Journal of Coastal Research

00

0

000–000

Coconut Creek, Florida

Month 0000

Modelling the Impacts of Mangrove Vegetation Structure on
Wave Dissipation in Ben Tre Province, Vietnam, under
Different Climate Change Scenarios
Nguyen Thi Kim Cuc†‡, Tomohiro Suzuki§††‡‡, Erik D. de Ruyter van Steveninck§§,
Hoang Hai†††

†

Department of Natural Resources
Management
Faculty of Water Resources Engineering
Water Resources University
175 Tay Son
Dong Da, Hanoi, Vietnam


‡
Mangrove Ecosystem Research Division
Centre for Natural Resources and

Delft, The Netherlands

†††

Faculty of Environment
Da Nang University of Technology,
Da Nang, Vietnam

Abstract
Cuc, N.T.K.; Suzuki, T.; Ruyter van Steveninck, E.D. de, and Hai, H., 0000. Modelling the impacts of mangrove
vegetation structure on wave dissipation in Ben Tre Province, Vietnam, under different climate change scenarios.
Journal of Coastal Research, 00(0), 000–000. Coconut Creek (Florida), ISSN 0749-0208.
Mangroves are widely distributed along the coastline of Vietnam, where they provide protection against sea waves
caused by extreme weather. Impacts of climate change, together with population growth and economic development, are
expected to exert pressure on these vulnerable systems. In this study the numerical wave-propagation model SWANVEG (Simulating Waves Nearshore–Vegetation) was used to simulate the possible impacts of climate change on the
wave-dissipation capacity of different types of mangrove vegetation. Vegetation characteristics were assessed in planted
plots (Rhizophora apiculata and a mix of R. mucronata, Sonneratia caseolaris, Avicennia alba, and Nypa fructicans) and
in natural regenerated areas (A. alba and S. caseolaris) in Thanh Phu Natural Reserve, Mekong Delta, Vietnam; these
assessments were used as model input. Different sea levels and mangrove vegetation characteristics were used to
simulate the potential impacts of climate change. Planted plots with a cover of 70% reduced the height of incoming waves
by 60%, compared with 40% for natural regenerated forest. Reducing the vegetation cover in planted plots from 70% to
50%, 35%, and 0% resulted in wave-height reductions of 51%, 42%, and À4%, respectively. A sea level rise (SLR) up to
0.96 m did not change the wave-dissipation potential of R. apiculata planted in the plots. However, an assumed decline in
the width of vegetation from 1.5 km to 0.5 km, e.g. as a consequence of coastal erosion, reduced the height of incoming
waves 21% (no SLR) and 29% (0.96 m SLR), as compared to 60% and 59%, respectively, without erosion.

ADDITIONAL INDEX WORDS: Mangrove, wave, climate change, SWAN-VEG, Thanh Phu.

INTRODUCTION
Tropical coastlines are under great pressure due to a rapid

and Cahalan, 1992). Based on field observations, Mazda,
Wolanski, et al. (1997) have shown the quantitative effects of
two mangrove species, Rhizophora stylosa and Kandelia
candel, on reduction of the impact of sea waves. Massel,
Furukawa, and Brinkman (1999) have discussed the wavedissipation capacity of Rhizophora spp. and Sonneratia spp.
based on a mathematical model. A predictive model of wave
propagation through a nonuniform forest in water of changing
depth in the mangrove forest of Can Gio, Vietnam, has been
developed by Vo-Luong and Massel (2008). Recently, Mendez
and Losada (2004) and Suzuki et al. (2011) have incorporated a
full frequency-direction wave spectrum in the numerical wave
model SWAN (Simulating Waves Nearshore) and additionally
included a layer-wise implementation of vegetation characteristics. Their results, however, cannot be applied directly to
other regions or species, as each mangrove species has a unique
configuration of trunks, prop roots, and pneumatophores that
produces a different drag force (Wolanski et al., 2001) and
therefore results in a different reduction rate of sea waves. For
example, with regard to Bruguiera spp., Sonneratia spp.,
Avicennia spp., and Nypa fruticans, no information on their
wave-dissipation capacity exists, either on their quantitative
hydrological functions or on their physical impact. Accordingly,
for useful and effective mangrove planting, there is a need for
quantitative knowledge of the physical impact of individual
mangrove species in relation to their wave-dissipation capacity.
Vietnam is located in the tropical region of Asia. With its long
coastlines and high concentration of population and economic
activities in coastal areas, its coastal ecosystems and communities could potentially be dramatically impacted by a rise in
sea level (CCFSC, 2001). According to Bates et al. (2008),
changes in the hydrological cycle are expected to be the most
significant aspect of climate change to affect the Mekong Delta.

