Middleton G.V. (Ed.) Encyclopedia of Sediments and Sedimentary Rocks
Подождите немного. Документ загружается.
RIPPLE,
RIPPLE MARK, RIPPLE STRUCTURE
565
'olymer
Class
la:
R
=
COjH, CHjOH,
0-Succlnyl,
CHj
Others
(?)
Class
Ib:
R
=
COjH, CHjOH,
CHj Others
(?)
No Succinic Acid
Class
Ic:
R-COjH. CHjOH,
CH3 Others
(?)
No Succinic Acid
Class
II
Polycadinene
Class
IV
Cedrane
sesquiterpenoid
Class
III
Polystyrene
Class V
., Abietane/Pimarane
*'^2"
Diterpenoids.
(Labdanoids nonnally
absent)
Figure
R5
Classification system
for
sedimentary resins based
on
structural characteristics
of
major organic constituents. Modified after
Anderson
and
Crelling,
1995.
common overall than Class
1
ambers
but are
nevertheless very
abundant
in
some regions, especially South East Asia
and
parts
of
western
and
southern North America. Class
III,
Class
IV, and
Class
V
ambers
are
generally rare
and
represent
resins
of
specific families
or
non-polymeric resins which
are
preserved only
in
unusual sedimentary circumstances.
Class
I
ambers
are by far the
best characterized ambers
at
this time,
but
even
for
these materials complete structural
characterization remains elusive.
It is
known, however, that
the structures
of
these materials undergo regular systematic
changes over time
in
response
to
(primarily) temperature. This
raises
the
possibility that these materials
be
useful
as
thermal
maturity indicators,
and if so
these materials would
not
suffer
from complicating factors such
as
multiple sources, migration,
or mixing, which
can be
problematic
in
other organic maturity
markers. Also,
in
most cases ambers occlude
a
variety
of
terpenes
in
addition
to the
common structural components
which make
up the
polymeric backbone.
In at
least some cases
these occluded compounds
may be
characteristic
of the
original botanical source, allowing investigators
to
determine
the presence
of
particular genera
or
species
at a
particular
location
at a
particular point
in
geologic time. Such informa-
tion
is
likely
to be
highly valuable
in
paleoenvironmental
reconstructions.
Inclusions
Finally,
it is not
possible
to
conclude this article without some
comment about inclusions, which
are
often found
in
ambers.
Due
to
their "sticky" nature when first exuded,
it is not
uncommon
for
extraneous materials
to
become embedded
within resins. These inclusions often include plant parts, small
insects,
and
(rarely) small vertebrates
and
other materials such
as feathers. Once enveloped,
the
quality
of
preservation
of
these included materials
is
unequaled
by any
other medium.
As
a result, these types
of
fossils
are
highly prized
by
collectors
and paleontologists
and are
sometimes
of
considerable value.
Insects,
for
example,
are
preserved
in
exquisite detail, often
allowing paleo-entomologists
to
study detailed characteristics.
This extraordinary degree
of
preservation extends
to
some
extent
to
soft tissues,
and
even cellular structure (Grimaldi,
1996).
Claims
of
preservation
of DNA, and
even viable
bacterial spores
in
amber have received considerable press,
but
for
the
most part remain controversial
at
this point.
Ken
B.
Anderson
Bibliography
Anderson, K.B.,
and
Crelling,
J.C,
1995. Introduction.
In
Anderson,
K.B.,
and
Crelling,
J.C.
(eds.). Amber, Resinite,
and
Eossil Resins.
ACS Symposium Series Volume 617.
pp.
ix-xvii.
Anderson,
K.B., 1996. The
nature
and
fate
of
natural resins
in the
geosphere.
VII. A
radiocarbon
('''C)
age
scale
for
description
of
immature natural resins.
An
invitation
to
scientific debate. Organic
Geochemistry,
25
(3/4): 251-253.
Andersoti,
K.B., and
Muntean,
J.V.,
2000.
The
nature
and
fate
of
natural resins
in the
geosphere.
X.
structural characteristics
of the
macromolecular constituents
of
modern Dammar resin
and
Class
II
ambers. Geochemical Translation. Article
No.
I.
http://www.rsc.org/
ej/gt/2000/b000495m/index.html
Grimaldi,
D.A., 1996.
Amber: Window to the Past. American Museum
of Natural History,
and
Harry
N.
Abrams.
Langenheim,
J.H.,
1995. Biology
of
Amber Producing Trees: Focus
on
Case Studies
of
Hymenaea
and
Agathis.
In
Anderson,
K.B., and
Crelling,
J.C.
(eds.). Amber, Resinite.
and
Eossil Resins.
ACS
Symposium Series, Volume 617,
pp. 1-31.
Langenheim,
J.H.,
2002. Plant Resins: their Value
to
Plants
and
Humans. Timber Press.
In
Press.
RIPPLE, RIPPLE MARK, RIPPLE STRUCTURE
Definition
Ripples, ripple marks,
or
ripple structures
can be
defined
as
small-scale, flow-transverse ridges
of
silt
or
sand produced
by
fluid shear
at the
boundary between moving water
or air and
an erodible sediment
bed.
Principal ripple types
are
current
ripples, formed
by
unidirectional water flows, wave ripples,
generated
by
oscillatory wave action,
and
wind ripples, formed
by eolian currents
(see
Desert Sedimentary Environments).
Large-scale equivalents
of
ripples
are
dunes
(see
Surface
Forms).
Basic concepts
Ripple marks
are
quasi-triangular
in
vertical cross-section
parallel
to
flow direction (Figures R6(A),(C-E))
or
wave
propagation (Figure
R6b).
Current ripples
are
asymmetric,
with gentle upstream face {stoss side)
and
steep downstream
face
{lee
side) approaching
or at
angle-of-repose. Individual
current ripples
can be up to
60 cm long
and
6
cm high,
but the
mean length
and
height
of a
field
of
current ripples
are
usually
<20cm
and
<2cm, respectively. Wave ripples
are
symmetric
or slightly asymmetric
in
cross-section (Figure R6(B)). They
can
be
200 cm long
and
23
cm high,
but
typical dimensions
are
an order
of
magnitude less.
566
RIPPLE,
RIPPLE MARK, RIPPLE STRUCTURE
suspension
settling
avalanching /,
(C)
BEDFORMS IN DEVELOPMENT
EQUILIBRIUM BEDFORMS
Figure
R6 (A)
Ripple
marks
in
vertical
profile
parallel
to
flow.
Terminology
is
based
on
Allen (1968)
and
Reineck
and
Singh
(1980).
Note
that
avalanching
and
suspension
settling
generate
cross-lamination depicted
in
(C)-(E).
vel. =
flow
velocity.
Water
depth
note
to scale. (B)
Plan
form
and
internal
structure
of
symmetrical
wave
ripples
with
rounded
crests.
Wave
ripple
crests
can
also
be pointed.
Wave
propagation
direction
is
perpendicular
to
crest
lines.
Note bundle-wise arrangement
of
cross-laminae.
