Journal of Hydro-environment Research 23 (2019) 1–9

Contents lists available at ScienceDirect

Journal of Hydro-environment Research

journal homepage: www.elsevier.com/locate/jher

Longshore propagation erosion of beach in the vicinity of tsunami-induced T

concave shoreline
⁎
Vo Cong Hoanga, , Hitoshi Tanakab, Yuta Mitobec
a
Division of Civil Engineering, Thuyloi University, Southern Campus, 2, Truong Sa Street, Binh Thanh District, Ho Chi Minh City, Viet Nam
b
Department of Civil Engineering, Tohoku University, 6-6-06 Aoba, Sendai 980-8579, Japan
c
Department of Civil and Environmental Engineering, Tohoku Gakuin University, 13-1, Chuo 1, Tagajo City, 985-8537, Japan

A R T I C LE I N FO A B S T R A C T

Keywords: The 2011 Great East Japan Earthquake and Tsunami caused significant changes in the morphology of the Sendai
Tsunami Coast, Miyagi Prefecture, Japan. These tsunami waves and the return flows resulted in the formation of concave
River mouth shorelines in areas like lagoon locations adjacent to river mouths or where sandy beaches located on pre-existent
One-line model (closed) river mouths were breached. This study investigates the erosion of sandy beaches adjacent to the
Diffusion coefficient
concave shoreline at the Nanakita River mouth and the Akaiko breaching areas after the tsunami through an
Erosion propagation
Concave shoreline
analysis of aerial photographs and analytical solutions of the one-line model. Soon after the tsunami, sandy
Analytical solution beaches on both sides of the concave shoreline experienced erosion that was propagating along the coast.
Morphological change Analysis of the analytical solution of the one-line model indicates that the erosion propagation distance is
proportional to square root of elapsed time. Measured data of erosion propagation distance from aerial photo-
graph is found to match well with theoretical results. An approach to estimate diffusion coefficient, which is an
important parameter in numerical simulation of shoreline change, is introduced.

1. Introduction
The tsunami of March 11, 2011, caused severe damages to the
coastal areas in northeastern Japan. At river mouth areas, it was
common to observe the disappearance the sand spit at the river mouth
induced by the tsunami. In many cases, the sand barrier in front of
lagoon adjacent to river mouths was also severely eroded by the tsu-
nami. At these river mouths, the concave shoreline was observed after
the tsunami (Tanaka et al., 2012). One of the factors leading to the fast
or slow recovery of morphology in these areas is the sediment supply
from adjacent sandy coasts. The Nanakita River mouth, which is located
in the northern part of Sendai Coast, Miyagi Prefecture, Japan, is a
typical case of concave shoreline formation after the tsunami. With
sufficient sediment supply from adjacent sandy beaches, the recovery of
concave morphology was comparatively fast. However, the erosion of
sandy beaches on both sides of the concave portion was also observed
during the recovery process. Sediment from adjacent beaches was
transported into the concave portion leading to the advancement of
shoreline in this area. The erosion first happened on the parts of the
beach which adjoining the concave portion subsequently propagating
along the coast. Moreover, during the recovery process, the intrusion of
sand spit into the river mouth area was also observed. Hoang et al.
(2018) indicated that the scouring of river mouth deeper than the depth
of closure was the reason leading to the intrusion of the sand spit. These
observations reveal that the sink effect at river mouths was induced by
the 2011 tsunami. That sink effect is similar to the sink effect of the
tidal inlet through the formation of flood shoal in the tidal inlet
(FitzGerald, 1988; Kraus and Wamsley, 2003), or the sink effect of river
mouth due to the dredging (Uda, 1997).
The significant changes and recovery process of coastal and es-
tuarine morphology after a huge tsunami has attracted the attention of
many researchers. After the 2004 Indian Ocean Tsunami, there were
several works by researchers such as Ali and Narayana (2015),
Choowong et al. (2009), and Liew et al. (2010). They presented the
damages and recovery process of beach morphology of the affected
areas in Andaman and Nicobar Islands of India, Indonesia, and Thailand
based on yearly satellite images. There have been similar studies on the
damage and recovery process of morphology on Sendai Coast and
surrounding areas after the 2011 tsunami. Tanaka et al. (2014a) in-
vestigated the recovery of morphology at the Nanakita River mouth and
its relevant phenomena such as closure or re-opening based on the
analysis of continuously measured data of water level and aerial pho-
tographs. Hoang et al. (2015) proposed a new analytical solution of the
one-line model, which describes the post-tsunami evolution of concave

⁎
Corresponding author.
E-mail addresses: hoangvc@tlu.edu.vn (V.C. Hoang), hitoshi.tanaka.b7@tohoku.ac.jp (H. Tanaka), y_mitobe@mail.tohoku-gakuin.ac.jp (Y. Mitobe).

