Applying Unstructured Grid Model for Tides in the Belgian Continental Shelf

Applying Unstructured Grid Model for Tides in the Belgian Continental Shelf
02.09.2015 / Brussels
Promotor : Professor Dr. ir. M. Chen
Advisor : Dr. Olivier Gourgue
By : Biniyam Sishah
02.09.2015 / Brussels

Content
Introduction
Method and Material
Results and Discussion
Conclusion and Recommendation
2

Content
Introduction
Method and Material
Conclusion and RecommendationConclusion and Recommendation
 General
 Modeling of Tides
 Preliminary Work
 Study Area
 Objectives
Introduction
Method and Material
3

General
 Tides are the vertical rise and fall of
the water surface in oceans due to
phases of the moon and sun.
Source: https://en.wikipedia.org/wiki/Tide
4

General
5
 Various interactions causes tidal
variations at a location.

General
6
 Various interactions causes tidal
variations at a location.
 Corresponding tidal variations(level
and Currents) are expressed by a
collection sum of simple sinusoid
curves called Tidal constituents
(𝑨, 𝝋,W).
𝑨 = Amplitude
𝝋 = Phase
w= angular velocity

Modeling of tides
 SWE derived from the 3D Navier-stokes equations are often used in models.
 These equations has no analytical solutions.
 FEM(finite elements method) with unstructured grids were used for this work.
7

 Generation of the Unstructured Grid
 Setting up of Telemac2D model
 Mesh refinement has been performed
in some locations.
Preliminary work
8
Source :Flanders Hydraulic research Center

 Generation of the Unstructured Grid
 Setting up of Telemac2D model
 Mesh refinement has been performed
in some locations.
Preliminary work
9
Source :Flanders Hydraulic research Center

Study area : what do we know?
The Belgium Continental Shelf (BCS)
 Surface area: 191 Km2
 Offshore length:76.2 Km
 Gullies and sand bank formations
 High turbidity near Coastlines
 Presence of Scheldt Estuary
The bathymetry of the Belgium shelf [source:(Anon, n.d.) ]
10

Objective
The main objective was to reproduce the tidal movement in the Belgium continental
shelf area. In the context of this work the following specific objectives were
accomplished:
 Selection of Calm and Stormy wind periods
 Calibration of the Telemac2D model for Tide levels
11

Content
Introduction
Method and Material
Introduction
 Methodology
 Materials and Data
Method and Material
12

Methodology
Matlab
13
1. Selection of wind periods:
2 . Calibration of Model
Criteria CFG-1 CFG-2
Coordinate of Unstructured grid mesh Cartesian coordinates
Telemac2D projection system Mercator UTM 31N
Tide generating force YES Not Available
Spherical coordinates YES Not Available
Coriolis coefficients Actual value at a Node Constant value
Telemac2D
(future work)

Materials and Data
Materials Data
 Telemac2D
 Matlab
 Others…
European shelf tidal dataset
(OTIS regional tidal solutions) Observations stations
14

Content
Introduction
Method and Material
Method and Material
Introduction
Results and Discussion  Wind period selection
 Boundary condition
 Bottom Friction
 Models’ Performance In the BCS
 Model performance (Local scale )
 Temporal scale of errors
15

Wind period selection
 Calm period: July => used for tidal calibration
16
Satellite Measurements Observation station

Wind period selection
 Calm period: July => used for tidal calibration
17
Satellite Measurements Observation station

18
Boundary condition
U-velocity

19
Boundary condition
U-velocityV-velocity

20
Boundary condition
U-velocityV-velocity
H-Sea surface level

Bottom friction
 Sensitivity analysis for Manning’s and chezy’s formulations
 CFG-2 had a better performance at Manning’s Coff.=0.024 and Chezy’s Coff. =56
Legend
Chezy's formulation
21
Manning's formulation

Bottom friction
Legend
Chezy's formulation
22

Bottom friction
Legend
Chezy's formulation
23

Bottom friction
Legend
Chezy's formulation
24

Bottom friction
 Spatial plots of Relative Harmonic errors over BCS( Sea-surface level of CFG-2)
 Except a difference in their weightage areas, convey similar information.
 Highest error value were achieved near the Scheldt Estuary ?
Manning's Coefficient Chezy’s Coefficient
25

Models’ Performance In the BCS
U-velocity H-Sea Surface Level
26
V-velocity

Model performance (Local scale)
 How is Model’s performance In different locations?
 Model performs is better in deep waters than Coastal locations.
 Model Performance is better in the BCS than other similar Coastal locations.
27

Model performance (Local scale)
 How is Model’s performance In different locations?
 Model performs is better in deep waters than Coastal locations.
 Model Performance is better in the BCS than other similar Coastal locations.
28

Temporal scale of errors (Goodness of fit tests)
29
Amplitudes=>(Observ. ≈CFG-2)
Phases=> (Observ. ≈ OTIS)

30

31

32

33

Content
Introduction
Method and Material
Method and Material
Conclusion and Recommendation
Introduction
34

Conclusion
 Boundary conditions for Sea Surface level were properly set.
 Optimum model setup: UTM 31N (CFG-2) projection with either Manning’s
(0.024) and Chezy’s (56) coefficient.
 CFG-2 has a good performance with tide levels but not tidal currents.
 Mesh refinement in the BCS have increased model performance
 There is high probability that the calibrated Teleamac2D model here, might
perform better than OTIS in the BCS. ( with further work)
35

Recommendation
Present
 Boundary nodes of U and V velocity should be modified.
 Simulations considering different parameter values for BCS and Scheldt Estuary
should be made in model.
 Telemac2D model setup of CFG-1 should be investigated for bugs.
Future
 Including downstream River influences near the Scheldt Estuary.
 Using lower model domain area.
 Simulation with Telemac3D
36

binibobibni@gmail.com
biniyambirhan.sishah@student.kuleuven.be
Biniyam.Sishah@vub.ac.be
Contact information

Applying Unstructured Grid Model for Tides in the Belgian Continental Shelf

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Applying Unstructured Grid Model for Tides in the Belgian Continental Shelf (20)

Recently uploaded (20)

Applying Unstructured Grid Model for Tides in the Belgian Continental Shelf