Upload
julia-diaz
View
229
Download
0
Embed Size (px)
Citation preview
8/11/2019 Transporte de Sedimento
1/62
Estimating Sediment TransportPeter Wilcock 9 August 2011
ore ac groun Critical and reference shear stresses
y t s ar to est mate transport rates Empirical vs formula estimates A calibrated method for estimating transport rates in
gravel-bed rivers
8/11/2019 Transporte de Sedimento
2/62
If a sediment rating curve is what we use in application, isnt there
Will 1,000 cfs in Silver Ck move the same amount of sediment as1,000 cfs in the Snake R?
One recent attempt:
s
sr r
Q Q
Q Q
b
s aQQ Its nice that a cancelled, but for this model to be general, therate of change of Qs with Q (i.e. the exponent b) must be thesame everywhere. But we will find that b varies from littlemore than one to more than ten!.
Basically , we are looking for a flow variable that can be
accurately scaled, such that we have a general model
8/11/2019 Transporte de Sedimento
3/62
A flow variable that does scale well (although it not always easy to.
transport models in common use.
What might a general transport model look like? First, lets get amore formal introduction to the la ers. The best wa to do this iswith a dimensional analysis which will produce a list of
dimensionless variables that represent the problem.This helps because* the number of dimensionless variables is (usually) 3 less than the
* the dimensionless variables often have a useful physical meaning* Any relation between the dimensionless variables should begenera , we ave s e a re evan var a es a e s ar .
8/11/2019 Transporte de Sedimento
4/62
( , , , , , , )s sq f h D g Our list:2L M M M M L
, L, L, , , ,
)/,*,*,(* h DsS f q
sq **
mens onaanalysis gives:
gDs
gDsgDs
)1(
)1(,
)1(
3
3
s /If s = const
& S* large
8/11/2019 Transporte de Sedimento
5/62
o(Flow force on bed)/(bed area): The Shields Number
3(Grain weight): ( 1)6
s g D
2
* Number of grains/area 1/
Flow Force D
e s um er:
Grain weight ( 1)
This is THE most im ortant variable in sediment trans ort
s gD
The Einstein Volumetric transport rate/width sq
TransportParameter *
Sediment fall velocity ( 1)
trans ort rate
w s gD
q
3fall velocity * grain size ( 1)
q
s gD
8/11/2019 Transporte de Sedimento
6/62
First, we define conditions for the onset of sediment transport;these should depend on the same variables as the transport rate
ot on o mot on , , ,s 3( 1)
* * ss gD D
( 1) /s gD h
The Shields curveIncipient motion defined in.
*c
terms of a critical value ofShields Number
Motion
* c Defined for unisize sediment No Motion
( 1)c s gD
0.0110 100 1000 10000 100000
*
ec oDimensionless plots must be
59.03.0* *35exp045.0*105.0 S S c
8/11/2019 Transporte de Sedimento
7/62
bcaq )*(** A typical transport model:
gDsc
c
)1( 2/3*** c-
3/ 2 3 ( 1) ( 1)( 1) s gD s gDs gD
0.1
1
10
3/ 23/ 28
( 1)s cq
s g
0.001
0.01
3/ 278.7for in (tons/day)
s c
s
q
q
1E-050.0001
0.01 0.1 1
* .and , in (Pa)c
3**q
8/11/2019 Transporte de Sedimento
8/62
bcaq )*(** A typical transport model:
gDsc
c
)1( 2/3*** c
)1(** s
gqsq
-
One more transport variable, W*2/32/3 )/(*)(
2/3
(sorry!)
One more stress, the reference stress r
*18*
cW
onvert t e - ormu a to anincorporate a reference shear stress. First,we divide M-PM by ( *)3/2 to get
At *= *r, W * = W *r = 0.002. Dividing by 8, raising both sides to the 2/3 power produces
** 996.0 r c *
1004.0r
c
*
*
996.018*
r W Meyer-Peter & Mller:
8/11/2019 Transporte de Sedimento
9/62
Reference shear stress? Critical shear stress?
