Presentation by Stuart Pearson, TU Delft/Deltares, at the XBeach User Day 2018, during Delft Software Days - Edition 2018. Thursday, 15 November 2018, Delft.
DSD-INT 2018 Assessment of runup reduction potential due to coral reef restoration - Pearson
1. 1
Assessment of Runup Reduction Potential
Due to Coral Reef Restoration
Stuart Pearson1,2, Marlies van der Lugt1, Ap van Dongeren1,
Gerben Hagenaars1, Andreas Burzel1, Boris Ton van Zanten3
2. Background
2
Reef-fronted tropical coastlines are faced with an increasing
threat of wave-induced flooding
Coral reef restoration may increase the resilience of the reef
Increases friction and reduces runup
Commissioned by the World Bank to investigate the runup
reduction potential on three islands in the Seychelles
(USGS, 2014)
5. Motivation
5
Catastrophic bleaching hit Seychelles reefs in 1998
Mean coral cover reduced to <11% of its historic extent
Coral bleaching reduces “structural complexity”
= hydraulic roughness
6. Objective
6
Identify priority sites with high coastal flood risk
and strong potential for flood risk reduction via
coral restoration
7. Methodology
7
1. Define ~300 transects around the Seychelles
2. Estimate reef geometry via remote sensing
3. Feed reef properties and hydrodynamic forcing into
the XBeach-based BEWARE system
4. Estimate runup as a function of changes in
coral cover
Runup potential = proxy for coastal flooding
9. Methodology: BEWARE
9
BEWARE (Bayesian Estimation of
Wave Attack in Reef Environments)
Generated synthetic dataset using
XBeach Non-Hydrostatic model
Expanded existing dataset to cover Seychelles
parameter space
(Pearson et al., 2017)
10. Methodology: BEWARE
10
To feed BEWARE, we used remote sensing-
derived bathymetry and hydrodynamic forcing
represented by 9 wave and water level conditions
Bayesian Network (Netica) then used to query database
Variable Values
Hs0
3, 4, 5, 6, 7
Hs0/L0
0.05, 0.1, 0.2
η0
0, 0.5, 1, 1.5, 2, 2.5, 3
Wreef
0, 50, 100, 150, 200, 250, 300,
350, 400, 500
βf
0.05, 0.1, 0.2, 1.0
βb
0.01, 0.05, 0.1, 0.5
cf
0.001, 0.01, 0.05, 0.5
(Pearson et al., 2017)
11. Methodology: Bayesian Network Example
11
What is the most likely runup for given conditions?
Prior prediction (no additional information):
Updated (Posterior) Prediction (with additional information):
High Tide
Hs=3.0 m
Tp=18 s
Wreef = 150 m
Cf = 0.05
Reef Slope = 1/2
Beach Slope = 1/10
1 to 2 m Runup
(83% chance)
All Possible
Hydrodynamic
Conditions
All Reefs in
Database
Equal Probability
(~20% chance)
Prunup
(%)
Runup
Prunup
(%)
Runup
12. Methodology: Bayesian Network Example
12
What is the most likely runup for given conditions?
Prior prediction (no additional information):
Updated (Posterior) Prediction (with additional information):
High Tide
Hs=3.0 m
Tp=18 s
Wreef = 150 m
Cf = 0.001
Reef Slope = 1/2
Beach Slope = 1/10
3 to 4 m Runup
(83% chance)
All Possible
Hydrodynamic
Conditions
All Reefs in
Database
Equal Probability
(~20% chance)
Prunup
(%)
Runup
Prunup
(%)
RunupWhat if roughness changes
due to coral die-off?
13. Methodology: Remote-Sensed Bathymetry
13
Near Infrared Red signal
Landward
Shoreline
Seaward
Methodology developed by Hagenaars et al. (2017)
Shoreline position Normalized Difference Water Index
Reef break Near Infra Red signal
Breakerline
Reef Width
14. Reef Width: 222 m
Depth: 1.5 m
Methodology: Remote-Sensed Bathymetry
14
Reef depth and offshore slope found from aerosol
(green and red bands)
Depths are referenced to MSL using tide information
Βf: 1/28
15. Interpreting the Results
15
A: Increasing roughness
strongly reduces runup
Higher restoration potential
B: Increasing roughness
has little effect on runup
Low restoration potential
C: Profile is highly
sensitive to changes in
roughness
Little effect for small
increases
Strong runup reduction for
high roughness
16. Results: La Digue – 4 m Waves
16
East Side
• Narrower reefs
• Medium cover
has little effect
West Side
• Wider reefs
• Medium cover has
some effect
• High cover is
most effective General
• Low cover has
no effect
Large
reduction at
priority site
17. Results: La Digue – 6 m Waves (+2 m)
17
Increasing Wave
Height
• Differences
between “no-high”
roughness become
more pronounced
18. Results: Mahe
18
• Runup reduction shows high spatial variation due to
alongshore differences in:
• Wave climate
• Reef geometry
• Depth
• Width
• Slopes
19. Limitations: Determining Roughness
19
Translating baseline coral cover/structural complexity to
hydrodynamic roughness
Tested sensitivity to range of roughness coefficients,
representing relative degrees of coral coverage
Cf = 0.001 (sandy beach) Cf = 0.1 (high coral coverage)
20. Reef Crest
????
Limitations: Remote Sensing
20
Remote sensing-derived bathymetry is less reliable for
reefs without a defined crest and reef flat
21. Conclusions
21
Generally, low coral cover offers little benefit
as flood protection when compared with medium or
high coral cover.
No point in partial restorations (from flood risk perspective)
As wave height increases, the absolute differences
between high and no (or low) roughness also
increase.
“Lost protection” of a dead reef will become more apparent
during storms
22. Conclusions
22
Shows the protective value of healthy, rough reefs
Demonstrates viability of BEWARE system for rapid
assessments of flood risk
23. What Next?
23
Site-specific data collection and modelling is necessary
Optimize cross-shore placement of restorations
Improve handling of complex bathymetry
Remote Sensing: Move beyond “classical” fringing reef
XBeach Modelling: Schematize 2D effects, diverse profile
shapes