Comparative Analysis of Hydrodynamic Characteristics off Shandong Under the Influence of Two Types of Storm Surges

Abstract

As China’s largest peninsula, the Shandong Peninsula faces recurrent threats from both tropical and extratropical cyclone-induced storm surges. Understanding the distinct mechanisms governing these surge types is critical for developing targeted coastal hazard mitigation strategies. This investigation employs the FVCOM-SWAVE coupled wave–current model to conduct numerical simulations and comparative analyses of two 2022 surge events, Typhoon Muifa (tropical) and the “221003” extratropical surge. The results demonstrate that hydrodynamic responses exhibit strong dependence on surge-generating meteorological regimes. Tropical surge dynamics correlate closely with typhoon track geometry, intensity gradients, and asymmetric wind field structures, manifesting rightward-biased energy intensification relative to storm motion. Conversely, extratropical surge variations align with evolving wind-pressure configurations during cold air advection, driven by synoptic-scale atmospheric reorganization. The hydrodynamic environmental response in the sea areas surrounding Jiaodong and Laizhou Bay is particularly pronounced, influenced by the intensity of wind stress on the sea surface, as well as the bathymetry and coastal geometry.

1. Introduction

Storm surge, defined as an abnormal sea-level fluctuation induced by severe meteorological disturbances (principally extreme winds and rapid barometric pressure variations), represents a critical coastal hazard that is categorized into tropical and extratropical types [1]. Tropical storm surges, generated by intense cyclonic systems, including typhoons and hurricanes, predominantly manifest in mid- to low-latitude coastal regions in summer and autumn. These events frequently exhibit catastrophic destruction and constitute one of the most severe marine disasters affecting global subtropical and tropical coastlines. In contrast, temperate storm surges, typically associated with extratropical cyclones and cold-air outbreaks, demonstrate more gradual water level variations, occasionally accompanied by negative surge. These phenomena primarily occur in mid–high-latitude zones during transitional seasons (spring and autumn), presenting lower magnitude impacts than their tropical counterparts [2].

As China’s largest peninsula, the Shandong Peninsula exhibits distinct storm surge occurrence characteristics, and its eastern and southern coasts are particularly vulnerable to tropical cyclone-induced typhoon surges from the northwest Pacific, whereas northern areas, including Laizhou Bay, experience higher exposure to extratropical storm surge systems [3]. Advancements in computational modeling have enabled systematic investigation of the storm surge dynamics of Shandong. Li et al. [4] implemented a coupled WRF-ROMS atmosphere–ocean model to comparatively analyze surface wind and current field evolution during two distinct extreme weather events in the Bohai Sea. Through numerical simulations of historical surge events (1985–2017), Du et al. [5] quantified the spatiotemporal patterns of water level anomalies and significant wave height variations across three major marine ranchings adjacent to the Shandong Peninsula. Li et al. [6] conducted sensitivity analyses on wind stress drag coefficient parameterizations and evaluated their impacts on typhoon surge simulation accuracy in the North China Sea under both wave-coupled and wave-uncoupled scenarios. Xu et al. [7] developed a three-dimensional typhoon surge model for the North Yellow Sea and Bohai Sea using the FVCOM-SWAVE, demonstrating that bidirectional wave–current coupling significantly improves simulation accuracy compared to uncoupled scenarios through validation against observational datasets. Their comparative analysis revealed enhanced consistency between the modeled and measured surge parameters under coupled conditions. Mo et al. [8] conducted sensitivity analyses of water level variability, current field dynamics, and significant wave height responses to nonlinear wave–current interactions during cold-air outbreaks in the northern East China Sea and Bohai-Yellow Sea systems, and the results show that the storm water level change triggered by wave action in the shallow water area is significantly larger than that in the central deep-water area of the East, Yellow, and Bohai Seas.

While existing research has predominantly focused on the generation mechanisms and water level anomalies of individual storm surge events along the Shandong coast, comparative analyses between tropical and extratropical surge types remain scarce. We implemented a three-dimensional real-time coupled model with bidirectional wave–current interactions using FVCOM-SWAVE and selected two representative events for systematic comparison: Typhoon Muifa (September 2022) as a tropical surge storm and the 221003 cold-air outbreak (October 2022) as an extratropical surge storm. Through high-resolution numerical simulations, we contrastingly analyzed the differential responses in sea surface elevation, sea current fields, and significant wave height evolution patterns between these meteorologically distinct storm types. These results provide critical insights for refining predictive models of marine hydrodynamics and enhancing operational storm surge warning systems along the coastline of Shandong.

2. Numerical Models

2.1. Holland-ERA5

The parametric empirical wind field demonstrates superior accuracy in typhoon core regions, whereas the background wind field provides more reliable simulations in peripheral zones beyond the radius of the maximum winds [9]. To leverage these complementary strengths, we used a hybrid methodology integrating both wind fields to generate the forcing wind field of the model. The background component was derived from the European Center for Medium-Range Weather Forecasts (ECMWF) ERA5 reanalysis dataset “https://www.ecmwf.int (accessed on 15 March 2025)”, featuring a 0.25° × 0.25° spatial resolution, hourly temporal resolution, and 10 m elevation above the mean sea level.