https://doi.org/10.1016/j.jher.2019.01.003
Received 19 January 2016; Received in revised form 9 January 2019; Accepted 22 January 2019
Available online 23 January 2019
1570-6443/ © 2019 International Association for Hydro-environment Engineering and Research, Asia Pacific Division. Published by Elsevier B.V. All rights
reserved.
V.C. Hoang et al.
shoreline bounded by solid boundaries at both ends. Tappin et al.
(2012) and Udo et al. (2012) gave an overview of morphology changes
induced by the tsunami in the Sendai Plain. In these studies, many sets
of bathymetry data and aerial photographs before and just after the
tsunami were utilized to reveal changes in morphology caused by the
tsunami. However, the erosion of beaches adjacent to the concave
shoreline and its propagation process have not been specifically in-
vestigated.
Similar to other studies that have been conducted on the recovery of
coastal morphology after the tsunami, this study would be also im-
portant for coastal management. Findings from this study would sup-
plement knowledge on the recovery of morphology after severe damage
which is still insufficient due to rare occurrence of such strong tsunami.
Such knowledge will also be elemental in the preparation of prevention
and recovery plan for similar disasters in the future.
This study attempts to investigate the erosion of sandy beaches
which are in the vicinity of the concave shoreline after a tsunami event
and relevant phenomenon through the analysis of aerial photographs
and analytical solutions of the one-line model.
2. Study areas and data collection
This study focuses on the coast around the Nanakita River mouth
which is located in the northern part of Sendai Coast, Miyagi Prefecture,
Japan (Fig. 1). There are long breakwaters at Sendai Port and Yuriage
Port where are located about 3 km north and about 8 km south of the
river mouth, respectively. In addition, there are 6 detached breakwaters
located in the middle of this area, about 3.8 km south of the River
mouth. Alongshore canal named Teizan connects the Nanakita River
and the Natori River which has the river mouth located next to Yuriage
Port. The predominant direction of longshore sediment transport on
this coast is from south to north, and it is completely blocked by the
long breakwaters mentioned above. Nanakita River has a length of
45 km, a basin area 229.1 km2 and an average annual river discharge of
about 10 m3/s. Total sediment supply from this river is about 2000 m3/
year.
In addition, Akaiko breaching, which is located approximately
16.5 km south of the Nanakita River mouth and on the east side of
Sendai Airport, is also taken as a study area. This breaching was formed
Fig. 1. Location map of the study area.
2
Journal of Hydro-environment Research 23 (2019) 1–9
by the tsunami waves and return flows at the location pre-existent river
mouth. More information on the mechanism of formation and map of
the former river mouth can be found in Tanaka et al. (2014b). Although
this area is not strictly river mouth, it was temporarily functioning as a
river mouth after the tsunami. There is a jetty located about 800 m on
the right side of the breaching.
Aerial photographs of the Nanakita River mouth area have been
taken frequently in every one or two months since 1990. All raw aerial
photographs are georeferenced to the World Geodetic System (WGS-
84). A line parallel to the shore, at an angle of 212° with the North
(measured clockwise) is taken as a baseline for shoreline position
measurement. Shoreline positions were extracted from aerial photo-
graphs in the alongshore direction. The tidal levels at Sendai Port,
which is calculated by the Japan Meteorological Agency (JMA), and the
average beach slope of 0.11 (Pradjoko and Tanaka, 2010) were used to
correct detected shoreline positions.
Aerial photographs of Akaiko breaching were collected from the
Geospatial Information Authority of Japan (GSI) and Google Earth. A
line parallel to the shore, at an angle of 163° with the North (measured
clockwise) is taken as a baseline for shoreline position measurement.
All the shoreline positions, which were extracted from photographs
collected from the third parties, have not been corrected to the tidal
level due to the lack of exact time of capture.
The characteristics of 20-min significant waves in the period from
January 1, 2009, to December 31, 2015, at Sendai Port are presented in
Fig. 2. The wave rose indicates that waves most come from the east and
south-east direction (90°–150° from the north). That result in the pre-
dominant direction from south to north of longshore sediment transport
along the coast including two study areas. The 20-min significant wave
heights, which are greater than 1 m, take 25.4 % of all the wave heights.
The characteristics of daily average wave data measured at the same
location in the period from 1991 to 2009 are shown in Pradjoko and
Tanaka (2010). Due to the damages of the facility at Sendai Port station
induced by the tsunami, the wave data in the period from after the
tsunami to the middle of 2012 is not available at this station. The tide at
Sendai Port Station is the semidiurnal type with the range of about
1.6 m.
The locations Nanakita River mouth and Akaiko breaching have
already been analyzed for studies such as Tanaka et al. (2014a,b), thus
more details on these areas can be found in those studies.
3. Results and discussion
3.1. Erosion propagation on beaches adjacent to the Nanakita River mouth
As mentioned above, the concave shoreline was commonly observed
at the river mouth area after the 2011 tsunami. A typical case is at the
Nanakita River mouth area. Concave shoreline was formed in this area
due to the erosion of sandy barrier in front of Gamo Lagoon adjacent to
the left bank of the river mouth and the flushing of sand spit at the river
mouth. The width of concave portion, which is the alongshore length of
the concave portion, is about 1050 m. This value is the distance be-
tween the right river bank and left the side of the concave portion,
measured just after the tsunami (Fig. 3).
In reality, the shape of the concave portion is not always truly
rectangular, and its bottom elevation is not uniform too. However, in
this study, the concave portion is assumed to be rectangular and of
uniform bottom elevation and the width is taken as the distance be-
tween the left and the right sides of the eroded area. Fig. 3 shows the
variation of shoreline position at the Nanakita River mouth area.
During the recovery process of morphology, sandy beaches on both
sides of the concave portion were being eroded. The erosion of shore-
line was propagating from the area bordering concave portion to the far
sides of sandy beaches.
Moreover, on the subsequent recovery process, the formation of
sand spit was observed, its location, however, shifted farther upstream
V.C. Hoang et al. Journal of Hydro-environment Research 23 (2019) 1–9