(C)
Straight-crested,
(D)
Sinuous-crested,
and (E)
Linguoid
current
ripples,
with
characteristic
cross-lamination.
Flow
is
from
left
to
right.
Note
that
current
ripples
evolve
from
straight-crested
via
sinuous-crested
to
linguoid,
independent
of
flow
velocity.
Ripples
in (C) and (D) are
therefore
developing, and (E)
shows
equilibrium
forms.
Figures
(B)-(E)
after
Reineck
and
Singh
(1980).
Copyright
© 1980 by
Springer-Verlag.
All
rights
reserved.
The geometry of ripples is expressed by the vertical form
index (or ripple index), L/H, the symmetry index,
LJ/LL,
and
the pianform (Figure R6). Most common vertical form indices
are 8-9 for current ripples and 3-8 for wave ripples, although
lower indices and indices up to 20 are no exception for both
types.
The mean symmetry index for a field of current ripples
is mostly >2.5, and dissimilar from typical indices of <2.5 for
wave ripples. However, co-existence of oscillatory and
unidirectional currents may produce combined-flow ripples,
whose symmetry index overlaps with current ripples, and thus
complicates genetic process reconstruction. Weak cross-sec-
tional asymmetry can also be produced by intrinsic resultant
mass transport of water in the direction of wave propagation,
mostly in shallow water. Current ripples comprise four
common styles of plan form: incipient, characterized by
mm-high solitary ripples or ripple patches on a flat bed;
straight-crested, with straight, continuous crest lines (Figure
R6(C)); sinuous-crested, with sinuous, continuous crestlines
(Figure R6(D)); and linguoid, with discontinuous, tongue-
shaped crestlines (Figure R6(E)). Although wave ripples can
have three-dimensional plan form, particularly in the presence
of unidirectional currents, wave interference, or under high
waves, straight crests with diagnostic bifurcation and abrupt
termination of crest lines are most common.
Flow over current ripples separates at the brinkpoint in an
upper zone of free flow and a zone of baekflow with separation
eddy in the lee of the ripple (Figure R6(A)). These zones are
divided by a zone of mixing, a shear layer along which Kelvin-
Helmholtz vortices generate high instantaneous shear. In the
ripple trough, the shear layer gradually trends downward until
the main flow becomes reattached to the bed. Current ripples
migrate in the flow direction by erosion of the stoss side and
deposition on the lee side. The migration rate is inversely
proportional to ripple size and proportional to local sediment
transport rate, which varies with the degree of sheltering by
adjacent ripples. Depositional processes associated with ripple
migration comprise suspension settling through the zone of
mixing and recurrent avalanching of bed load material from
the brinkpoint onto the lee slope (Figure R6(A)). Each
avalanche-settling cycle produces a cross lamina. Sets of
cross-laminae are diagnostic of migrating current ripples.
Preservation of cross-lamination is enhanced by net deposition
in flows that are oversaturated with sediment and produce
dimbing
ripple
cross-lamination (see Cross-Stratiflcation).
Flow over wave ripples comprises three stages: (1) During
passage of a water wave crest, flow separation occurs over the
ripple crest and a separation eddy and avalanche deposit are
produced; (2) During subsequent decelerating flow in the wave
cycle, the eddy rises, and may take suspended sediment with it;
(3) During passage of the water wave trough (with reversed
flow direction), the suspended sediment moves over the
ripple crest, a new separation eddy develops and avalanching
occurs on the opposite ripple face. Wave ripple cross-
lamination produced by net deposition from successive waves
is characterized by bundles of oppositely inclined laminae
(Figure R6(B)). For combined-flow ripples, the dominant
inclination of cross-laminae is parallel to the unidirectional
flow component.
Historical development
Ripple marks have been described extensively since the
beginning of modern observational geology. Sorby (1859),
Darwin (1884), Hunt (1904), Kindle (1917) and Bucher (1919)
established the foundations for modern ripple analysis. Allen
(1968),
Potter and Pettijohn (1977), Reineck and Singh (1980)
RIPPLE,
RIPPLE MARK, RIPPLE STRUCTURE
567
and Allen (1984) summarized further benchmark papers,
which include studies on ripples in recent and ancient
sediments, ripples in experimental flumes, and the theory of
ripple formation and stability. In the last two decades, ripple
marks have been used primarily in facies analysis. Yet,
innovative work continued in experimental flumes and
through mathematical modeling (e.g.. Miller and Komar,
1980;
Diem, 1985; Arnott and Southard, 1990; Gyr and
Schmid, 1989; Nelson and Smith, 1989; Best, 1992; Baas, 1994,
1999;
Werner and Kocurek, 1999; Baas et al., 2000). Best
(1996) and Raudkivi (1997) presented excellent reviews of
recent current-ripple literature.
Applications
Ripples and cross-lamination have been used as indicators of
stratigraphic younging and to reconstruct (paleo)flow direction
or (paleo)wave crest orientation. Moreover, ripples can be
used as estimators of local flow properties, notably flow type,
flow strength and sediment accumulation rate. Flow strength
can be constrained from bedform stability diagrams (see Surface
Forms).
Current ripples form at relatively low flow velocities
between the threshold for sediment movement and upper-stage
plane bed (for particles <~ 0.12 mm) or dunes (for particles
~0.12-0.7 mm). They do not form in cohesive clay and in
sand coarser than ~ 0.7 mm. Wave ripples exist in all silt- and
sand-grades, and between maximum oscillatory velocities for
the threshold of sediment movement and upper-stage
plane bed.
Unidirectional flows occur in many depositional settings,
which renders current ripples of limited use in paleoenviron-
mental reconstruction. Nevertheless, hydraulic conditions are
most favorable in sandy rivers, on intertidal flats, and in
turbidites. Wave ripples require relatively weak near-bed
oscillatory flow, such as on lake beaches, intertidal flats and
the marine shoreface. They are therefore better suited for
paleoenvironmental analysis than current ripples. Combined-
flow ripples have been found on intertidal flats and in other
shallow marine and coastal environments.
Ripple marks provide the sediment bed with form rough-
ness.
Form roughness is essential for calculating flow
resistance, bed shear stress and velocity distribution in
sediment transport analysis.
lit 1 • .
Current investigations and gaps in knowledge
Reflnement of existing models and development of new ideas
continually improve our knowledge of ripples. Present-day
work on current ripples concentrates on the mechanisms
controlling their formation (Best, 1992; Raudkivi, 1963;
Williams and Kemp, 1971), their rate of development toward
equilibrium morphology (Baas, 1999) and the feedback
relationships between ripple geometry and flow conditions
(e.g., based on "wave instability" theory by Richards, 1980).
A fleld of current ripples can be generated from a single,
artificial, or flow-induced, bed defect if the height of the defect
is large enough to induce flow separation at its lee side.
