reproduced with permission of
DigitalGlobe. Fig 1.16 Schematic ¬gure illustrating the coupling between
the hydrologic and eolian systems on playa surfaces.

of systems is the length and time scales involved
in particle motion. Silt and clay transport in- alluvial fan terraces of Eagle Mountain piedmont
volves travel distances on the order of 1 km or are sourced from Franklin Lake Playa. The dust of
greater and residence times on the order of 1 Franklin Lake Playa is, in turn, sourced from the
hour or more in the atmosphere. Sand, in con- Amargosa Valley drainage system, of which Eagle
trast, moves primarily in saltation. Saltation in- Mountain piedmont is a part. As the Amargosa
volves trajectories on the order of 10--1000 cm and River drains into Franklin Lake Playa, its low gra-
residence times on the order of 1 s. In this sec- dient and distributary channel geometry causes

<< . .

. 3
( : 51)



. . >>