Abstract

With the construction of coastal infrastructure in the complex environment of nearshore straits, accurate estimation of wave conditions during the passage of tropical cyclones is essential for reliable structural design. In this study, internal and external dependence relationships between height- and period-related parameters for typhoon-induced mean wave conditions in nearshore straits are established. Field measurements and Simulating WAves Nearshore (SWAN) + ADvanced CIRCulation (ADCIRC) simulations of waves in two straits on the east coast of China under historical typhoon conditions were used. In reference to field data, the developed external relationships provide closer estimates of the lower, expected, and upper bounds of waves than do the standard-recommended and literature-proposed models. The comparison of newly proposed internal relationships with standard-recommended empirical models indicates that the use of relationships currently recommended in practice can yield up to 2% taller heights and 14–38% shorter periods in nearshore straits prone to typhoon metocean conditions. The results of this study can be used to enhance analyses of the typhoon-generated wave environment and the design of coastal infrastructure in nearshore strait areas.

1 Introduction

Nearshore waterways, such as straits, channels, and estuaries, accommodate various types of coastal transportation infrastructure [1]. Large waves generated by typhoons in the Pacific Ocean or hurricanes in the Atlantic Ocean [2] cause excessive hydrodynamic loads that threaten structural safety and significantly shorten the lifetime of such structures [3]; for example, severe damage to coastal bridges was reported during Hurricane Ivan in 2004 [4], Hurricane Katrina in 2005 [5], and Hurricane Ike in 2008 [6]. Therefore, a reliable assessment of wave conditions in nearshore waterways prone to tropical cyclones is highly important for designing typhoon-resistant coastal and sea-crossing infrastructure and mitigating the impacts of potential disasters.

In the current design practice, site-specific wave conditions corresponding to a specific range of mean return periods (MRPs), such as 1, 5, 50, and 100 years, are used to characterize storm-induced wave loads for the design and performance assessment of coastal and offshore structures [710]. To determine the correlation among different wave parameters, it is necessary to perform a rigorous statistical analysis of geographically referenced long-term measured or simulated wave data. Therefore, dependence analysis and database information are two fundamental factors in estimating typhoon-induced design wave conditions.

The data from in situ buoys are typically geographically sparse due to a lack of instrumentation and are limited in temporal coverage because of the relatively infrequent occurrence of typhoons. The data collected during measurement campaigns are normally contaminated with missing values and outliers due to a low signal-to-noise ratio, instrumental failures, or transmission errors during extreme events [11,12]. Therefore, data obtained via numerical hindcasting have become a common supplemental source for performing dependence analysis and long-term extrapolation [13].

Early definitions of the dependence between the wave height and period were characterized by either an empirical scatter table or a joint probability distribution of the relevant parameters [1315]. The former definition provides the cumulative occurrences of the wave parameters over different intervals, which is fundamental for joint probability formulation. However, detailed values for specific sea states are not always available if only an associated scatter table is provided without a conversion method [16]. The latter definition is based on both mathematical and probabilistic techniques, which are used to estimate the correlations; thus, this approach is infeasible for the early phases of design at a site with insufficient data. Consequently, this definition is generally used as a rule of thumb in real-time practical circumstances [17]. To avoid certain difficulties, the current design standards and recommended practices for coastal and offshore structures [710] stipulate regression-based internal and external parametrizations of governing variables. The internal scheme relates the bulk measures (extreme/significant/mean) within each defining characteristic (height or period), e.g., T¯Tp or HmaxHs. The external approach expresses the relationship between the bulk measures (extreme/significant/mean) of height and period, e.g., T¯Hscoeff. This provides a convenient basis for the gross estimation of the expected or variation range of design characteristics in early design stages in areas with insufficient data. Following this approach, scholars have established internal [1820] and external [2123] empirical relationships for different locations worldwide. A comparison of the recommended and proposed relationships reveals high sensitivity to geographic and metocean factors. Such variability has been shown to have a significant effect on the performance of marine structures that are dynamically sensitive, i.e., structures with natural frequencies very close to the frequencies of waves [24]. This highlights the importance of defining site-specific internal and external relationships for coastal waterways prone to typhoon metocean conditions, where dynamically sensitive coastal infrastructure, such as large and long-span sea-crossing bridges [3], are planned or currently in operation.

The objective of this study is to identify nearshore straits susceptible to typhoons by developing and validating the internal and external relationships between wave height and period. The remainder of the article is organized as follows. Section 2 introduces detailed information for two operational sites. Section 3 provides the numerical background, configuration for the studied areas, setup, and validation of the coupled SWAN (Simulating WAves Nearshore) + ADCIRC (ADvanced CIRCulation) model as well as the measured data and their postprocessing procedure. Section 4 presents the methods used to determine the internal and external relationships between bulk parameters and comparisons with the relations developed in the literature and recommended by design codes. Additionally, the validation of the developed relationships is performed, and their application in the design process is explored. The summary and conclusions are given in Sec. 5.

2 Reference and Validation Sites

Two operational nearshore straits along the east coast of China were selected as the study areas, as shown in Fig. 1. The Pingtan Strait, where a 16.32-km-long sea-crossing cable-stayed bridge connects Fuzhou city to Pingtan Island, was selected as the “reference” site to determine the relationship between the wave height and period. The Xihoumen Strait, which contains a newly constructed 2.664-km sea-crossing cable-stayed suspension bridge, was selected as the “validation” site to examine and validate the relationships developed at the reference site. The seabed at the selected straits is rugged and complex, with slopes and valleys. Both sites are surrounded by many islands and reefs, which implies a limited fetch in most directions. Figure 2 shows that the dominant fetch window at the validation site is oriented between west‒northwest and north‒northeast, whereas it is oriented between east and north‒northeast at the reference site. This makes the validation site more sheltered against remotely propagated destructive typhoon-generated waves from the western Pacific Ocean; hence, a less severe wave environment is present at the validation site than at the reference site under certain conditions (see Sec. 4.3). Overall, wave generation and propagation processes and, as a result, wave characteristics can be largely affected by the complex seabed and geographical terrain differences in the study areas.

Fig. 1
(a) Layout and topography of the studied operational nearshore straits (dashed-dotted rectangles) along the east coast of China. A zoomed-in view of the topography over the (b) reference (Pingtan Strait) and (c) validation (Xihoumen Strait) sites, with square dots denoting the location of the measurement points and solid lines indicating the approximate straight representation of the axis of the operating coastal long-span bridge at each site.
Fig. 1
(a) Layout and topography of the studied operational nearshore straits (dashed-dotted rectangles) along the east coast of China. A zoomed-in view of the topography over the (b) reference (Pingtan Strait) and (c) validation (Xihoumen Strait) sites, with square dots denoting the location of the measurement points and solid lines indicating the approximate straight representation of the axis of the operating coastal long-span bridge at each site.
Close modal
Fig. 2
Directional fetch distribution at the measurement points at each site given an imposed upper limit (typically, 250–500 km) [13,25,26]
Fig. 2
Directional fetch distribution at the measurement points at each site given an imposed upper limit (typically, 250–500 km) [13,25,26]
Close modal

3 Description of Wave Data

In this section, two sets of short-term measured and long-term simulated wave datasets are presented. The objective is to use these data to develop and validate internal and external relationships between wave height and period so that one can define typhoon-generated design mean wave conditions in nearshore straits with limited fetch and shallow water conditions on the east coast of China.

3.1 Measured Data.

The first set of data includes short-term in situ measurements taken at both sites. These data are used to establish internal relationships between height- and period-defining parameters. On the basis of the historical typhoon information provided by the Fujian Marine Forecasting Observatory2 and China National Meteorological Center,3 more than 100 typhoons have influenced the studied areas [13,25,26], 5 of which significantly affected the selected sites during the measurement campaigns. These events were Typhoon Soudelor (No. 201513) and Dujuan (No. 201522) at the Pingtan Strait (solid track lines in Fig. 3) and Typhoon In-Fa (No. 202106), Typhoon Chanthu (No. 202114), and Typhoon Muifa (No. 202212) at the Xihoumen Strait (dashed track lines in Fig. 3). The time histories of the water surface elevation were obtained from two field wave measurement campaigns that were conducted in 2015 at the Pingtan Strait [11] and between 2021 and 2022 at the Xihoumen Strait. These data were recorded hourly for 17 min with a sampling interval of 0.5 s by SBY2-1 ultrasonic wave gauges, which have spatial and temporal measurement accuracies of ±0.2 m and ±0.25 s, respectively. The time histories were preprocessed through an iterative preprocessing framework developed in a previous study [11], which included error data cleaning, baseline trend removal, and missing data reconstruction. The wave data were subsequently obtained through spectral and zero-crossing analyses. The pre- and postprocessing procedure is briefly summarized in Fig. 4.

