Upload
letruc
View
229
Download
3
Embed Size (px)
Citation preview
NOAA NTHMP Mapping & Modeling
Subcommittee Benchmarking Workshop:
Tsunami Current
Yefei Bai, Yoshiki Yamazaki, and Kwok Fai Cheung
Department of Ocean and Resources Engineering
University of Hawaii at Manoa
Portland, Oregon
February 9 - 10, 2015
NEOWAVE Validation
• NEOWAVE Model Introduction
• Prior Model Validation - 2011 Tohoku Tsunami
– Honolulu Coast
• Benchmark Cases
– BM1
– BM2
– BM5
• Conclusions
Outline
Governing Equations
● Depth-integrated, Non-hydrostatic Equations Weakly dispersive waves through non-hydrostatic pressure
(Stelling and Zijlema, 2003; Yamazaki et al., 2009 & 2011)
Numerical Schemes
● Semi-implicit, Finite Difference (FD) Model Explicit hydrostatic solution
Implicit non-hydrostatic solution
● Two-Way, Grid-Nesting Scheme
Standard grid refinement scheme for FD tsunami models
● Momentum Conserved Advection (MCA) Scheme Shock capturing scheme for FD models
(Stelling and Duinmeijer, 2003; Yamazaki et al., 2009 & 2011)
NEOWAVE
Non-hydrostatic Evolution of Ocean Wave
Non-hydrostatic/Hydrostatic Hybrid Scheme
Approximation of Energetic Breaking Waves as Bores
Dispersive wave model generates artificial spike when wave
front becomes steep at wave breaking or bore development
Breaking region may consider hydrostatic for bore approximation
- Hydrostatic pressure is main driving force
- Non-hydrostatic pressure effects are negligible
Non-hydrostatic solution(TURN ON non-hydrostatic)
Hydrostatic solution(TURN OFF non-hydrostatic)
Breaking initiation
gDvu 5.022
Breaking termination
gDvu 15.022
Check status of
each cell every
time step
NO wave breaking
Prior Validation with 2011
Tohoku Tsunami Tide Gauge
Kilo Nalu
Surface elevation time series (cm) Amplitude Spectrum
Entrance ADCPKilo Nalu
Velocity components time series (cm/s) Amplitude Spectrum
NEOWAVE
High Sampling Rate Mea.
Low Sampling Rate Mea.
BM NO.1Shallow-Water Flow around Submerged Conical Island
(Lloyd and Stansby, 1997)
RIGHT
boundary
Open B.C.
27 cm
5.4 cm
500 cm 102 cm
976 m (975cm)
152 cm
U0 = 11.5 cm/s
Manning coefficient =0 .01
1.0-cm grid
2.0-cm grid
4.0-cm grid
BM NO.1Close View of Conical Island with Three Resolutions
75 cm
102 m
5 cm
4.9 cm 5.4 cm
27 cm
center point
left point
1cm 2cm 4cmExp.
center point
left point
BM NO.1 – Velocity Comparison
BM NO.1 – Vortex Field Comparison
Δx=1cm (977x153)
Δx=2cm (485x77)
Δx=4cm (245x39)
Δt=0.0025s
Δt=0.005s
Δt=0.01s
BM NO. 1 – Vortex
Field Closeview 1cm
2cm 4cm
• Vortices are formed faster in
finer grid
• Clear boundary between
clockwise and
counterclockwise vortex field
BM NO.2Shallow-Water Flow around Hilo Bay
RIGHT boundary
Wall B.C.
TOP boundary
Initiates solitary wave.
: tide gauge or ADCP
: initial wave extraction point
BOTTOM boundary
Wall B.C.
LEFT boundary
Wall B.C.
Control Point
HA1125HA1126
Tide Gauge
u
v
BM NO.2: DEM Data ModificationOriginal NTHMP Hilo DEM Data - Induce Instability
BM NO.2: DEM Data ModificationModified NTHMP Hilo DEM Data - Remove Instability
5m 10m 20mrecord
ed.
BM NO.2 – Tide Gauge Comparison
(hour)
5m 10m 20mMea.
8 9 10 11 12 13 10 100
spectrum amplitudeHilo Tide Gauge - Nested Grid
NEOWAVEMea.
BM NO.2 – Velocity Comparison
5m 10m 20mMea.
5m 10m 20mMea.
BM NO.2 – Velocity Comparison
BM NO.2 – Vortex Field Comparison
Δt=0.025s
Δx=5m (1401x1029)
Δt=0.05s
Δx=10m (701x515)
Δt=0.1s
Δx=20m (351x258)2m/s2m/s 2m/s
(a) Overview
(b) Side view
BM NO. 5Shallow Shelf with a Conical Island
10 m 7.5 m
45 m
8.4 m
6.6 m 12.5 m12.5 m 12.5 m
25.9 m 19.1 m
26 m
r=3 m
17 m
1 / 3.57
1 / 15
1 / 250
78 cm
22 cm
28 cm
45 cm
3 cm
8 cm
• Inundation Science & Engineering Cooperative (ISEC)
Community Workshop
• Oregon State University, Corvallis, Oregon, July 8 - 10,
2009 (http://isec.nacse.org/workshop/isec_workshop_2009/)
Gird spaces:
Δ x = Δ y = 5cm, 10cm, 25cm
Time step:
Δ t = 0.002s, 0.003s, 0.01s
Manning coefficients:
n = 0.012 (for finished concrete)
Initial condition
Solitary wave
h = 78cm
A = 0.5 h
BOTTOM boundary
Wall B.C.
RIGHT boundary
Wall B.C.
LEFT boundary
Initiates solitary wave.
TOP boundary
Wall B.C.
: gesture gauge
: wave gauge and ADV
BM NO. 5Model Setup
x
y
BM NO.5Surface
Elevation
Comparison at
Gesture Gauges
5cm
10cm
25cm
Exp.
5cm
10cm
25cm
Exp.
BM NO.5Velocity Components Comparison at ADVs
Central Front Point
Central Back Point
Right Back Point
BM NO. 5Vortex Field Comparison
5cm
10cm 25cm
• Vortices are generated in the wake and
around the island
• Vortex strength is weaker in coarser grid
but general pattern remains in all three
grid sizes
• Runup process also involve vortex field
Conclusions
• NEOWAVE can reproduce the mean flow
which is less sensitive to resolution
• Numerically generated vortex field depends on
– Spatial and temporal resolution
– Bottom friction
– Numerical scheme
– Generation mechanism
• Relation with physical vortex field