Significant W ave Height (m) Wave Rose

H 5 N (0°)
S
2 H <5
S
1.2 H <2 33%
S
1 H < 1.2 26.4%
S
0.5 H <1
S
0 H < 0.5 19.8%
S

13.2%

6.6%

W (270°) 0% E ( 90°)

S (180°)

Fig. 2. Characteristics of significant waves in the period from January 1, 2009 to December 31, 2015, at Sendai Port.

compared to its location before the tsunami. This phenomenon was

commonly observed at many river mouths after the 2011 tsunami. The
mechanism of this intrusion was presented by Hoang et al. (2018). The
tsunami waves and its return flows flushed the sand spit at the river
mouth scouring the river bed deeper than the depth of closure (8 m,
Uda, 1997).
Fig. 4 presents the erosion of shoreline and its propagation occur-
ring on the right bank of the Nanakita River mouth. It is noted that
aerial photographs, which shows in detail the changes of morphology at
the Nanakita River mouth after the tsunami, can be found in Tanaka
et al. (2012, 2014a). In order to evaluate the propagation of erosion
along the beach, a parameter which is called erosion propagation dis-
tance, xe, is utilized. It is defined as the distance from the concave
portion to the point where shoreline position becomes 99 % of initial
shoreline position, wherein initial shoreline position is shoreline posi-
tion soon after the tsunami. The illustration of this distance is presented Fig. 4. Variation of shoreline on the right bank of the Nanakita River mouth
in Fig. 4. after the tsunami (background image: May 3, 2011, Google Earth).
Fig. 5 shows the relationship between the eroded length of the
beach on the right bank of the Nanakita River mouth and elapsed time.

Fig. 3. Deformation of beach morphology around the Nanakita River mouth after the tsunami (background image: March 14, 2011, Google Earth).

3
V.C. Hoang et al.
1000
100
1
2 et al. (2015).
xe (m)
10
1 meas.
Eq. (14)
0.1
0.1 1 10 100 1000
t (days)
Fig. 5. Erosion propagation process on the right bank of the Nanakita River
mouth.
The results obtained from data are shown in Fig. 4. The results shown in
Fig. 5 confirm that the distance of erosion propagation increases in
proportion to the square root of elapsed time.
Although similar investigation the erosion propagation of sandy
beach from the north side of Gamo Lagoon to breakwater of Sendai Port
has been done, a clear relationship similar to which observed in Fig. 5
was not obtained. The existence of concrete blocks functioning as de-
tached breakwaters (Fig. 6), installed in this area to prevent coastal
erosion, is believed to be the main reason.
The Nanakita River mouth was completely closed in September
2011 as shown in Fig. 3, subsequently, a new river mouth opened at
Gamo Lagoon by the flooding flow from the Nanakita River. The long-
term recovery process of morphology at the Nanakita River mouth was
described by Tanaka et al. (2014a).
3.2. Erosion propagation of the beaches adjacent to Akaiko breaching
On the occasion of the 2011 Great East Japan Earthquake Tsunami,
breaching of sandy beach was observed at many places. At the Akaiko
area, breaching formed at the area where used to be a river mouth. The
mechanism of this breaching and others was presented in Tanaka et al.
(2014b). Due to the existence of breaching, after the tsunami, Teizan
Canal in the Akaiko area was connected to the ocean, and the shoreline
was in concave form. The width of concave portion (breaching) was
about 120 m.
Fig. 7 shows the changes in morphology at Akaiko breaching in-
duced by the tsunami and subsequent recovery. The concave shoreline
Fig. 6. Concrete blocks on the left side of the concave portion at the Nanakita River mouth (background image: April 6, 2011, Google Earth).
4
Journal of Hydro-environment Research 23 (2019) 1–9
at Akaiko breaching was closed and recovered much faster than the one
at the Nanakita River mouth. The reason is that the concave portion
width is much smaller and the total length of adjacent sandy beaches is
longer compared to the case of concave shoreline at the Nanakita River
mouth. More details on the relationship between the recovery process
and the total length of adjacent sandy beaches for concave shorelines at
the Nanakita River mouth and Akaiko breaching can be found in Hoang
The temporal variation of shoreline at the Akaiko breaching is ex-
panded and shown in Fig. 7. According to this figure, with the progress
of the recovery process, the beaches on both sides of the breaching were
eroded. The erosion extended along beaches on both sides. This erosion
propagation is similar to the erosion propagation at the Nanakita River
mouth shown in Fig. 4.
The similar process, which was applied for the case of erosion
propagation at the Nanakita River mouth, is also applied for the Akaiko
breaching. The relationship between the erosion propagation distances,
xe, and the elapsed time, t, for the beach on the right side of the
breaching can also be obtained based on the data in Fig. 7. According to
the results shown in Fig. 8, the erosion propagation distance is pro-
portional to square root of elapsed time. This result confirms the result
shown in Fig. 5.
Although the similar analysis was done for the beach on the left side
of the breaching, a relationship between the erosion propagation dis-
tances, xe, and the elapsed time, t, which observed in Fig. 7 for the right
side, was not obtained. The complicated initial geometry of the coast
after the tsunami, V-shaped channel and the existing of an embankment
on the left side (Tappin et al., 2012), are considered as the main rea-
sons.
3.3. Investigating the erosion of sandy beach using analytical solutions of
the one-line model
As discussed in the previous parts, during the recovery process of
morphology, sediment was transported into the concave portion,
leading to the advancement of shoreline; however, shoreline positions
on adjacent sandy beaches were found to retreat. In addition, mor-
phology in study areas was highly imbalanced due to the formation of
concave shoreline induced by the tsunami; highly oblique incident
waves transported sediment from adjacent beaches to the concave
portion. Accordingly, it can be said that during the recovery process of
concave shoreline at the Nanakita River mouth and Akaiko Breaching,
longshore sediment is predominant. Therefore, the evolution of shore-
line can be described by the theory of the one-line model. One-line
shoreline change model, which beach profile is considered to move
parallel to itself out to a limiting depth of closure, was introduced by
Pelnard-Considere (1956). The governing equation of the one-line
model is given as Eq. (1),
V.C. Hoang et al. Journal of Hydro-environment Research 23 (2019) 1–9