Fig. 3
Tracks of typhoons affecting reference (Pingtan Strait) and validation (Xihoumen Strait) sites during relevant field measurement campaigns in 2015 and between 2021 and 2022
Fig. 3
Tracks of typhoons affecting reference (Pingtan Strait) and validation (Xihoumen Strait) sites during relevant field measurement campaigns in 2015 and between 2021 and 2022
Close modal
Fig. 4
Flow diagram of the steps used in pre- and postprocessing the raw water surface elevation measurements
Fig. 4
Flow diagram of the steps used in pre- and postprocessing the raw water surface elevation measurements
Close modal

3.2 Simulated Data.

The second set of data is wave data from historical typhoons simulated via a coupled wave and circulation model, SWAN + ADCIRC [27]. The simulated data encompass a total of 58 typhoons [13] that affected the Pingtan Strait from 1990 to 2018 (Fig. 5(a)) and 49 typhoons [25,26] that affected the Xihoumen Strait from 1987 to 2018 (Fig. 5(b)); notably, these events either crossed or passed through these straits. These data are used to describe the external relationship between the mean wave period and the significant wave height. Next, the theoretical background and details of the numerical simulations are presented, with a focus on the SWAN and ADCIRC modeling schemes, the coupled simulation procedure, and the model configuration.

Fig. 5
Historically simulated typhoon tracks influencing the (a) reference (Pingtan Strait) and (b) validation (Xihoumen Strait) sites
Fig. 5
Historically simulated typhoon tracks influencing the (a) reference (Pingtan Strait) and (b) validation (Xihoumen Strait) sites
Close modal

3.2.1 ADCIRC Model.

ADCIRC is a time-dependent hydrodynamic circulation model that solves the depth-integrated, nonlinear momentum, and continuity equations on an unstructured grid with the continuous-Galerkin finite element method [28]. The flexibility of the grid allows the model to resolve issues associated with complex bathymetry and topography in the process of simulating ocean water levels and currents in coastal areas. In two-dimensional mode, the governing equations in spherical coordinates are as follows:
(1)
(2)
(3)
where t is the time; ζ is the free surface elevation relative to the mean; H=ζ+d is the total water depth; d is the bathymetric water depth relative to the mean; U and V are the depth-integrated current velocity components in the west–east and south–north directions, respectively; R is the radius of the Earth; f is the Coriolis parameter; ps is the atmospheric pressure at the surface; ρ0 is the reference density of water; g is gravitational acceleration; α is the effective Earth elasticity factor; η is the Newtonian equilibrium tidal potential; vT is the depth-averaged horizontal eddy viscosity coefficient; τsλ and τsφ are the surface wind stresses in the longitudinal and latitudinal directions, respectively; and τb is the bottom friction, which is defined as follows:
(4)
where Cf is the bottom friction coefficient. More details on the theoretical background and implementations can be found in the online ADCIRC modeling manual4 and the references therein.

3.2.2 SWAN Model.

SWAN is a third-generation phase-averaged spectral wave model that simulates the evolution of wave action density, N(σ,θ;λ,φ,t), in time t, spatial space (with longitude λ and latitude φ), and spectral space (with relative frequencies σ and directions θ). This evolution is governed by the spectral action balance equation, which can be expressed in an Eulerian-based numerical framework as follows:
(5)

The left-hand side includes the kinematic terms that simulate the spatial propagation of wave action density in the speed of wave group velocity, (cg,λ,cg,φ), during which frequency (cg,σ) and direction (cg,θ) dispersion occur. The right-hand side contains the source/sink terms S(σ,θ;λ,φ,t), which simulate the total wave energy density generated by wind; dissipated due to whitecapping, bottom friction, and depth-induced breaking; and redistributed owing to nonlinear quadruplet and triad interactions between waves. This conserved form allows the integration of currents and water levels into the numerical framework, thus accounting for their effects on waves. More detailed specifications regarding the theoretical and numerical backgrounds can be found in the SWAN scientific and technical documentation [29] and the relevant literature therein.

3.2.3 SWAN + ADCIRC-Coupled Simulation Procedure.

The simulations were performed on the basis of an integrated SWAN and ADCIRC scheme [27]. Unstructured finite element mesh grids were used in both models so that information can be exchanged between these two models without interpolation, enhancing both accuracy and efficiency (see Fig. 6). The coupling in each computational interval is carried out through a three-step sequential dynamic procedure. First, the ADCIRC model is initiated and run with meteorological drivers, astronomical boundary conditions and geographical data (coastline, topography, bathymetry, etc.). Second, the water levels and currents computed with ADCIRC along with the initial inputs are passed to the SWAN model to update all the relevant physical processes in the action balance equation and obtain the wave spectra. Third, the wave radiation stress gradients calculated on the basis of the wave energy density spectra from SWAN are transferred back to ADCIRC to update and drive circulation calculations for the next computational interval.

Fig. 6
Schematic representation of the SWAN + ADCIRC-coupled simulation procedure
Fig. 6
Schematic representation of the SWAN + ADCIRC-coupled simulation procedure
Close modal

3.2.4 Coupled Model Configuration.

The drivers of the coupled model are geographical data (coastline, bathymetry, and topography), unstructured mesh grids, oceanographic boundary conditions, spatially and temporally free surface winds, and atmospheric pressures. The modeled domain for the reference site spans from 18.6°N to 30.9°N latitude and from 114.2°E to 127.5°E longitude, encompassing the whole Taiwan Strait and extending outward to the western North Pacific. The modeled domain extends from 25°N to 41°N latitude and from 115°E to 127°E longitude for the validation site, encompassing the Yellow Sea, the Bohai Sea, and the East China Sea, which opens up to the western North Pacific. The simulated domains were selected to be sufficiently large to account for the appropriate formation of remotely propagated swell waves from typhoons inside the computational domain but far from the studied straits.

The bathymetric and topographic models were configured with the NOAA ETOPO1 (1 arc-minute global relief model), which was enhanced by nearshore bathymetry data from the China Maritime Safety Administration [13] and considering a minimum water depth of 4 m [30] to improve the quality in the nearshore zone and provide a correction for tides. Flexible unstructured triangular mesh grids were employed. The resolutions of the meshes were approximately 50 m and 150 m in the shallow water coastal regions and gradually increased to approximately 32 km and 60 km toward the open ocean at the reference and validation sites, respectively. Thus, fine-resolution meshes were established in the studied straits and near the surrounding islands, accounting for the complex geometry throughout both regions that can influence wave generation, dissipation, and propagation processes in SWAN, as well as the accurate exchange of wave radiation stress gradients in ADCIRC. This approach also provides a well-balanced tradeoff between computational accuracy and convergence efficiency. The mesh grid of the modeled domain for the reference site included 110,788 elements and 56,928 nodes, and that for the reference site contained 70,144 elements and 37,064 nodes. Figure 7 shows the mesh grids for the respective simulated domains.

Fig. 7
Unstructured mesh of the computational domain for the (a) reference (Pingtan Strait) and (b) validation (Xihoumen Strait) sites
Fig. 7
Unstructured mesh of the computational domain for the (a) reference (Pingtan Strait) and (b) validation (Xihoumen Strait) sites
Close modal

The oceanographic boundary conditions were set on the basis of the most significant astronomical tide constituents (K1, K2, M2, N2, O1, P1, Q1, and S2) from the TPXO9_atlas model [31]. Meteorological driving forces were established via empirical wind and pressure fields. The empirical pressure field was set according to the Holland model [32]. The empirical wind field was defined on the basis of a combination of parametric and background wind fields to improve the parametric wind field far from the typhoon center. This hybrid wind field reflects both the central and peripheral wind field characteristics of typhoons and was obtained through a distance-based superposition method [33]. The parametric wind field was assumed to be composed of translational and gradient components. The translational component was calculated on the basis of the Miyazaki Masae model of geostrophic wind velocity [34]. The gradient component was estimated according to the modified dynamic Holland model [35] incorporating the empirical formula proposed in a previous study [36,37] for the maximum wind speed radius and the Holland model [32] to calculate the B parameter. The surface gradient wind at 10 m above sea level was converted to the wind at the top of the atmospheric boundary layer, with a boundary layer adjustment factor of 0.9 [13,25]. The key inputs for the parametric wind field were set on the basis of adjusted historical typhoon data retrieved from the China Meteorological Administration [13], and the background wind field used to enhance the quality and precision of the wind field was set on the basis of the cross-calibrated multiplatform reanalysis product.

