1 River patterns GEO3-4305, 2014 Introduction 1. Channel and bar

This course: the morphodynamic system
Research group
River and delta morphodynamics
River patterns
GEO3-4305, 2014
Dr. Maarten Kleinhans
[email protected]
www.geo.uu.nl/fg/mkleinhans
Introduction
1. Channel pattern classification
•
•
•
bars
channels
channel pattern prediction
2. What really determines bar pattern?
3. What really determines channel pattern?
flow
sediment
transport
morphology
•
•
•
•
Introduction
River flooding
Hydraulic roughness and
bedforms
•
•
Sediment transport
Mixture effects
•
•
•
•
•
Hydraulic geometry
Bars, bends, islands
Overbank sedimentation
Channel patterns
vegetation
1. Channel and bar patterns
Many classifications and
phenomenologies…
A few theories…
Meandering:
initiating from alternating bars (vertical)
initiating from bend instability (planform)
BIG QUESTIONS
 What determines bar pattern?
 What determines channel pattern?
Alternating bars
braided river
transitional
meandering river
W/h < 20
no bars
20 < W/h < 30 stable alternating bars
W/h > 30
dynamic alternating bars
1
Andre Ermolaev
8/30
Channel pattern stability diagram
0.42
 v,t  900D50
Kleinhans & van den Berg, cond. Acc ESPL
After Nanson & Knighton (1996)
9/30
van den Berg 1995
The power of patterns
Explanation??
Why valley stream power?
Streampower? ω = τu = ρgQS/W
Van den Berg (1995):
potential specific valley-related streampower
Why median grain size?
where
Braided separated from meandering by
12/30
2
The data
Kleinhans & Van den Berg (2010, ESPL)
• No human impact
• No entrenchment
or confinement
• Data from literature
• Google Earth
• Mean annual Q
indept. of morph
• Valley gradient
unpolluted by
sinuosity
14/30
Why does it work??
Deliberate misprediction of channel width
wide shallow river
→ narrower, deeper, higher ω
→ braided river
Bank strength!
→ floodplain formation and destruction
(see review paper Kleinhans 2010)
Kleinhans & Van den Berg (2010, ESPL)
16/30
2. Explaining bar patterns
Ships don’t like bars...
Bar pattern
interaction flow and sediment
in a channel
width-depth ratio!
Channel pattern
interaction flow (and bars) with banks
out of the channel:
bank erosion and floodplain formation
3
Bar theory (1)
Forced bars and free bars
(Struiksma et al. 1985)
Forced bars: stationairy,
initiated and forced by channel curvature
flow and sediment interact
qs ~ mun (m=constant)
■ n>3 for theoretical reasons
■ n=3 for Meyer-Peter & Mueller
■ n=5 for Engelund & Hansen
Free bars: migrating,
initiated by perturbation that grows
(=unstable)
slope effects on sediment:
downslope easier
■ transverse slope in bends >> river gradient
secondary flows in bends: upslope
■ sharper bends → stronger secondary flow
Damping and exciting bars (1)
Bar theory (2)
flow needs length to adapt to bed (bars…)
relaxation length λw:
Bar types:
forced bars: forced by flow curvature; static
position
free bars: spontaneously develop and migrate;
initiated at perturbation (groyne, bend, tree…)
imagine: momentum!
sediment too!
relaxation length λs:
Analogy:
spring-mass-‘damper’
damper can also excite
imagine: (transverse) slope effect
→ bed cannot suddenly jump
22/51
Three regimes for free (alternating) bars
Overdamped
graphics: after Mosselman et
al. 2006
Underdamped
Excited
23/51
24/51
4
Bar theory (4)
Bar theory (3)
from theory:
(just accept this…
full derivation in Struiksma et al.!)
bar damping length LD
so, most important parameter (spring-damper!) is
Interaction Parameter: λs/λw
s
2
g W 
 2 f   2  
w 
C h
2
IP depends mostly on W/h
and a bit on friction and on the slope effect f(θ)
bar length Lp
Bend flow
conservation of momentum AND
logarithmical flow velocity profile
 helical (spiral) flow
 bed shear stress towards inner bend
s: longitudinal coordinate
n: transverse coordinate
v: velocity (u for us)
z: elevation, s: surface, b: bed
h: depth
R: bend radius
Bend flow (2)
Koen
Blanckaert
inner-bend bar
 main flow forced towards outer-bend
 transverse movement of momentum
and net transverse flow velocity
with:
Think! infinitely long bend?