Fig. 7. Temporal variation of shoreline position around the Akaiko breaching after the tsunami (background image: May 3, 2011, Google Earth).

1000 of water; n is sediment porosity. It is noted that by utilizing the one-line

model, the following assumptions need to be made.

i. The beach profile is in equilibrium and moves parallel to it. This

100 assumption is not valid after a strong storm; however, it is rea-
sonable for long-term consideration.
1
ii. Longshore sediment movement takes place uniformly over the
2 beach profile down to the depth of closure.
xe (m)

10 iii. No cross-shore sediment transport is considered.

iv. The nearshore circulation is neglected (except the longshore due to
wave diffraction from coastal structures).
v. The longshore sand transport rate is proportional to the angle of
1 incidence of breaking wave crests to the shoreline.
meas.
vi. The angle between the breaking wave crests and the shoreline is
Eq. (14)
small.
vii. The angle of the shoreline with respect to the x-axis is small.
0.1
0.1 1 10 100 1000
t (days) By solving Eq. (2) with corresponding initial and boundary condi-
tions, several analytical solutions can be obtained such as those by
Fig. 8. Erosion propagation process around Akaiko breaching.

∂y 1 ∂Q
=−
∂t (DB + DC ) ∂x (1)

where Q is the longshore sediment transport rate; DB is the berm height;

DC is the depth of closure; x is the coordinate along the x-axis which is
parallel to the trend line of the beach; y is the measure of change in
shoreline position on the cross-shore direction; t is the time.
The simplified governing equation of the one-line model, Eq. (2),
can be obtained based on the assumptions that the angle of wave
breaking crests to the local shoreline is supposed to be small, and
breaking wave characteristics are assumed to be constant along the
coast (independent of x and t). More details on how to derive the
simplified governing equation of one-line model, Eq. (2), from Eq. (1)
can be found in Larson et al. (1987),

∂y ∂ 2y
=ε 2
∂t ∂x (2)

where ε is the diffusion coefficient and is given by the following

equation,

K (H 2Cg )b ρ 1 1
ε=
8 ρs − ρ (1 − n) (DB + DC ) (3)

where K is the dimensionless empirical coefficient in the longshore

sediment transport rate formula; H is the wave height; Cg is the wave
group celerity; the subscript “b” denotes the quantity at the breaking Fig. 9. Initial conditions of morphology. (a) Semi-infinite rectangular beach cut
line; ρs is the mass density of the sediment grains; ρ is the mass density – Case (1). (b) Rectangular cut in an infinite beach – Case (2).