3.2.5 Simulation Setup.

The coupled model was executed in nonstationary mode with spherical coordinates. In SWAN, the wave action densities were computed in 36 equally spaced propagation directions, 34 logarithmically spaced frequency bins between 0.04 Hz and 1 Hz, and with a time step of 10 min. The governing equation was integrated in time, spatial space, and spectral space via an advantageous implicit upwind finite difference scheme known as the backward space, backward time scheme, which is simple, robust, and economical for modeling waves in coastal regions. The solution was considered to have converged if 90% of the vertices in the modeled domain converged after a maximum of 10 iterations. The convergence criteria for terminating the iterative procedure were set on the basis of the default absolute and relative changes in the significant wave height as well as its normalized curvature of the iteration curve. Additionally, the default directional refraction limiter was used to avoid potential local instabilities from spreading throughout the simulated domain [38]. The parameterization of the wind input and whitecapping dissipation terms in the SWAN model was performed on the basis of Janssen's (WAM-Cycle4) formulation with default values for linear and exponential growth terms [3942], as this approach has been shown to minimize the combined error in the prediction of the significant wave height and zero-upcrossing wave period under cyclonic metocean conditions [4345]. The nonlinear quadruplet and triad wave‒wave interactions were computed via a fully explicit per-sweep computational scheme, the discrete interaction approximation method [46] and the lumped triad approximation method [47], respectively, with relevant default tuning values. Depth-induced wave breaking was specified via the BJ78 model with a breaker index of 0.73 and a dissipation rate proportionality coefficient of 1 [48]. The bottom friction was also included in the computations via the JONSWAP formulation [49], with a bottom friction coefficient of 0.067m2/s3, which is typically recommended for waves in storms [50,51]. The threshold depth was set equal to 0.1 m and any positive depth smaller than this value was discarded and replaced with 0.1 m. All other physical and numerical settings were set to the default values. Further details are available in the SWAN scientific and technical documentation [29]. The time step for integration was set to 10 min. In ADCIRC, the lateral eddy viscosity coefficient was set to a constant of 5m/s2. The hybrid bottom friction relationship was used to allow the bottom friction to change with respect to water depth. The relationship was formulated on the basis of the minimum bottom friction coefficient, breaking depth, a parameter that reflects the tendency of the hybrid bottom friction relationship to approach the upper or lower limit when the water depth is greater than or less than the breaking height, and a parameter that reflects the rate at which the friction factor increases with decreasing water depth. These parameters were set to 0.0022, 1 m, 10, and 1/3, respectively, in this study. Barotropic dynamics were included, but baroclinic effects have been shown to be relatively small during typhoon events [52]. Finite amplitude and advective effects were also accounted for in the simulations. The computational time step for the ADCIRC model was also set to 2 s and 1.5 s for the reference and validation sites, respectively.

3.2.6 Simulation Performance.

The established numerical models were successfully validated through comparisons with measured wind, wave, and sea surface elevation data at both sites [13,25,26]. A detailed wave-only validation of the Pingtan Strait is presented here for illustrative purposes. The computed wave parameters from the aforementioned spectral analysis were compared with concurrent hourly numerically simulated data at grid points close to the monitoring locations during Typhoons Soudelor and Dujuan. Figures 8(a) and 8(b) show that the simulation results for both typhoons agree well with the simulated data. For quantitative evaluation of the model, error indices, namely, bias and the root mean square error (RMSE), were calculated according to Eqs. (6) and (7) and the results are shown in Fig. 8.
(6)
(7)
where x and y are the simulated and measured values, respectively; x¯ and y¯ are their associated mean values; and N is the total number of data. The error indicators reveal that the mean and residual discrepancies in the simulated significant wave heights for both typhoons are small compared with the measured data. The reason for such a disparity (with the same trend but underestimated) could be inadequate calibration of the model for the simulation of typhoon-induced wind-sea and swell waves in SWAN, which has been shown to underestimate wave period statistics [53,54], particularly during the landfall period of typhoons [23,5557]. According to the model validation results, simulations based on the SWAN + ADCIRC-coupled model are thought to be reasonable and reliable for constructing a database of typhoon-induced waves for nearshore straits [58,59].
Fig. 8
Comparison of simulated and measured data within Pingtan Strait for Typhoons (a) Soudelor and (b) Dujuan. The starting time was 00:00 CST on August 6, 2015, and September 26, 2015, for Soudelor and Dujuan, respectively.
Fig. 8
Comparison of simulated and measured data within Pingtan Strait for Typhoons (a) Soudelor and (b) Dujuan. The starting time was 00:00 CST on August 6, 2015, and September 26, 2015, for Soudelor and Dujuan, respectively.
Close modal

4 Development of Relationships Between Wave Parameters

In this section, the measured and simulated data are used to develop the internal and external relationships between wave parameters. To estimate the design wave conditions, the correlation between the wave climate variables that characterize a sea state needs to be considered and modeled with an acceptable degree of accuracy [14,24]. In practice, a sea state can be characterized by various wave height- and period-related characteristic measures depending on the design purpose and the particular wave theory selected. For the wave height, the mean wave height, H¯, is defined as the representative height, whereas the significant wave height, Hs, and extreme wave height, He, are defined and used in structural analyses. For the wave period, the peak Tp, significance Ts, and mean T¯ are the three most commonly used period values in engineering practice. Therefore, the relationships defining the sea state entail not only the dependence relationships between different characteristic measures of height and period, i.e., internal relationships, but also the correlations between height and period, i.e., external relationships [19,60]. In this section, both sets of relations are developed using the measured and simulated data described in Sec. 3. For this purpose, the aforementioned wave statistics are estimated through spectral and zero-crossing analyses, as shown in Fig. 4, and formulated as follows:
(8)
(9)
(10)
(11)
(12)
(13)
where Hi is the individual wave height obtained via zero-upcrossing analysis and rearranged in descending order from the largest wave to the smallest wave, THi is the corresponding individual wave period, and mn,n{0,1} is the moment of the spectrum estimated via spectral analysis:
(14)
where S(f) is the spectral density of the wave surface elevation and f1 and f2 are the lower and upper cutoff frequencies, respectively.

4.1 Relationships Among Wave Height- and Period-Related Parameters.

The internal relationships between height- and period-defining parameters were determined through linear regression analyses of measured data. Tables 1 and 2 present the ensembled average ratios of the wave height- and period-related characteristic measures and their corresponding linear fitting coefficients and prediction confidence bounds. Table 1 indicates a strong correlation between the height parameters and the mean wave height, with the significant wave height displaying the highest correlation, as it is directly obtained from the mean wave height. A comparison of the results with the current design practice (see Table 3) reveals that wave height-related measures provide a very close estimate of actual values, with less than 2% underestimation, compared with those based on the Chinese Code of the Hydrology for Harbor and Waterway [10]. Similar slight underestimations of wave characteristic heights have also been reported in offshore and coastal areas of China, e.g., the South China Sea [61,62], Zhoushan Island [63], Hangzhou Bay [64], and Sanmen Bay [65]. In addition, a comparison with the literature in different sea areas indicated close agreement with the empirical estimates derived on the basis of the analysis of field wave data. In this case, the ratio Hs/H¯, with an overall mean and standard deviation of 1.59 and 0.06, respectively, ranges from 1.40–1.75, and the ratio He/Hs, with an overall mean and standard deviation of 1.65 and 0.26, respectively, lies mainly within the range of 1.1–2.4 [18]. Similar observations from field wave data at different locations of the Pacific and Atlantic coasts have been reported by other scholars [6669]. Compared with the theoretical estimates based on the assumption of the Rayleigh distribution of wave heights, a very close prediction of significant wave height, Hs=1.6H¯, yet fairly comparable estimates of extreme wave height, He=(1.93.9)H¯ and He=(1.22.5)Hs, were observed [20]. This discrepancy might be due to the deviation of the wave height distribution from the Rayleigh distribution in shallow waters [70,71]. On the basis of the relationships between representative wave periods, estimations are relatively higher than the prescribed values in the design practice. This may be partly due to the strong influence of swell waves on the overall wave statistics at the reference site [11,72]. Similarly, a positive and high bias for wave period characteristic measures has also been observed in other coastal areas of China, e.g., Zhoushan Island [63] and Sanmen Bay [65], whereas observations in the southern offshore areas of China [61,62] have indicated negative yet similar highly biased estimations. In addition, a comparison with field measurements at several locations with less limited fetch and water depths deeper than that at the reference site [18,22,23,73] indicated a lower range of variations or expected values for the extreme (maximum) wave period, Te=(0.541.82)T¯, and significant wave period, Ts=(0.91.4)T¯, whereas a close average value for Ts/Tp0.8 was reported by Ref. [23]. Interestingly, the maximum wave period at the reference site can be considered approximately equal to the significant wave period, i.e., Te/Ts1.1, which is somewhat consistent with the assumption of Ref. [19] that characteristic periods do not vary much for large waves. Contrary to what was observed for the wave characteristic heights, the discrepancies in the characteristic wave periods from the recommended estimations indicate relatively high sensitivity of the wave period to the complex physical processes in nearshore straits, such as land shedding, shallow water effects, wave diversion, fetch restrictions, and other factors.