3/ 210
*10
*
Meyer-Peter & Mller
** 8 1 cW
0.11
0.1
1
3/ 2*
or
.996 0.01
*W
0.01
*
.0.01 0.1 1
*
*
.0.01 0.1 1
* *c
r
8/11/2019 Transporte de Sedimento
10/62
How the reference shear stress gets used
100
/ ( m s
) ]
1 r a
t e W *
1 r a
t e W *
1
o r t r a t e q
b i [
0.1
s t r a n s p o r
t
0.1
s t r a n s p o r
t
0.001
0.01
.
r a v e l
t r a n s p
0.001
.
i m e n s i o n
l e
0.001
.
J06
J14 i m e n s i o n
l e
*
r W
0.00001
0.0001
G
0.00001
0.0001
0.00001
0.0001J27BOMC
0.1 1 10 100
Bed Shear Stress (Pa)
0.1 1 10 100
Bed Shear Stress (Pa)
0.1 1 10 100
r /r
8/11/2019 Transporte de Sedimento
11/62
The Difference Between c
and r
c : boundary between motion & no motion.Definable exactly for an individual grain;
10
*W
definable statistically for a river bed Hard to measure, tracer grains are yourbest shot
0.1
r : the stress associated with a small,agreed-upon transport rate ( W* = 0.002) 0.001
0.01
0.01 0.1 1
*r W
provides a threshold for transport estimate
Easy to determine from transport
*
* *observations c
tracers (painted rocks, magnetic rocks, radio rocks, rock scum) answer the question did the
11
, , ,flows, develop relation between %entrained, grain size, and flow)
Need large numbers of grains for reliable sample Need to place naturally
8/11/2019 Transporte de Sedimento
12/62
cWill sediment move? Is > ?
As increases, up go stage & (controlled by resistance and continuity).o
Q R
The incipient motion problem
For a given , we can find boundary stress using momentum =
and compare too
c
S gRS
100
1.5
2.0
2.5 Lazy River Section 2
a t i o n
( m )
10
a r s
t r e s s
( P a
)
c Motion
0.0
0.5
.
0 2 4 6 8 10 12 14 16Station (m)
E l e
1 S h e
No Motion
At what does ?o cQ
0.10.1 1 10 100
Grain Size D (mm)quartz in water
2 / 3 2 / 3Combining and :S S
Q UA U R Q ARn n
If we know and and , we can find from
and then calculate . Or, we can simply adjust untilc c c c
c c
n S R gR S
Q Q R R
8/11/2019 Transporte de Sedimento
13/62
Discharge ( Q ): 70.0 cfs
Enter values only in green cells Channel Velocity ( U ): 3.31 ft/sec 2 / 31.49 S
1. CALIBRATION. For the slope, flow depth, and cross
section you have specified, use Manning's eqn to find
X/S.xls illustrates the back-calculation of n
Manning's n : 0.054
Channel Slope ( S ): 0.0120 Manning's n : 0.047Flow Depth: 1.64 ft Channel Velocity ( U ): 3.82 ft/sec
Selected Flow Depth: 1.64 ft Discharge ( Q ): 80.8 cfsCross-Sectional Area ( A ): 21.1 sq.ft. 2.3 cms
Hydraulic Radius ( R ): 1.2 ft Grain Size D 60 mm nD = 0.026Critical Shields Number t*c 0 06 t' = 16 8
2 / 31.49 S Q ARn
nQ
2. FIND DISCHARGE. For the slope, flow depth, andcross section you have specified, use Manning's eqn tofind the discharge Q for a value of the roughness n incell F53. FIND CRITICAL DISCHARGE. You can also
the roughness n for a value of the discharge Q in cellF1
Apply . .