5
V.C. Hoang et al. Journal of Hydro-environment Research 23 (2019) 1–9

Bakker and Edelman (1965), Bakker (1969), Grijm (1961, 1965), rectangular beach fill (convex shoreline) in an infinite beach. Moreover,
Walton and Chiu (1979) and Larson et al. (1987). that modified equation also can be used to evaluate the sediment ex-
With the initial conditions as shown in Fig. 9(a) and (b) (Cases (1) pansion rate after a rectangular planform is created for nourishing the
and (2)), analytical solutions of the one-line model are given as Eqs. (4) beach (Dean, 2003).
and (5), respectively by Larson et al. (1987), Hereafter, the erosion propagation distance which is shown in
Figs. 4 and 7, is investigated based on the Eqs. (4) and (5). According to
1 ⎡ x ⎞⎤
y (x , t ) = Y0 1 + erf ⎛ the theoretical characteristics of Eqs. (4) and (5), shoreline position is
2 ⎢⎣ ⎝ εt ⎠⎥
2 ⎦ (4) described by the error function and error function complementary;
hence the erosion propagation distance, xe, which is the length of an
1 ⎡ B − 2x ⎞ B + 2x ⎞ ⎤
y (x , t ) = Y0 erfc ⎛ + erfc ⎛ erosion propagation from the beach cut region to the tip of erosion area
2 ⎢⎣ ⎝ 4 εt ⎠ ⎝ 4 εt ⎠ ⎥ ⎦ (5) on the far side, is infinite. In order to obtain this value, it is redefined as
where Y0 is the cross-shore distance of beach cut from the initial the distance from beach cut region to the point where shoreline be-
shoreline. This distance is estimated based on the actual condition of comes 99 % of Y0. This rectification is analogous with the definition of
shoreline after the tsunami; erfc is the complementary error function; B boundary layer thickness in the boundary layer flow which has similar
is the width of rectangular beach cut. characteristics (e.g., Sana and Tanaka, 2007).
Eqs. (4) and (5) can be expressed in term of dimensionless para- According to Eq. (4) and the definition of erosion propagation dis-
meters as following, tance, xe, the following equation is obtained,

1⎡ x ∗ ⎞⎤ 1⎡ x ⎞⎤
y∗ (x ∗, t ∗) = 1 + erf ⎛ 0.99 = 1 + erf⎛
⎢
⎜ ⎟
⎢
2⎣ ⎥
⎝ 2 εt ⎠⎦ (12)
⎝ 2 t ⎠⎥
2⎣ ∗
⎦ (6)
or
1⎡ B∗ − 2x ∗ ⎞ B∗ − 2x ∗ ⎞ ⎤
y∗ (x ∗, t ∗) = erfc ⎛ ⎜ + erfc ⎛
⎟ ⎜ ⎟
⎢
2⎣ ⎝ 4 t ⎠
∗
⎝ 4 t ⎠⎥
∗
(7) x ⎞
⎦ erf⎛ = 0.98
⎝ 2 εt ⎠ (13)
where the dimensionless parameters are defined as bellow,
x by using the error function table for Eq. (13), the following equation
x∗ = can be obtained,
Y0 (8)
y x e = 3.3 εt (14)
y∗ =
Y0 (9)
Eq. (14) reveals the relationship between erosion propagation dis-
εt tance and elapsed time. The erosion propagation distance is propor-
t∗ = 2
Y0 (10) tional to the square root of elapsed time.
According to Fig. 10, the erosion propagation process is the same
B with the deposition propagation process in the beach cut region. This is
B∗ =
Y0 (11) valid for all time for Case (1), while it is valid only in the early stage of
Case (2). Hence, if similar definition and procedures, which have been
Fig. 10 presents the evolution of shoreline positions from analytical
used to determine the relationship between erosion propagation dis-
solutions of one-line model for cases of semi-infinite rectangular beach
tance and elapsed time, are applied then the relationship between the
cut (Case (1)) and rectangular cut in an infinite beach (Case (2)). The
deposition propagation distance, xd, and elapsed time, t, also can be
evolution of shoreline positions of these two cases above is represented
obtained.
by dash lines with circle markers and solid lines, respectively. An ar-
The time, in which the deposition propagation distance excesses B/
bitrary value of B* of 2 is selected for the computation of shoreline
2, is considered as the critical elapsed time point. Within the critical
positions evolution. According to Fig. 10, in the early stage, the evo-
elapsed time point, TCR, the evolution of shoreline position on the left
lution of shoreline positions on the left and the right sides of Case (2) is
and the right-hand sides of beach cut region has no influence on each
independent, and hence the shoreline positions obtained from two
other, and analytical solution for the case of a rectangular, Case (2), cut
analytical solutions for two cases are the same. With the passage of time
in an infinite beach can be used for the other case, Case (1).
the evolution of shoreline positions on both sides has interacted on each
Alternatively, above relationship between erosion propagation dis-
other, the positions of shorelines obtained from two analytical solutions
tance and elapsed time can be expressed in term of dimensionless
are in different trend that of Case (1) tends to advance faster in the
parameters as following,
beach cut region and retreat earlier on the adjacent beach compared to
that of Case (2). x *e=3.3 t* (15)
It is also noted that if the complementary error function, erfc, in Eq.
(7) is replaced by the error function, erf, it can describe the evolution of where the dimensionless parameter is defined as below,

t*=0 Fig. 10. Evolution of shoreline positions from ana-

t*=0.01 t*=0.05 t*=0.1 t*=0.2 lytical solutions of one-line model for cases of semi-
1 t*=0.4
infinite rectangular beach cut (Case (1)) and rectan-
t*=0.6 gular cut in an infinite beach (Case (2)). (It is noted
that in order to show the evolution of shoreline po-
t*=1
0
y/Y