Table 1

The ensemble average ratios of wave height-related statistical measures and their corresponding linear fitting coefficients

RatioVariation rangeMeanSTDaFitting coefficientbCI coefficientAdj. R-square
He/H¯1.93–3.782.680.352.63 (2.55, 2.71)1.020.92
Hs/H¯1.50–1.761.600.051.60 (1.58, 1.62)0.270.98
He/Hs1.23–2.401.670.221.64 (1.60, 1.69)0.950.93
RatioVariation rangeMeanSTDaFitting coefficientbCI coefficientAdj. R-square
He/H¯1.93–3.782.680.352.63 (2.55, 2.71)1.020.92
Hs/H¯1.50–1.761.600.051.60 (1.58, 1.62)0.270.98
He/Hs1.23–2.401.670.221.64 (1.60, 1.69)0.950.93
a

STD stands for standard deviation.

b

99% coefficient intervals of the fitting coefficient are also included.

Table 2

The ensemble average ratios of wave period-related statistical measures and their corresponding linear fitting coefficients

RatioVariation rangeMeanSTDaFitting coefficientbCI coefficientAdj. R-square
Ts/T¯1.16–1.511.330.071.33 (1.32, 1.34)0.820.96
Tp/T¯0.93–2.231.680.271.66 (1.61, 1.71)3.170.70
Te/T¯0.90–2.951.460.341.45 (1.39, 1.50)3.730.59
RatioVariation rangeMeanSTDaFitting coefficientbCI coefficientAdj. R-square
Ts/T¯1.16–1.511.330.071.33 (1.32, 1.34)0.820.96
Tp/T¯0.93–2.231.680.271.66 (1.61, 1.71)3.170.70
Te/T¯0.90–2.951.460.341.45 (1.39, 1.50)3.730.59
a

STD stands for standard deviation.

b

99% coefficient intervals of the fitting coefficient are also included.

Table 3

Comparison of fitted and JTS [10] standard prescribed values of wave height- and wave period-related measures

ValueHe/H¯Hs/H¯He/HsTs/T¯Tp/T¯Te/T¯
Fitted2.631.601.641.341.661.45
Prescribed2.661.601.661.151.211.05
Bias (%)a−1.120−1.2016.5237.1938.10
ValueHe/H¯Hs/H¯He/HsTs/T¯Tp/T¯Te/T¯
Fitted2.631.601.641.341.661.45
Prescribed2.661.601.661.151.211.05
Bias (%)a−1.120−1.2016.5237.1938.10
a

A negative value of error indicates underprediction because the bias is estimated by subtracting the fitted value from the standard prediction.

The comparisons show that the use of the prescribed internal relationships in current practice and the literature provides conservative estimates of wave conditions, i.e., taller heights and shorter periods that could increase the wave loadings on intensity-dependent and dynamically sensitive marine structures. Figures 9(a)9(c) depict a nearly linear dependence among the wave height-related parameters, and a trend that becomes more scattered for moderate to high sea states. A relatively scattered trend over the entire range of waves is also observed among the wave period-related characteristic measures in Figs. 9(d)9(f), except for the significant wave period, which is directly estimated from the mean wave period. Moreover, the standard prescribed relationships vary at the lower edge of the data, providing a lower bound for the dependence on the characteristic periods, which again highlights the conservatism of these recommendations compared with measurements. A comparison with the lower 99% confidence bounds of the fitted curves further confirms this finding. These relationships provide an average basis for estimating variables based on the values of other variables and could be essential if one has access to only certain design-defining variables, e.g., significant wave height or peak wave period are available, while an equivalent form, like extreme (maximum) wave height or extreme (maximum) wave period, is needed for the desired analysis form or design situation. This application is illustrated in Sec. 4.4. It should be noted that the duration of the measurements, the accompanying uncertainties, and the nonintercept linear fitting process chosen for simplicity may impact the degree of accuracy of the presented results.

Fig. 9
Scatter plots of (a–c) wave height-related and (d–f) wave period-related statistical characteristic values. Measurements are from the Pingtan Strait, and JTS refers to the Chinese Code of the Hydrology for Harbor and Waterways [10].
Fig. 9
Scatter plots of (a–c) wave height-related and (d–f) wave period-related statistical characteristic values. Measurements are from the Pingtan Strait, and JTS refers to the Chinese Code of the Hydrology for Harbor and Waterways [10].
Close modal

4.2 Relationship Between the Significant Wave Height and Mean Wave Period.

A conditional modeling approach [9] and linear regression analysis were used to describe the external relationship between the mean wave period T¯ and the significant wave height Hs. The conditional modeling was performed in three steps. In the first step, a scatter table of T¯ for a certain bin size of the significant wave height was created so that a suitable probability model could be fitted to the data in each bin in the second step. In terms of the probability density distribution, different models that can reasonably well define the distribution of the wave period exist [74,75]. Notably, the log-normal distribution can be used to effectively model the data in the study area. Therefore, the distribution of T¯ given Hs is defined as
(15)
where μln(T¯) and σln(T¯) are the log-normal distribution parameters, i.e., the mean and standard deviation of ln(T¯), which are calculated for different binned wave classes. These parameters can be directly related to the expected value of the wave period or its confidence limit, CI.
(16)
(17)
When defining the log-normal distribution parameters as a function of wave height, σln(T¯) often tends to be negligible at wave heights approaching extreme values [14,15]. As such, the expected value of the wave period becomes increasingly dependent on μln(T¯) and can be estimated for extreme wave conditions as follows:
(18)

Therefore, estimating the μln(T¯) on the basis of the wave height is critically important for providing a reliable wave period range for extreme conditions, i.e., large MRPs for which the stability and integrity of a structure are at great risk. In the third step, a physically interpretive dependence function is fitted to E(T¯|E[Hs]). Several functions, including a two- or three-parameter power function, fμln(T¯)(x)=c1+c2xc3, an exponential function, fμln(T¯)(x)=c1+c2ec3x, or an arctangent function, fμln(T¯)(x)=c1×arctan[c2(x+c3)], have been used to model the conditionality [9]. Although the latter expressions may provide reliable estimates for extreme wave conditions [76], the power function seems to be the most physically consistent with actual conditions [77], expressing a generic yet physically interpretive form of wave steepness, s, if the shape parameter is fixed and set as T¯=2π/gsHs, where [Hs]=m and [g]=m/s2 result in [T¯]=s. On the basis of such a physically consistent dependence function, several empirical functions of the relationship between the significant wave height and corresponding mean wave period under extreme wave conditions have been developed in the literature [2123], whereas others are recommended in current design standards, such as the DNV [8] and the Chinese Code of the Hydrology for Harbor and Waterway [10]. The relationships considered in this study are presented in Table 4. The selected cases mainly represent wave environments in regional seas the northeast Asia that are close to those in this study. The function for establishing these relations is reported mainly in the form of Ts=aHsb, where a and b are the scale and shape parameters, respectively [10,2123]. This necessitates the internal conversion between significant and mean wave periods to define the physically interpretive conditionality in the third step and estimate the expected mean wave period given a significant wave height via Eq. (18). This conversion was performed based on the internal relationships that were presented in the relevant studies, as summarized in Table 4.

