Hydraulic Modelling of
wetland flow
Data collection and problem solving
Prof. dr. ir. Ronny Verhoeven
Hydraulics Laboratory
Ghent University
Belgium
WETHYDRO WORKSHOP – 13 – 14 June 2003
Hydraulic Modelling of
wetland flow
•
•
•
•
•
•
Introduction
Hydraulic modelling of open channel flow
Extension to wetlands
Data collection – problems, questions, solutions
Input data – problems, questions, solutions
Conclusions - questions
WETHYDRO WORKSHOP – 13 – 14 June 2003
Introduction
Engineer >>
translates reality
into formula
Deterministic approach is
what he likes: p = r g h
Stochastic representation is
what he needs to live with
Introduction
Hydraulic Modelling of
open Channel Flow
Supositions
•Uniform velocity distribution: Q = A . U
•Prismatic bed – constant cross-section
•Hydrostatic cross sections
•Constant bottom slope
•Constant friction factor
Hydraulic Modelling of
open Channel Flow
Steady state
Continuity:
Q=A.U
Motion – Bresse equation:
So  Sf
h 
 x 1 So2  BQ ²
gA³
Uniform flow
>> Manning <<
U = 1/n.R2/3.S01/2
Hydraulic Modelling of
open Channel Flow
Unsteady state
Saint Venant equations
Continuity:  Q  B  h  0
x
Motion:
t
Q
h
  Q2 

  g. A..S o  S f 

t  x A 
x

Hydraulic Modelling of
open Channel Flow
Unsteady state
Saint Venant equations – solved by implicit finite
difference Preismann scheme
>> choise of Q is important > stability
>> choise of Ds and Dt also > accuracy
50
45
40
Q [m3/s]
35
30
25
20
15
10
5
0
0
10
20
30
Time [h]
40
50
Extension to wetlands
Quasi 2D modelling
>> Network structure - flow
>> Cells - exchange of volumes
>> Combination - what to choose?
Input data – what do we need?
Topographical
• Cross-sections of river and floodplain
• Longitudinal profile (Thalweg)
• Water levels ( f(t) )
Hydraulic
• Discharge ( f(t) ) – lateral discharges
• Friction coefficients
• Sediment transport (bottom / suspended)
Data collection
Topographical
• ? Distance between 2 cross-sections
• ? Boundaries of flood plains
• Altitude measurements should be the most accurate ones
• Accuracy of measurements is influenced by:
- mud
- vegetation
- obstacles in cross-section
- soft bottom
Data collection
Hydraulic data – discharge measurements
• Integration of velocity field over cross-section
• Propeller meter or electromagnetic, acoustic velocity meter
• From bridge or from boat
Data collection
Hydraulic data – discharge measurements
> Problems <
• Velocity distribution – horizontal / vertical
1,6
0
5
10
15
20
25
0
1,4
0,2
Depth [m]
1,2
0,4
0,6
0,8
1
1,2
1
0,8
0,6
0,4
1,4
0,2
1,6
0
0
1,8
Depth y (m)
Mean velocity (m/s)
0,1
0,2
0,3
0,4
velocity [m/s]
0,5
Data collection
Hydraulic data – discharge measurements
> Problems <
Influence of
• Vegetation
- velocity fluctuation as a function of time
- slowing down propeller
- block the propeller
- local influence on velocity meter
• Stones or rocks
• Soft bottom
• Wind while measuring from a boat
• Measuring errors
Bieb
rza
R.
0.48
ie
Deb
owo
3.43
Dol
isto
wo
6.88
Gon
iad
z
Oso
w ie
c
Bialogrady
1.15
Ost
row
Elk
R
.
Ch .
Rud
zki
6.88
Aug
Jag
usto
low
wsk
o
i Ch
.
Przechody
7.00
0.89
Biebrza R.
5.98
3.94
sk
0.24
R.
wka
R.
Measurement Campaigns: discharges in m3/s
a
Sidr
0.91
0.86
o
Broz
Karpowicze
1.33
Lip
8.06
bin
Szta
9.17
2000
za R
.
Bieb
r
Aug
Jag
usto
low
wsk
o
i Ch
o
0.52
3.04
Ost
row
ie
Deb
owo
3.48
6.14
2.67
Dol
isto
w
Gon
iad
z
Oso
wie
c
Bialogrady
7.18
10.77
Elk
R
Rud
zki
8.58
.
Ch .
.
Przechody
9.15
1.79
Biebrza R.
13.63
sk
R.
wka
R.
Measurement Campaigns: discharges in m3/s
a
Sidr
1.90
0.85
o
Broz
Karpowicze
2.36
Lip
16.57
bin
Szta
18.10
2003
Input data
How to determine the cross-section?
Input data
How to determine the cross-section?
?
depth [m]
0
0
0,5
1
1,5
2
2,5
-0,05
-0,1
width [m]
Solution: define cross-section with A, P and R
equal to the average value of all cross-sections
depth [m]
0
0
0,5
1
1,5
2
2,5
-0,05
>> Calibration of friction coefficient becomes
very important !!!
-0,1
width [m]
depth [m]
0
0
0,5
1
1,5
-0,05
-0,1
width [m]
2
2,5
Input data
How to determine the longitudinal profile?
Effect of friction!
Input data
How to determine the bottom slope?
Input data
How to determine the friction coefficient
• n = f (bottom roughness, shape cross-section, vegetation,
obstacles, meandering, velocity distribution, …)
• n = f (time, location, interaction of previous parameters)
• n must be determined from measurements
Q
hd
hu
Bresse
n
Biebrza R.
R.
R.
Manning n determination by calibration
a
Sidr
wka
Determination of n using:
• Uniform flow principle (Manning formula)
• Bresse equation
0.069 –
0.651
k
o
Broz
Karpowicze
0.060
0.030 –
0.276
s
Lip
0.040
bin
Szta
0.045
Ost
0.090 –
0.369
row
ie
Bieb
rza
R.
.
Aug
Jag
usto
low
wsk
o
i Ch
0.061
owo
Gon
Oso
wie
iad
c
z
Dol
isto
wo
Bialogrady
D eb
Rud
z
Elk
R
.
ki C
h.
Przechody
119,0
measurement 02.06.2002
118,5
n by Manning formula
water level [m aBsl.]
118,0
n by calibration
117,5
117,0
116,5
116,0
115,5
115,0
32
37
42
47
52
57
distance along river [km]
62
67
72
BUT!!!
Hydraulic Modelling of
open Channel Flow
Hydraulic Modelling of
open Channel Flow
Discharge variation in Goniadz : B-14
57
UNET
measurement
Q [m3/s]
52
47
42
37
32
27
0
5
10
15
time [days]
20
25
30
Conclusions and Questions
• Flood-routing theory is quite simple
• Numerical solution methods are well developed
• Practical application is confronted with many inaccuracies
• Good simulation results thanks to well considered calibration
• ? Definition of cross-section?
• ? Determination of longitudinal profile?
• ? Best way to determine the friction coefficient?
• ? Suggestions to improve measurements quality?
Acknowledgements
T. Okruszko, S. Ignar, R. Michalowski, J. Chormanski,
D. Swiatek, I. Kardel
SGGW, Warsaw
L. Van Poucke, M. Huygens, R. Banasiak
Hydraulics laboratory, Ghent University
Universities of Brussels and Antwerp
Funding from Polish and Flemish government
bilateral cooperation projects
Biebrza National Park Authorities
1998
2002