ArticlePDF Available

The Responses of Storm Surges to Representative Typhoons under Wave–Current Interaction in the Yangtze River Estuary

January 2024
Journal of Marine Science and Engineering 12(1):90

January 2024
12(1):90

DOI:10.3390/jmse12010090

License
CC BY 4.0

Authors:

Jie Wang

Tongji University

Cuiping Kuang

Tongji University

Daidu Fan

Tongji University

Show all 6 authorsHide

Storm surge is one of the most remarkable natural calamities, which is shown as the abnormal sea level changes in the coastal waters during a typhoon event. To investigate the responses of storm surges to the typhoon paths, intensities and coastal dynamics, a coupled wave–current model is used to study the impacts of strong winds, considerable waves and complex currents on storm surges in the Yangtze River Estuary (YRE) during three representative typhoons of Fongwong (2014), Ampil (2018) and Lekima (2019) with different intensities and paths. The model is verified using the measured data on significant wave height and period, water level and current velocity and performs well in modeling real conditions. The numerical results demonstrate that (1) the maximum storm surge occurred in the South Channel (SC) during Fongwong and Lekima while in the North Branch (NB) during Ampil due to the typhoon path and the estuarine terrain. Among the three typhoons, Lekima presented the highest surge, with a maximum value of 1.17 m at SC2 (the inner point of the SC). There was a negative surge during Ampil, which reached −0.42 m at SC2, due to the representative path (SE to NW) and offshore wind action. (2) Tide is the main influencing factor of storm surge as the maximum or minimum value always occurs at the low or high tidal level, respectively. Meanwhile, typhoon intensity is important as it influences the variation rate of surge with higher intensity leading to a sudden increase in surge while the tidal intensity primarily affects the peak value. (3) The wave setup can counteract the wind-induced negative surge. The peak differences between storm surge isoline and wave setup isoline are 0.15, 0.2 and 0.2 m during Fongwong, Ampil and Lekima, respectively, which illustrates the impacts of the combined actions of the typhoon path and intensity on the wave setup. This research emphasizes the influences of wave–current interaction on estuarine storm surge during typhoon events and reveals the potential risks for oceanic disasters like coastal inundation.

Available via license: CC BY 4.0

Content may be subject to copyright.

Citation: Wang, J.; Kuang, C.; Cheng,

S.; Fan, D.; Chen, K.; Chen, J. The

Responses of Storm Surges to

Representative Typhoons under

Wave–Current Interaction in the

Yangtze River Estuary. J. Mar. Sci. Eng.

2024,12, 90. https://doi.org/

10.3390/jmse12010090

Academic Editors: Jinyu Sheng and

Harshinie Karunarathna

Received: 31 August 2023

Revised: 23 December 2023

Accepted: 29 December 2023

Published: 1 January 2024

Licensee MDPI, Basel, Switzerland.

This article is an open access article

distributed under the terms and

conditions of the Creative Commons

Attribution (CC BY) license (https://

creativecommons.org/licenses/by/

4.0/).

Journal of

Marine Science

and Engineering

Article

The Responses of Storm Surges to Representative Typhoons

under Wave–Current Interaction in the Yangtze River Estuary

Jie Wang 1, Cuiping Kuang 1,* , Subin Cheng 1, *, Daidu Fan 2, * , Kuo Chen 1,3 and Jilong Chen 1

1College of Civil Engineering, Tongji University, Shanghai 200092, China; 2011141@tongji.edu.cn (J.W.);

chenkuo@ecs.mnr.gov.cn (K.C.); 2232335@tongji.edu.cn (J.C.)

2State Key Laboratory of Marine Geology, Tongji University, Shanghai 200092, China

3Information Center of East China Sea, East China Sea Bureau of Ministry of Natural Resources,

Shanghai 200136, China

*Correspondence: cpkuang@tongji.edu.cn (C.K.); 02107@tongji.edu.cn (S.C.); ddfan@tongji.edu.cn (D.F.)

Abstract: Storm surge is one of the most remarkable natural calamities, which is shown as the

abnormal sea level changes in the coastal waters during a typhoon event. To investigate the responses

of storm surges to the typhoon paths, intensities and coastal dynamics, a coupled wave–current

model is used to study the impacts of strong winds, considerable waves and complex currents on

storm surges in the Yangtze River Estuary (YRE) during three representative typhoons of Fongwong

(2014), Ampil (2018) and Lekima (2019) with different intensities and paths. The model is veriﬁed

using the measured data on signiﬁcant wave height and period, water level and current velocity and

performs well in modeling real conditions. The numerical results demonstrate that (1) the maximum

storm surge occurred in the South Channel (SC) during Fongwong and Lekima while in the North

Branch (NB) during Ampil due to the typhoon path and the estuarine terrain. Among the three

typhoons, Lekima presented the highest surge, with a maximum value of 1.17 m at SC2 (the inner

point of the SC). There was a negative surge during Ampil, which reached

−

0.42 m at SC2, due

to the representative path (SE to NW) and offshore wind action. (2) Tide is the main inﬂuencing

factor of storm surge as the maximum or minimum value always occurs at the low or high tidal

level, respectively. Meanwhile, typhoon intensity is important as it inﬂuences the variation rate of

surge with higher intensity leading to a sudden increase in surge while the tidal intensity primarily

affects the peak value. (3) The wave setup can counteract the wind-induced negative surge. The

peak differences between storm surge isoline and wave setup isoline are 0.15, 0.2 and 0.2 m during

Fongwong, Ampil and Lekima, respectively, which illustrates the impacts of the combined actions

of the typhoon path and intensity on the wave setup. This research emphasizes the inﬂuences of

wave–current interaction on estuarine storm surge during typhoon events and reveals the potential

risks for oceanic disasters like coastal inundation.

Keywords: wave–current interaction (WCI); storm surge; hydrodynamic; typhoon; Yangtze River

Estuary (YRE)

1. Introduction

Estuaries are important for local dynamic interactions among land, river and ocean,

where hydrodynamic conditions are sensitive to strong winds, waves and currents. As

the most frequent disaster in coastal areas, typhoons act shortly and strongly with high

wind speeds and may alter the current structure and salinity distribution [

–

]. During the

occurrence of typhoon events, winds, waves and currents play crucial roles as the primary

environmental factors within estuarine systems. Strong winds can generate considerable

waves and complex current structures, which then aggravate local destratiﬁcation, estuarine

circulation and matter exchange [4–9].