Baas (1994, 1999) demonstrated that current ripples
developing on flne- and very fine sand-grade flat beds evolve
from incipient, via straight-crested and sinuous-crested, to
linguoid equilibrium plan form with constant mean height and
length. Contrary to earlier work (e.g., Allen, 1968), the
equilibrium dimensions were found to be independent of flow
velocity. Velocity governs merely the time needed to reach
equilibrium geometry, which may range from several minutes
to hundreds of hours. Baas (1993) proposed relationships
between sediment size, D, and equilibrium current ripple
height, //,., and length, L^ {cf. Raudkivi, 1997):
3.41ogZ)-f 18
75.4 log Z)+197
mmm
(Eq. I)
Although the above model has been applied successfully to
current ripple dynamics in tidal environments (Oost and Baas,
1994) and used to reconstruct sedimentation rates in turbidites
(Baas etal., 2000), several gaps in our knowledge remain. First,
the role of turbulence in the growth rate of current ripples has
never been assessed in detail. Second, aggradation promotes
current ripple preservation, but no comprehensive study of its
influence on ripple dynamics exists, particularly under condi-
tions of turbulence modulation. This knowledge would
strengthen the use of current ripples in (paleo)hydraulic
recqnstructions, and in the determination of drag coefficients
in sediment transport calculations.
The latest studies on wave-induced bedforms concentrate on
the sediment entrainment under oscillatory and combined
flows, and the dynamics of such flows over wave and
combined-flow ripples (e.g., Traykovski etal., 1999; Paphitis
etal., 2001). This work will remain important, considering the
complexity of the feedback relationships between flows and
bedforms. A better process-based distinction between com-
bined-flow and current ripples is needed, as the use of indices is
unsatisfactory. This could be achieved by exploring potential
differences in grain shape fabric of current, wave, and
combined-flow ripples.
Jaco H. Baas
Bibliography
Allen, J.R.L., 1968. Current
ripples:
their
relation
to patterns of water and
sediment motion. Amsterdam: North-Holland Publishing Com-
pany.
Allen, J.R.L., 1984. Sedimentary struetures: their
character
and physical
basis. Elsevier.
Arnott, R.W., and Southard, J.B., 1990. Exploratory flow-duct
experiments on combined-flow bed configurations, and some
implications for interpreting storm-event stratification. Journalof
Sedimentary Petrology, 60:211-219.
Baas,
J.H., 1993. Dimensional analysis of current ripples in recent and
ancient depositional environments. Geologica Ultraieetina, 106:
199pp.
Baas,
J.H., 1994. A flume study on the development and equilibrium
morphology of small-scale bedforms in very fine sand. Sedimentol-
ogy, 41: 185-209.
Baas,
J.H., 1999. An empirical model for the development and
equilibrium morphology of current ripples in fine sand. Sedimen-
tology, A6: 123-138.
Baas,
J.H., Van Dam, R.L., and Storms, J.E.A., 2000. Duration of
deposition from decelerating high-density turbidity currents. Sedi-
mentary Geology, 136: 71-88.
Best, J.L., 1992. On the entrainment of sediment and initiation of bed
defects: insights from recent developments within turbulent
boundary layer research. Sedimentology, 39:
797-811.
Best, J.L., 1996. The fluid dynamics of small-scale alluvial bedforms.
In Carling, P.A., and Dawson, M.R. (eds.), Advanees in Eluvial
DynamiesandStratigraphy. John Wiley & Sons, pp. 67-125.
Bucher, W.H., t919. On ripples and related sedimentary surface forms
and their paleogeographic interpretations. American Journal of
Science, 47: 149-210, 241-269.
568
RIVERS AND ALLUVIAL FANS
Darwin. G.H,, 1884. On the formation of ripple-mark. Proeeedingsof
the Royal Soeiety of London, 36,
18-43.
Diem, B., 1985. Analytical method for estimating palaeovvave climate
and water depth from wave ripple marks. Sedimentologv, 32: 705-
720.
Gyr, A., and Schmid, 1989. The different ripple formation mechanism.
Journatof Hydraulic Research, 27: 61-74.
Hunt, A.R., 1904. The descriptive nomenclature of ripple-mark. Geo-
logical Magazine, 1: 410-418.
Kindle, E.M., t917. Recent andEossit Ripple-Marks. Museum Bulletin
ofthe Geological Survey of Canada, Volume 25,
pp.121.
Miller, M.C., and Komar, P.D., 1980. Oscillation sand ripples
generated by laboratory apparatus. Journal of Sedimentary
Petrol-
ogy, 50: 173-182.
Nelson, J.M., and Smith, J.D., 1989. Mechanics of flow over ripples
and dunes. Journal of
Geophysical
Research, C, Oceans, 94: 8146-
8162.
Oost, A.P., and Baas, J.H., 1994. The development of small-scale
bedforms in tidal environments: an empirical model and its
applications. Sedimentologv, 41: 883-903.
Paphitis, D., Velegrakis, A.F.,'Collins, M.B., and Muirhead, A., 2001.
Laboratory investigations into the threshold of movement of
natural sand-sized sediments under unidirectional, oscillatory and
combined flows. Sedimentotogy, 48: 645-659.
Potter, P.E., and Pettijohn, F.J., 1977.
Paleocurrents
and Basin Analy-
sis.
Second, corrected, and updated edition. Springer-Verlag.
Raudkivi, A.J., 1963. Study of sediment ripple formation, Journalof
the Hydraulics
Division,
Proeeedingsof ASCE, 89:
15-33.
Raudkivi, A.J., 1997. Ripples on stream bed. Journal of Hydraulic En-
gineering, 123: 58-64.
Reineck, Fi.E., and Singh, I.B., 1980. Depositionat Sedimentary Envir-
onments, with Reference
to Terrigenous
Clasties. Second, revised and
updated edition. Springer-Verlag.
Richards, K.J., 1980. The formation of ripples and dunes on an
erodible bed. Journal of Eluid Meehanics, 99: 597-618.
Sorby, Pl.C, 1859. On the structures produced by the currents present
during the deposition of stratified rock. The
Geologist,
2: 137-147.
Traykovski, P., Hay, A.E., Irish, J.D., and Lynch, J.F., 1999.
Geometry, migration, and evolution of wave orbital ripples at
LEO-15.
Journal of
Geophvsical
Research, C, Oceans, 104: 1505-
1524.
Werner, B.T., and Kocurek, G., 1999. Bedform spacing from defect
dynamics. Geology, 27: lH-l'iQ.
Williams, P.B., and Kemp, P.H., 1971. tnitiation of ripples on flat
sediment beds. Journal of the Hydraulics Division, Proceedings of
ASCE, 97: 505-522.
Cross-references
Angle of Repose
Bedding and tnternal Structures
Bedset and Laminaset
Cross-Stratification
Desert Sedimentary Environments
Flow Resistance
Flume
Paleocurrent Analysis
Sediment Transport by Tides
Sediment Transport by Unidirectional Water Flows
Sediment Transport by Waves
Sedimentologists
Surface Forms
Turbidites
RIVERS AND ALLUVIAL FANS
Introduction
A river is a natural stream of water that under the influence of
gravity flows regularly or intermittently in a channel or
channels toward a receiving basin, commonly an ocean or
lake.