0.5 t*= t/ Y2 sitions for the Case (1) corresponded to the origin as
0 in this figure, the term x* in Eq. (6) is replaced by the
B/ Y =2 t*=1 term x* − B*/2).
0

0
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
x/Y
0
*
t*=0.6 t*=0.4 t*=0.2 t*=0.1 t*=0.05 t =0.01

6
V.C. Hoang et al. Journal of Hydro-environment Research 23 (2019) 1–9

Fig. 11. Analytical solutions on the propagation of erosion. (a) The relationship between t* and x *e . (b) The relationship between t* and x *e for cases of the Nanakita
River mouth and Akaiko breaching and measured data.

xe propagation distance obtained from Eq. (17) is asymptotic to the one

x *e=
Y0 (16) obtained from Eq. (15). Moreover, if the dimensionless elapsed time is
Similar to the case of semi-infinite rectangular beach cut, if the small, the dimensionless erosion propagation distance obtained from
analytical solution of the one-line model, Eq. (5), which describes the Eqs. (15) and (17) are in good agreement. Hence, the measured data of
evolution of shoreline positions for the case of a rectangular cut in an erosion propagation distance, which is shown in Figs. 5 and 8, indicates
infinite beach, is used, the dimensionless erosion propagation distance high agreement with the theoretical erosion propagation distance, Eq.
on the beaches on both sides of the beach cut region will be given by the (15), which is obtained from the analytical solution of the one-line
following equation, model. Moreover, as discussed above when B* is large, the dimension-
less erosion propagation distances obtained from analytical solutions of
1⎡ 2x * 2B*+2x *e ⎞ ⎤ two cases are proportional to the square root of dimensionless elapsed
⎢erfc ⎛⎜- e ⎞⎟+erfc ⎛⎜ ⎟ =0.99
⎥ time. It is common to observed similar processes of physical phenom-
2 ⎝ 4 t ⎠
* *
⎝ 4 t ⎠⎦ (17)
⎣
enon which is described by diffusion equations, such as heat conduction
The relationship between the dimensionless erosion propagation phenomenon (e.g., Carslaw and Jaeger, 1959), viscous fluid (e.g.,
distance and dimensionless elapsed time obtained from Eq. (17) cor- Schlichting and Gersten, 2000), and substance diffusion phenomenon
responding to several values of B* is illustrated in Fig. 11(a). According (e.g., Crank, 1975).
to these results, when B* is small, the recovery process is completed in a The dimensionless erosion propagation time for two study areas are
shorter time, thus x *e is also shorter than others. When B* is increasing estimated based on Eq. (15) and measured data of Y0 and xe. Values of
(e.g., B* = 5, 10, 20 in Fig. 11(a)), the dimensionless erosion