Storm surge is a remarkable meteorological event caused by typhoon events. According

to the sixth appraisal report of the IPCC (The Intergovernmental Panel on Climate Change),

J. Mar. Sci. Eng. 2024,12, 90. https://doi.org/10.3390/jmse12010090 https://www.mdpi.com/journal/jmse

J. Mar. Sci. Eng. 2024,12, 90 2 of 24

global warming and sea level rising raise the risk of typhoons in terms of frequency and

intensity, as well as coastal inundation. A large number of studies have investigated the

mechanism underlying water level variation and nonlinear interaction in estuaries. For

instance, Guo and Lin [

] analyzed historical typhoon data over

20 years

and obtained the

concrete temporal and spatial distribution characteristics of storm surges. Yu [

] categorized

storm surges into four typical forms, including normalized form, unnormalized form,

undulated form and symmetric similarity form. The mechanism of storm surge is relatively

complex owing to its multiple factors. Musinguzi and Akbar [

] conducted an investigation

on surge responses to wind speed, atmospheric pressure, and storm passage and found

that wind speed is the predominant factor. Wang et al. [

] focused on storm surges in

the Pearl River mouth and observed that typhoons may cause contrary effects inside and

outside the bay, which is related to their approaching speed. Thuy et al. [

] discovered a

consistent trend of increasing water levels in the Beibu Gulf with the strengthened intensity

of typhoons. Wang and Sheng [

] simulated three typhoon events in Canada and examined

the mechanism of waves and currents during different types of storm surges. Yang et al. [

]

concluded that bottom roughness and advection are prominent factors affecting the tide–

storm surge interaction by analyzing nonlinear elements.

Numerical modeling has been proven to be an efﬁcient approach in investigations

of storm surges since the 1970s [

–

]. Among various dynamic factors, waves affect

the coastal circulation and coordinate the current structure. Hence, the wave–current

interaction (WCI) exhibits pronounced intensity during storms. The inﬂuence of WCI leads

to a notable enhancement in the interference coefﬁcient and, consequently, redistributes

the local ﬂow ﬁeld [

–

]. Dietrich et al. [

] were the ﬁrst to apply ADCIRC coupled

with the SWAN model to simulate storm surge. This model garnered signiﬁcant attention

and, subsequently, found applications in storm surge studies in the United States, Korea,

and China’s Zhujiang Estuary and Hangzhou Bay (HB). Mao and Xia [

] utilized the

FVCOM coupled with the SWAN model to study the storm surge in a single-inlet lagoon

system and further applied the model to oceanic systems with multiple inlets.

Estuaries in China are faced with potential hazards of tropical cyclones [

]. The

influences of typhoons present two aspects. One involves meteorological calamities of strong

wind and torrential rainfall induced by typhoon passage, and the other involves oceanic

forcings of considerable waves and coastal floods [

]. The Yangtze River Estuary (YRE)

is located between Zhejiang and Jiangsu provinces, straightly facing the East Sea (shown in

Figure 1), which is considerably susceptible to frequent typhoon impacts. The main channel

lengthof the YRE spans 232 km from Jiangyin to the river mouth and consists of two prominent

branches known as the North Branch (NB) and the South Branch (SB). As it flows downstream,

the SB is divided by the Changxing and Hengsha Islands into the North Channel (NC) and

the South Channel (SC). According to the data from the China Meteorological Administration,

typhoons generated from 5

◦

N~22

◦

N on the Northwest Pacific Ocean are more likely to incur

coastal disasters in China, especially in summer and autumn [

–

]. During most typhoon

events, the prevailing wind typically originates SW and possesses a subdued intensity below

5 m/s when in close proximity to the YRE [

]. Nonetheless, upon making landfall, the wind

velocity may increase significantly to over 20 m/s [40].

This study aims to investigate the responses of storm surges to the three representative

typhoons of Fongwong (2014), Ampil (2018), and Lekima (2019) with different intensities

and paths under WCI conditions in the YRE. Figure 1contains the major bathymetric

features of the study and the paths of three typhoons. A two-dimensional hydrodynamic

model is established by coupling with a wave model. The main focus of this study primarily

examines the processes and distributions of storm surge in the YRE under WCI conditions.

In addition, the nonlinear relationships between storm surge and tide, wind and wave

are discussed by distinguishing the respective effects. The conclusions of this study have

practical implications for comprehending the potential hazards of storm surges in coastal

estuaries, such as the YRE. Furthermore, this study provides a valuable reference for

regional risk management strategies.

J. Mar. Sci. Eng. 2024,12, 90 3 of 24

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 3 of 25

and wave are discussed by distinguishing the respective eﬀects. The conclusions of this

study have practical implications for comprehending the potential hazards of storm

surges in coastal estuaries, such as the YRE. Furthermore, this study provides a valuable

reference for regional risk management strategies.

Figure 1. (a) The study area and the observation stations (the pink triangles represent the water level

stations; the black circles are the current velocity stations, and the blue cross shows the position of

the wave station); (b) the analyzed area of this study with the two sections and six study points in

the three branches; (c) the typhoon path and moving direction.

2. The Three Typical Typhoon Processes

A tropical cyclone is deﬁned as a low-pressure eddy generated above the sea surface

of the tropical ocean, such as the Northwest Paciﬁc Ocean. The value of wind center speed

is used to divide the speciﬁc intensity grades of cyclones, which is as formulated by the

China Meteorological Administration, as shown in Table 1. A typhoon is a kind of tropical

cyclone once the maximum average center wind speed is above 32.7 m/s.

Table 1. Cyclone grades deﬁned by China Meteorological Administration.

Maximum Center Speed (m/s) Grade Definition

10.8~17.1 6~7 Tropical Depression

17.2~24.4 8~9 Tropical Storm

24.5~32.6 10~11 Strong Tropical Storm

32.7~41.4 12~13 Typhoon

41.5~50.9 14~15 Strong Typhoon

≥51.0 ≥16 Super Typhoon

Figure 2 presents the variations of water level, wind velocity and signiﬁcant wave

height and period during Fongwong. The data on water level and wind velocity were

121 122 123 124 125

Longitude (E)

e (N)

120.5 121 121.5 122 122.5

Longitude (E)

30.8

31.3

31.8

Latitude (N)

(a)

TD1

TD1:Sheshan Station

TD2:Dajishan Station

TD3:Luchaogang Station

a.NGN4SD

b.CS3SD

c.NC6D

Jiangsu

Shanghai

Hangzhou Bay

-70-60-50-40-30-20-10010

Bathymetry (m)

TD2

TD3

Chongming Island

NB1 NB2

NC1

NC2

SC1

SC2

(b)

Yangkougang Station

Changxing Island

Jiangyin

Study point

121 122 123 124 125

Longitude (E)

Latitude (N)

1416# Tyhoon Fongwong

1810# Typhoon Ampil

1909# Typhoon Lekima

(c)

Figure 1. (a) The study area and the observation stations (the pink triangles represent the water level

stations; the black circles are the current velocity stations, and the blue cross shows the position of

the wave station); (b) the analyzed area of this study with the two sections and six study points in the

three branches; (c) the typhoon path and moving direction.

2. The Three Typical Typhoon Processes

A tropical cyclone is deﬁned as a low-pressure eddy generated above the sea surface

of the tropical ocean, such as the Northwest Paciﬁc Ocean. The value of wind center speed

is used to divide the speciﬁc intensity grades of cyclones, which is as formulated by the

China Meteorological Administration, as shown in Table 1. A typhoon is a kind of tropical

cyclone once the maximum average center wind speed is above 32.7 m/s.