Rivers are open systems in which energy and matter are
exchanged with the external environment (Knighton, 1998).
They are supplied with water almost entirely from precipita-
tion (meteoric water) routed to the channel via overland flow,
groundwater, swamps, lakes, snowfields, and glaciers, and they
acquire most of their sediment load by dissecting uplands or by
reworking unconsolidated debris from previous erosional
events. Sediment deposited by rivers in subaerial settings is
called alluvium. Consequently, rivers are divisible into bedrock
reaches with rigid boundaries, and alluvial
reaches
that have a
mobile boundary of relatively unconsolidated material, often
with vegetation assisting channel stabilization. Rivers drain
catchments (drainage basins or watersheds), are organized into
complex patterns of trunk, tributary, and distributary chan-
nels,
and commonly support adjacent floodplains that are
inundated when the channels exceed bankfull capacity. Their
headwaters are principally in upland areas and they may flow
throughout the year (perennial) or cease for part of each year
(intermittent if flow is seasonal and ephemeral if flow is
irregular). Due largely to tributary contributions, river systems
generally increase in discharge and channel dimensions down-
stream, however, in areas of permeable sediment or aridity,
percolation, and evaporation may cause channels to reduce or
disappear.
While only about 0.005 percent of continental water is
stored in rivers at any one time (Knighton, 1998), rivers
transport to depositional sinks the great majority of sub-
aerially weathered and eroded sediment as well as dissolved
material. River drainage networks occupy about 69 percent of
the land area, transporting an estimated 19 billion tonnes of
material annually, about 20 percent of which is in solution.
Denudation rates and river load reflect primarily the effects of
climate,
relief,
and tectonic uplift, as well as rock, and soil
type,
catchment size, vegetation cover, and human influence.
Upland rivers are commonly bedrock systems that form a
sediment source zone (Schumm, 1981). They are typically
erosional with very limited sediment storage. Intermediate
reaches usually form a transfer zone with disjunct areas of
bedrock erosion and temporary deposition. Lowland rivers are
dominantly alluvial. They lose competence to form a sinkzone,
depositing their load in floodplains, alluvial fans, lakes, desert
basins (internal sinks), or in marine deltas, estuaries, lagoons,
continental shelves, and ocean basins (external sinks). Many
rivers have responded to sea-level fluctuations by cutting and
filling their lowermost valleys.
Fluvial strata constitute about 10-20 percent of the
Phanerozoic rock record, with many predominantly alluvial
deposits in excess of
10
km thick. The oldest known fluvial
deposits are in the Archean of southern Africa, where they
date back to more than 3.5 Ga and reflect the emergence of
continental crust and establishment of subaerial weathering.
Alluvial detritus is formed principally of silicate gravel, sand,
and mud, with volcanic, carbonate, and evaporite grains, and
particles of gold, tin-bearing minerals, and other economic
materials (see Placers, Fluvial). Some deposits contain huge
boulders derived from rock falls and catastrophic floods in
steeplands. Rivers that traverse peatlands and densely vege-
tated regions transport large amounts of plant material and
dissolved and colloidal organic matter, and these blackwater
streams contribute significantly to the carbon content of
sediments in sink zones. Alluviai strata include a wide range of
sedimentary structures mostly generated by unidirectional
RIVERS
AND
ALLUVIAL FANS
569
flow. Floodplain deposits
are
usually associated with soil
formation
and
consequently,
in
combination, alluvial strati-
graphy
can
reveal
a
detailed story
of
variable flow eonditions
and changing terrestrial en\ironmenls over spaec
and
time.
Historical developments
The importance
of
river erosion
and
landscape formation
was
elearly recognized
in
classical times. Whereas Thales
of
Miletus
(/».
-624 BC)
thought that deltas building into
the sea
were
evidenee that water could change into sediment. Herodotus {b.
484
BC)
identified seditnentation
and
de.seribed
the
alluvial
plains
of
Egypt
as the
"'gift"
of the
Nile. Plato
(/>. 428 BC)
clearly reeognized fluvial erosion
and
transportation. Aristotle
(/'.
3S4
BC)
was the
first seholar known
to
have eoncepttialized
the modern hydrologieal cycle
and he
also identified river
erosional
and
depositional proeesses, partieularly those
on the
Nile
and the
silting
up of
river channels entering
the
Black
Sea.
Strabo
(^.
64/63
BC)
described floodplains
as
having been
formed
of
sediment from their adjacent rivers. However,
between Roman times
and the
Renaissanee
a
long scientific
interregnum prevailed
in
Europe, with
a
widespread belief that
ihe Earth's surface
had
been shaped
by
eatastrophie proeesses
associated with
the
biblical Great Flood
(the
Diluyialist
iheory).
After
the
geologieal importance
of
hydraulies
and
river
proeesses
was
rediseovered
by
Leonardo
da
Vinci. Agrieola,
Bernard Palissy, George Bauer
and
others
in the
i5th
and 16th
Centuries,
the
importance
of
fluvial processes through
geological time
was
questioned again
in the 18th
Century
due
to a
group lead
by a
German .scholar, Abraham Werner.
Known
as the
Nepiunisis. ihey proposed that
the
tock record
represented itepositicm
in
deep oceans where powerful currents
had earved valleys
and
ridges, restricting alluvial influence
to
modern rivers
and
their sediments. James Hutton
in the 18th
Century
was the
first post-elassieal seholar
to
elearly reeognize
and articulate
the
importance
of
modern-day processes
to
explain features visible
in
ancient roeks,
the
basis
of his
principle
of
Unilbrmitarianism.
He saw
that long-term fluvial
erosion
was the
erosive mechanism active
in the
repetitive
cycles
of
landscape denudation: No trace of a
hegittnittg,
ttopro-
speetofanend.
He
argued that fluvially eroded debris from
the
land created most
of the
detritus appearing
in
ancient
sedimentary strata,
and
henee
he and his
followers becatiie
known
as the
Fluyialist.s. Hiitton's concepts were clearly
espoused
by
Charles Lyell
in the
many editions
of
his seminal
book
Tbe
Principles of Geology
(1830
'|833).
ll
was not
until
the
lale
19th and
early 20th Centuries that
modern
and
ancient river deposits
and
fluvially eroded
landscapes received widespread
and
detailed investigation,
espeeially from North Amerieans stich
as J.W.
Powell during
exploration
and
settlement
of
that continent.
G.K.
Gilbert
(1877) was
the
first
to
apply
the
concept
of
dynamieequdibriiitn
to rivers,
and W.M.
Davis
in the
late 1890s proposed
the
normal
or
fluvial cyele
of
erosion
and
landscape change that,
although
now not
wholly aeeepted. fostered
a
widespread
interest
in
modern geology
and
geotiiorphology. Davis
was
probably
the
first person
to
recogni/c braiding
as a
river
pattern distinctly different from meandering.
In 1925. J.
Barrell demonstrated that seditnentary strata, hitherto widely
ihought
to be
deposited
in
marine environments, were
in
many
eases terrestrial ineluding fluvial.