Table 4

Relationships between the mean wave period and significant wave height outlined in the standards and proposed in previous studies for extreme wave conditions

Standards/authorsTs/T¯T¯=aHsb
Chinese Code of the Hydrology for Harbor and Waterway (cf., Secs. 6.2.4 and 6.4.2.1 in Ref. [10])1.153.72Hs0.5
DNV-RP-H103 (cf., Sec. 4.3.2 in Ref. [8])2.84Hs0.5
Goda (cf., Fig. 4 in Ref. [21] and Sec. 2.2.3 in Ref. [19])1.22.75Hs0.63
Suh, Kwon, and Lee (cf., Fig. 8 in Ref. [22])1.14Lower bound2.46Hs0.63
Upper bound3.33Hs0.63
Chun and Suh (cf., Fig. 9 in Ref. [23])1.12Lower bound4.42Hs0.25
Upper bound5.35Hs0.32
Standards/authorsTs/T¯T¯=aHsb
Chinese Code of the Hydrology for Harbor and Waterway (cf., Secs. 6.2.4 and 6.4.2.1 in Ref. [10])1.153.72Hs0.5
DNV-RP-H103 (cf., Sec. 4.3.2 in Ref. [8])2.84Hs0.5
Goda (cf., Fig. 4 in Ref. [21] and Sec. 2.2.3 in Ref. [19])1.22.75Hs0.63
Suh, Kwon, and Lee (cf., Fig. 8 in Ref. [22])1.14Lower bound2.46Hs0.63
Upper bound3.33Hs0.63
Chun and Suh (cf., Fig. 9 in Ref. [23])1.12Lower bound4.42Hs0.25
Upper bound5.35Hs0.32

On this basis, a comparison with the simulated data, as shown in Fig. 10, was first performed to examine the extent to which practical relationships are applicable in the case of typhoon-affected nearshore straits with similar geographical and metocean conditions in the East China Sea. Figure 10 shows that all of the proposed formulas tend to coincide with low (Hs<0.6m) and medium (0.6m<Hs<1.5m) wave height trends up to approximately Hs=2.6m and deviate from the data as the wave height increases, except for the one recommended by DNV-RP-H103 [8]. This is because these formulas were developed mostly for extreme wave conditions along coastal regions in the western Pacific Ocean, where moderately limited fetch incorporates long-period swells into locally generated wind waves. Such a combined wave system was also reported in the study area in previous works [11,72], within which high waves were associated with a narrow range of periods, whereas low (Hs<0.6m) and medium (0.6m<Hs<1.5m) waves were associated with a relatively wide range of periods.

Fig. 10
Comparison of Pingtan Strait simulated data with practice-recommended and literature-proposed external relationships. The literature-proposed and practice-recommended relationships can be found in Table 4.
Fig. 10
Comparison of Pingtan Strait simulated data with practice-recommended and literature-proposed external relationships. The literature-proposed and practice-recommended relationships can be found in Table 4.
Close modal

Figure 10 also shows that the lower bound of high waves tends to be linear, which indicates a strong correlation among the variables owing to the breaking of waves beyond a certain steepness. However, none of the recommended relationships effectively model the lower bound of the data, meaning that there might be sea states in nearshore strait areas with the same wave height but a lower wave period that can be discarded. According to Ref. [78], the same wave height with a lower wave period can induce a larger wave load on an offshore jacket-type structure. Moreover, the fundamental natural period of coastal infrastructure in nearshore straits usually ranges from 2 to 5 s, which makes the assessments of structural dynamics sensitive to the selection of wave periods. This was demonstrated in an earlier study [24], in which the use of a generic dependence relation led to the exclusion of critical sea states within the lower bound of the period and resulted in a less conservative design that was not consistent with the language used on specifications or industrial practice. Therefore, the abovementioned points highlight the importance of characterizing the upper and lower bounds of the data. Additionally, they emphasize the potential uncertainties that could lead to damage or uncertainty in the ultimate design of infrastructure in nearshore straits along East China Sea coastlines if practical and literature-based external relationships are used without adequate justifications.

After the efficiency of the standard-recommended and literature-proposed relationships was evaluated, new scale and shape parameters for the external relationship were derived on the basis of the described probabilistic approach. To this end, the logarithmic conversions of the mean wave periods were sampled within a bin size of a one-meter significant wave height (first step). The binned data were then used for probability model fitting, assuming a log-normal distribution (second step). The mean values and standard deviations within each group were calculated and fitted by a power function through the least squares method (third step). Two scenarios were considered to establish a conditional power function relationship. In the first one, the shape parameter was set to 0.5, so that the form of the function was consistent with the theory and specifications, T¯Hs, whereas in the second one, it was set as an unknown value to be estimated. This modeling approach supports further comparisons on the basis of the suitable form of the dependence function. Figure 11 shows that both scenarios provide a close estimate of the lower bound (LB), whereas using a predefined shape parameter (Fig. 11(a)) results in an upper bound (UB) that is underestimated for low (Hs<0.6m) and medium (0.6m<Hs<1.5m) sea states up to approximately Hs=2.6m and relatively poor approximations for extreme waves. Therefore, for the lower bound that is of particular concern in nearshore strait areas, one can use either of the following formulas: T¯Fitted,1LB=1.95Hs or T¯Fitted,2LB=1.74Hs0.6. For the upper bound, however, it is recommended that the result of the second scenario shown in (Fig. 11(b)), i.e., T¯Fitted,2UB=5.96Hs0.09, be considered because it displays high agreement with the measurements within all ranges of the data. From the first scenario (Fig. 11(a)), it can also be concluded that the wave steepness 2πHs/(gT¯2) is approximately limited to steepness1/23, whereas on average, steepness1:11.63, which is very close to the value recommended by DNV-RP-C205 [9].

Fig. 11
Simulated data of Pingtan Strait and corresponding best-fit bounds with (a) predefined and (b) estimated shape parameters. TFittedLB, TFittedMean, and TFittedUB refer to the fitted curves on the lower bound, mean, and upper bound of the data, respectively.
Fig. 11
Simulated data of Pingtan Strait and corresponding best-fit bounds with (a) predefined and (b) estimated shape parameters. TFittedLB, TFittedMean, and TFittedUB refer to the fitted curves on the lower bound, mean, and upper bound of the data, respectively.
Close modal

The results of the analysis of the simulated data were also compared with data obtained on the basis of the specifications and previous studies, as shown in Fig. 12. Compared with the first scenario-driven relationships, the relation in the Chinese Code of the Hydrology for Harbor and Waterway (JTS) [10] yields results that closely follow the upper bound curve, with slight underestimations, whereas the DNV-RP-H103 relation [8] slightly overestimates the expected value of this scenario and hence may be applicable for determining mean wave conditions in nearshore areas prone to typhoon metocean conditions in the East China Sea. Among the proposed formulas in the literature, the lower and upper bounds of the relationships that were developed by Ref. [22] yield close yet slightly high estimations of the expected and upper bounds in the first scenario-driven relationships. A comparison with the second scenario-driven relationships shows that the DNV-RP-H103 [8] model and lower bound of Ref. [22] model seem to yield slightly lower estimates than the expected values in the second scenario for low (Hs<0.6m) and medium (0.6m<Hs<1.5m) sea states, whereas they yield higher estimates than the expected values in the second scenario in cases with extreme sea states.

Fig. 12
Comparison of the newly fitted, literature-proposed, and practice-recommended relationships with the simulated data from the Pingtan Strait. TFittedLB, TFittedMean, and TFittedUB refer to the fitted curves on the lower bound, mean, and upper bound of the data, and subscripts 1 and 2 depict the two fitting scenarios of the power function that were used to establish the dependence relationship, i.e., T¯=aHs0.5 (predefined shape parameter) and T¯=aHsb (estimated shape parameter), respectively. The literature-proposed and practice-recommended relationships can be found in Table 4.
Fig. 12
Comparison of the newly fitted, literature-proposed, and practice-recommended relationships with the simulated data from the Pingtan Strait. TFittedLB, TFittedMean, and TFittedUB refer to the fitted curves on the lower bound, mean, and upper bound of the data, and subscripts 1 and 2 depict the two fitting scenarios of the power function that were used to establish the dependence relationship, i.e., T¯=aHs0.5 (predefined shape parameter) and T¯=aHsb (estimated shape parameter), respectively. The literature-proposed and practice-recommended relationships can be found in Table 4.
Close modal

The close agreement but slight over- or underestimation of values with the standard-recommended and literature-proposed relationships is due mainly to potential differences in depth and site-specific geography or metocean conditions. For example, the formulas presented by Ref. [23] were derived on the basis of hindcasting data along the east coast of the Korean Peninsula and the west coast of Japan, areas with greater depths and more uniform and less limited fetch conditions than those at the reference site. Similarly, the formulas of Ref. [22] were developed on the basis of hindcasting data from the Korean Peninsula area for Hs>3m, which explains the similarity between the estimations of this model and those obtained based on the first scenario-driven relationships in cases with Hs=3m. Such differences also exist in Refs. [8,10], where an averaged formula in Ref. [10] and greater depths and more uniform and less limited fetch in Ref. [8] could explain the slight under- and overestimations, respectively.