Table 1. Cyclone grades deﬁned by China Meteorological Administration.

Maximum Center Speed (m/s) Grade Deﬁnition

10.8~17.1 6~7 Tropical Depression

17.2~24.4 8~9 Tropical Storm

24.5~32.6 10~11 Strong Tropical Storm

32.7~41.4 12~13 Typhoon

41.5~50.9 14~15 Strong Typhoon

≥51.0 ≥16 Super Typhoon

Figure 2presents the variations of water level, wind velocity and signiﬁcant wave

height and period during Fongwong. The data on water level and wind velocity were

collected from the Sheshan Station. The data on wave height and period were measured at

Yangkougang station. Fongwong was generated close to the Philippines at the beginning,

and it gradually became the intensity of a tropical storm (grade 8~9) when it landed in the

YRE. The intensity of Fongwong was relatively low, and its maximum speed was 16.8 m/s

(measured at the Sheshan station), but its variable movement directions (Figure 1) deserve

a more detailed study. The exact landing moment was at 14:00 on 23 September 2014.

J. Mar. Sci. Eng. 2024,12, 90 4 of 24

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 4 of 25

collected from the Sheshan Station. The data on wave height and period were measured

at Yangkougang station. Fongwong was generated close to the Philippines at the begin-

ning, and it gradually became the intensity of a tropical storm (grade 8~9) when it landed

in the YRE. The intensity of Fongwong was relatively low, and its maximum speed was

16.8 m/s (measured at the Sheshan station), but its variable movement directions (Figure

1) deserve a more detailed study. The exact landing moment was at 14:00 on 23 September

2014.

Figure 2. Variations in the main dynamics (wind direction and speed, water level, signiﬁcant wave

height and period) during Fongwong measured at Sheshan and Yangkougang station. The red rec-

tangle is the period when typhoon eﬀect is the strongest.

Figure 3 presents the variations in the main dynamics collected at Yangkougang sta-

tion during Ampil. Ampil was generated over the Northwest Paciﬁc Ocean, where the

typhoon center was about 1360 km away from Zhejiang Province. Ampil landed at Chong-

ming Island in Shanghai at 12:00 on 22 July 2018, and the maximum measured wind speed

near the typhoon center was about 28 m/s. After it landed, the direction of Ampil gradu-

ally turned NW with declining wind velocity. Ampil was the strongest typhoon event to

land directly in the YRE since 1990. The path of Ampil formed a cross with Fongwong

(Figure 1), which is another meaningful process.

Figure 2. Variations in the main dynamics (wind direction and speed, water level, signiﬁcant wave

height and period) during Fongwong measured at Sheshan and Yangkougang station. The red

rectangle is the period when typhoon effect is the strongest.

Figure 3presents the variations in the main dynamics collected at Yangkougang station

during Ampil. Ampil was generated over the Northwest Paciﬁc Ocean, where the typhoon

center was about 1360 km away from Zhejiang Province. Ampil landed at Chongming

Island in Shanghai at 12:00 on 22 July 2018, and the maximum measured wind speed near

the typhoon center was about 28 m/s. After it landed, the direction of Ampil gradually

turned NW with declining wind velocity. Ampil was the strongest typhoon event to land

directly in the YRE since 1990. The path of Ampil formed a cross with Fongwong (Figure 1),

which is another meaningful process.

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 5 of 25

Figure 3. Variations in the main dynamics (wind direction and speed, water level, signiﬁcant wave

height and period) during Ampil measured at the Yangkougang station in 2018. The red rectangle

is the period when typhoon eﬀect is the strongest.

Figure 4 presents the variations in the main dynamics measured at Yangkougang sta-

tion during Lekima. Lekima was generated in the Philippines and moved towards the

north. It was a super typhoon landing in Taizhou with a wind intensity over Grade 16 (the

wind speed of the typhoon center was ~52 m/s). As the strongest typhoon event in 2019,

although it merely passed the YRE, it still caused serious damage in the adjacent regions.

Meanwhile, the path of Lekima was away from the main branches of the YRE without

direction changes during the process (Figure 1), so it was selected as another representa-

tive path.

Figure 4. Variations in the main dynamics (wind direction and speed, water level, signiﬁcant wave

height and period) during Lekima measured at the Yangkougang station in 2019. The red rectangle

is the period when typhoon eﬀect is the strongest.

Figure 3. Variations in the main dynamics (wind direction and speed, water level, signiﬁcant wave

height and period) during Ampil measured at the Yangkougang station in 2018. The red rectangle is

the period when typhoon effect is the strongest.

J. Mar. Sci. Eng. 2024,12, 90 5 of 24

Figure 4presents the variations in the main dynamics measured at Yangkougang

station during Lekima. Lekima was generated in the Philippines and moved towards the

north. It was a super typhoon landing in Taizhou with a wind intensity over Grade 16

(the wind speed of the typhoon center was ~52 m/s). As the strongest typhoon event in

2019, although it merely passed the YRE, it still caused serious damage in the adjacent

regions. Meanwhile, the path of Lekima was away from the main branches of the YRE

without direction changes during the process (Figure 1), so it was selected as another

representative path.

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 5 of 25

Figure 3. Variations in the main dynamics (wind direction and speed, water level, signiﬁcant wave

height and period) during Ampil measured at the Yangkougang station in 2018. The red rectangle

is the period when typhoon eﬀect is the strongest.

Figure 4 presents the variations in the main dynamics measured at Yangkougang sta-

tion during Lekima. Lekima was generated in the Philippines and moved towards the

north. It was a super typhoon landing in Taizhou with a wind intensity over Grade 16 (the

wind speed of the typhoon center was ~52 m/s). As the strongest typhoon event in 2019,

although it merely passed the YRE, it still caused serious damage in the adjacent regions.

Meanwhile, the path of Lekima was away from the main branches of the YRE without

direction changes during the process (Figure 1), so it was selected as another representa-

tive path.

Figure 4. Variations in the main dynamics (wind direction and speed, water level, signiﬁcant wave

height and period) during Lekima measured at the Yangkougang station in 2019. The red rectangle

is the period when typhoon eﬀect is the strongest.

Figure 4. Variations in the main dynamics (wind direction and speed, water level, signiﬁcant wave

height and period) during Lekima measured at the Yangkougang station in 2019. The red rectangle is

the period when typhoon effect is the strongest.

3. Model Setup

A two-dimensional hydrodynamic model was established by coupling the wave

model in MIKE21 for simulations of storm surges [

]. The governing equations and other

parameter-setting rules can be found in the MIKE Manual [

]. The details of the present

model are discussed below.

3.1. Study Area and Mesh

The domain covers the YRE from Jiangyin to the river mouth. The study area ranges

from 26.9

◦

N to 34.4

◦

N and 120.2

◦

E to 125.6

◦

E, spanning 810 km in the N-S direction

and 500 km in the W-E direction (Figure 1). This extensive study area covers the main

center positions in typhoon paths and can reﬂect the regional water exchanges and current

structures. The unstructured triangular mesh contains 46,066 nodes and 89,710 elements

(Figure 5). The element scales decrease gradually from the east open boundary to the

channel of the YRE, where the maximum value is 33,000 m in the sea boundary and

the minimum value is 10 m in the inner channel. Reﬁnement in the core research area

can result in superior efﬁciency in simulations. Meanwhile, the position of the Deep-

Water Channel Project is reﬁned to elaborately present the dynamics variation around the

complex structure.