However, deforestation due to war activities and conversion to
agricultural land and (more recently) shrimp ponds (Hong,
1999) has resulted in a decline from 408,500 ha in 1943
(Maurand, 1943) to 290,000 ha in 1962 (Rollet, 1981), 252,000
ha in 1982 (FIPI, 1982), 156,608 ha in 2001 (FIPI, 2001),
209,741 ha in 2006 (FIPI, 2007), and 139,955 ha in 2010
(MARD, 2011).
In this paper, the SWAN model has been used to analyze the
impact of different mangrove species and densities on wavedissipation capacity and wave-energy dissipation characteristics in Thanh Phu Natural Reserve, Mekong Delta. The SWAN
model is a third-generation wave model developed at Delft
University of Technology, the Netherlands (TU Delft, 2011).
The SWAN model computes the processes of wave generation
by wind, quadruplet wave-wave interactions, white-capping,
bottom friction, depth-induced breaking interactions, and triad
wave-wave interactions; it is thus capable of estimating wave
propagation from offshore to nearshore. In addition to the
fundamental features of waves, the effects of vegetation have
been recently incorporated into the model. Using data about
the present mangrove vegetation in Thanh Phu Natural
Reserve, a comparison has been made between a planted area
dominated by Rhizophora sp. and a naturally regenerated area
dominated by Avicennia sp. To analyse the impacts of tree
density, sea level rise (SLR) associated with climate change,
and coastal erosion, only data about the planted forest have
been used.

MATERIALS AND METHODS
Study Area
The study was carried out in Thanh Phu Natural Reserve,
Ben Tre Province, Vietnam (Figure 1). Ben Tre is a coastal

swamps. Thanh Phu Natural Reserve contains a narrow strip
(about 0.8–5.0 km) of mangroves along the coastline between
two of the mouths of the Mekong River: the Co Chien and Ham
Luong estuaries (Figure 1). As is the case with other sites on
the eastern coastline of the Mekong Delta, Thanh Phu Natural
Reserve is strongly affected by both erosion as well as accretion
(Sub-FIPI II, 1998, 2003).
In 2009, Thanh Phu had a population of 127,574 inhabitants.
The district covers an area of 44,350 ha, of which 17,300 ha is
used for aquaculture and 15,000 ha as paddy fields. Another
2000–3000 ha is used for industrial crops, fruit farming, etc.
(Ben Tre Statistics Office, 2010).

Mangrove Vegetation Structure
In the study area, planted Rhizophora apiculata is the
predominant species, representing more than 80% of the
mangrove vegetation. Three transects perpendicular to the
coastline were laid out with a length of 1–2 km, depending on
width of the vegetation. In each transect three 10 3 10 m plots
were established—one close to the shore, one in the middle, and
one at the end of the transect—according to accessibility. In
each plot all trees were counted, while a separate 1 3 1 m
subplot was established within each plot to count seedlings
(less than 1.0 m in height) and saplings (1.0–4 m in height).
Trees taller than 4 m were identified to the species level, and for
these the following parameters were measured as described by
English et al. (1994):
(1)
(2)
(3)