Cross section of Minebank Run stream channel, 41.4 Pa t'/tc = 0.2928 feet downstream of gage 0158397967 December 2002. c 58.3 Pa
Survey Data: c 0.71Grade Rod (ft) Distance (ft)
4.0 0.04.4 0.04.4 2.0
3
3.5Left Bank
critical stress c needed to entrain grains of size Dspecified in cell F9 (given a value of critical ShieldsNumber given in cell F10
0
*( 1)
cc
gRS
s gD
Enter inform ation only in
green cells!!No cuttin and astin !
n c
. .4.7 8.04.8 12.04.9 14.05.5 18.05.7 20.05.9 21.06.7 22.3
4
4.5
5
5.5 d e R o d ( f t )
No inserting or deleting
rows!
7.0 23.07.2 24.07.1 26.07.3 28.07.3 30.07.0 32.07.1 34.06.8 36.0
6
6.5
7
7.5
G r a
Input
7.1 36.35.7 37.0 0 10 20 30 40 50Distance (ft)
13
8/11/2019 Transporte de Sedimento
14/62
Q BhU continuity
/
b B aQ
h gS
For a simple, wide,
hydraulic geometry
momentum
2 / 3
5 / 3
soS
U hn
,find Q c directly Mannings eqn.
1
or bQ aQgS n
a S Q
n gS
*
15 / 3 1*
( 1) soc c
b
s gD
a
0.045( 1)
cc s gD
7 / 6c c s
nS
8/11/2019 Transporte de Sedimento
15/62
Q BhU
too sure aboutsome of the values/
b B aQ
h gS
needed todetermine Q c ?2 / 3
5 / 3
soS
U hn
Like n, D , and * c
what do you do?1
or bQ aQgS n
a S Q
n gS
*
15 / 3 1*
( 1) soc c
b
s gD
a
7 / 6c c
snS
8/11/2019 Transporte de Sedimento
16/62
Suppose your best estimate of Mannings n is 0.035and that you are pretty sure that the real value falls n c
y0.035
0.0025n
n
e ween . an . .
We could approximate your assessment of thevalue of n with a normal distribution with
F r e q u
2 n
mean = . s an ar ev a on = . .
95% of this distribution falls between 0.03 and 0.04,as can be seen in the cumulative fre uenc lot, so
n n 0.02 0.03 0.04 0.05Manning's n
0.8
1
q u e n c y
2 n
we are saying that the real value of n is 95% likely tofall between 0.03 and 0.04 and that it is more likely
to be around the center of the distribution (0.035) 0.20.4
0.6
u m u
l a t i v e
F r e
.values of n in our Monte Carlo simulation.
How does that work? We use a random number
00.02 0.03 0.04 0.05
Manning's n
1 n c y
generator to pick a number between 0 and 1 andthen use this number to find a value of n for thecumulative frequency distribution. For example,
=0.4
0.60.8
u l a t i v e
F r e q u
. , .for 0.23, n = 0.0332 0
.
0.02 0.03 0.04 0.05
C u
Manning's n
8/11/2019 Transporte de Sedimento
17/62
n D*c
1 .
The Monte Carlo simulation
*. c va ues o , , an rom spec e requency s r u ons.
2 . Calculate critical discharge and transport rate.
3 . Repeat 1000 times.
c
b
Q BhU
B a2 .
4 . Distribution of calculated values givesestimate of the effect of input uncertaintyon calculated critical discharge and transport rate.
2 / 3
/
so
h gS
S U h
n
1
5 / 3 1
or b
b
S Q aQ
gS n
a S
100200
300
400Manning's n(a)
50
100
150
200*c (b)
4 .