J. Mar. Sci. Eng. 2024,12, 90 6 of 24

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 6 of 25

3. Model Setup

A two-dimensional hydrodynamic model was established by coupling the wave

model in MIKE21 for simulations of storm surges [39]. The governing equations and other

parameter-seing rules can be found in the MIKE Manual [39]. The details of the present

model are discussed below.

3.1. Study Area and Mesh

The domain covers the YRE from Jiangyin to the river mouth. The study area ranges

from 26.9° N to 34.4° N and 120.2° E to 125.6° E, spanning 810 km in the N-S direction and

500 km in the W-E direction (Figure 1). This extensive study area covers the main center

positions in typhoon paths and can reﬂect the regional water exchanges and current struc-

tures. The unstructured triangular mesh contains 46,066 nodes and 89,710 elements (Fig-

ure 5). The element scales decrease gradually from the east open boundary to the channel

of the YRE, where the maximum value is 33,000 m in the sea boundary and the minimum

value is 10 m in the inner channel. Reﬁnement in the core research area can result in su-

perior eﬃciency in simulations. Meanwhile, the position of the Deep-Water Channel Pro-

ject is reﬁned to elaborately present the dynamics variation around the complex structure.

Figure 5. The mesh and the reﬁned area of the Deep-Water Channel Project.

3.2. Hydrodynamic Model

The 2D hydrodynamic model is mainly driven by the continuity equation and hori-

zontal momentum equations.

The continuity equation is obtained as follows:

hhuhv

tx y

∂∂ ∂

++=

∂∂ ∂ (1)

where t is the time; x and y are the coordinates; h is the total water depth which is calcu-

lated based on the surface elevation and still water depth; S is the magnitude of the dis-

charge due to point sources; and u and v are the depth-averaged velocities in the x-

direction and y-direction, respectively.

The horizontal momentum equations are obtained as follows:

Figure 5. The mesh and the reﬁned area of the Deep-Water Channel Project.

3.2. Hydrodynamic Model

The 2D hydrodynamic model is mainly driven by the continuity equation and hori-

zontal momentum equations.

The continuity equation is obtained as follows:

∂h

∂t+∂hu

∂x+∂hv

∂y=hS (1)

where tis the time; xand yare the coordinates; his the total water depth which is calculated

based on the surface elevation and still water depth; Sis the magnitude of the discharge

due to point sources; and

and

are the depth-averaged velocities in the x-direction and

y-direction, respectively.

The horizontal momentum equations are obtained as follows:

∂hu

∂t+∂hu2

∂x+∂hvu

∂y=f vh −gh ∂η

∂x−h

ρ0

∂pa

∂x−gh2

2ρ0

∂ρ

∂x

+τsx

ρ0−τbx

ρ0−1

ρ0∂sxx

∂x+∂sxy

∂y+∂

∂x(hTxx )+∂

∂yhTxy (2)

∂hv

∂t+∂huv

∂x+∂hv2

∂y=−f uh −gh ∂η

∂y−h

ρ0

∂pa

∂y−gh2

2ρ0

∂ρ

∂y

+τsy

ρ0−τby

ρ0−1

ρ0∂syx

∂x+∂syy

∂y+∂

∂xhTxy +∂

∂yhTyy(3)

where fis the Coriolis parameter;

ρ0

is the reference density of water;

is the density of

water, p

is the atmospheric pressure;

τsx

and

τsy

are the surface stress components;

τbx

and

τby are the bottom stress components; and Txx,Txy and Tyy are the lateral stresses.

The three open boundaries are driven by the time-varying tidal level data from a

harmonic model called the Global Tidal Model [

], which accounts for eight main astro-

nomic components (M

, K

, S

, N

, K

, P

, O

, and Q

) and covers the entire China Sea.

Apart from the open boundaries, river boundaries are driven by constant discharges for

simulations. In this model, Jiangyin (as the representative station of the Yangtze River) and

Cangqian (stand for Qiantangjiang) are taken as the upstream river boundaries with the

discharges of 28,484 m

/s and 1000 m

/s, respectively. The threshold depths of drying,

ﬂooding and wetting are set as 0.005, 0.05 and 0.1 m, respectively. The bathymetry of the

YRE and HB has been conducted by Kuang et al. [

] and Chen et al. [

], whose studies

were based on the measured data obtained from the high-resolution charting of the People’s

Liberation Army Navy and data obtained from remote sensing of the National Oceanic

J. Mar. Sci. Eng. 2024,12, 90 7 of 24

and Atmospheric Administration (NOAA). The bed resistance is described by the Manning

number, which is related to the sediment grain size and the water depth as follows:

M=2.5√gln(30h

kse)

h1/6 (4)

where his the water depth; gis the gravitational acceleration; and k

is the bottom roughness

height. The grain size in the YRE is 44.9~94.9

m, with the Manning number in the range

of 68~94 m1/3/s varying in the domain.

The wind stress is obtained using the following empirical relation:

τs=ρacd|uw|uw(5)

where

ρa

is the density of air; c

is the drag coefﬁcient of air; and

uw= (uw

vw)

is the

wind speed at 10 m above the sea surface. The wind data from 2014 to 2019 were down-

loaded from the ECMWF (European Centre for Medium-Range Weather Forecasts) with a

resolution of 0.25

◦×

0.25

◦

, which covers the initialization periods and simulation periods.

The shallow water equation requires the CFL (Courant–Friedrichs–Lewy) number to

be within 1, so the time step was set as 30 s with the self-adaptive CFL number as 0.8.

3.3. Wave Model

The spectral wave module was launched to establish the wave model. The governing

equations can be found in the MIKE 21 manual [

] and the parameter setting was in

accordance with Wang et al. [

]. The relationship between the action density

N(σ,θ)

and

the energy density E(σ,θ)is be deﬁned by the following equation:

N(σ,θ) = E(σ,θ)

σ(6)

where the σis relative angular frequency and θis the wave direction.

The energy source term Sis constituted for ﬁve representative components: the

wave energy generated by wind (

Sin

), the wave energy transferred due to nonlinear wave

interaction (

Sn1

), the dissipation of wave energy due to white capping (

Sds

), the dissipation

of wave energy due to bottom friction (

Sbot

) and the dissipation of wave energy due to

surface wave breaking (Ssurf). The calculation is as follows:

S=Sin +Sn1+Sds +Sbot +Ssur f (7)

To realize the coupling process, the wave radiation stresses are the key inputs of

the hydrodynamic model, which can be obtained from the numerical results of the wave

model. The equations of wave radiation stresses in the three directions during the wave

propagation are as follows:

Sxx =ρgZncos2θ+n−1

2Edσdθ(8)

Sxy =Syx =ρgZnsin θcos θEdσdθ(9)

Syy =ρgZnsin2θ+n−1

2Edσdθ(10)

where S

and S

are the wave radiation stresses in the xx-, xy- and yy- directions,

respectively, and nis the group velocity. The wave radiation stresses are signiﬁcant in

coastal wave propagation, such as wave reﬂection and wave breaking.