4.3 Validation of the Developed Relations.

Since the relationships presented in Secs. 4.1 and 4.2 were developed on the basis of data from the Pingtan Strait, to assess the efficiency of these relationships in similar nearshore straits, the developed internal and external relation functions were compared to measured and simulated datasets from the Xihoumen Strait. The comparisons of internal relationships are presented in Table 5, wherein, in most cases, the relationships yield positively biased yet very close estimates. The bias errors for the wave height-related parameters are generally less than ∼3%, whereas those for the wave period-related variables are less than ∼14%. This positive bias is not unexpected, as the Xihoumen Strait is much more limited in terms of fetch, as shown in Fig. 2; as a result, less severe waves than those in the Pingtan Strait are encountered in the Xihoumen Strait. A data-over-threshold analysis of the data collected at the Xihoumen Strait revealed that the change in biases for wave height-related measures is insignificant, up to 1.2%, whereas the discrepancies between wave period-related measures of observed data and developed relationships can be reduced by ∼2%–5%. This confirms the dominance of less severe waves in the Xihoumen Strait, with only up to an approximately 10% discrepancy between the actual data and the data obtained from wave period-related relationships. It should be noted that the data above various thresholds were analyzed because typhoon and nontyphoon data were unavailable at the validation site. The external relationships are compared in Fig. 13. Reasonable agreement between the simulated data from the Xihoumen Strait and the data-based external relationships was observed. Like those based on internal relationships, the results obtained on the basis of external relationships are positive yet slightly biased in terms of the lower and upper bounds, although the bias of the upper bound seems to increase in extreme sea states. These comparisons increase the confidence that the proposed relationships in Secs. 4.1 and 4.2 can reasonably reproduce typhoon-generated mean wave conditions in nearshore straits. However, one should consider the potential uncertainties that may influence the calculations, especially under extreme wave conditions, when used in the design of infrastructure in nearshore straits in the East China Sea.

Fig. 13
Comparison of the simulated data from the Xihoumen Strait and the fitted relationships to the Pingtan Strait data. TFittedLB, TFittedMean, and TFittedUB refer to the fitted curves on the lower bound, mean, and upper bound of the data, and subscripts 1 and 2 depict the two fitting scenarios of the power function that were used to establish the dependence relationship, i.e., T¯=aHs0.5 (predefined shape parameter) and T¯=aHsb (estimated shape parameter), respectively.
Fig. 13
Comparison of the simulated data from the Xihoumen Strait and the fitted relationships to the Pingtan Strait data. TFittedLB, TFittedMean, and TFittedUB refer to the fitted curves on the lower bound, mean, and upper bound of the data, and subscripts 1 and 2 depict the two fitting scenarios of the power function that were used to establish the dependence relationship, i.e., T¯=aHs0.5 (predefined shape parameter) and T¯=aHsb (estimated shape parameter), respectively.
Close modal
Table 5

Comparison of linear regression analyses of wave height- and wave period-related parameters in the Xihoumen and Pingtan straits

ValueHe/H¯Hs/H¯He/HsTs/T¯Te/T¯
Xihoumen Strait2.581.561.661.161.29
Pingtan Strait2.631.601.641.341.45
Error (%)a−1.90−2.501.22−13.43−11.03
ValueHe/H¯Hs/H¯He/HsTs/T¯Te/T¯
Xihoumen Strait2.581.561.661.161.29
Pingtan Strait2.631.601.641.341.45
Error (%)a−1.90−2.501.22−13.43−11.03
a

A negative error value indicates underprediction because the error is estimated by subtracting the predicted values from the actual values.

4.4 Engineering Application of the Developed Relationships.

After defining the relationship between the mean wave period and the significant wave height and establishing the relationships between wave height- and period-related statistical characteristic measures, a logical question is, can the newly developed formulas be applied in the design and performance assessment of infrastructure located in nearshore strait areas? In practice, different combinations of height and period definitions, depending on the regular/irregular linear/nonlinear wave theory, are used to simulate the wave field. These combinations are not all available, meaning that one can usually only access certain combinations of design-defining variables, e.g., significant wave height or peak wave period, whereas an equivalent form of this combination, e.g., extreme (maximum) wave height or extreme (maximum) wave period, might be needed for a certain analysis or design situation.

To illustrate the engineering application of the developed relationships for estimating typhoon-generated mean wave conditions in nearshore straits, an integrated framework of internal and external relationships was established, as summarized in Fig. 14 and as follows:

  • Given any wave height-related characteristic variable, such as the extreme (maximum) wave height, He, or average wave height, H¯, the internal relationships in Table 1 provide a basis for estimating the significant wave height, Hs.

  • The external relationships presented in Sec. 4.2 provide a predictive model to determine the average/mean wave period, T¯, with the upper and lower boundaries corresponding to the estimated significant wave height, Hs.

  • Given the mean wave period, other wave period-related characteristic variables, such as the extreme (maximum) wave period, Te; significant wave period, Ts; and peak wave period, Tp, can be estimated via the internal relationships in Table 2.

  • This process can be used inversely if one desires wave height-related characteristics for a specific wave period-related range.

Fig. 14
Application of the developed relationships between wave parameters (predictive models) for estimating typhoon-generated mean wave conditions in nearshore straits
Fig. 14
Application of the developed relationships between wave parameters (predictive models) for estimating typhoon-generated mean wave conditions in nearshore straits
Close modal

5 Summary and Conclusions

A relationship representing the external dependence between the mean wave period and significant wave height was established to provide the basis for estimating the mean wave conditions in nearshore straits affected by typhoons. This relationship was developed via long-term numerically simulated typhoon-generated waves in the relatively shallow waters of the Pingtan Strait on the east coast of China. The simulations included 58 typhoons that hit this site between 1990 and 2018; these events were validated in comparison to wave data from a measurement campaign that was conducted in 2015 in the Pingtan Strait. The developed relationships, including the lower and upper bounds and mean curves, were established through a probabilistic method using a physically interpretive function that represents a generic form of wave steepness: T¯=aHsb. The internal relationships between the representative wave height and wave period variables were also determined by performing a regression analysis of the wave data obtained from the aforementioned measurements. This approach provides a statistical basis within which one can use external relationships to obtain basic characteristic measures of mean wave conditions, namely, the mean wave period or significant wave height, and then exchange them with desired relevant representative variables for any design and analysis purpose via internal relationships. The developed relationships were then tested and compared to practice-recommended and literature-proposed external relationships. Moreover, the results were compared to field wave data from typhoon-affected straits and numerically simulated typhoon-generated wave data from the Xihoumen Strait. The major findings of the study are as follows:

  1. Linear regression analyses of wave height- and period-related characteristic measurements revealed that the prescribed internal relationships in current practice and in the literature provide conservative estimates of typhoon-generated mean wave conditions in the Pingtan Strait, indicating that taller heights and shorter periods could be used in design calculations if the recommended internal relationships were used.

  2. The preceding analysis of the simulated data revealed that among the empirical relationships given in specifications and the literature, the combination of the external relationships suggested by the Chinese Code JTS-145 [10] and the formula recommended by the DNV Code H103 [8] provides a close estimate of the typhoon-generated mean wave conditions in the Pingtan Strait. Similarly, the lower bound of the formulas in Ref. [22] provides a reasonable estimate of the expected extreme waves at the reference site.

  3. None of the external relationships recommended on the basis of design standards and the literature satisfactorily predict either the lower bound, T¯=1.95Hs, or the upper bound, T¯=5.96Hs0.09, of typhoon-generated wave conditions in the Pingtan Strait.

Notably, even though the newly derived relationships from the reference site agree well with the measured and simulated data at the validation site, the generality of the developed internal and external relationships is limited to the East China Sea. Therefore, the use of derived equations beyond the described environmental conditions and water depths in this study must be validated. Moreover, the presented results might be affected by statistical, modeling, and climatic uncertainties. The assumptions, limitations, and methodologies that were used in the derivation process should therefore be considered when the established relationships are used. In future studies, the applicability of these relationships to other coastal areas of China could be explored through studies of infrastructure in nearshore straits prone to tropical storms and the assimilation of numerical models.

Footnotes

Acknowledgment

This work was supported by the National Natural Science Foundation of China (Grant Nos. 52222804 and 52221002).

Conflict of Interest

There is no conflict of interest

Data Availability Statement

Some or all data and codes that support the findings of this study are available from the corresponding author upon reasonable request.