*
1
( 1) soc c
n gS
s gD
0
0 . 0 2 0
0 . 0 2 4
0 . 0 2 8
0 . 0 3 2
0 . 0 3 6
0 . 0 4 0
0 . 0 4 4
0 . 0 4 8
0
0 . 0 3 0
0 . 0 3 4
0 . 0 3 8
0 . 0 4 2
0 . 0 4 6
0 . 0 5 0
0 . 0 5 4
0 . 0 5 8
200
250 Grain Size D(mm)(c)
200
250
300 Critical DischargeQc (m^3/s)(d)
*7 / 6 ( 1)c caQ s DnS 0
50
100
1 4 2 0 2 6 3 2 3 8 4 4 5 0 5 60
50
100
150
0 3 6 9 1 2 1 5 1 8 2 1 Monte Carlo
8/11/2019 Transporte de Sedimento
18/62
LEGENDSpecify Mean St. Deviation pdf
Manning's n 0.03 0.002 normal
cMonte Carlo Simulation
. .Grain size 5 0.25 normal
D (mm) 32 lognormal95% Pred.
o Qc ca cu at on, us ng1000 trials
300
400Manning's n(a)
150
200* (b)
a cu a e ean . ev a on n erva
Qc (m^3/s) 8.1 3.4 6.7
100
200
50
100
(uncertain)
solution to
0 . 0
2 0
0 . 0
2 4
0 . 0
2 8
0 . 0
3 2
0 . 0
3 6
0 . 0
4 0
0 . 0
4 4
0 . 0
4 8
0 . 0
3 0
0 . 0
3 4
0 . 0
3 8
0 . 0
4 2
0 . 0
4 6
0 . 0
5 0
0 . 0
5 4
0 . 0
5 8
250 Grain Size D 300 Critical Discharge
thresholdchannel
roblem
100
150
200mm
(c)
150
200250 Qc (m 3/s) Need a
better
0
50
1 4
2 0
2 6
3 2
3 8
4 4
5 0
5 6 0
50
0 3 6 9 1 2
1 5
1 8
2 1 tracergravels
8/11/2019 Transporte de Sedimento
19/62
Application of Monte Carlo approach for assessing risk:
Given a channel with a design discharge. The channel isintended to be static. You would like to be 90% sure that the
grains in the channel bed will not move at the designdischarge.
Solution: specify possible range in n , D, and * c , use MonteCarlo to determine the range of possible discharges. Adjust
are larger than the design discharge.
300 Critical Discharge
DesignDischarge
100
150
200250 Qc (m 3/s)
0
50
0 4 8 1 2 1 6 2 0 2 4 2 8
8/11/2019 Transporte de Sedimento
20/62
Garcia, Vanoni Bed Load Transport Relations
8/11/2019 Transporte de Sedimento
21/62
100,000100,000 D = 45 mm
Sediment Transport Why its hard to estimate 0
60,000
70,000
80,000
,
t e ( k g
/ h r )
10,000
t e ( k g
/ h r )
30,000
40,000
50,000
T r a n s p o r t
R
100
1,000
T r a n s p o r t
R
0
10,000
20,000
0 20 40 60 80 100 120
D = 45 mm
D = 30 mm
1010 100
DischargeDischarge
Example calculation using Meyer-Peter and Muller for a channel= = =
Discharge (cms) Discharge (cms)
. , . , .solid curve uses = 0.045 and D = 45 mm. The dashed curve uses
= 0.045 and D = 30 mm.
*c
*c
Note that at discharge Q = 55 cms, one curve indicates zerotransport and the other a transport rate of 80,000 kg/hr.
8/11/2019 Transporte de Sedimento
22/62
10 M-P&M2t*c
Sediment Transport Why its hard to estimate I
1ranspor ra e s a s eep, non nearfunction of bed shear stress
0.1
q* s
Three things make it difficult to
0.001
.accura e y es ma e :
1. Accelerations in unsteady,
0.0001
. .
2. Only a portion of the total acts totrans ort sediment
1E-050.01 0.1 1
varies locally. Because transportis a nonlinear fn of estimates
* based on total will be in error
8/11/2019 Transporte de Sedimento
23/62
t g xg xS gR0
Three things make it difficult toh U U accura e y es ma e :
1. Accelerations in unsteady,0
0 f
g x g x
gRS
. .
2. Only a portion of the total acts totrans ort sediment
NOTE:
varies locally. Because transportis a nonlinear fn of estimates
compute S f
based on total will be in error
8/11/2019 Transporte de Sedimento
24/62
The Drag Partition
, 0, per area, that the flow exerts on the boundary of the channel.