According to previous studies, the wind velocity obtained from ERA5 is generally

lower than the real values [

–

]. Fan conducted an investigation to determine the

coefﬁcient between the ECMWF data and the measured data in the Bohai Sea [

]. That

J. Mar. Sci. Eng. 2024,12, 90 8 of 24

study obtained a correction coefﬁcient distribution ranging from 1.10 to 1.55. In the present

study, the correction coefﬁcient was set as 1.3, and the wind data were corrected before being

input into the wave model. Other important parameters in the wave module, including

the wave breaking (Gamma), the bottom friction (sand grain size, d50), and the dissipation

coefﬁcient, were set as defaults with values of 0.8, 0.00025, and 0.15, respectively [43,47].

3.4. Model Validation

Model validations are quantiﬁed based on the RMSE (Root Mean Squared Error) and

the Skill number [

]. RMSE is a statistical method to judge the difference between the

measured and simulated data, while the Skill number provides an index of the agreement

between the model data and the real data. The equations of the two kinds of assessment

are deﬁned as follows:

RMSE =

∑

(ηi

measured −ηi

simulated)2

n(11)

where

ηi

measured

is the ith measured data;

ηi

simulated

is the ith simulated data; and nis the total

number of observations. A lower value of RMSE stands for better accuracy of the model.

In general, 0.5 is taken as the critical value.

Skill =1−

∑

i=1|Mi−Di|2

∑

i=1

(

Mi−D

+

Di−D

)2(12)

where M

and D

are the ith model result and the in situ measured data, respectively;

the mean value of the in situ measured data; and Nis the number of observations.

Figures 6–8show the validations of water level, current velocity, and wave based on

our previous work [

], with the assessment values presented in the ﬁgures. The station

positions are shown in Figure 1. All simulated data on the hydrodynamic factors match

well with the measured data in the following ﬁgures.

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 9 of 25

Figures 6–8 show the validations of water level, current velocity, and wave based on

our previous work [46], with the assessment values presented in the ﬁgures. The station

positions are shown in Figure 1. All simulated data on the hydrodynamic factors match

well with the measured data in the following ﬁgures.

Figure 6. The veriﬁcations of water level at the three measured stations (Sheshan, Luchaogang and

Dajishan stations, positions shown in Figure 1) during three typhoon events: (A) Fongwong (2014);

(B) Ampil (2018); and (C) Lekima (2019).

Figure 7. The veriﬁcations of current velocity in the three stations of NGN4SD, CS3SD and NC6D

(the stations positions are shown in Figure 1, and the RMSE and Skill numbers are shown in this

ﬁgure). (a) –(c) are the current speeds, and (d) –(f) are the current directions

Figure 8. The veriﬁcations of the signiﬁcant wave height and wave period at the Yangkougang sta-

tion during three typhoon events: (a,b): Fongwong; (c,d): Ampil; and (e,f): Lekima.

Figure 6. The veriﬁcations of water level at the three measured stations (Sheshan, Luchaogang and

Dajishan stations, positions shown in Figure 1) during three typhoon events: (A) Fongwong (2014);

(B) Ampil (2018); and (C) Lekima (2019).

J. Mar. Sci. Eng. 2024,12, 90 9 of 24

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 9 of 25

Figures 6–8 show the validations of water level, current velocity, and wave based on

our previous work [46], with the assessment values presented in the ﬁgures. The station

positions are shown in Figure 1. All simulated data on the hydrodynamic factors match

well with the measured data in the following ﬁgures.

Figure 6. The veriﬁcations of water level at the three measured stations (Sheshan, Luchaogang and

Dajishan stations, positions shown in Figure 1) during three typhoon events: (A) Fongwong (2014);

(B) Ampil (2018); and (C) Lekima (2019).

Figure 7. The veriﬁcations of current velocity in the three stations of NGN4SD, CS3SD and NC6D

(the stations positions are shown in Figure 1, and the RMSE and Skill numbers are shown in this

ﬁgure). (a) –(c) are the current speeds, and (d) –(f) are the current directions

Figure 8. The veriﬁcations of the signiﬁcant wave height and wave period at the Yangkougang sta-

tion during three typhoon events: (a,b): Fongwong; (c,d): Ampil; and (e,f): Lekima.

Figure 7. The veriﬁcations of current velocity in the three stations of NGN4SD, CS3SD and NC6D

(the stations positions are shown in Figure 1, and the RMSE and Skill numbers are shown in this

ﬁgure). (a–c) are the current speeds, and (d–f) are the current directions.

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 9 of 25

Figures 6–8 show the validations of water level, current velocity, and wave based on

our previous work [46], with the assessment values presented in the ﬁgures. The station

positions are shown in Figure 1. All simulated data on the hydrodynamic factors match

well with the measured data in the following ﬁgures.

Figure 6. The veriﬁcations of water level at the three measured stations (Sheshan, Luchaogang and

Dajishan stations, positions shown in Figure 1) during three typhoon events: (A) Fongwong (2014);

(B) Ampil (2018); and (C) Lekima (2019).

Figure 7. The veriﬁcations of current velocity in the three stations of NGN4SD, CS3SD and NC6D

(the stations positions are shown in Figure 1, and the RMSE and Skill numbers are shown in this

ﬁgure). (a) –(c) are the current speeds, and (d) –(f) are the current directions

Figure 8. The veriﬁcations of the signiﬁcant wave height and wave period at the Yangkougang sta-

tion during three typhoon events: (a,b): Fongwong; (c,d): Ampil; and (e,f): Lekima.

Figure 8. The veriﬁcations of the signiﬁcant wave height and wave period at the Yangkougang station

during three typhoon events: (a,b): Fongwong; (c,d): Ampil; and (e,f): Lekima.

According to the RMSE and the Skill numbers, the indictors are within the appropriate

ranges. Meanwhile, the corresponding trend in the validations of the water level, current

speed, current direction, and signiﬁcant wave height and wave period illustrate the excel-

lent performances of both the hydrodynamic model and the wave model. Therefore, the

coupled model can be applied to the simulations.

4. Results

As a middle tide-dominated estuary, the typical position and coastline of the YRE

are propitious to the accumulation and development of storm power. The historical

maximum storm surge in the YRE exceeded 3.0 m, which caused signiﬁcant risks to life and

property safety in the nearby coastal cities, such as Shanghai and Hangzhou. Based on the

simulations of the three typical typhoon events, the spatial and temporal distributions of

storm surges were obtained for qualitative and quantitative analysis. A common calculation

for storm surge is HWCI minus HAT, where HWCI is the surface elevation simulated under

WCI conditions and HAT is the result when astronomic tide acts independently.