The empirical wind field was simulated using the Holland model [10], which incorporates the radial pressure distribution parameter B (dimensionless shape coefficient) scaled by the radius of maximum wind (R_max) to optimize the barometric profile configuration with fewer calibrated parameters. The model’s governing equations defining the pressure field and gradient wind velocity are expressed as:

$$p_r=p_c+\left(p_n-p_c\right)exp\left[-{\left({\displaystyle \frac{R_{max}}{r}}\right)}^B\right],$$

(1)

$$V_g\left(r\right)={\left[{\displaystyle \frac{B}{{\rho }_{\alpha }}}{\left({\displaystyle \frac{R_{max}}{r}}\right)}^B\left(p_n-p_c\right)exp\left[-{\left({\displaystyle \frac{R_{max}}{r}}\right)}^B\right]+{\left({\displaystyle \frac{rf}{2}}\right)}^2\right]}^{{\displaystyle \raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.}}-{\displaystyle \frac{rf}{2}},$$

(2)

where $p_r$ represents the atmospheric pressure at a distance r from the typhoon center, $V_g\left(r\right)$ denotes the gradient wind speed at the radial distance of r, $p_c$ and $p_n$ correspond to the central and peripheral atmospheric pressures of the typhoon ($p_n$ sets at 1010 hPa), R_max signifies the radius of the maximum wind speed, B is the radial pressure distribution parameter, r is the distance from the calculation point to the center of the typhoon, ${\rho }_{\alpha }$ is the air density (the value is 1.2 kg/m³) and f represents the Coriolis parameter.

Among them, there are several different empirical formulas for the maximum typhoon wind radius R_max, and Willoughby’s formula [11] is used here, as follows:

$$R_{max}=51.6exp\left(-0.0223V_f+0.0281\left|\phi \right|\right),$$

(3)

where φ represents the latitude of the typhoon center, $V_f$ denotes the maximum wind speed.

The hybrid wind-field synthesis employed a radial distance-dependent weighting coefficient e to dynamically blend the Holland parametric wind field with the ERA5 reanalysis wind field. This blending scheme prioritizes the Holland gradient wind representation in the typhoon core region (r < 2R_max) and transitions to ERA5 dominance in the peripheral zones (r ≥ 2R_max). The distance-weighted coefficient ensures a smooth transitional behavior between the two datasets while maintaining physical realism in the synthesized wind field [12]. The governing formulation for this wind-field fusion is expressed as [9]:

$$\overrightarrow{V}=\left\{\begin{array}{c}{\overrightarrow{V}}_{Holland}\left(1-e\right)+e{\overrightarrow{V}}_{ERA5}, r<2R_{max}\\ {\overrightarrow{V}}_{ERA5} , r\ge 2R_{max}\end{array}\right.,$$

(4)

where $\overrightarrow{V}$ represents the composite wind field, ${\overrightarrow{V}}_{Holland}$ denotes the empirical wind field derived from the Holland model, ${\overrightarrow{V}}_{ERA5}$ signifies the ERA5 background wind field, $e$ is a weighting coefficient, defined as $e=C^4/\left(1+C^4\right)$, where the coefficient $C=r/\left(\mathrm{n}\times R_{max}\right)$, and n is assigned a value of nine.

The radial pressure distribution parameter B in the Holland model critically modulates the typhoon intensity metrics and wind field structure. Under controlled conditions with a fixed central pressure ($p_c$) and radius of maximum winds (R_max), sensitivity analyses revealed that elevated B values produced progressively steeper wind speed profiles (Figure 1). This steepening gradient manifests as intensified barotropic pressure differentials between the typhoon core and transitional zone, particularly enhancing the wind speed variability of the storm center in blended wind field simulations.

Comparisons between hybrid wind field simulations and in situ anemometer measurements at Xiaomai Island (XMD) and Zhifu Island (ZFD) demonstrated different capabilities in capturing dynamic wind speed/direction variations under varying parameter B configurations (Figure 2). Calibration at B = 0.4 achieved optimal alignment with observed wind speeds (RMSE < 4.0 m/s) while maintaining directional consistency (mean angular deviation < 35°). However, early-phase typhoon simulations exhibited systematic overestimation, particularly at the XMD station, where the blended wind speeds exceeded the measurements. This discrepancy stems from inherent limitations in the capacity of the Holland model to replicate the asymmetric wind structure characteristics of real-world typhoons, which often generate intensified peripheral flows [12].