But onl a ortion of the
Three things make it difficult to total 0 acts on thesediment grains to produce
accura e y es ma e :
1. Accelerations in unsteady,
.transport rates, we need to
partition 0 into the part. .
2. Only a portion of the total acts totrans ort sediment
that acts only on thesediment grains. We callthis the skin friction or
varies locally. Because transportis a nonlinear fn of estimates
grain stress .There is no direct way to
based on total will be in error .
an approximate method,using Mannings equation.
8/11/2019 Transporte de Sedimento
25/62
Solve the Manning eqn for the boundary stress:
2 / 3 1/ 6 2 / 3( ) ( )g S nU gRS 2 / 3SRU
0gS nU
3/ 2 1/ 4 3/ 2 '
1/ 6 .
1/ 4
Manning-Strickler Roughnessfor D in mm
.
65 65where we use 2 D D
1/6
*8.1
s
U hu k
3 21000 kg/m & = 9.81 m/sg
1/6
8.1
s D
k n
g
8/11/2019 Transporte de Sedimento
26/62
Using both the grain roughness and total roughness, we have anestimate of the portion of the total stess acting on the grains the
0.06
0.07
0.08
typical range of n values
. 3/ 21/ 4o gS nU
0.04
0.05
a n n i n g s n
3 21/ 4
' DgS n U
0.01
0.02
.
Manning-Strickler n
D
o
n
n
1 10 100 1000Charactersitc Bed Material Grain Size (mm)
3/ 2' D o
n
nBasically the same drag partition suggested by Meyer-Peter & Muller in the . ,application in HEC-RAS
8/11/2019 Transporte de Sedimento
27/62
Sediment Transport Why its hard to estimate II
W at s or a reac ?
Critical shear stress depends linearlyon grain size, but the grain size of thetransport is poorly known!
100
)
10
e a r s t r e s s
( P a
c Motion
Transport rate is a steep, nonlinearfunction of bed shear stress
1 S
h
No Motion
Small error in c can produce bigerror in qs
0.1
0.1 1 10 100Grain Size D (mm)quartz in water
8/11/2019 Transporte de Sedimento
28/62
Estimating Sediment Transport: Why its hard
Sediment transport is a local process: between individual grainsand the near-bed flow they encounter
The transport process is highly nonlinear: a very steep functionw w
Real streams have considerable variability in topography, flow,and bed composition The local flow actin on rains is hard to estimate The distribution of bed grain sizes is hard to determine You never know either local flow or local bed composition in
detail, so you have to estimate Errors in estimated transport rates are typically very large
(often well in excess of an order of magnitude!) Next, we examine an approach to provide reliable transport
estimates of reater accurac with acce table effort
8/11/2019 Transporte de Sedimento
29/62
If estimating transport rates using shear stress is so hard, remind me again why we cant find some general formula using Q.
t an est mate or , ets return tothe sediment rating curve problem. sQ Q
relation based on discharge Q can beeneral.
sr r
10 M-P&M
1
2t*c(t*)^3We approximate transport with a
power function of
* *q 3'sq 0.01.
q*
3'
0.0001
0.001
Write the relation twice, take ratio
'sr r q
1E-050.01 0.1 1
*
8/11/2019 Transporte de Sedimento
30/62
3
' 'sr r q
''
r
U U
We use our grain stress relation:
4.5
sq U
Which givessr r q U
.sQ U For simplicity, we assumethe width of transport does
sr r not change
8/11/2019 Transporte de Sedimento
31/62
We can relate U to Q using the hydraulicgeometry:
4.5 4.5 mbaQ B
r r U Q
mk U
c 4.5 msQ Q sr r
For 0.2 < m < 0.8, 1.0 < < 3.6.s no a cons an s no a su a y generavariable for a transport model.
Even larger values of found in steep mountainstreams (typically very small transport rates,
>
W find the exponent to vary between 1 and 10 (!!)