r¯ ¼

k¯ ¼

EÀ1
tot ¼

SWAN-VEG Model
The SWAN-VEG model consists of the original SWAN model
with a vegetation module added. This module consists of a
variable for energy dissipation due to vegetation. The dissipated energy is subtracted from the incoming wave energy, which
results in less wave energy behind the vegetation field and thus
a lower wave height. Vegetation is modelled as cylindrical
obstacles; vegetation characteristics such as height, width,
density, and drag coefficient are used to determine the
magnitude of the dissipation term. In addition, wave characteristics like significant wave height and peak period influence
the energy-dissipation term (de Oude et al., 2010). The SWANVEG module for wave dissipation by vegetation (Suzuki et al.,
2011) is based on Mendez and Losada (2004) and de Oude et al.
(2010). The model assumes a group of cylinders as representation of vegetation. It includes vertical-layer schematization,
making it possible to calculate multilayer structures such as
mangroves. The energy dissipation term in SWAN-VEG is
described by Suzuki et al. (2011):
Sds;veg ¼

n
X

Sds;veg;i ;

ð1Þ

aih is the plant height, and E(r,h) is the wave-energy density.
The bulk drag coefficient value 1.0 is used for simplicity in this
study as per de Oude et al. (2010) and Narayan et al. (2010). As
far as we are concerned, the bulk drag coefficient value of
mangrove forest has not been very clear up to now due to the
complexity of the vortex effect around cylindrical structures (e.g.
stems and roots) under wave motion. For instance, different
cylinder densities and arrays, which generate different vortex
patterns, give different bulk drag coefficient values under
waves. Furthermore, even drag coefficient values of a single
cylinder range from 0.5 to 2.5 (Sarpkaya and Michael, 1981) with
changes in Keulegan–Carpenter (KC) value and beta value. The

EÀ1
tot



EÀ1
tot

Z
0

2P

Z

2P
0

ð3Þ
À2
ð4Þ

‘

Eðr; hÞdrdh

ð5Þ

0

The wave-dissipation term associated with vegetation,
Sds,veg, is implemented in the source term Stot, as shown in
Equation (6):
Stot ¼ Sin þ Snl3 þ Snl4 þ Sds;b þ Sds;wc þ Sds;br

ð6Þ

where Sds,b is bottom friction, Sds,wc is white-capping, Sds,br is
depth-induced breaking, Sin is wave growth due to wind input,
Snl4 and (Snl3) are energy transfer within the spectrum due to
nonlinear wave-wave interactions such as quadruplets (Snl4)
and triads (Snl3). These six processes contribute to the source
term Stot. The evolution of the wave spectrum is described by
the spectral action balance equation, which for Cartesian
coordinates is given by (e.g. Hasselmann et al., 1973)
d
d
d

Water Level, Wave Height and Wave Peak Period
According to the Centre for Meteorology Forecast in Ben Tre
and the Vietnam Institute of Meteorology, Hydrology, and
Environment, the maximum recorded value of significant wave
height on the coast of Ben Tre during the past 100 years was 0.9
m at a point 2.5 km from the shoreline with a depth ofÀ1 m. The
shoreline refers to the line behind the 1.5 km wide mangrove
vegetation (see Figure 2). The peak periods range from 7 s to 10
s. This value occurred when the water level was at 4.1 m (total
depth at À1 m is 5.1 m). This value was used as the extremeevent condition in the simulation.

Vegetation Type and Density
The local forest management system allows households to
harvest or convert up to 30% of mangrove areas that are
managed by them to water surface for aquaculture activities.

Journal of Coastal Research, Vol. 00, No. 0, 0000


Modelling the Impacts of Mangrove Vegetation Structure on Wave Dissipation

0

Table 1. Summary of data used as input for the SWAN-VEG model to analyse the impact of mangrove type (planted vs. natural regenerated) and coverage,
SLR, and coastal erosion with and without SLR.
Vegetation

Vegetation type and density

SLR

P
P
P
P

70
0
50
35
70

1.5
1.5

70

1.0
0.5
1.0
0.5

70

WL (m)

Hs (m)

Tp (s)

4.1

0.9

9.0

SLR (m)
0
0

1.5

Abbreviations: Width ¼ width of mangrove forest, SLR ¼ possible sea level–rise scenarios based on MONRE (2010), WL ¼ water level, Hs ¼ significant wave
height, Tp ¼ wave peak period, P ¼ planted mangrove type, N ¼ natural regenerated mangrove type. The simulation codes are also used in Figures 3–6.
Note: mangroves coverage(0–70% cover); sea level rise (0.39-0.96 m) and coastal erosion (from 1.5-0.5 km width) with/without sea level rise (0.96 m).