4.1. Temporal and Spatial Variation Characteristics of Fongwong

According to the measured data on typhoon Fongwong, the simulation period (20–26

September in 2014) covered the entire process, consisting of the developing, maturing and

weakening stages. Figure 9shows the variations of storm surge during Fongwong, in

which the typhoon landing moment is marked by a blue arrow. It is clear that the storm

surge was dominated by the tide when the typhoon was far from the YRE, and the value

J. Mar. Sci. Eng. 2024,12, 90 10 of 24

almost vibrated around 0. When the typhoon moved close to the estuary, all study points

experienced an increase in storm surge simultaneously and gained their peak at 6:00 on

23 September,

which was earlier than the landing moment due to the S-N moving direction.

The maximum surge value of the inner channel was 0.90 m at NB1, which was higher than

the offshore value of 0.62 m at NB2. Such a difference resulted from the narrow width

and low current speed inside the NB. The maximum surge values at SC2 and NC2 were

about 0.97 and 0.68 m, respectively, while the SC1 and NC1 presented values of 0.90 and

0.76 m,

respectively. As the typhoon moved away from the YRE and the intensity declined,

the storm surge gradually recovered and its value vibrated around 0 after 25 September.

According to the typhoon process, the storm surges at different points lasted for 3–4 days.

The inner point suffered a more intensive storm surge than the offshore points in the NB

and NC, but the difference in the SC was relatively small.

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 10 of 25

According to the RMSE and the Skill numbers, the indictors are within the appropri-

ate ranges. Meanwhile, the corresponding trend in the validations of the water level, cur-

rent speed, current direction, and signiﬁcant wave height and wave period illustrate the

excellent performances of both the hydrodynamic model and the wave model. Therefore,

the coupled model can be applied to the simulations.

4. Results

As a middle tide-dominated estuary, the typical position and coastline of the YRE are

propitious to the accumulation and development of storm power. The historical maxi-

mum storm surge in the YRE exceeded 3.0 m, which caused signiﬁcant risks to life and

property safety in the nearby coastal cities, such as Shanghai and Hangzhou. Based on the

simulations of the three typical typhoon events, the spatial and temporal distributions of

storm surges were obtained for qualitative and quantitative analysis. A common calcula-

tion for storm surge is H

WCI

minus H

, where H

WCI

is the surface elevation simulated un-

der WCI conditions and H

is the result when astronomic tide acts independently.

4.1. Temporal and Spatial Variation Characteristics of Fongwong

According to the measured data on typhoon Fongwong, the simulation period (20–

26 September in 2014) covered the entire process, consisting of the developing, maturing

and weakening stages. Figure 9 shows the variations of storm surge during Fongwong, in

which the typhoon landing moment is marked by a blue arrow. It is clear that the storm

surge was dominated by the tide when the typhoon was far from the YRE, and the value

almost vibrated around 0. When the typhoon moved close to the estuary, all study points

experienced an increase in storm surge simultaneously and gained their peak at 6:00 on

23 September, which was earlier than the landing moment due to the S-N moving direc-

tion. The maximum surge value of the inner channel was 0.90 m at NB1, which was higher

than the oﬀshore value of 0.62 m at NB2. Such a diﬀerence resulted from the narrow width

and low current speed inside the NB. The maximum surge values at SC2 and NC2 were

about 0.97 and 0.68 m, respectively, while the SC1 and NC1 presented values of 0.90 and

0.76 m, respectively. As the typhoon moved away from the YRE and the intensity declined,

the storm surge gradually recovered and its value vibrated around 0 after 25 September.

According to the typhoon process, the storm surges at diﬀerent points lasted for 3–4 days.

The inner point suﬀered a more intensive storm surge than the oﬀshore points in the NB

and NC, but the diﬀerence in the SC was relatively small.

Figure 9. Variations in storm surge at the six study points (position shown in Figure 1) during Fong-

wong from September 19 to 26, 2014 (six days), where the blue arrow points at the landing moment.

Figure 9. Variations in storm surge at the six study points (position shown in Figure 1) during Fongwong

from 19 to 26 September 2014 (six days), where the blue arrow points at the landing moment.

To visualize storm surge variations directly, the distribution of storm surge at different

time points (with an interval of 3 h) is shown in Figure 10. Figure 10 illustrates that during

the developing stage, a strong wind initially caused high surge in the HB (Hangzhou

Bay). As the typhoon moved toward the north, the surge center also moved, accompanied

by an increasing value. Meanwhile, the surge center led to a complex current direction,

particularly at the estuarine mouth. Additionally, the intensiﬁed ﬂood tide current brought

about more storm surge than the ebb tide current. When the typhoon moved toward NE,

the surge center shifted with the wind to Jiangsu province along the coastline and decreased

with declined wind speed. When the typhoon moved away from the estuary, most areas

recovered to common hydrodynamic conditions. This synchronous behavior between

typhoon path and storm surge variation illustrates the consistent response and recovery of

surge in the channel and the offshore area. However, the storm surge in the NB failed to

fade immediately due to the narrow channel, shallow depth and lower current speed.

4.2. Temporal and Spatial Variation Characteristics of Ampil

The simulation period of Ampil ranged from July 20 to 26 in 2018, and the landing

moment was at 11:00 on July 22 with a center wind speed of 28 m/s. Figure 11 shows the

variations in storm surge during Ampil. Similar to Fongwong, the surge vibrated around 0

before the typhoon landed and rapidly increased to the peak. The maximum value at NB1

was 0.83 m, but the SC showed a lower value due to the farther distance away from the path,

so the typhoon path is important to storm surge in terms of distance. After the typhoon

landing, the storm surge suddenly dropped to a negative value, and it decreased to

−

0.42 m

(at SC2) once because all study points were located on the left side of the path and affected

J. Mar. Sci. Eng. 2024,12, 90 11 of 24

by the anticlockwise cyclone. After the typhoon landed, the dominant wind shifted to the

offshore direction. Comparing the two typhoon events (Fongwong and Ampil), the surge

variation during Ampil experienced a rapid increase and, subsequently, a rapid decrease.

However, the surge peak showed a slight difference. These results indicate that typhoon

intensity and astronomic tide are two signiﬁcant factors among all the primary elements

impacting the rise in water levels. The impacts of typhoon intensity primarily occur in the

surge process, while the tide primarily affects the peak surge value.

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 11 of 25

To visualize storm surge variations directly, the distribution of storm surge at diﬀer-

ent time points (with an interval of 3 h) is shown in Figure 10. Figure 10 illustrates that

during the developing stage, a strong wind initially caused high surge in the HB (Hang-

zhou Bay). As the typhoon moved toward the north, the surge center also moved, accom-

panied by an increasing value. Meanwhile, the surge center led to a complex current di-

rection, particularly at the estuarine mouth. Additionally, the intensiﬁed ﬂood tide current

brought about more storm surge than the ebb tide current. When the typhoon moved to-

ward NE, the surge center shifted with the wind to Jiangsu province along the coastline

and decreased with declined wind speed. When the typhoon moved away from the estu-

ary, most areas recovered to common hydrodynamic conditions. This synchronous be-

havior between typhoon path and storm surge variation illustrates the consistent response

and recovery of surge in the channel and the oﬀshore area. However, the storm surge in

the NB failed to fade immediately due to the narrow channel, shallow depth and lower

current speed.