References

1.
Wei
,
K.
,
Zhou
,
C.
,
Zhang
,
M.
,
Ti
,
Z.
, and
Qin
,
S.
,
2020
, “
Review of the Hydrodynamic Challenges in the Design of Elevated Pile Cap Foundations for Sea-Crossing Bridges
,”
Adv. Bridge Eng.
,
1
(
1
), p.
21
.
2.
Zhang
,
Y.
,
Wei
,
K.
,
Shen
,
Z.
,
Bai
,
X.
,
Lu
,
X.
, and
Soares
,
C. G.
,
2020
, “
Economic Impact of Typhoon-Induced Wind Disasters on Port Operations: A Case Study of Ports in China
,”
Int. J. Dis. Risk Reduct.
,
50
(
9
), p.
101719
.
3.
Fang
,
C.
,
Tang
,
H.
, and
Li
,
Y.
,
2020
, “
Stochastic Response Assessment of Cross-Sea Bridges Under Correlated Wind and Waves via Machine Learning
,”
J. Bridge Eng.
,
25
(
6
), p.
04020025
.
4.
Huang
,
W.
, and
Xiao
,
H.
,
2009
, “
Numerical Modeling of Dynamic Wave Force Acting on Escambia Bay Bridge Deck During Hurricane Ivan
,”
J. Waterway, Port, Coast., Ocean Eng.
,
135
(
4
), pp.
164
175
. doi:10.1061/(ASCE)0733-950X(2009)135:4(164)
5.
Robertson
,
I. N.
,
Riggs
,
H. R.
,
Yim
,
S. C.
, and
Young
,
Y. L.
,
2007
, “
Lessons From Hurricane Katrina Storm Surge on Bridges and Buildings
,”
J. Waterway, Port, Coast., Ocean Eng.
,
133
(
6
), pp.
463
483
. doi:10.1061/(ASCE)0733-950X(2007)133:6(463)
6.
Murphy
,
R. R.
,
Steimle
,
E.
,
Hall
,
M.
,
Lindemuth
,
M.
,
Trejo
,
D.
,
Hurlebaus
,
S.
,
Medina-Cetina
,
Z.
, and
Slocum
,
D.
,
2009
, “
Robot-Assisted Bridge Inspection After Hurricane Ike
,”
Proceedings of 2009 IEEE International Workshop on Safety, Security & Rescue Robotics (SSRR 2009)
,
Denver, CO
,
Nov. 3–6
, pp.
1
5
. http://dx.doi.org/ 10.1109/SSRR.2009.5424144
7.
SPM
,
1977
, “Shore Protection Manual,” Coastal Engineering Research Center, US Army, 1.
8.
DNV-RP-H103
,
2011
,
Modelling and Analysis of Marine Operations
,
Det Norske Veritas Group
,
Norway
.
9.
DNV-RP-C205
,
2017
,
Environmental Conditions and Environmental Loads
,
Det Norske Veritas Group
,
Norway
.
10.
JTS-145
,
2015
,
Chinese Code of the Hydrology for Harbor and Waterway
,
Ministry of Transport of the People’s Republic of China
,
Beijing, China
.
11.
Wei
,
K.
,
Imani
,
H.
, and
Qin
,
S.
,
2021
, “
Parametric Wave Spectrum Model for Typhoon-Generated Waves Based on Field Measurements in Nearshore Strait Water
,”
ASME J. Offshore Mech. Arct. Eng.
,
143
(
5
), p.
051201
.
12.
Mazaheri
,
S.
, and
Imani
,
H.
,
2019
, “
Evaluation and Modification of JONSWAP Spectral Parameters in the Persian Gulf Considering Offshore Wave Characteristics Under Storm Conditions
,”
Ocean Dyn.
,
69
(
5
), pp.
615
639
.
13.
Wei
,
K.
,
Shen
,
Z.
,
Ti
,
Z.
, and
Qin
,
S.
,
2021
, “
Trivariate Joint Probability Model of Typhoon-Induced Wind, Wave and Their Time Lag Based on the Numerical Simulation of Historical Typhoons
,”
Stoch. Environ. Res. Risk Assessm.
,
35
(
2
), pp.
325
344
.
14.
Haghayeghi
,
Z. S.
,
Imani
,
H.
, and
Karimirad
,
M.
,
2020
, “
Directional Dependence of Extreme Metocean Conditions for Analysis and Design of Marine Structures
,”
Appl. Ocean Res.
,
100
(
7
), p.
102181
.
15.
Imani
,
H.
,
Abbaspour
,
M.
,
Tabeshpour
,
M. R.
, and
Karimirad
,
M.
,
2020
, “
Effects of Motion and Structural Vibration–Induced Loadings on the Coupled Dynamic Response of a Mono-Column Tension-Leg-Platform Floating Wind Turbine
,”
Proc. Inst. Mech. Eng., Part M: J. Eng. Mar. Environ.
,
234
(
2
), pp.
426
445
.
16.
Longuet-Higgins
,
M. S.
,
1983
, “
On the Joint Distribution of Wave Periods and Amplitudes in a Random Wave Field
,”
Proc. R. Soc. London. Ser., A, Math. Phys. Sci.
,
389
(
1797
), pp.
241
258
.
17.
Feng
,
X.
,
Li
,
H.
,
Feng
,
X.
,
Zhao
,
J.
, and
Feng
,
W.
,
2021
, “
Dependence of Ocean Wave Return Levels on Water Depth and Sampling Length: A Focus on the South Yellow Sea
,”
Ocean Eng.
,
219
(
1
), p.
108295
.
18.
Goda
,
Y.
,
1974
, “
Investigation of the Statistical Properties of Sea Waves With Field and Simulation Data
,”
Rept. Port Harbour Res. Inst.
,
13
(
1
), pp.
3
37
. https://cir.nii.ac.jp/crid/1572261550733569536
19.
Goda
,
Y.
,
2010
,
Random Seas and Design of Maritime Structures
, 3rd ed., Vol.
33
,
World Scientific Publishing Company
.
20.
Longuet-Higgins
,
M. S.
,
1953
, “
On the Statistical Distribution of the Heights of Sea Waves
,”
J. Mar. Res.
,
11
(
3
), pp.
245
266
. https://elischolar.library.yale.edu/journal_of_marine_research/774
21.
Goda
,
Y.
,
2003
, “
Revisiting Wilson’s Formulas for Simplified Wind-Wave Prediction
,”
J. Waterway, Port, Coast., Ocean Eng.
,
129
(
2
), pp.
93
95
. doi:10.1061/(ASCE)0733-950X(2003)129:2(93)
22.
Suh
,
K.-D.
,
Kwon
,
H.-D.
, and
Lee
,
D.-Y.
,
2010
, “
Some Statistical Characteristics of Large Deepwater Waves Around the Korean Peninsula
,”
Coast. Eng.
,
57
(
4
), pp.
375
384
.
23.
Chun
,
H.
, and
Suh
,
K.-D.
,
2018
, “
Estimation of Significant Wave Period From Wave Spectrum
,”
Ocean Eng.
,
163
(
17
), pp.
609
616
.
24.
Valamanesh
,
V.
,
Myers
,
A. T.
,
Arwade
,
S. R.
, and
Hajjar
,
J. F.
,
2014
, “
The Impact of Peak Spectral Period in the Design of Offshore Wind Turbines for the Extreme Sea State
,”
Struct. Congr.
,
2014
(
1
), pp.
1684
1693
.
25.
Wei
,
K.
,
Shang
,
D.
, and
Zhong
,
X.
,
2023
, “
Integrated Approach for Estimating Extreme Hydrodynamic Loads on Elevated Pile Cap Foundation Using Environmental Contour of Simulated Typhoon Wave, Current, and Surge Conditions
,”
ASME J. Offshore Mech. Arct. Eng.
,
145
(
2
), p.
021702
.
26.
Zhong
,
X.
,
Wei
,
K.
, and
Shang
,
D.
,
2023
, “
An Improved Azimuth-Dependent Holland Model for Typhoons Along the Zhejiang Coast Prior to Landfall Based on WRF–ARW Simulations
,”
Nat. Hazards
,
117
(
3
), pp.
2325
2346
.
27.
Dietrich
,
J. C.
,
Zijlema
,
M.
,
Westerink
,
J. J.
,
Holthuijsen
,
L. H.
,
Dawson
,
C.
,
Luettich
,
R. A.
,
Jensen
,
R. E.
,
Smith
,
J. M.
,
Stelling
,
G. S.
, and
Stone
,
G. W.
,
2011
, “
Modeling Hurricane Waves and Storm Surge Using Integrally-Coupled, Scalable Computations