8/11/2019 Transporte de Sedimento
32/62
Estimating Bed-Material
8/11/2019 Transporte de Sedimento
33/62
Bed/Suspended Load vsWash/Bed Material Load How many sizes?
8/11/2019 Transporte de Sedimento
34/62
Washloadclay, silt, fine
suspension ?? uplands,banks,
true washload:(a) Transport not predictable
san , oo e n e o a ectransport of other fractions
Bed Material Fine
bed load orsuspension
interstices,stripes and
grain path in near-bed regiondominated by larger grains;
med-crs sand, pea gravel
dunes,subsurface
hard to sample & model
Bed Material bed load bed framework displacements generally rareCoarse
med-crs gravel,cobble
Hugeboulder
typical high flows
grains to exclude from thetransportable population
We focus on fine and coarse bed material
8/11/2019 Transporte de Sedimento
35/62
What about more fractions?
Many-fraction models available: including ones basedon the grain size of the bed surface, which allow for the
.
These models are fragile output is sensitive to the.
To simulate transport at a particular location, at aarticular time
need large amounts of detailed bed and flow data
Use these models to ask more general questions:
e.g. change in bed composition & transport witha change in flow releases from dams or
We will apply such a model later
8/11/2019 Transporte de Sedimento
36/62
Estimating Sediment Transport: Three overarchingconstraints bound the problem
Large spatial & temporal variabilityNeed BIG samples
Strongly nonlinear processese -mater a transport s a oca a a r
Modeling with spatial/temporal averages produces large error Sparse information available relative to that needed to model the
Modeling with high spatial/temporal resolution requires vast amounts of data Generally, we need a robust approach
8/11/2019 Transporte de Sedimento
37/62
Target: sediment rating curve Q s = ( Q )
Predict from a flow & transport model We cant measure ever thin but
Approaches
Models too fragile
Sample transport , Accuracy poor and a full sampling program too
burdensome
Tracer gravels and scour chains Measure volumetric change in local topography
1. Dealing with grain size2. Calculating transport rates
Issues:
3. Sampling bed material load
4. Best approach?
8/11/2019 Transporte de Sedimento
38/62
.
motion: What is Q c ? -
rate: What is Q se ave trou e est mat ng c, as we as y rau c parameters
8/11/2019 Transporte de Sedimento
39/62
Calculating cumulativeQ
h
sediment transport in asimple, prismatic channel,
b B aQ
over a hydrograph U h
nnQ
2 / 33 / 5
baQ Sha or uncer a n y nn, D , and * c
nh
a S
consequences?
3/ 23/51 0.73 *2650
bnQ S
ep ace n
ghS Replace in Meyer-Peter & Muller:
1000 ( 1)s o ca s D 0.7nQg S
a
8/11/2019 Transporte de Sedimento
40/62
LEGENDSpecify Mean St. Deviation pdf
Mannin 's n 0.03 0.002 normalMonte Carlo Simulation
c 0.045 0.0050 normalGrain size 5 0.25 normal
D (mm) 32 lognormal
of Qs calculation, using1000 trials
400'
200 Cumulative
.Calculate Mean St. Deviation Interval
Cumulative transport (tons) 14,997 2,806 5,509
0
100
200
300 a
0
50
100
150 c (b)
100150200250 Transport
(metric tons)(f)
0 . 0
2 0
0 . 0
2 4
0 . 0
2 8
0 . 0
3 2
0 . 0
3 6
0 . 0
4 0
0 . 0
4 4
0 . 0
4 8
0 . 0
3 0
0 . 0
3 4
0 . 0
3 8
0 . 0
4 2
0 . 0
4 6
0 . 0
5 0
0 . 0
5 4
0 . 0
5 8
200
250 Grain Size D(mm)(c)
60
80
100
( m ^ 3 / s )
(d)
050
2 0 0 0
6 0 0 0
0 0 0 0
4 0 0 0
8 0 0 0
2 0 0 0
6 0 0 0
0 0 0 0
0
50
100
1 4
2 0
2 6
3 2
3 8
4 4
5 0
5 6
0
20
40
0 20 40 60 80 100Time (hrs)
D i s c h a r g e 1 1 1 2 2 3
Useful for evaluating uncertaintyin sediment balance between
MonteCarloTransport.xls
8/11/2019 Transporte de Sedimento
41/62
8/11/2019 Transporte de Sedimento
42/62
2. Estimating Transport: Sampling
Option 1: trap all the transport in a weir, slot, or pool
-
samplers installed on the bed.