2.2. FVCOM-SWAVE

The Finite Volume Coastal Ocean Model (FVCOM), initially developed by Chen et al. [13], employs a finite volume discretization scheme to solve three-dimensional primitive hydrodynamic equations on unstructured triangular meshes. Subsequent advancements by Qi et al. [14] integrated spectral wave dynamics through the FVCOM-SWAVE module, which adapted the third-generation SWAN wave model to FVCOM’s unstructured grid architecture. This coupled system maintains horizontal grid consistency between the hydrodynamic and wave modules, achieving real-time bidirectional coupling via shared-memory variable exchange. Key wave parameters calculated after model initialization, including the significant wave height (H_S), wavelength (L), peak spectral period (T_p), and radiation stress components ($S_{xx}$, $S_{xy}$, $S_{yx} \mathrm{a}\mathrm{n}\mathrm{d} S_{yy}$), are dynamically transferred from SWAVE to FVCOM. Concurrently, the hydrodynamic outputs (current velocity u and sea surface elevation η) were synchronized back to the wave module. These coupled fields jointly inform the bottom boundary layer model, where iterative calculation of the bed shear stress (τ_b) occurs until the hydrodynamic convergence criteria are met [15]. The governing equations incorporating wave–current interactions in σ-coordinates are formulated as follows:

$${\displaystyle \frac{\partial \mathrm{ƞ}}{\partial t}}+{\displaystyle \frac{\partial uD}{\partial x}}+{\displaystyle \frac{\partial vD}{\partial y}}+{\displaystyle \frac{\partial w}{\partial \sigma }}=0,$$

(5)

$$\begin{gathered} {\displaystyle \frac{\partial uD}{\partial t}}+{\displaystyle \frac{\partial u^{2}D}{\partial x}}+{\displaystyle \frac{\partial uvD}{\partial y}}+{\displaystyle \frac{\partial uw}{\partial \sigma }}-fvD+D{\displaystyle \frac{\partial }{\partial x}}(g\mathrm{ƞ}+P_a)={\displaystyle \frac{1}{D}}{\displaystyle \frac{\partial }{\partial \sigma }}(K_m{\displaystyle \frac{\partial u}{\partial \sigma }})+D{F}_u- \\ {\displaystyle \frac{gD}{{\rho }_0}}\left[{\displaystyle \frac{\partial }{\partial x}}\left(D{\int }_{\sigma }^0{\rho }_{w}d{\sigma }^{\prime }\right)+\sigma {\rho }_w{\displaystyle \frac{\partial D}{\partial x}}\right]-\left({\displaystyle \frac{\partial D{S}_{xx}}{\partial x}}+{\displaystyle \frac{\partial D{S}_{yx}}{\partial y}}\right)+\sigma \left({\displaystyle \frac{\partial D}{\partial x}}{\displaystyle \frac{\partial S_{xx}}{\partial \sigma }}+{\displaystyle \frac{\partial D}{\partial y}}{\displaystyle \frac{\partial S_{yx}}{\partial \sigma }}\right), \end{gathered}$$

(6)

$$\begin{gathered} {\displaystyle \frac{\partial vD}{\partial t}}+{\displaystyle \frac{\partial uvD}{\partial x}}+{\displaystyle \frac{\partial v^{2}D}{\partial y}}+{\displaystyle \frac{\partial vw}{\partial \sigma }}+fuD+D{\displaystyle \frac{\partial }{\partial y}}(g\mathrm{ƞ}+P_a)={\displaystyle \frac{1}{D}}{\displaystyle \frac{\partial }{\partial \sigma }}(K_m{\displaystyle \frac{\partial v}{\partial \sigma }})+D{F}_v- \\ {\displaystyle \frac{gD}{{\rho }_0}}\left[{\displaystyle \frac{\partial }{\partial y}}\left(D{\int }_{\sigma }^0{\rho }_{w}d{\sigma }^{\prime }\right)+\sigma {\rho }_w{\displaystyle \frac{\partial D}{\partial y}}\right]-\left({\displaystyle \frac{\partial D{S}_{xy}}{\partial x}}+{\displaystyle \frac{\partial D{S}_{yy}}{\partial y}}\right)+\sigma \left({\displaystyle \frac{\partial D}{\partial x}}{\displaystyle \frac{\partial S_{xy}}{\partial \sigma }}+{\displaystyle \frac{\partial D}{\partial y}}{\displaystyle \frac{\partial S_{yy}}{\partial \sigma }}\right), \end{gathered}$$

(7)

where D = H + η is the total water depth, H is the static water depth, η is the undulating height of the sea surface, u, v and w are the velocity components in x, y and σ directions, t is time, f represents the Coriolis parameter, g is the acceleration of gravity, $P_a$ is the atmospheric pressure on the sea surface, $K_m$ is the vertical eddy viscosity coefficient, $F_u$, $F_v$ are horizontal momentum diffusion terms, ${\rho }_0$, ${\rho }_{\mathrm{w}}$ are the reference density and the total density of seawater. Because the difference in seawater density is not considered in this study, ${\rho }_{\mathrm{w}}={\rho }_0$. $S_{xx}$, $S_{xy}$, $S_{yx} \mathrm{a}\mathrm{n}\mathrm{d} S_{yy}$ are the horizontal wave radiation stress terms.