The
outstanding study ofthe
Mississippi Ri\er
by H.
Fisk (1944) deseribed
the
fluvial
geomorphology. sedimentology.
and
stratigraphy
of a
very
large river system. Miall (1996) regarditig this work
as the
first
major advance
in
modern fluvial sedimentology.
Post-war fluvial faeies analysis, strongly linked
to
drilling
and exploitation
of
hydrocarbon reservoirs,
led in the
I96()s
lo
J.R.L. Allen's eomprehensive model
for
meandering-river
deposits,
the
first
fbr any
environmental setting. Undoubtedly
the most significant modern advanees
in
understanding river
form
and
process originated from research
by L.B.
Leopold.
M.G. Wolman.
S.A.
Sehumtn.
J. M.
Miller
and
their assoeiates
in
the
1950s
and
1960s.
R.A.
Bagnold
(see
Fliysies of Sediment
FransportiThe Coniributiotisof
R.
A.
Hagttold) greatly advanced
seditnent transport theory,
and A.
Sundborg described
in
detail tneandering ri\er processes
and
floodplain sedimentol-
ogy. providing important Kuropean conlribuliotis. .Sediment
dynamic e.vperiments. following studies
in the 19th
Century
by
H.C. Sorby. allowed detailed faeies interpretation
in
hydro-
dynamie ternis
(see
Flumes: Sediment Transport
by
Unidiree-
tionat
Flows).
With
the
development
of
radioearbon dating
in
the late l'-)40s
and
luminescence dating
in the
1970s,
the
means
of dating Quaternary alluvium became available, allowing
fluvial sediments
and
paleoflow rcgitnes
to be
related more
precisely
to
wider controls, principally tectonic aetivity,
eustasy.
and
clitnate ehange.
Theoretical underpinnings
II
has
been tecognized since
the 18th
Century that quantitative
analysis
in
fluvial studies
has
been
of
necessity
a
compromise
between theory (based
on
"ideal nonviscous fluids")
and
etnpiricism (based
on
real viseous lluids).
a
problem known
as
the d"Aietnbert paradox. Despite major advances,
the
devel-
opment
of
unifying theories
in
fluid mechanics
and
sediment
transport remain elusive. Empirical research into stoehastie
relationships
has
shown that,
as
flows vary, rivers construct
highly predictable channel forms
and
sedimentary structures.
However,
a
truly rational
or
deterministic explanation
for
such
consistency
has not
been possible
due to a
lack
of
mathema-
tical closure,
for
while there
are
four flow \ariabies (width,
deplh. veloeity,
and
slope) fhere
are
only ihree detertnining
eq tuitions (eotitiiniity, resistance,
and
sediment transport).
Solutions have been sought
by
adopting extremalbypothe.se.i.
If
fluvial and thermodynamic systems
are
compared, rivers
may
be visualized
as
adopting
a
most probable state
by
minimizing
both
the
variance
of the
system's
How
components
and the
total work done. More strongly physically-based extremal
hypotheses have been proposed
by
imposing assumed eondi-
tions sueh
as
maximum sediment transport rate
or
minimum
stream power.
In a
reeent reassessment
of
sotne
of
these
approaehes. Huang
and
Nanson (2000) have demonstrated
mathematically that straight reaehes
of
alluvial rivers appear
to operate
at
maximum efficiency
and
illustrate
the
basic
physical Frineiple
of
Lea.it Aetion. However, such theoretieal
analyzes remain contentious.
Drainage networks
Drainage basins
are
usually characterized
by
interconnected
lengths
of
river ehannel that converge frotn
the
elevated basin
margins toward
a
trunk stream that exits
at the
lowest point.
The resulting eombination
of
ehannels forms
a
drainage net-
work (Figure
R7).
Drainage networks refer
to the
pattern
of
the network
of
stream ehannels wilhin
a
basin
or
sub-basin.
570
kiVERS AND ALLUVIAL FANS
Order - Formative streams
Order - Excess Streams
- - ' Exterior links
• --' Interiof links
Figure R7 Drain.ige-basin s!ream networks with stream ordering: (A) as
by Horton (194'j) (modified by Slr<ihler); (B) channel links nrdered by
magnitude (atter Knighton, f 998).
and should not be confused with .siteatnpatterns that refer to
the planform appearanee of individual reaehes of river (e.g.,
tneandering and braiding, see below). They are usually
erosional features and have been mapped in the strafigraphic
reeord where unconformable surfaces separate bedroek from
overlying alluvium. Headwater streams are normally relatively
small, steep, often bedrock controlled with uneven gradients,
and the bases of the hillslopes are eonneeted direetly to the
stream ehannels such that rates of ehannel incision and
sediment removal relate closely to slope proeesses. Down-
stream, the basin channels are of lower and tnore uniform
gradient, and are cotnmonly self-forttted. being shaped within
their own alluvium. Although in places they may connect with
the hillslope. it is more eommon that alluvial terraces and
floodplains act as a temporal and spatial buffer between
channel and slope proeesses. Channel links are a length of
channel from the source to the first tributary {externallink) or
between two suecessive tributaries (iniernal
link).
In
f945.
Horton developed a system of ordering chainiel links within
drainage networks in a downstream direction (Figure R7(A)).
This provides a useful classification of the position of streams
within a network, although link order is not a reliable
indicator of channel size or discharge. To determine link mag-
nitude, eaeh link is assigned the value of the sum of all the
external links supplying it, providing a useful surrogate to link
discharge (Figure R7(B)).
In terrain of homogeneous or alternatively highly complex
geological structure, drainage networks tend to take on a
strongly irregular dendritic pattern. However, where well-
defined and relatively simple geologieal struetures are present
(faults, folds, domes, volcanic cones, rectangular joints etc.).
visually distinctive and sotiiewhat regular drainage patterns
can form. Drainage networks are highly amenable to
numerical analysis and in consequence they have been studied
in great detail in attempts to interpret their controlling
variables, internal relationships, sediment yields and their
relationship to both catchment and overall landscape evolu-
tion. However, because most networks are dendritic, their
essentially irregular pattern means that few physically defini-
tive relationships have been identified, except perhaps those
from topofogieal studies that have pointed out the spatially
nonrandom elements present.
The most widely used and probably useful statistic derived
from drainage network analysis is that of drainagedensity (the
total length of ehannels per unit area ofthe basin). It expresses
the degree of basin disseetion and sediment produetion and
transport by surfaee streams, and is influenced by sueh
variables as elimate, roek, and soil type, vegetation density,
and landuse. Highly impermeable, soft, and poorly vegetated
roeks can produce high drainage densities and sediment yields
(e.g.. badlands). Drainage density is also broadly eorrelated
with preeipitation. It is low in very arid areas, greatest in semi-
arid areas where there is suftieient rainfall to eause erosion but
insufficient vegetation to itnpede it, and lower again in more
humid areas with denser vegetation. It appears to partially rise
again in very wet areas where the stabilizing effeets of even
dense vegetation are overeome.