Figure 10. The distributions of storm surge with current velocity during Fongwong (time interval is

3 h).

4.2. Temporal and Spatial Variation Characteristics of Ampil

The simulation period of Ampil ranged from July 20 to 26 in 2018, and the landing

moment was at 11:00 on July 22 with a center wind speed of 28 m/s. Figure 11 shows the

variations in storm surge during Ampil. Similar to Fongwong, the surge vibrated around

0 before the typhoon landed and rapidly increased to the peak. The maximum value at

NB1 was 0.83 m, but the SC showed a lower value due to the farther distance away from

the path, so the typhoon path is important to storm surge in terms of distance. After the

typhoon landing, the storm surge suddenly dropped to a negative value, and it decreased

to −0.42 m (at SC2) once because all study points were located on the left side of the path

and aﬀected by the anticlockwise cyclone. After the typhoon landed, the dominant wind

shifted to the oﬀshore direction. Comparing the two typhoon events (Fongwong and

Figure 10. The distributions of storm surge with current velocity during Fongwong (time interval is 3 h).

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 12 of 25

Ampil), the surge variation during Ampil experienced a rapid increase and, subsequently,

a rapid decrease. However, the surge peak showed a slight diﬀerence. These results indi-

cate that typhoon intensity and astronomic tide are two signiﬁcant factors among all the

primary elements impacting the rise in water levels. The impacts of typhoon intensity pri-

marily occur in the surge process, while the tide primarily aﬀects the peak surge value.

Figure 11. Variations in the storm surge at the six study points (positions shown in Figure 1) during

Ampil from July 20 to 26, 2018 (six days), where the blue arrow points at the landing moment.

Ampil was a special typhoon that developed during a neap tide and caused an obvi-

ous negative surge. To further investigate this rare negative surge, the distribution of

storm surge from July 21 to 23 in 2018 is shown in Figure 12. At 12:00 on July 22, the surge

achieved its peak (Figure 12g) in the main branches and decreased to a negative value

within 9 h. As the typhoon gradually moved landwards, the regional negative surge oc-

curred, except at the upstream site, because the seaward wind dominated (corresponding

with Figure 11), which intensiﬁed the ebb tide current further. At 12:00 on July 23, the sea

level began to recover. Additionally, the surge center path was consistent with the ty-

phoon path, which stresses the importance of the path. The high value of the upstream

surge resulted from the lifting actions of runoﬀ.

Figure 11. Variations in the storm surge at the six study points (positions shown in Figure 1) during

Ampil from 20 to 26 July 2018 (six days), where the blue arrow points at the landing moment.

J. Mar. Sci. Eng. 2024,12, 90 12 of 24

Ampil was a special typhoon that developed during a neap tide and caused an obvious

negative surge. To further investigate this rare negative surge, the distribution of storm

surge from 21 to 23 July in 2018 is shown in Figure 12. At 12:00 on 22 July, the surge achieved

its peak (Figure 12g) in the main branches and decreased to a negative value within 9 h. As

the typhoon gradually moved landwards, the regional negative surge occurred, except at

the upstream site, because the seaward wind dominated (corresponding with Figure 11),

which intensiﬁed the ebb tide current further. At 12:00 on 23 July, the sea level began to

recover. Additionally, the surge center path was consistent with the typhoon path, which

stresses the importance of the path. The high value of the upstream surge resulted from the

lifting actions of runoff.

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 13 of 25

Figure 12. The distribution of storm surge with current velocity during Ampil (time interval is 3 h).

Ampil caused a limited surge in the inner part of the estuary, and the maximum value

mainly occurred in the NB. The negative surge was evident in the SB, owing to the landing

position that led to a lower water level on the left side of the typhoon path where the YRE

was. However, this typhoon event also resulted in signiﬁcant damage to the coastal cities.

A lower water depth of the Deep-Water Channel Project causes diﬃculties in mooring.

Moreover, the SB plays a vital role in fresh water supplies because it links the

Dongfengxisha, Chenxing and Qingcaosha reservoirs. If the negative surge period lasted

a longer time, the water supplies might have faced considerable challenges. Therefore, the

investigation of the storm surge during Ampil provides meaningful information for estu-

arine risk prevention.

4.3. Temporal and Spatial Variation Characteristics of Lekima

The simulation period of Lekima was from August 7 to 13 in 2019. This event never

landed in the YRE, so the approaching moment was set as the moment when it crossed

the upstream river of the YRE (at 15:00 on August 10 with a center wind speed of 23 m/s).

Figure 13 shows the variations in storm surge during Lekima. Compared with Fongwong

and Am pil, alth ough Le kima just cr ossed t he upst ream rive r of t he YR E, it caus ed the m ost

serious surge in terms of super typhoon intensity. As shown in Figure 13, the surge value

before the typhoon vibrates above 0. The maximum surge presents at SC2, and the peak

might have reached 1.50 m. This event proves that typhoon intensity is one of the most

signiﬁcant dynamics aﬀecting storm surge. The storm surge lasted 2–3 days, which was

shorter than the Fongwong-induced surge and longer than the Ampil-induced surge, so

the high surge was mainly maintained by the tide.

Figure 12. The distribution of storm surge with current velocity during Ampil (time interval is 3 h).

Ampil caused a limited surge in the inner part of the estuary, and the maximum

value mainly occurred in the NB. The negative surge was evident in the SB, owing to the

landing position that led to a lower water level on the left side of the typhoon path where

the YRE was. However, this typhoon event also resulted in signiﬁcant damage to the

coastal cities. A lower water depth of the Deep-Water Channel Project causes difﬁculties in

mooring. Moreover, the SB plays a vital role in fresh water supplies because it links the

Dongfengxisha, Chenxing and Qingcaosha reservoirs. If the negative surge period lasted

a longer time, the water supplies might have faced considerable challenges. Therefore,

the investigation of the storm surge during Ampil provides meaningful information for

estuarine risk prevention.

4.3. Temporal and Spatial Variation Characteristics of Lekima

The simulation period of Lekima was from 7 to 13 August in 2019. This event never

landed in the YRE, so the approaching moment was set as the moment when it crossed

the upstream river of the YRE (at 15:00 on 10 August with a center wind speed of 23 m/s).

Figure 13 shows the variations in storm surge during Lekima. Compared with Fongwong

and Ampil, although Lekima just crossed the upstream river of the YRE, it caused the most

serious surge in terms of super typhoon intensity. As shown in Figure 13, the surge value

before the typhoon vibrates above 0. The maximum surge presents at SC2, and the peak

might have reached 1.50 m. This event proves that typhoon intensity is one of the most

signiﬁcant dynamics affecting storm surge. The storm surge lasted 2–3 days, which was

shorter than the Fongwong-induced surge and longer than the Ampil-induced surge, so

the high surge was mainly maintained by the tide.