3. Model Setup

The 2022 “221003” extratropical storm surge event inflicted direct economic losses of RMB 1.198 billion in Shandong Province, representing 50.35% of China’s total storm surge-related damages that year. And the severe Typhoon Muifa (international designation: 2212) caused provincial losses exceeding RMB 67 million through its tropical surge impacts [16]. It is clear that the 2022 co-occurrence of tropical and extratropical storm surges inflicted substantial socioeconomic impacts across Shandong’s coastal urban areas. A comparative analysis of these meteorologically distinct surge types provides critical insights for optimizing coastal defense strategies, enhancing real-time emergency response protocols, and refining risk assessment frameworks in cyclone-prone regions. Therefore, given factors such as the occurrence time, scope of action, and intensity of the disasters, these events were selected as archetypal cases for comparative analysis: (1) Typhoon Muifa (14–16 September 2022), exemplifying tropical cyclone-induced surges, and (2) Cold-Air Outbreak 221003 (3–5 October 2022), representing extratropical surge mechanisms.

The Typhoon Storm Surge Model (TSSM) and Extratropical Storm Surge Model (ESSM) were constructed using the Surface Water Model System (SMS) with the Global Self-Consistent, Hierarchical, High-Resolution Geography (GSHHG) dataset to generate horizontally unstructured triangular grids (Figure 3). The computational domain spans 117–129° E and 30–41° N, employing σ-coordinates vertically with nine layers. Open boundaries were positioned 750 km beyond the study area to adequately resolve the surge generation and propagation dynamics. Bathymetric inputs adopted the General Bathymetric Chart of the Oceans (GEBCO_2023) “https://www.gebco.net (accessed on 15 March 2025)” 15-arcsecond grid (IHO/IOC standard) and high-resolution electronic navigational charts (ENCs) for nearshore zones (<80 m depth), all of which referenced the mean sea level. The bottom friction was parameterized using a quadratic drag coefficient (C_d = 0.0012).

Table 1. Parameter configuration in the model.

Parameter	Value
Mesh domain	117° E~129° E, 30°–41° N
Mesh resolution	Maximum triangular mesh is 20 km; minimum triangular mesh is 450 m
Mesh nodes and elements	107298 nodes and 209024 elements
Vertical stratification	Nine layers
Run duration (UTC)	TSSM: from 7 September 2022 00:00 to 16 September 2022 18:00; ESSM: from 28 September 2022 00:00 to 7 October 2022 00:00
Time step (FVCOM)	The external mode is 0.5 s, the internal mode is 5 s, and the time splitting mode ratio is 10
Time step (SWAVE)	5 s
Spectral resolution (frequency × angle)	20 × 72
Wave spectrum	JONSWAP
Boundary stage	Four diurnal tides (K1, O1, P1, Q1) and four semidiurnal tides (M2, S2, N2, K2) simulated by TPXO9
Starting conditions	Cold start (TSSM initial temperature is 20 °C, ESSM is 18 °C; the salinity was set to 30%)
Wind and air pressure fields	TSSM: Holland-ERA5wind field; ESSM: ECMWF-ERA5 reanalysis wind field data

summarizes the critical TSSM/ESSM configuration parameters, including mesh resolution, time step, and run duration.

The tide level validation data are from the measured water level data of six tide gauge stations: Qingdao (QD), Longkou (LK), Tanggu (TG), Huanghua harbor (HHG), Rizhao harbor (RZG), and Weifang harbor (WFG). The validation data of tidal current, effective wave height, and wind speed were obtained from the National Ocean Science Data Center “https://mds.nmdis.org.cn (accessed on 15 March 2025)”, a national platform for sharing scientific and technological resources, of which the current flow data were the tidal current prediction results of the three stations, namely, Bohai Sea Route (C1), Chengshanjiao (C2), and Laotieshan Waterway (C3). The significant wave height data were the measured data from Xiaomai Island (XMD), and the wind speed measurement stations were the two stations from Xiaomai Island (XMD) and Zhifudao Island (ZFD). The locations of these stations are shown in Figure 4.

The evaluation coefficient S [17,18,19] and root mean square error R were used to describe the degree of model accuracy using the following formula:

$$S=1-{\displaystyle \frac{{\sum }_{i=1}^n{\left|X_{\mathrm{m}\mathrm{o}\mathrm{d}}-X_{\mathrm{o}\mathrm{b}\mathrm{s}}\right|}^2}{{\sum }_{i=1}^n{\left(\left|X_{\mathrm{m}\mathrm{o}\mathrm{d}}-{\overline{X}}_{\mathrm{o}\mathrm{b}\mathrm{s}}\right|+\left|X_{\mathrm{o}\mathrm{b}\mathrm{s}}-{\overline{X}}_{\mathrm{o}\mathrm{b}\mathrm{s}}\right|\right)}^2}},$$

(8)

$$R=\sqrt{{\displaystyle \frac{1}{n}}{\sum }_{i=1}^n{\left(X_{\mathrm{m}\mathrm{o}\mathrm{d}}-X_{\mathrm{o}\mathrm{b}\mathrm{s}}\right)}^2},$$

(9)

where n is the total number of samples for the corresponding variable, Xmod and Xobs are the simulated and measured values, respectively, and ${\overline{X}}_{\mathrm{o}\mathrm{b}\mathrm{s}}$ is the average value of the measured variables. The value range of S is [0, 1]; 0 indicates that the simulated value is completely inconsistent with the measured value, and 1 indicates that the simulated value is completely consistent with the measured value.