Thus coverage of 70% of the actual measured figures was used
to compare the impact of planted forests with naturally
regenerated forests (codes A1 and A3, respectively; see Table
1). The impact of vegetation density was analysed by reducing
the coverage in the planted forest from 70% (A1) to 50% (A5),
35% (A6), and 0% (A4) (Table 1).

caseolaris and A. alba were mainly found on the alluvial coastal
and river mud flats, while N. fruticans was planted in the
inland brackish aquaculture ponds. In the planted area, the
tree density was 0.16 trees mÀ2, with a mean tree height of 12.7

Sea Level Rise
Sea level rise scenarios for Vietnam have been developed
based on different emission scenarios: low (B1), medium (B2),
and high (A1F1) (MONRE, 2009). MONRE (2010) recalculated

A3, A4, A5, and A6: vegetation type and density (see Table 1 footnote for
explanation of simulation codes). Wave: Hs denotes significant wave
height; Tp denotes wave peak period; WL denotes water level. The dotted
arrow shows an SLR of 0.96 m (scenario 3); the dashed arrow shows an SLR
of 0.65 m (senario 2); the dotted and dashed arrow shows an SLR of 0.39 m
(senario 1).

Journal of Coastal Research, Vol. 00, No. 0, 0000


0

Cuc et al.

Table 2. Mean values (significant difference) of mangrove vegetation
characteristics in both planted Rhizophora (n ¼ 9) and natural
regenerated (n ¼ 3) plots in Thanh Phu Natural Reserve.
Parameter
À2

Tree density (m )
Diameter at 1.3 m (m)
Height branch (m)
Height leaf (m)
Height top (m)
Stilt roots
Density (mÀ2)
Diameter (m)
Height root (m)


—
—

m and a diameter (measured at a height of 1.3 m) of 0.12 m
(Table 2).
Vegetation in the naturally regenerated site consisted
mainly of A. alba and S. caseolaris. Avicennia alba grows
naturally on mud flats with average tidal levels of 1.0 m to 1.5
m and is directly exposed to waves and wind from the sea. In
some areas, A. officinalis appears landwards from the A. alba
zone, protected from the direct impact of waves and wind.
Sonneratia caseolaris develops on soft mud at depths of 1.0–1.5
m, affected by tides, waves and wind, and a mixture of seawater
and freshwater from rivers. Although tree density in these
plots was higher than in the planted plots (0.24 trees mÀ2), their
average size was smaller (3.9 m high and a diameter of 0.11 m).
Besides these differences, another major difference between
the planted and naturally regenerated sites that could affect
wave-dissipation capacity is caused by the different root
structures for the various tree species, i.e. stilt roots for
Rhizophora and pneumatophores for Avicennia (Figure 3).
Based on the characteristics of the planted Rhizophora trees
(h top, 12.73 m; h leaf, 7.20 m; h branch, 6.67 m) and the
maximum water level (5.06 m), it was assumed that the stems
and roots of the trees will play a major role in wave dissipation;
therefore the influence of the canopy was not considered.

Figure 3. Root structures for the different tree species: stilt roots for
Rhizophora sp. and pneumatophores for Avicennia sp. and Sonneratia sp.
(Hong and San, 1993).

35% resulted in a reduction of wave height of 51% and 42%,
respectively (A5 and A6; see Figure 2, Table 3). When the
vegetation was removed completely (A4), the wave height even
slightly increased after passing the 1.5 km of no forest (now
cleared) (Figure 2, Table 3).