7
V.C. Hoang et al.
Y0 are 95 m and 270 m, respectively for the two cases. The measure-
ment of Y0 is based on the actual condition of morphology right after
the tsunami. Although there are some difficulties in estimating the
value of cross-shore distance beach cut, Y0, in the case of Akaiko
breaching, it was clearer to obtain it rather than the other case. Some
sandy islands still remained in the area of the Nanakita River mouth
after the tsunami, thus Y0 is taken from the location of tips of those
islands on seaside to the line which is parallel and corresponded with
shoreline position on the right side of the river mouth. It is noted that
the applicability of the one-line model theory is not valid in the beach
cut region. And, the value of Y0 does not affect the estimation diffusion
coefficient values, I µ̂ , of two study areas which are estimated based on
the measured xe, elapsed time t and Eq. (14) (presenting in Section 3.4).
Values of xe mentioned above are estimated from edge between the
concave portion and sandy beach to the rigid boundary on the right side
of those two cases; In the case of the Nanakita River mouth, there are
six detached breakwaters on the right side, whereas there is the jetty for
the case of Akaiko breaching. Those structures can be seen in (Fig. 1).
Based on Eq. (11) and the values of B and Y0 presented in Sections
3.1–3.3, the B* values of two study areas, concave shorelines at the
Nanakita River mouth and at Akaiko breaching, are obtained as 11 and
0.44, respectively. Results obtained above and measured data of erosion
propagation of two study areas are plotted together in Fig. 11(b).
As the recovery process of concave shoreline at Akaiko breaching is
shorter than the case of concave shoreline at the Nanakita River mouth,
hence its measured data of erosion propagation is also shorter.
According to the results shown in Fig. 11(b), both measured data sets
are in good agreement with the theoretical relationship t* ∼ x *e obtained
from the analytical solution of the one-line model for the case of a
rectangular cut in an infinite beach, especially in the case when t* is still
small. When t* becomes large, the agreement between measured data
and theoretical result for the case of concave shoreline at the Nanakita
River mouth becomes worse. Measured xe* tend to be smaller than
theoretical xe*.
3.4. Estimation of the diffusion coefficient of littoral drifts
By substituting the measured data of erosion propagation distance,
xe, and elapsed time, t, which is shown in Figs. 5 and 8, into Eq. (14),
the average values of diffusion coefficient, ε, for two study areas are
obtained. These values for the case of the Nanakita River mouth and
Akaiko breaching areas are 13 m2/h and 14 m2/h, respectively. Ac-
cording to Pradjoko and Tanaka, 2010, the uncertainty of a shoreline
position when using the method of image analysis, which is the same in
this study, is about 6 m. It means the uncertainty in the estimation of
propagation distance is such amount. Eq. (14) indicates that the diffu-
sion coefficient is proportional to erosion propagation distance square.
Therefore, the uncertainty in estimating of diffusion coefficient value as
above is about 1.0 m2/h.
The values of diffusion coefficient found in these study areas are
consistent with the values of this parameter found in the previous
studies. Mano et al. (1996) carried out the value of dimensionless em-
pirical proportionality coefficient in longshore sediment transport rate
formula, K = 0.03, for the area around Yuriage Port using the 6-year
long observed shoreline position data and the characteristics of mean
wave. Kang and Tanaka (2006) calibrated the value of K (K = 0.03) for
Sendai Coast from Sendai Port to Yuriage Port based on yearly raw
measured shoreline position data. Average wave conditions were also
used in this study (H = 1.62 m; Cg = 3.64 m/s). Based on Eq. (3) with
ρs − ρ = 1.65 T/m3; n = 0.4 (Mano et al. 1996); DB = 3 m (Tinh and
Tanaka, 2011); DC = 8 m (Uda, 1997), the values of the diffusion
coefficient of these two studies are 12 m2/h. This value is consistent
with values carried out in the current study. In addition, Dean (2003)
documented the values of diffusion coefficient for the beaches in the
State of Florida, the U.S. The value of diffusion coefficient varies from
10 m2/h to 40 m2/h for the beach facing the Gulf of Mexico, whereas it
8
Journal of Hydro-environment Research 23 (2019) 1–9
is in wider range for the beach facing the Atlantic Ocean (6.7–50 m2/h).
Although the characteristics of waves and sediment for the two coasts
would be different, values of diffusion coefficient for Sendai Coast
carried out in this study are in the range of values of the diffusion
coefficient of Florida Coast.
4. Conclusions and recommendation
This paper investigates the erosion of sandy beaches adjacent to
river mouth with flushing of sand spit and breaching of sandy beach at
the area of former river mouth induced by the 2011 Great East Japan
Earthquake Tsunami. The following conclusions have been made.
(1) After the severe damage to river mouth morphology and forming of
breaching of sandy beach induced by the tsunami, the erosion
propagation occurred on the adjacent sandy beaches. Erosion pro-
pagation distance is proportional to the square root of elapsed time.
This result is in agreement with the common physical phenomenon
described by the diffusion equation. This appears to be a valuable
finding of the current study.
(2) The value of diffusion coefficient of the beach at the Nanakita River
mouth and Akaiko breaching areas are obtained based on the
analytical solutions of the one-line model and the relationship be-