4. Results

4.1. 2212 Typhoon Storm Surge

Typhoon Muifa’s track was obtained from the China Meteorological Administration’s (CMA) Tropical Cyclone Best Track Dataset [20,21], with the genesis occurring at 21.5° N, 136.7° E [22] in the western North Pacific (Figure 5a). The system made landfall as a tropical storm (TS) at Laoshan, Qingdao (36.15° N, 120.65° E), on 16 September 2022, at 0000 UTC (0800 Beijing Time), featuring a central pressure of 990 hPa and a maximum sustained wind speed of 23 m/s (Beaufort Force 9). During its Shandong coastal transit (15 September 0800-1600 UTC), the storm induced a steepened pressure gradient, generating 8–10 Beaufort scale gusts (17–24 m/s) and localized precipitation exceeding 150 mm. The system continuously propagated northeastward, with cyclonic energy concentrated over the Bohai Sea (37.0–41.0° N, 118.7–122.3° E). By 0400 UTC on 16 September, it had again made landfall in Liaoning (39.6° N, 122.5° E) and was gradually weakening (Figure 5b–f). Wind vectors (10 m elevation, black arrows) and isobars (red contours) illustrate evolving synoptic patterns.

The Typhoon Storm Surge Model (TSSM) was validated by comparing the hourly simulated and observed total water levels, current velocities and directions, significant wave heights (H_S), and wind speeds across the monitoring stations (Figure 6). Typhoon storm tide simulations demonstrated strong agreement with observations at six tide gauges (S > 0.93), although phase deviations led to a 1 h lead in tidal predictions at Qingdao (QD) and Rizhao (RZG). Current velocity root mean square errors (R) remained ≤ 0.15 m/s across stations, with directional variations maintaining coherence. At Xiaomai Island (XMD), HS simulations accurately replicated the observed rapid growth-slow decay pattern, achieving S = 0.98 and R = 0.26 m. The full validation metrics (S and R) for all parameters are systematically tabulated in

Table 2. Error analysis of simulation results of TSSM.

Verification Elements	Observation Stations	S-Value	R-Value
Water level	QD	0.94	0.53
	LK	0.96	0.17
	TG	0.96	0.33
	HHG	0.97	0.31
	RZG	0.97	0.42
	WFG	0.97	0.24
Current velocity	C1	0.90	0.15
	C2	0.98	0.12
	C3	0.81	0.14
Current direction	C1	0.94	-
	C2	0.93	-
	C3	0.86	-
Significant wave height	XMD	0.98	0.26

The unit of tidal elevation and significant wave height R is meter, and the unit of current velocity R is m/s. ‘-’ indicates that the value has not been obtained.

, delineating the model performance across hydrodynamic variables.

Typhoon-induced storm surges exhibited significant amplification along the Shandong Peninsula, particularly along the southwestern coastline and Laizhou Bay (Figure 7). During the pre-landfall phase (8 h prior to landfall), maximum surge amplitudes (~1.0 m) developed seaward of the Rizhao-Qingdao coastal zone, corresponding to the typhoon’s radius of maximum wind over the sea. The surge intensity subsequently decayed as the translational speed of the typhoon decreased. Concurrently, Laizhou Bay experienced a progressive water level rise during the typhoon’s northward trajectory, culminating in a 1.3 m surge at landfall (Figure 7d). Following the typhoon’s northward migration into the Bohai-Yellow Sea confluence, persistent northerly winds (orthogonal to the shoreline) drove a substantial seawater influx into southern Laizhou Bay. This wind-forced setup interacted synergistically with the bay’s concave coastal morphology and gradual bathymetric gradient, resulting in resonant water accumulation at the bayhead. The combined wind-topography forcing ultimately generated a maximum surge of 1.4 m at the WFG station at the top of Laizhou Bay (Figure 7f).