Sea Level Rise
Figure 4 shows the transmitted wave heights in the planted
forest for three different sea levels (B1, 4.49 m; B2, 4.75 m; and
B3, 5.06 m). Although the water levels increased, the impact of
the mangrove forest on the incoming wave height was the

Parameter
Vegetation type
A1
A3
Density
A1
A5
A6
A4
SLR
A1
B1
B2
B3
Erosion
A1
C1
C2

Figure 4. Spatial variation of significant wave height for simulations B1, B2,
and B3 according to SLR (see Table 1 footnote for explanation of simulation
codes).

Figure 5. Spatial variation of significant wave height in erosion simulations
C1/C3 and C2/C4 (see Table 1 footnote for explanation of simulation codes).

DISCUSSION AND CONCLUSIONS

the planted Rhizophora forest. The resulting wave height of
almost 0.4 m is not different in the four tested water depths.
This probably could be explained by the high density of stems
and aboveground roots distributed throughout the whole water
depth, even with a rise in sea level.
The width of the vegetation, however, seems to have more
impact on the wave-dissipation capacity of the planted
mangrove forest. Although a reduction from 1.5 km to 1.0 km
resulted in a slight increase in wave height from 0.4 m to 0.5 m,
further reducing the forest width to 0.5 km resulted in a wave
height of 0.7 m. The actual sea level (present level and an
increase of 0.96 m) did not affect these results. In addition to
vegetation density, the width of the area to be planted is an
important factor in wave attenuation for protecting tropical
coastlines.
Quartel et al. (2007) found that wave heights were depth
limited. Small water depths corresponded with small wave
heights and large water depths with higher incident waves.
With the expected SLR as a consequence of climate change,
coastal areas will face the impact of higher wave height. The
existence of mangrove forests in coastal areas in general, and in

Besides reducing the height of incoming waves, mangrove
forests probably also reduce wind-driven and tidal currents due
to their dense network of stems, branches, and aboveground
roots (Quartel et al., 2007). The mangrove trees in the study
area rise quite high above the water level (12.7 m on average)
and might efficiently reduce wind energy. Further studies on
this role of mangroves should be carried out, especially in the
Mekong Delta area, where the windy season has a strong
influence on the coastline. Increasing the sea level in the model
up to 0.96 m, thus anticipating a possible SLR due to climate
change, did not reduce the wave-energy dissipation potential of

ACKNOWLEDGMENTS
We thank the Department of Agriculture and Rural
Development of Ben Tre Province and the Management
Board of Thanh Phu Natural Reserve for their unstinting
support for data collection and field survey. The work
reported here was undertaken as part of the research
programme ‘‘PRoACC—Postdoctoral Research Programme
on Climate Change Adaptation in the Mekong River Basin’’.
The project is funded by the Netherlands Ministry of
Development Cooperation (DGIS) through the UNESCOIHE Partnership Research Fund. This research project is a

Journal of Coastal Research, Vol. 00, No. 0, 0000


0

Cuc et al.