Stream gradients
Stream gradients are the result of two broad eauses. An
independent gradient is imposed on the stream by the gross
valley form, the produci of geologieal and drainage history.
However, an adjustable and therefote dependent eomponent
of gradient develops frotn the interactton of st rea tn dtseharge,
width, depth, velocity, seditncnl size, sediment load, boundary
roughness, and path sinuosity. Mackin (1948) stated that a
gradedtirequilibrium stream (see below) is one in which, over a
period of years, slope is delicately adjusted to provide, with
available discharge and prevailing channel characteristics, just
the veloeity required for the transportation of the seditnent
supplied from upstream. Confined upland streams are
generally not equilibrium alluvial systems beeause their
gradients are largely imposed by a bedroek valley that is the
produet of ancient stream erosion or forces such as glaciation
and teetonism. It is in the middle or lower reaehes where the
valley is wider that the stream can more readily adjust its
gradient by altering its sinuosity and hence its path length in
response to contetnporary eonditions. Change in ehannel slope
accounts for only part of the adjusttnent of alluvial rivers to
their controlling factors and, more realistically, the ehatinefs
variotis morphological and hydraulic parameters tend to show
mutual adjustment.
Over their full distance, the longitudinal profiles of natural
rivers show a strong tendency to become concave upward. This
eoncavity is assoeiated with at least three possible causes,
although the direetions of cause and effeet are difficult to
resolve. F'irstly. there is a tendency iti most rivers fbr discharge
to increase downstream but for velocity to remain fairly
constant. Discharge (a volume) increases as a cubic function
whereas the resisting boundary of the channef (the area of the
bed and banks) inereases only as a squared function. If the
gradient did not deeline substantially an imbalance between
impelling and resisting forces would result and the Mow would
accelerate rapidly downstream. Profiles tend to be most
concave where discharges increase downstream most rapidly.
Secondly, grain size cotntnonly declines downstream due to
RIVERS AND ALLUVIAL FANS
571
elast abrasion (see .Attrition, Fluvial) and size sorting, so the
gradients required for sediment entrainment and transport will
decline downstream. Stream profiles are more concave where
eiast sizes deeline in size downstream tnost rapidly and show
little concavity where grain size is constant or increases. Slope
decreases tnost rapidly with a deeltne tn gravel stze and least
wilh a decline in sand size. Because ofthe production of finer
clasis.
streams over shale tend to have gentler gradients than
those over sandstone or limestone, although concavity depends
on the eotnbined effeets of grain size and discharge. A third
possible reason for profile coneavity is that antecedent relief
and potential energy conditions along a river from headwaters
to mouth eause random-walk models to develop concavity as
the tnostprt>b(tNe profile. In relation to this proposal it has
been Ibund that shorter streams tend to have less eoncave
profiles and streams with greater relief exhibit greater
eoncavity.
Marked ehanges in river gradient are termed knickpoints.
and may refleet the presence of more resistant bedrock.
changes in sediment load (fbr example, from tributary
streams), tectonie activity, or base-level changes in the past.
Fustatic or other base level ehanges tend nol to be propagated
over long distances upstream unless the stream is eonfined
and ean nol readily adjust other parameters such as
sinuosity.
Biota and soils
Prior to the tnid Paleozoic, subaerial erosion was dominantly
physical and prodticed abundant coarse material, with river
forms probably reminiscent of today's active glacioiluvial or
volcanic landscapes. In the late Silurian and Devonian the
evolution of terrestrial plant eotnnutnities and associated soils
greatly enhanced chetnical weathering and the production of
clays.
The development of cohesive stream banks and
stabilizing root systems must have ehanged rivers dramatically
in the late Paleozoic, promoting the accumulation of
substantial argillaceous and organie deposits in terrestrial
environments and strongly affeeting global patterns of earbon
producf ion and storage. The development of terrestrial
vegetation also redueed runoff and allowed the diversifieation
of a complex array of lloodplains and associated soil types in
environments ranging from rainforests to grasslands (af\er the
mid Cenozoic) and deserts. The first peats (precursors to coal)
were deposited in the Devonian and peat has subsequently
been an important component of fluvial strata. The rapid
succession of plant communities in response to climate ehange.
termed lltnvl overturn, should have exerted an itnportant
influence on sedinient flttx to depositional basins.
There is now appreciation ofthe importanee and eomplexity
of river-vegetation interactions. Particular attention has been
given to the efTeet of vegetation growing within channels on
flow resistanee, of bankline vegetation and debris dams on
bank strength and ehannel tnorphology, and of riparian
\egetation generally on floodpfain and ehannel seditnentation
atid erosion. Wildfires destroy vegetation and protnote land-
scape degradation, greatly increasing sediment yield and
adding ehareoal to the fossil record.
!-loodplain soils ean be diagnostic of particular environ-
ments, eontaining evidetice in the form of poifeii and spores,
invertebrate remains, burrows, rhizotnorphs, evaporites, and
desiccation cracks (see Weathering, Soils, and
Paleo.sols).
For
this reason and beeause soils develop through close interaction
with the atmosphere and the groundwater prism, they are
important repositories of information about past climates.
Large animals, including hippopotamus, beaver, and perhaps
dinosaurs, have affeeted rivers by trampling sediments,
damming ehannels, and ereating paths aeross banks that can
lead to avulsion.
River hydraulics and sediment entrainment
Flow m open channels is subject to tuo roughly opposing
prineipal forces. Acceleration due to gravity {g) acts to move
water and seditnent downslope with its effectiveness modilied
by stream gradient [S). while flow resistanee (channef rough-
ness) (see Fhny Resistanee) opposes this downslope motion.
The interaction of these two forees ultimately determines the
ability of flowing water to erode and transport sediment, and
thereby generate fandforms. Flow in rivers is almost always
turbulent. The flow boundaries including the water surface are
deformable. thereby changing the shape, roughness, and
energy conditions in the ehannel as the flow changes.
Frictional losses are the product of boundary roughness,
internal distortion, and changing flow phases (super- fo sub-
eritieal). All these factors make fluid tneehanies much more
complicated than solid mechanies, however, relatively simple
empirieal formulas have been developed to obtain approx-
imate relationships suitable fbr most straightforward analy-
tieal procedures.
River discharge (Q)\
Q = wdy
(Eq.
where tt= ehannel width; (/=mcan depth;
v
= mean velocity.
This is known as the
flow
continuity eqttatiott.
Mean boundary shear stress (r):
z^/RS
(Eq.2)
where 7 = speeifie weight of water; /? = hydraulie radius ^d.
{•;
=
!>
g, where /) is water density). This is known as the Du
Boys
equation.