During typhoon transit, the sea surface current fields are governed by the synergistic effects of astronomical tides, typhoon wind stress, and bathymetric constraints. During the pre-landfall phase, southeasterly winds perpendicular to the shoreline dominated the south-eastern Shandong coastal region (Figure 5). As the typhoon center migrated northward, wind-driven surface currents developed pronounced cyclonic circulation patterns, with directional shifts from wind-aligned flows to counterclockwise rotations centered on the typhoon eye (Figure 8a,b). Upon landfall at Qingdao, current field trajectories radiated asymmetrically from Jiaozhou Bay, exhibiting enhanced flow velocities (>0.8 m/s) along the typhoon’s right-front with distinct cyclonic shear characteristics (Figure 8c,d). This asymmetric cyclonic structure persisted through the typhoon’s northward propagation across Shandong, gradually dissipating only after the Liaoning landfall at 0400 UTC on 16 September (Figure 8e,f). Wind stress amplification during surge conditions elevated nearshore current magnitudes to approximately 1.0 m/s along critical coastal sectors.

The spatial distribution of the significant wave heights (H_S) around the Shandong Peninsula exhibited a strong correlation with the radius of the typhoon’s maximum winds. As Typhoon Meihua propagated northward through the South Yellow Sea, the HS magnitudes increased progressively, displaying rightward-biased asymmetry relative to the storm track (Figure 9). A persistent crescent-shaped zone of extreme waves (H_S ≥ 8 m) developed near the Changmenyan and Chiyan Islands, maintaining quasi-stationary growth patterns. While deep-water wave energy typically concentrates nearshore owing to decreasing bathymetric gradients, subsequent irregularities in shoaling and coastal areas normally induce wave breaking and energy dissipation, causing rapid post-peak height reduction [5]. The anomalous maintenance of elevated HS near Chiyan Island suggests potential garden sprinkler effects (GSE) which require further investigation, including (1) insufficient spectral resolution for swell evolution modeling or (2) delayed wave breaking dynamics allowing paradoxical intensification. Post-typhoon conditions revealed contrasting trends: the southern coastal HS decreased markedly with wind relaxation, while the northern coasts (north of Shidao) experienced moderate wave reinforcement (H_S ~3 m) due to residual energy propagation (Figure 9e,f).

4.2. 221003 Extratropical Storm Surge

From 2 to 4 October, a southward-propagating cold air mass originating from Siberia and Mongolia triggered a nationwide cold wave event, inducing temperature descent ranges exceeding 12 °C with 9–11 Beaufort scale northeast gales across the Bohai-Yellow Sea coastal zone [22,23]. This synoptic pattern generated severe extratropical storm surges along Bohai Bay, Laizhou Bay, and the northern coast of Shandong. Between 0000 and 1800 UTC on 3 October, the eastward-migrating East Asian trough transitioned from zonal to meridional orientation, driving sustained coastal pressure rises in Shandong accompanied by persistent 7–9 Beaufort northeast winds (gusting 9–11 Beaufort) (Figure 10).

The Extratropical Storm Surge Model (ESSM) was validated by comparing the hourly simulated and observed total water levels, current velocities and directions, significant wave heights (H_S), and wind speeds across the monitoring stations (Figure 11). Extratropical storm tide simulations showed strong agreement with observations at six tide gauges (S ≥ 0.93), although phase deviations led to a 1 h lead in tidal predictions at Qingdao (QD) and Rizhao (RZG). Current speed overestimations (R > 0.20 m/s) occurred at Station C3, with directional discrepancies during peak flow phases. From 0000 UTC 3 October to 0000 UTC 4 October, H_S simulations exhibited systematic overestimations (R~0.25 m), potentially linked to ERA5 wind speed biases in the nearshore boundary layer. Comprehensive validation metrics (S and R) are quantified in

Table 3. Error analysis of simulation results of ESSM.

Verification Elements	Observation Stations	S-Value	R-Value
Water level	QD	0.93	0.51
	LK	0.94	0.27
	TG	0.94	0.49
	HHG	0.94	0.48
	RZG	0.95	0.47
	WFG	0.96	0.36
Current velocity	C1	0.89	0.15
	C2	0.77	0.36
	C3	0.84	0.21
Current direction	C1	0.91	-
	C2	0.63	-
	C3	0.92	-
Significant wave height	XMD	0.64	0.25

The unit of tidal elevation and significant wave height R is meter, and the unit of current velocity R is m/s. ‘-’ indicates that the value has not been obtained.

, highlighting the parameter-specific model performance variations.

During the “221003” extratropical storm surge event, the most pronounced water level anomalies along the Shandong coast occurred in Laizhou Bay, attributable to its semi-enclosed geomorphology and bathymetric configuration (Figure 12). Northeast gales associated with the cold surge drove southward transport of Liaodong Bay surface waters, converging along the southern Bohai Sea coastline to generate 1.0–1.6 m surge elevations from Bohai Bay to Laizhou Bay. The maximum surge of 1.9 m recorded at Weifang Port coincided with astronomical high tides, inducing intensified hydrodynamic activity through combined wind stress, barometric forcing, and tidal amplification. Surge propagation exhibited temporal lag effects, with Laizhou Bay peak levels occurring 2–3 h later than Bohai Bay due to (1) southeastward migration of the atmospheric pressure system altering wind stress patterns, and (2) topographic retardation. The southern Shandong coast experienced attenuated surge impacts (<0.5 m) compared to northern sectors, demonstrating reduced vulnerability to extratropical surge penetration from a northern approach.