J. Mar. Sci. Eng. 2024,12, 90 13 of 24

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 14 of 25

Figure 13. Variations in storm surge at the six study points (positions shown in Figure 1) during

typhoon Lekima from August 7 to 13, 2019 (six days), where the blue arrow points at the landing

moment.

Figure 14 shows the surge distributions with current velocity during Lekima. As

shown in the ﬁgure, the surges in the HB and SB were relatively high before the typhoon

approached due to the strong wind. When the typhoon moved close to the upstream river

of the YRE, the storm surge center remained in the YRE for a long time (Figure 14c–h).

When the typhoon left far away, the surge tardily decreased, especially in the branches,

revealing that the typhoon accumulated a signiﬁcant amount of water inside the channel.

The surge center path was consistent with the typhoon path from S to N. From 18:00 on

Aug ust 11, the o ﬀshore area faced a negative surge, while the channel and mouth areas

presented a high surge due to the weak ebb tidal actions during the neap tide.

Figure 13. Variations in storm surge at the six study points (positions shown in Figure 1) dur-

ing typhoon Lekima from 7 to 13 August 2019 (six days), where the blue arrow points at the

landing moment.

Figure 14 shows the surge distributions with current velocity during Lekima. As

shown in the ﬁgure, the surges in the HB and SB were relatively high before the typhoon

approached due to the strong wind. When the typhoon moved close to the upstream river

of the YRE, the storm surge center remained in the YRE for a long time (Figure 14c–h).

When the typhoon left far away, the surge tardily decreased, especially in the branches,

revealing that the typhoon accumulated a signiﬁcant amount of water inside the channel.

The surge center path was consistent with the typhoon path from S to N. From 18:00 on

11 August, the offshore area faced a negative surge, while the channel and mouth areas

presented a high surge due to the weak ebb tidal actions during the neap tide.

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 14 of 25

Figure 13. Variations in storm surge at the six study points (positions shown in Figure 1) during

typhoon Lekima from August 7 to 13, 2019 (six days), where the blue arrow points at the landing

moment.

Figure 14 shows the surge distributions with current velocity during Lekima. As

shown in the ﬁgure, the surges in the HB and SB were relatively high before the typhoon

approached due to the strong wind. When the typhoon moved close to the upstream river

of the YRE, the storm surge center remained in the YRE for a long time (Figure 14c–h).

When the typhoon left far away, the surge tardily decreased, especially in the branches,

revealing that the typhoon accumulated a signiﬁcant amount of water inside the channel.

The surge center path was consistent with the typhoon path from S to N. From 18:00 on

Aug ust 11, the oﬀshore area faced a negative surge, while the channel and mouth areas

presented a high surge due to the weak ebb tidal actions during the neap tide.

Figure 14. The distribution of storm surge with current velocity during Lekima (time interval is 3 h).

J. Mar. Sci. Eng. 2024,12, 90 14 of 24

4.4. Summaries of Storm Surges Variations during Typical Typhoon Processes

The three representative typhoons show evident differences in terms of intensity and

path. By comparing the surge characteristics during these three representative typhoons,

both similarities and differences were obtained and are summarized in Table 2. Although

the typhoon intensity of Fongwong was much lower than the other two high-intensity

typhoons, it showed a longer-lasting time of storm surge for about 3–4 days at all study

points due to the intensive spring tidal actions. Ampil landed north of the YRE and moved

toward NW (from sea to land). It brought about a rapid increase and then a rapid decrease

in storm surge. The storm surge lasted only two days, while the negative surge remained

even for 3 days, and it caused the lack of water supplies. As the strongest typhoon among

the three events, Lekima only passed the upstream river of the YRE, but induced the highest

storm surge, which demonstrates that the storm surge is more sensitive to typhoon intensity

than the landing position.

Table 2. Summaries of storm surges during the three typhoon events.

Typhoon Fongwong Ampil Lekima

Year 2014 2018 2019

Landing Directly Directly Passed

Path

J. Mar. Sci. Eng. 2024, 12, x FOR PEER REVIEW 15 of 25

Figure 14. The distribution of storm surge with current velocity during Lekima (time interval is 3

h).

4.4. Summaries of Storm Surges Variations during Typical Typhoon Processes

The three representative typhoons show evident diﬀerences in terms of intensity and

path. By comparing the surge characteristics during these three representative typhoons,

both similarities and diﬀerences were obtained and are summarized in Table 2. Although

the typhoon intensity of Fongwong was much lower than the other two high-intensity

typhoons, it showed a longer-lasting time of storm surge for about 3–4 days at all study

points due to the intensive spring tidal actions. Ampil landed north of the YRE and moved

toward NW (from sea to land). It brought about a rapid increase and then a rapid decrease

in storm surge. The storm surge lasted only two days, while the negative surge remained

even for 3 days, and it caused the lack of water supplies. As the strongest typhoon among

the three events, Lekima only passed the upstream river of the YRE, but induced the high-

est storm surge, which demonstrates that the storm surge is more sensitive to typhoon

intensity than the landing position.

Overall, tidal intensity, typhoon intensity (wind speed), landing position and ty-

phoon path are of great importance during a storm surge process. Their combined func-

tions characterize the storm surge processed in the YRE.

Table 2. Summaries of storm surges during the three typhoon events.

Typhoon Fongwong Ampil Lekima

Year 2014 2018 2019

Landing Directly Directly Passed

Path

Tide Spring tide Neap tide Neap tide

Intensity Low Middle High

Storm surge All positive Positive to negative Positive

Lasting time 3–4days 2 days 2–3days

Peak value (m) 0.97 (SC2) 0.79 (NB1) 1.17 (SC2)

Variation rate Slow Fast Slow–fast

5. Discussion

The interactions among wind, wave and tide have a vital inﬂuence on the mechanism

underlying storm surge variations. Previous investigations have indicated that the non-

linear interaction between tide and wind is related to the landing moment, typhoon path,

tidal intensity, water depth and location [49–51]. Thomas et al. [52] concluded that storm

surges have a strong nonlinear relation with tide, which was consistent with the conclu-

sion drawn by Peng et al. [53]. Furthermore, the wave–current nonlinear interaction is one

of the most meaningful factors. Waves can alter the tidal level by changing the surface and

boom stresses. Meanwhile, wave radiation stresses also have distinct eﬀects on storm

surge, with a contribution of 2–5%. In this section, to ﬁgure out the key interactions be-

tween the main dynamics and storm surge, four diﬀerent dynamic combinations were

examined, as shown in Table 3. Previous studies [54–56] named the storm water level

(HWCI) as the total water level (Condition 4). The storm surge HSS is the storm water level

(HWCI) excluding the astronomic tide actions HAT (Condition 1), which can be described as

HSS = HWCI − HAT. The wave setup HWS can be obtained from the water level under the com-

bined actions of tide and wave (Condition 3), while excluding the astronomic tide actions