de Oude, R. Augustijn, D. Dekker, F. Wijnberg, K. de Vries, M., and
Suzuki, T., 2010. Modelling wave attenuation by vegetation in
SWAN to determine a vegetation field for decreasing dike height in
the Noordwaard, the Netherlands. Environmental Hydraulics, 1,
253–258.
English, S. Wilkinson, C., and Baker, V., 1994. Survey Manual for
Tropical Marine Resources. Townsville, Queensland, Australia:
Australian Institute of Marine Science, 368p.
Fabrice, G.R. and Claudia, K. (eds.), 2012. The Mekong Delta System
Interdisciplinary Analyses of a River Delta. Dordrecht, The
Netherlands: Springer Environmental Science and Engineering,
463p.
FIPI (Forest Inventory and Planning Institute) Staff, 1982. Results of
Review and Planning of Coastal Protection Forests. Hanoi City,
Vietnam: Forest Inventory and Planning Institute, 50p. [in
Vietnamese].
FIPI Staff, 2001. Results of National Forest Survey Following
Decision No. 03/2001 QD/TTG of the Prime Minister, dated 5
January 2001. Hanoi, Vietnam: Forest Inventory and Planning
Institute, 48p. [in Vietnamese].
FIPI Staff, 2007. Results of National Forest Survey Following
Decision No 405/TTg-KTN of the Prime Minister, dated March
16, 2009. City, Vietnam: Forest Inventory and Planning Institute,
64p. [in Vietnamese].
Fonseca, M.S. and Cahalan, J.A., 1992. A preliminary evaluation of
wave attenuation by four species of seagrass. Estuarine Coastal
Shelf Science, 35, 565–576.
Hasselmann, K. Barnett, T. Bouws, E. Carlson, H. Cartwright, D.
Enke, K. Ewing, J. Gienapp, H. Hasselmann, D. Kruseman, P.
Meerburg, A. Muller, P. Olbers, D. Richter, K. Snell, W., and

Mangroves Salt Marshes, 1, 127–135.
Mazda, Y. Magi, M. Ikeda, Y. Kurokawa, T., and Asano, T., 2006.
Wave reduction in a mangrove forest dominated by Sonneratia sp.
Wetlands Ecology and Management, 14, 365–378.
Mazda, Y. Wolanski, E. King, B. Sase, A. Ohtsuka, D., and Magi, M.,
1997. Drag force due to vegetation in mangrove swamps.
Mangroves Salt Marshes, 1, 193–199.
Mendez, F.M. and Losada, I.J., 2004. An empirical model to estimate
the propagation of random breaking and nonbreaking waves over
vegetation fields. Coastal Engineering, 51, 103–118.
MONRE (Ministry of Natural Resources and Environment), 2009.
Climate Change, Sea Level Rise Scenarios for Viet Nam. Hanoi:
Ministry of Natural Resources and Environment, Hanoi Press, 34p.
MONRE, 2010. Sea Level Rise Scenarios and Possible Disaster Risk
Reduction. Hanoi: Viet Nam Institute of Meteorology, Hydrology
and Environment, Ministry of Natural Resources and Environment, Hanoi Press, 80p.
Narayan, S. Suzuki, T. Stive, M. Verhagen, H. Ursem, W., and
Ranasinghe, R., 2010. The effectiveness of mangrove in attenuating
cyclone induced waves. In: Proceedings of the 32nd International
Conference on Coastal Engineering, Shanghai, China, (in press).
Ngo An, 2004. Contributing to the Research on Mangrove Land Use
Planning of Ben Tre Province on a Sustainable Development Basis.
Ho Chi Minh City, Vietnam: Ho Chi Minh Press, 38p. [in
Vietnamese].
Quartel, S. Kroon, A. Augustinus, P.G.E.F. van Santen, P., and Tri,
N.H., 2007. Wave attenuation in coastal mangroves in the Red
River Delta, Viet Nam. Journal of Asian Earth Sciences, 29, 576–
584.
Rollet, B., 1981. Bibliography on Mangrove Research 1600–1975.
UNESCO, Paris, France, 479p.

mangrove forests of arbitrary depth. Journal of Marine Systems,
74, 603–622.

0

Vu, T.T., 1994. Estuarine Ecosystems of Viet Nam. Hanoi, Vietnam:
Scientific and Technical Publishing House, 271p.
Wamdi Group, 1988. The WAM model—a third generation ocean
wave prediction model. Journal of Physical Oceanography, 18,
1775–1810.
Wolanski, E. Mazda, Y. Furukawa, K. Ridd, P. Kitheka, J. Spagnol,
S., and Stieglitz, T., 2001. Water circulation in mangroves, and its
implications for biodiversity. In: Wolanski, E. (ed.), Oceanographic
Processes of Coral Reefs. London: CRC, pp. 53–76.

Journal of Coastal Research, Vol. 00, No. 0, 0000