The passage of cold waves induced intense wind-driven processes and significant barometric variations offshore Shandong, altering both water level dynamics and current field structures (Figure 13). The initial phase currents exhibited a complete flow reversal: Bohai Sea surface currents shifted from shoreward inflow to southward outflow through Liaodong Bay, with velocities reaching 0.8 m/s (peak > 1.0 m/s) (Figure 13a–d). The synoptic pressure transition from zonal to meridional trough orientation generated barotropic pressure gradients, driving seawater redistribution (Figure 13e,f). Concurrently, rapid thermal stratification from abrupt sea surface cooling modifies the vertical current shear. Southeastern Shandong waters exhibited weaker winds, resulting in minimal wind-driven currents and persistent dominance of western Yellow Sea tidal flows. ERA5 wind field limitations introduce directional biases in nearshore current simulations during peak wind events.

Persistent northerly winds induced spatiotemporal variability in wave energy across the Bohai-Yellow Sea system, with significant wave height (H_S) exhibiting a positive correlation with the wind speed magnitude (Figure 14). Wave intensification progressed southeastward following the trajectory of cold air. Under combined wind stress and barometric forcing, the HS of the Bohai Sea increased markedly, peaking at 5.0 m in the northern Laizhou Bay. Concurrent southeastward migration of the wind core enhanced the Yellow Sea wave activity, yielding HS values of 3–4 m in the northern sectors and 2–3 m in the southern regions. Bathymetry produced distinct nearshore–offshore gradients, with Shandong’s coastal shallows (<20 m depth) demonstrating 40–60% lower HS compared to adjacent deeper waters (>50 m depth), confirming the inverse proportionality between wave heights and proximity to the shoreline.

5. Discussion

Significant disparities exist between tropical and extratropical storm surge wind fields in Shandong’s offshore waters, corresponding to distinct surge magnitudes and spatial ranges. Typhoon Muifa landed in Qingdao from south to north, resulting in a storm surge near the southwest coast of Shandong. Subsequently, the intensity of the typhoon did not weaken and continued to move forward, leading to a more serious water increase in northern Laizhou Bay. Conversely, the “221003” extratropical surge event produced no obvious water-level anomalies along the southwestern coast. Under the influence of wind stress, Laizhou Bay exhibited the most pronounced hydrodynamic responses, which were attributed to its unique bathymetric profile and coastal configuration. The maximal surge magnitudes in this bay zone aligned with earlier findings by Yu et al. [24], confirming regional vulnerability to both surge types. Typhoon and extratropical storm surges induce distinct surface current responses in Shandong’s offshore waters. Counterclockwise circulation develops around the typhoon core, while cold wave events maintain unidirectional current dominance in the Bohai Sea and Yellow Sea. This study advances prior research by comparing wave field responses to both surge types, revealing critical differences in significant wave height (H_S) distributions. During typhoon passage, HS exhibits rightward-skewed asymmetry relative to the track of the typhoon, with maximum wave energy concentrated within the radius of maximum winds (RMW). Wave energy propagation shows strong spatial coupling with the typhoon trajectory from the central Yellow Sea through the Bohai Sea. Conversely, extratropical surges merely generate the most pronounced H_S variations (>2.5 m) in deeper offshore zones (>50 m depth) of the North Yellow Sea, despite southerly cold air penetration steadily.

To examine the most critical characteristics of storm surges, this study concentrates on the variation patterns of storm surge elevation across two distinct types of storm surges. Both storm surge types exhibit analogous hydrodynamic signatures in the intensity of storm surge elevation; strong winds tend to bring large increases in water, wherein wind-driven surge magnitudes are amplified during astronomical high-tide phases, likely culminating in extreme water level anomalies. Coastal embayments with complex morphologies demonstrate heightened susceptibility to onshore wind-induced setups compared to open-ocean environments, leading to pronounced water level elevations through geomorphic amplification. Both regions exhibited a similarity in the extent of the sea area affected by water increase, both along northern Laizhou Bay. Following its landfall in Qingdao, Shandong Province, Typhoon Muifa progressed northeastward, traversing the entire Jiaodong Peninsula through its center. Consequently, the waters of Laizhou Bay, and even portions of Bohai Bay, fell within the radius of the typhoon’s maximum wind speed, resulting in significant disturbance of the surface seawater due to intense northward winds. Additionally, cold waves originating from the north and moving southward traversed Shandong, the entire Bohai Sea, and the northern Yellow Sea and were driven by persistent strong northerly winds. This phenomenon caused seawater to be displaced southward, with the Laizhou Bay waters being the most affected. In addition, the difference between the two is that the typhoon storm surge also affected Qingdao, Rizhao, and other coastal cities on the southwest coast, the surrounding waters in Typhoon Muifa from south to north gradually approaching Shandong and landing in Qingdao during the existence of a significant rise in water level. The reason for this cannot be separated from the direction of movement of the typhoon and the landing location. Stormwater increase did not occur in these areas during the cold wave, which is considered to be related to the special geographic location and shoreline morphology of the Shandong Peninsula. Owing to blockage by the landmass of the Jiaodong Peninsula extending seaward, the winds at sea were weakened, and the wind direction was almost parallel to the southwest coast, which made it easier to blow coastal seawater away from the sea. From the point of view of the time sequence of the occurrence of storm surge, in the process of a typhoon, when the sea approaches the maximum wind speed radius, the storm surge phenomenon also gradually begins to appear. Cold wave cold air has unique barometric pressure–trough characteristics. When the barometric pressure field gradually moved from horizontal to vertical, the scope of action of the wind stress gradually moved southeast. The original strong wind area affecting northwestern Bohai Bay gradually acted over Laizhou Bay, thus leading to a time difference in the occurrence of storm surges. From the above perspectives, the main factors affecting the degree of water augmentation in different types of storm surges were the strength and structure of the wind field.