Channel roughness or How resistanee is commonly obtained
from one of two equations;
Manning equatkni:
n=\/vR-''S^^' (Eq. 3)
Darey
IVeisbaeli
eqtiatiott:
f = SgRS/v- (Eq. 41
The work done by a river ean be expressed in terms of stream
power in two fbrms:
Power per unit length ofthe channel:
Q^VQS (Eq. 5)
Power per area of the bed:
ff.! = ,' QS/w (Eq. 6)
While suspended and dissolved sedimetit load can be relatively
easily measured the more eomplex problem of determining bed
load is usually attempted using one of many available
transport equations, none of whieh are reliable over a wide
range of conditions (see Nutverical Models and Simulation of
Sediment
Transport
and
Depositloti;
Seditnent
Flu.xes
attd Rates
ofSeditventation). These equations ean
be
classified info three
S72 RIVERS AND ALLUVIAf FANS
basie types (Knighton. 1998):
Excess shear stress
(r,,-r,~,):
a
h
= X' r (r - r 1
Excess discharge
Iq-q,-,)'-
q^H
=
X".s'[q
- q,,)
Excess stream power
((»-<
Du Buys type
Schoklitsch type
(Fq. 7)
(Fq. 8)
q,i,
^ ivj -
oj,,)-^'-
d-^ /?-'•'= Bagnold type (Fq. 9)
In the above, ty^/, is bed load per unit channel width. X' and X"
are sediment coefficients. D is grain size, q is How discharge per
unit width of channel and the subscript "CT" denotes the
eritieal eondition fbr sediment motion.
In a straight river eross section, the flow velocity is usually
fastest at or just below the surface near the eenter of the
channel and declines toward the bed and banks, the How field
def\>rniing through river bends (Figure R8). A narrow deep
ehannel usually directs a relatively gentle veloeity gradient to a
fine-grained bed that requires low shear stresses for transport.
and directs relatively steep gradients to banks that are often
cohesive, well vegetated, and erosion resistant. Wide shallow
ehannels tend to exhibit erodible banks and coarse and/or
abundant bed load that requires high shear stress for
transport, braided rivers being a classic type.
The erosion of cohesive material such as bedrock or
partially indurated alluvium occurs in three ways. Corrosion
is the chemical dissolution of roek in water and ean be
important where rocks are readily soluble (e.g.. limestone) or
where water is chemically active. Corrasion occurs when rock
is mechanically braided by water armed with partieies cither in
transport or stationary but vibrating on the roek surfaee. Ca-
vitation results from the produetion of shock waves due to the
formation and implosion of vapor bubbles on rock surfaces at
very high velocities. This is a rare proeess in tnost subaerial
streams but can be common in subglacial or underground
streams where flow is eonfined.
The entrainment of noncohesive sediment from the bound-
ary is a eomplex process but largely a function of flow velocity
and grain size. In the ease of gravels, the sorting, and packing
arrangements ofthe elasts on the bed are also important, for
an armored or itnbrieated bed can be diffictilt to entrain (see
Arnu>r).
As flow veloeity increases two dominant fbrees act to
entrain individual clasts. A drag force caused by friction
between the flow and the surfaee of the particle causes it to
(B)
Figure R8 Velocity fields in cross sections ot; (A) a wide shallow
channel (note tho steep velocity gradient to the bed): (B) a narrow deep
channel bend viewed downstream and curving to the left (note ihe
steep velocity gradient against the outer
move downstream and to slide or roil (or pivot) over
any obstructing particle. However, as velocity increases a
pressure gradient develops between the top and bottom of the
grain setting up a lift force that acts to lift tbe partiele out of its
recess on the bed. The lift force decreases rapidly away from
the bed but once the partiele is elevated it is far rnore
suseeptible to the drag foree and to turbulent eddying within
the body ofthe flow that acts to keep partieies in motion. The
entrainment of sediment on the boundary ean be expressed in
terms of mean flow velocity and mean grain size, but the
problem is to define the threshold velocity for motion on an
imbricated or armored bed, or whete grain size is highly
variable.
The competence of historic or prehistoric flows is commonly
judged from the size of the largest partiele in the resulting
deposit, although such estimates may suffer from the limited
availability of coarse sediment. Fine-grained, eohesive deposits
require mueh higher flow velocity for entrainment and
transport than the size of individual grains would suggest.
Sueh materials inelude bed and bank tnaterial. sand-sized mud
aggregates reworked from soils, and floes of aggregated elay.
organistns. and detrital organie material in rivers and estuaries
(e.g.. Gibling eial.. 1998). Beeause aggregates and lines are
comtnonly destroyed by eompaetion during burial, flow
competence tnay be diffieult to assess fbr some fine-grained
material, whieh may have originated as sand-sized bed-load
rather than from settling of suspended flakes.
Equilibrium and threshold theories
Because alluvial rivers are open systems with mobile and
deformable boundaries, they have the ability to self regulate, lf
perturbed, they will often return to something like their
original eondition (homeostasis). or they will adopt a new form
but one that minimizes the effects of the original change. This
reflects dynamie equilibrium, a eondition first described for
rivers by Gilbert (1877) and applied widely in seienee.
especially ehemistry (Le Chatelier's Principle). Because rivers
are relatively slow to adjust and beeause their morphologieal
parameters adjust al different rates, it ean be difficult to
determine if a partieufar reach is in an equilibrium or balaneed
eondition. Nevertheless, the recognition that alluviai rivers
adjust toward equilibrium eonditions has becotne a basie tenet.
almost a paradigm of modern fluvial research. Richards (1982)
has suggested that equilibrium conditions in rivers can be
identified on the basis of four criteria: (1) essentially stable
relationships between form and process; (2) continuity of
sediment transport in a reach over time and space: (3) strong
correlations between systetn variables: (4) an adjustment to
maximum operational efficiency. In tnany studies roughly
balanced eonditions have been shown to exist over short
periods between erosion and deposition in a single reach, or
between width, depth, and velocity, tf one variable is altered,
the others adjust in a way that minimizes the overall change
and maintains a balaneed condition. Beeause flow conditions
and sedimentary bedforms are usually in dynamic equilibritim.
ancient flow structures permit an accurate interpretation of
paleoflow conditions.
Most ehanges in rivers are the result of external factors such
as changes in base level, clitnate. vegetation, or sediment
supply, and occur gradually. Rivers in dynamie equilibrium
generally resist change, however, Sehumm (1973) has shown
that an extrlttsie thresbold can be reached when a progressive
RIVERS AND ALLUVIAL TANS
573
change in an external variable triggers a sudden change in Ihe
system's response. At an imposed critical ehange of slope or
sediment load, a meandering channel can change abruptly to a
braided channel. Similarly, a gradual and progressive increase
in the llow velocity will suddenly achieve the threshold tor
sediment enlrainmeiit. after whieh the whole streambed
becomes mobile. There are numerous thresholds above critical
enlraiiiment when bedforms change from a motionless bed. to
ripples, lo dunes, to traitsitional. and to super-critical flow
bedfonns (see Surface Forma). Schumm (1973) also showed
that changes may be initiated intrinsically when, with no
external change, one of ihe variables reaches a eritieal
condition (an iiurinsic thrcslmld). While operating wilhin its
normal range
ot"
waler discharge and sediment load variation,
a river may reach a point of incipieni instability when a
threshold is crossed and a sudden internal ehange is initiated.