Transverse slope effect
Transverse bed slope effect
Gravity acts on particles on transverse
slope
 particles pulled towards outer-bank
counteracts spiral flow
Schuurman, Marra & Kleinhans 2013, JGR
5
How to reduce transverse slope?
fixed layers
longitudinal dams
Underdamping  overshoot!
bendway weirs
Struiksma et al (1985)
bottom vanes
Adaptation and overshoot (Delft3D)
Examples: a bit overshoot and much excitation
Damping and
exciting bars (2)
three regimes:
profiles along outer-bend bank
bit overshoot
overdamped
■ just forced bars in
reaction to curvature
straight
bend
straight
long bend
straight
underdamped
■ overshoot superimposed
on forced bars
unstable, exciting
excitation
■ free bars grow spatially;
‘spread like a disease’
■ damping length negative
Wider channels: braid bars
Narrow channel
Wide channel: stability of higher wave modes
mode m=2, etc.!
6
Is this getting chaotic?
chaotic forcing
violin:
■ horsetail on string
river:
■ many perturbations
■ turbulence
dominant modes
violin:
■ length of the string, +overtones
river:
■ width
Andre Ermolaev
Andre Ermolaev
Andre Ermolaev
3. From bar pattern to channel pattern
Width-depth ratio!!
remember hydraulic geometry:
width depends on bank strength
(depth depends on width and roughness)
■ vegetation
■ cohesive sediment
Andre Ermolaev
7
Yep!
John Bridge
Courtesy Jim Best
Remember: bank strength!?
Imagine a river in sand without mud or
vegetation
Bars and bank erosion
pools between the bars:
velocity high
bank deeply undercut
celerity and size of bars:
θc,banks < θc,bed (slope of bank!)
initially: θ >> θc
so, erosion of bank toe and collapse of
bank!
eroded sediment deposited on bed 
shallowing
continue until θ = θc in bankfull conditions
‘threshold channel’
wide river: fast, small bars
→ everywhere bank erosion
→ straight planform
narrow river: slow, large bars
→ alternating bank erosion
→ meandering planform
Alternate bars and bank erosion
low flow conditions!
Bridge (2003) Rivers and Floodplains
8
But, does it really work like this?
Ongoing work!
Wout van Dijk PhD finished
Filip Schuurman PhD study ongoing
work in Delft, Illinois and Japan (with us)
bank
sedimentation
floodplain
sedimentation
Effect of mud in rivers
Meander migration
Wout van Dijk, Wietse van de Lageweg
9
1977
1989
Ganges River
1985
1997
Simulation
bend instability
Sun et al 2001
1999
10
Camporeale et al 2007
These models
look better
than they are!
One pattern explanation
interaction between pools (bars…) and banks
Nevengeul van de Vecht,
Junne, bij Dalfsen
stronger banks
→ narrower channels
→ slower, alternating bars
→ meandering
weaker banks
→ braiding
Bank strength derived from
self-formed floodplains
vegetation
So far state of the art… much work in progress!
Major points
1. What determines bar pattern?
2. What determines channel pattern?
Let’s overheat the brain...
Example calculation for 18th century Rhine
river upstream of Lobith:
S_valley = 0.00016, S_channel = 0.00013
Qmaf = 5580 m3/s, Qbf = 3370 m3/s
D50 = 2 mm, D90 = 8 mm
Width = 520 m, depth = 5 m
calculate:
θ’, Fr, Re, λBW, ωpv, LD, Lp, λs, λw, IP
11