This study acknowledges the limitations in elucidating the mechanistic differences in water level anomalies between tropical and extratropical storm surges. Nearshore surge dynamics inherently involve complex wave–current–tide interactions, where water-level fluctuations represent the synergistic effects of multiple forcing agents, often incorporating storm surge, wave surge, and water-level changes due to astronomical tide-storm surge nonlinear interactions. Previous investigations employing diverse methodologies have advanced our understanding of surge mechanisms, and Ignacio et al. [25] quantified the relative contributions of atmospheric pressure gradients and wind stress to surge variability along Spain’s heterogeneous coastline. You et al. [26] statistically analyzed long-term water level data and found that significant astronomical tide-storm surge nonlinear interactions exist in estuaries and bays where the water depth is shallow and the tidal power is strong. Therefore, sensitivity tests can be further developed around different driving factors to consider the mechanisms of wind stress, barotropic gradient force, wave–current interaction, and astronomical tide–storm surge nonlinear interaction on the water-enhancing effects of the two types of storm surges offshore Shandong. At the moment of the maximum water increase caused by the two types of storm surges, the wave field, barometric pressure field, and wind field in the study area are also characterized by a very high concentration of energy (Figure 15), but now it is impossible to directly extract information on the amount of water increase caused by each driving factor and the contribution rate to the extreme value of water increase. Therefore, a subsequent quantitative comparison of the water increase caused by each influencing factor by strictly controlling the driving field variables is conducive to further improving the understanding of the mechanism of the two types of storm surge water increase along the Shandong coast. At the same time, the sample size of the actual tide gauge stations used to validate the storm surge model in this study was relatively small, which does not fully cover the wide range of the neighboring sea area of the Shandong Peninsula, and thus may have overlooked the details of the wave changes in a short period of time in some sections. In the future, we should focus on collecting more tide gauge stations to improve the reliability of the simulation data.

6. Conclusions

This study built a coupled storm surge modeling system for the Shandong Peninsula, successfully simulating two distinct events: the tropical cyclone Muifa (2212) and the “221003” extratropical surge. The TSSM simulations demonstrate that the total water level anomalies, surface currents, and significant wave height (HS) distributions exhibit strong dependence on typhoon trajectory, wind field morphology, and wind intensity, with characteristic right-biased energy asymmetry relative to the typhoon track. For extratropical surge scenarios (ESSM), hydrodynamic parameter variations align closely with evolving wind/pressure field configurations during cold air advection.

As a leading maritime province in China, Shandong experiences annual storm surge events, with extratropical surges constituting the predominant type in both frequency and socioeconomic impacts. This study investigated two 2022 events: the economically devastating “221003” extratropical surge and tropical cyclone Muifa’s impact. Given the rarity of direct typhoon landfalls in Shandong (only two occurrences between 2013 and 2022: Lekima (2019) and Muifa (2022)), these findings provide critical insights into tropical surge mechanisms within temperate zones, informing localized disaster mitigation strategies. Comparative analyses revealed Laizhou Bay’s heightened vulnerability to both surge types, particularly Weifang City at the bay’s apex, corroborating prior studies on coastal geomorphic susceptibility. Wave dynamics analyses demonstrated that while storm-induced water level anomalies correlated with wave radiation stress, nearshore shoaling processes significantly attenuated wave energy contributions (generally <15% of the total setup), suggesting that wave forcing plays a secondary role in surge amplification during extreme events.

Although this work elucidates the key hydrodynamic responses of the Shandong coastal system to both surge types, critical limitations persist. Future efforts should prioritize enhancing wave-tide-current coupling fidelity, resolving spectral energy propagation deficiencies in Hs simulations, and implementing controlled sensitivity experiments to quantify the individual forcing contributions. These refinements will advance the predictive capabilities of compound meteorological–hydrodynamic interactions in Shandong coastal systems.