A meander eutofl" is an example where gradual, ongoing
adjustments to equilibrium eonditions prevail until a threshold
is reached, with breaching of the meander neck and a dramatic
chanye in loeal sinuosifv and gradient (Kiitiihton. iy9S).
Crossing an intrinsic threshold will normally produce only a
localized change recognizable in the stratigraphic record as. for
example, an avtilsion ehannel or erevasse splay.
Hydraulic geometry, regime theory and
dominant discharge
Chiinnel geomcliv is the eross-sectional form of a stream
channel {width, mean, or maximum depth, cross-sectional
area) lashioned over a period of time in response to formative
diseharges (approximately bankfull) and sediment character-
istics (Figure R9). Beeause the abo\e three geometric para-
meters and the additional four flow parameters (velocity,
water-surface slope, tlow resistance, and sediment concentra-
tion) vary with diseharge. the term hyilraulicficomtiry is used
to describe the relationships of all seven parameters lo
discharge as the independent variable. Discharge changes can
be measured at-a-station as the channel fills during a flood, or
at bankfull in the downstream direction. There are significantly
different relationships for til-a-sldlioii and
Discharge
N.6.
AN
scales are logarithmic
Change downstteam for
discharge of given frequency
_ Change at gagitig station fot
discharges of different frequencies
Figure R9 I Ivdr.iulic geometry relationships ol river (han nek, compjrinp variations in width, depth, velocity, suspended
load,
tnughness, and
sidpc lo varijlions in discharge, Ixilh <it-<i-station and downslretim (alter Leopold and Maddock, lO'i^).
574
RIVFRS AND ALLUVIAL FANS
hydraulic geometry (Figure R9). Holding diseharge constant,
at-a-station hydraulic geometry is largely conlrolled by
variations in bank, strength and available sediment
load.
Streams with low sediment loads and cohesive or
well-
vegetaled banks lend to be relatively narrow and deep whereas
those with abundant loads and weak banks tend to be wide
and shallow. However, beeause bank strength has only a
moderate range but river discharges vary by many orders of
magnitude, hydraulie geometry is remarkably eonsistent aeross
the full range of river discharges (Figure R9). Beeause channel
depth is greatly restrieted by limited bank strengths, rivers
increase in width relative to their depth as their size and
discharge inereases—a prominent downstream tendency
(Chureh.
1992).
Research during the construction and maintenance of
irrigation canals with near-constant diseharge in India and
Pakistan in the late 19th and early 20th Centuries indepen-
dently determined very similar hydraulic geometry relation-
ships to those in natural rivers, engineering findings that have
been termed reginwiheory. Consequently, stable alluvial rivers
exhibiting consistent and predictable hydraulic geometries are
said to be in eiiuilihriiini or in regime.
Hydraulic geometry shows that river ehannel dimensions are
closely adjtisted to water discharge. However, discharge varies
from perhaps no flow in droughts through to catastrophic
tlood events: so which discharge(s) define the channel's
characteristics? Woiman and Leopold (1957) showed that in
the USA, bankfull flows oecur with the surprising regularity of
about once every I 2 years aeross a diverse range of rivers,
something that would be an extraordinary coincidence if
bankfull flows did not in themselves play a large part in
determining channel dimensions. It has also been shown that
with increasing at-a-station discharge, flow velocity tends to
increase until near bankfull llow conditions and then stabilizes
at higher discharges because ofa marked increase in roughness
near the bank crests and over the floodplains. In other words.
most flows beyond bankfull are not notably more effective in
altering the ehannel and transporting sediment than is bankfull
flow. Furthermore, while in some cases exceptional floods may
undertake signilicant work in the form of sediment transport
and channel reconstruction, they are sufficiently rare that on
an average annual basis, they usually achieve far less than do
smaller but more frequent events of about bankfull
(Figure RIO). However, there is evidence thai in a limited
number of environments
(e.g..
pro-glacial streams, confined
gorges, and some dryland rivers), catastrophic events ean
achieve substantial and long-lasting changes. In other words,
here these become the dominant diseharge. Extreme, high-
veloeity events cause eonsiderable channel enlargement,
followed by a long period of "recovery" from smaller flows.
Thus, channel dimensions at a given time in such an
environment may reflect the period that has elapsed since the
last extreme event, whieh also terminated '"memory" of the
pre-existing channel geometry.
Sediment transport
As discharge increases, detrital sediment transport in alluvial
channels mostly occurs when the flow conditions (iargeiy
velocity) exceed the critical condition for entrainment of
sediment at rest on the bed and banks (see Numcrleal Models
andSimulalionof Sedimcnl TrunsporlandDepo.silion). The flow
conditions required to entrain stationary sediment are greater
t "O
3 .3
-.,
O CD
ro
o
E
t o o
O
C .„
D.
dJ
O
(0
=1
=)
c
crX)
m
IU £
H ul D.
Discharge
Qe Most effective transporting discharge
Qt Threshold discharge
Figure
RIO
Dominant
discharge.
Curve
A is the
Iransport rate
of
supenrleH load rising with
discharge-
Curve B
is the
frequency
of (he
tuli range
of
possible
discharges.
Curve
C is the
product
of
curves
A
and
B and
shows thai
the
most effective discharges
tor
transporting
a
river's load
are
moderate
floods,
generally occurring aboul once every
1
to 5
years.
than those required to keep sediment moving. Rivers transport
their sediment load in essentially four ways.
Bed
load or uac-
rionload']^ almost constantly in contact with the bed. saluiiing
load bounces or skips over the bed (effeetively intermediate
between bed load and suspended load),
.suspendedload
is held
in the water column by turbulence and dissolved load is
transported in solution. The first three are composed of
detrital sediment, and the eoncentrations and relative propor-
tions of the first two are highly dependent on the energy of the
flow, whieh inereases with water diseharge. Suspended load is
energy dependent but can also be strongly influeneed by the
rate of sediment supply to the river. The concentration of
dissolved material depends largely on water temperature,
catchment geology, groundwater chemistry, and vegetation. In
contrast to detrital
load,
dissolved load commonly decreases in
concentration due to dilution elTects with increasing flood
discharge. The three detrital types are the primary constituents
of alluvial strata, with bed. and saltating load commonly
forming sedimentary flow structures diagnostic of variable
flow conditions. Soluble matter, once in solution, is mostly
flushed from the catchment, being precipitated only under
unusual geochemical conditions. Because of the difficulty of
separating dilTerent sediment types on the basis of transport,
another eommonly used classification defines the detrital
sediment load in terms of the bed composition prior to
entrainment. Bed-uunerial load is that which is contained
largely within the bed material at rest, whereas
xva.shload
is the
muddy suspended component not found in significant quan-
tities in the bed (it is derived largely t>oni fines in the stream
banks, floodplains, and valley side-slopes).
Yields of suspended sediment and dissolved material are
usually closely related to water discharge and are therefore
relatively easily measured. In contrast, bed load (especially
coarse bed load) is transported in a highly variably t^ashion.