Option 1 is best, butgenerally not practical
Option 2 is practical,but involves larger
error, some r s , someluck, and lots of effort
8/11/2019 Transporte de Sedimento
43/62
Bi River Sam lin 1991
8/11/2019 Transporte de Sedimento
44/62
8/11/2019 Transporte de Sedimento
45/62
Big River Sampling 2007
R ll Bi Ri S li 2007
8/11/2019 Transporte de Sedimento
46/62
Really Big River Sampling 2007
8/11/2019 Transporte de Sedimento
47/62
8/11/2019 Transporte de Sedimento
48/62
Sampling ITransport field highly variable in space & time
Need LARGE samples!
Define Sampling Intensity for a 6.5m stream
Transport. = ,
Hand-held 13 * 0.075m X 120s = 117 msSamplingTransect
.
rapSampler . m s = ,ms (23.5%)
So, pit or net-frame traps look pretty good
Pit samplers fill up
8/11/2019 Transporte de Sedimento
49/62
Pit samplers fill up
Photo J Pizzuto, U Delaware, home of big samples
S d t f l
8/11/2019 Transporte de Sedimento
50/62
So do net-frame samplers
Bunte, K., 1998. Development and ield testing o a stationary net- rame bedload sampler ormeasuring entrainment of pebble and cobble particles, Report prepared for the StreamSystems Technology Center, USDA Forest Service, Rocky Mountain Research Station, Fort
Collins, CO, 74 pp.
8/11/2019 Transporte de Sedimento
51/62
3 E ti ti g S di t T t
8/11/2019 Transporte de Sedimento
52/62
3. Estimating Sediment TransportDirect sampling+ gives a direct relation between Q and Qs- requ res a g e or
- cannot predict future conditions- error: is it random or s stematic? - Handheld samplers: Scooping, perching, limited
grain size & TINY SAMPLES- - ,rates
+ can predict future changes- highly inaccurate
ow ar o sca eboundary conditions poorly known
The alternative? Join forces.
8/11/2019 Transporte de Sedimento
53/62
The alternative? Join forces.
the future is a combination of simple robustmodels and efficient measurement . A first cut, using today s technology
If pit/trap samplers can collect good samplesof small transport rates, why not use these tocalibrate a model of coarse bed-materialranspor
GOOD samples! (long duration, spatially
Two sizes: sand and gravel om ne n ro us ramewor a s
insensitive to major sources of error
T t F l (G l)
8/11/2019 Transporte de Sedimento
54/62
Transport Formula (Gravel)
r r 5.4
846.012.11
iW 2.14*
r r .
r Calibrate **
( 1) bs gqqW
* ( / )
Formula provides the trend,but the samples provide the accuracy
Transport formula based on shear stress so
8/11/2019 Transporte de Sedimento
55/62
Transport formula based on shear stress , so
stress from discharge. We return to ourMannin -Strickler formula:
2/34/16517 U SD
For Cub River: .
D65 = 90 mm = 0.42 .
8/11/2019 Transporte de Sedimento
56/62
0.01
0.1 ObservedTransport Funct ionW*r
W* 0.0001
0.001 ObservedTransport FunctionW*r
Qs(m^3/s)
* 0.039c
0.001 1E-05
0.00001
.