tween the erosion propagation distance and elapsed time. This
study supplements a new technique to estimate the diffusion coef-
ficient which so far is usually obtained through the empirical
coefficient in longshore sediment transport rate formula, K.
The high waves during typhoon combining with the big floods could
flush the sand spit, even the sand barrier in front of the lagoon and
scour the river mouth area deeper than the depth of closure. Such event
can form the morphology in concave shape. The erosion of the adjacent
sandy beach and its propagation could happen similar to the cases in
this study. Therefore, it would be interesting if such morphology
changes can be evaluated using the same approach proposed in this
study. This is a recommendation for future study.
Acknowledgment
The authors would like to thank the anonymous reviewers who have
helped greatly in the revision of the originally unacceptable manu-
script.
References
Ali, P.Y., Narayana, A.C., 2015. Short-term morphological and shoreline changes at
Trinkat island, Andaman and Nicobar, India, after the 2004 tsunami. Mar. Geod. 38
(1), 26–39.
Bakker, W.T., 1969. The dynamics op a coast with a groyne system. In: Proceedings of the
11th Coastal Engineering Conference, pp. 492–517.
Bakker, W.T., Edelman, T., 1965. The coastline of river deltas. In: Proceeding of the 9th
Coastal Engineering Conference, pp. 199–218.
Carslaw, H., Jaeger, J., 1959. In: Conduction of Heat in Solids, second ed. Clarendon
Press, pp. 510.
Choowong, M., Phantuwongraj, S., Charoentitirat, T., Chutakositkanon, V., Yumuang, S.,
Charusiri, P., 2009. Beach recovery after 2004 Indian Ocean tsunami from Phang-
nga, Thailand. J. Geomorphology 104, 134–142.
Crank, J., 1975. In: The Mathematics of Diffusion, second ed. Clarendon Press, pp. 414.
Dean, R.G., 2003. In: Beach Nourishment – Theory and Practice. World Scientific, pp.
420.
FitzGerald, D.M., 1988. Shoreline erosional-depositional processes associated with tidal
inlets. Hydrodyn. Sediment Dyn. Tidal Inlets 29, 186–225.
Grijm, W., 1961. Theoretical forms of shoreline. In: Proceeding of the 7th Coastal
Engineering Conference, pp. 197–202.
Grijm, W., 1965. Theoretical forms of shoreline. In: Proceeding of the 9th Coastal
Engineering Conference, pp. 219–235.
Hoang, V.C., Tanaka, H., Mitobe, Y., 2015. Theoretical study on the recovery process of
the concave landform after the tsunami. J. JSCE, Ser. B1 (Hydraulic Eng.) 71 (4),
I_31–I_36.
Hoang, V.C., Tanaka, H., Mitobe, Y., 2018. Estuarine morphology recovery after the 2011
Great East Japan earthquake tsunami. Mar. Geol. 398, 112–125.
Kang, H., Tanaka, H., 2006. Influence of cross-shore sediment movement on long-term
V.C. Hoang et al.
shoreline change simulation. In: Parker, G., Garcia, M.H. (Eds.), River, Coastal and
Estuarine Morphodynamics. Taylor & Francis, pp. 343–348.
Kraus, N., Wamsley, T.V., 2003. Coastal barrier breaching, Part 1: Overview of breaching
processes. ERDC/CHL CHETN-IV-56.
Larson, M., Hanson, H., Kraus, N.C., 1987. Analytical solutions of the one-line model of
shoreline change. Technical Report CERC-87-15. U.S. Army Engineer Waterways
Experiment Station.
Liew, Gupta, A., Wong, P.P., Kwoh, L.K., 2010. Recovery from a large tsunami mapped
over time: the Aceh coast, Sumatra. J. Geomorphol. 114 (4), 520–529.
Mano, A., Muhammad, H., Sawamoto, M., 1996. Annual mean waves for the littoral drift
formula. Proc. Civ. Eng. Ocean 12, 55–59 (in Japanese).
Pradjoko, E., Tanaka, H., 2010. Aerial photograph of Sendai Coast for shoreline behavior
analysis. In: Proceedings of the 32nd International Conference on Coastal
Engineering.
Pelnard-Considere, R., 1956. Essi de theorie de l’evolution des formes de ravage en plages
de sables et de galets. Societe Hydrotechnique de France, IV’eme Journee de
L’Hydraulique Question III,” Rapport 1, 74-1-10.
Sana, A., Tanaka, H., 2007. Full-range equation for wave boundary layer thickness. Coast.
Eng. 54, 639–642.
Schlichting, H., Gersten, K., 2000. Translated by Mayes, C. In: Boundary-Layer Theory.
Springer, pp. 801.
Journal of Hydro-environment Research 23 (2019) 1–9
Tanaka, Tinh, N.X., Umeda, M., Hirao, R., Pradjoko, E., Mano, A., Udo, K., 2012. Coastal
and estuarine morphology changes induced by the 2011 Great East Japan earthquake
tsunami. Coastal Eng. J. 54 (1), 1250010 (25 pages).
Tanaka, H., Adityawan, M.B., Mano, A., 2014a. Morphological changes at the Nanakita
River mouth after the Great East Japan Tsunami of 2011. Coast. Eng. 86, 14–26.
Tanaka, H, Adityawan, M.B., Udo, K., Mano, A., 2014b. Breaching and tsunami water
drainage at old river mouth locations during the 2011 tsunami. In: Proceedings of the
34th International Conference on Coastal Engineering.
Tappin, Evans, H.M., Jordan, C.J., Richmond, B., Sugawara, D., Goto, K., 2012. Coastal
changes in the Sendai area from the impact of the 2011 Tōhoku-oki tsunami:
Interpretations of time series satellite images, helicopter-borne video footage and
field observations. Sed. Geol. 282 (30), 151–174.
Tinh, N.X., Tanaka, H., 2011. Discussion of overwash prevention construction on the
northern part of Sendai Coast. In: Proceedings of 6th International Conference on
Coastal Structures, Yokohama, CD-ROM.
Uda, T., 1997. In: Beach Erosion in Japan. Sankaido Press, pp. 442 (in Japanese).
Udo, Sugawara, D., Tanaka, H., Imai, K., Mano, A., 2012. Impact of the 2011 Tohoku
Earthquake and Tsunami on beach morphology along the northern Sendai Coast.
Coastal Eng. J. 54 (01), 1250009 (15 pages).
Walton, T.L., Chiu, T.Y., 1979. A review of analytical techniques to solve the sand
transport equation and some simplified solutions. Proc. Coastal Struct. 79, 809–837.