1E-07
-
0.0000010.1 1 10
/ r
1E-080.1 1 10
Q (m^3/s)
SRC.xls
8/11/2019 Transporte de Sedimento
57/62
1 A B C D E F G H I J K L M N O P
Lonely R. NR Whirlpool Observed Observed23456789
s wor s ee or ranspor Slope S 0.015D65 90 (mm) D65 of bed surfaceD50 42 (mm) D50 of bed subsurfacefg 0.50 proportion fravel in bedb 15.0 (m) width of bed material, not channelrg 30 (Pa) Adjust the reference shear stress*rg 0.044 Reference Shields Number; is it reasonable?
0.001
0.01
.Transport FunctionW*r
W*
0.00001
0.0001
. Transport Function*
Qs(m^3/s)
10111213141516
Qrg 10.2 cmsObserved Observed
Q Qsg ' '/ rg qsg W*g(m^3/s) (m^3/s) (Pa) - (m^2/s) -
3.37 7.07E-09 17.0 0.57 4.71E-10 6.87E-064.90 8.80E-08 20.6 0.69 5.87E-09 6.42E-056.71 2.76E-08 24.2 0.81 1.84E-09 1.59E-05
- - - 0.000001
0.00001
0.0001
1E-08
0.0000001
0.000001
181920212223
24
. . . . . .8.78 4.35E-06 27.7 0.92 2.90E-07 2.03E-03
0.00 1.00E-20 1.00E-200.00 1.00E-20 1.00E-200.00 1.00E-20 1.00E-200.00 1.00E-20 1.00E-200.00 1.00E-20 1.00E-20
0.00 1.00E-20 1.00E-20
0.00000010.1 1 10
'/ r
1E-091 10 100
Q (m^3/s)
Lonely R. NR Whirlpool
Enter parameters of velocity rating curve herek = 0.63 m = 0.34 Qc = 0.00
( )mt U k Q Q 1.21.4
1.6
1.8
U ( m / s )
The cells below contain values of k and m
fitted to the (Q, U) values for specified Qck = 0.63 m = 0.34
0.3382 -0.1983 LINEST resultsQ U Q-Qt
(m^3/s) (m/s) (m^3/s)1.568
0.4
0.6
0.8
1
M e a n
V e l o c i
t y
. . .4.225 1.04 4.2255.946 1.17 5.9467.891 1.29 7.891
12.340 1.47 12.34013.326 1.50 13.326
0.0000.000
0
0.2
0 5 10 15 20
Discharge Q (cms)
Transport samples provide accuracy by
8/11/2019 Transporte de Sedimento
58/62
p p p y y
(1) Establishing grain size D of the transport &
2 Scalin the flow
Calculate grain stress: 0.25 1.5
6517 SD U 1.5for two different flows & take the ratio:
Suppose
1 1
2 2U b
U a1.5b
1.5b
Q en
2 2Q r r
If you calibrate, the transport formula is needed only to
give the change in transport with the change in discharge.
8/11/2019 Transporte de Sedimento
59/62
high flows & you base yourestimate on samples at low
, , .
in gravel-bed rivers, Earth Surface Processes & Landforms 26, 1395-1408.
Estimating bed-material transport in gravel-bedi
8/11/2019 Transporte de Sedimento
60/62
rivers Conceptual basis
- fine and coarse bed material- (supply of one affects the transport of the other)
Sampling- standard needed for minimum sample size- fine bed material Helle -Smith or ?- coarse bed material pit or net frame samplers- big rivers ???
- 2-fraction model captures essential interaction between fine &coarse
& facilitates inte ral measure of reach rain size
- But it can t do everything (armoring; change in sand or gravel size) Future
,- can be done now in wadeable streams, although effort is non-trivial
SRC & BAGS
8/11/2019 Transporte de Sedimento
61/62
-transport rates in gravel-bed rivers. Stream Systems Technology Center, (the StreamTeam), US Forest Service
B GS Yantao Cu , St water Sc ence : supports var ety o transport ormu as, a ows
variety of input, includes calibrated approachUsers Manual by John Pitlick
Primer by Peter Wilcock
8/11/2019 Transporte de Sedimento
62/62