,”
Coast. Eng.
,
58
(
1
), pp.
45
65
.
28.
Luettich
,
R. A.
, and
Westerink
,
J. J.
,
2004
,
Formulation and Numerical Implementation of the 2D/3D ADCIRC Finite Element Model Version 44. XX
,
R. Luettich
,
Chapel Hill, NC, USA
.
29.
SWAN team
,
2019
,
SWAN Scientific and Technical Documentation
,
Delft University of Technology
,
Netherlands
. http://swanmodel.sourceforge.net/online_doc/swantech/swantech.html.
30.
Yu
,
X.
,
Pan
,
W.
,
Zheng
,
X.
,
Zhou
,
S.
, and
Tao
,
X.
,
2017
, “
Effects of Wave-Current Interaction on Storm Surge in the Taiwan Strait: Insights From Typhoon Morakot
,”
Contin. Shelf Res.
,
146
(
15
), pp.
47
57
.
31.
Egbert
,
G. D.
, and
Erofeeva
,
S. Y.
,
2002
, “
Efficient Inverse Modeling of Barotropic Ocean Tides
,”
J. Atmos. Ocean. Technol.
,
19
(
2
), pp.
183
204
. doi:10.1175/1520-0426(2002)019<0183:EIMOBO>2.0.CO;2
32.
Holland
,
G. J.
,
1980
, “
An Analytic Model of the Wind and Pressure Profiles in Hurricanes
,”
Monthly Weather Rev.
,
108
(
8
), pp.
1212
1218
. doi:10.1175/1520-0493(1980)108<1212:AAMOTW>2.0.CO;2
33.
Pan
,
Y.
,
Chen
,
Y.
,
Li
,
J.
, and
Ding
,
X.
,
2016
, “
Improvement of Wind Field Hindcasts for Tropical Cyclones
,”
Water Sci. Eng.
,
9
(
1
), pp.
58
66
.
34.
Jakobsen
,
F.
, and
Madsen
,
H.
,
2004
, “
Comparison and Further Development of Parametric Tropical Cyclone Models for Storm Surge Modelling
,”
J. Wind Eng. Ind. Aerodyn.
,
92
(
5
), pp.
375
391
.
35.
Fleming
,
J. G.
,
Fulcher
,
C. W.
,
Luettich
,
R. A.
,
Estrade
,
B. D.
,
Allen
,
G. D.
, and
Winer
,
H. S.
,
2008
, “
A Real Time Storm Surge Forecasting System Using ADCIRC
,”
Estuarine Coast. Model.
,
2007
(
1
), pp.
893
912
.
36.
Zhao
,
X.
,
Yao
,
Y.
,
Huang
,
S.
,
Xie
,
Y.
, and
Lou
,
H.
,
2009
, “
Forecast of Super Typhoon “Saomai” and “Weipa”
,”
Mar. Forecast
,
26
(
1
), pp.
19
28
.
37.
Zhou
,
T.
,
Tan
,
Y.
,
Chu
,
A.
, and
Zhang
,
C.
,
2018
, “
Integrated Model for Astronomic Tide and Storm Surge Induced by Typhoon for Ningbo Coast
,”
Proceedings of the 28th International Ocean and Polar Engineering Conference
,
Sapporo, Japan
,
June 10–15
, p. ISOPE-I-18-265.
38.
Dietrich
,
J. C.
,
Zijlema
,
M.
,
Allier
,
P.-E.
,
Holthuijsen
,
L. H.
,
Booij
,
N.
,
Meixner
,
J. D.
,
Proft
,
J.
,
Dawson
,
C. N.
,
Bender
,
C. J.
, and
Naimaster
,
A.
,
2013
, “
Limiters for Spectral Propagation Velocities in SWAN
,”
Ocean Model.
,
70
(
10
), pp.
85
102
.
39.
Günther
,
H.
,
Hasselmann
,
S.
, and
Janssen
,
P. A.
,
1992
, “The WAM Model Cycle 4 (Revised Version),” Deutsch. Klim. Rechenzentrum, Techn. Rep. No. 4, Hamburg, Germany.
40.
Cavaleri
,
L.
, and
Rizzoli
,
P. M.
,
1981
, “
Wind Wave Prediction in Shallow Water: Theory and Applications
,”
J. Geophys. Res. Oceans
,
86
(
C11
), pp.
10961
10973
.
41.
Janssen
,
P. A. E. M.
,
1989
, “
Wave-Induced Stress and the Drag of Air Flow Over Sea Waves
,”
J. Phys. Oceanogr.
,
19
(
6
), pp.
745
754
. doi:10.1175/1520-0485(1989)019<0745:WISATD>2.0.CO;2
42.
Janssen
,
P. A. E. M.
,
1991
, “
Quasi-Linear Theory of Wind-Wave Generation Applied to Wave Forecasting
,”
J. Phys. Oceanogr.
,
21
(
11
), pp.
1631
1642
. doi:10.1175/1520-0485(1991)021<1631:QLTOWW>2.0.CO;2
43.
Rahimian
,
M.
,
Beyramzadeh
,
M.
,
Siadatmousavi
,
S. M.
, and
Allahdadi
,
M. N.
,
2023
, “
Simulating Meteorological and Water Wave Characteristics of Cyclone Shaheen
,”
Atmosphere
,
14
(
3
), p.
533
.
44.
Jiang
,
Y.
,
Rong
,
Z.
,
Li
,
P.
,
Qin
,
T.
,
Yu
,
X.
,
Chi
,
Y.
, and
Gao
,
Z.
,
2022
, “
Modeling Waves Over the Changjiang River Estuary Using a High-Resolution Unstructured SWAN Model
,”
Ocean Model.
,
173
(
5
), p.
102007
.
45.
Siadatmousavi
,
S. M.
,
Jose
,
F.
, and
Stone
,
G. W.
,
2011
, “
Evaluation of two WAM White Capping Parameterizations Using Parallel Unstructured SWAN With Application to the Northern Gulf of Mexico, USA
,”
Appl. Ocean Res.
,
33
(
1
), pp.
23
30
.
46.
Hasselmann
,
S.
,
Hasselmann
,
K.
,
Allender
,
J. H.
, and
Barnett
,
T. P.
,
1985
, “
Computations and Parameterizations of the Nonlinear Energy Transfer in a Gravity-Wave Specturm. Part II: Parameterizations of the Nonlinear Energy Transfer for Application in Wave Models
,”
J. Phys. Oceanogr.
,
15
(
11
), pp.
1378
1391
. doi:10.1175/1520-0485(1985)015<1378:CAPOTN>2.0.CO;2
47.
Eldeberky
,
Y.
,
1996
, “
Nonlinear Transformation of Wave Spectra in the Nearshore Zone
,”
Ph.D. thesis
,
Delft University of Technology, Department of Civil Engineering
,
Netherlands
.
48.
Battjes
,
J. A.
, and
Janssen
,
J. P. F. M.
,
1978
, “
Energy Loss and Set-Up Due to Breaking of Random Waves
,”
Coast. Eng.
,
6
(
1
), pp.
569
587
.
49.
Hasselmann
,
K.
,
Barnett
,
T. P.
,
Bouws
,
E.
,
Carlson
,
H.
,
Cartwright
,
D. E.
,
Enke
,
K.
,
Ewing
,
J.
,
Gienapp
,
A.
,
Hasselmann
,
D.
, and
Kruseman
,
P.
,
1973
, “
Measurements of Wind-Wave Growth and Swell Decay During the Joint North Sea Wave Project (JONSWAP)
,”
Ergänzungsheft zur Deutschen Hydrographischen Zeitschrift, Reihe A
,
Nr. 12
. https://hdl.handle.net/21.11116/0000-0007-DD3C-E
50.
Bouws
,
E.
, and
Komen
,
G. J.
,
1983
, “
On the Balance Between Growth and Dissipation in an Extreme Depth-Limited Wind-Sea in the Southern North Sea
,”
J. Phys. Oceanogr.
,
13
(
9
), pp.
1653
1658
. doi:10.1175/1520-0485(1983)013<1653:OTBBGA>2.0.CO;2
51.
Zijlema
,
M.
,
van Vledder
,
G. P.
, and
Holthuijsen
,
L. H.
,
2012
, “
Bottom Friction and Wind Drag for Wave Models
,”
Coast. Eng.
,
65
(
7
), pp.
19
26
.
52.
Liu
,
W.-C.
, and
Huang
,
W.-C.
,
2019
, “
Influences of Sea Level Rise on Tides and Storm Surges Around the Taiwan Coast
,”
Contin. Shelf Res.
,
173
(
2
), pp.
56
72
.
53.
Baoshu
,
Y.
, and
Dezhou
,
Y.
,
2004
, “
Application Study of Shallow Water Wave Model (SWAN) in Bohai Sea
,”
Proceedings of the Twelfth Workshop on Ocean Models (WOM-12)
,
Dalian, China
, China National Marine Environment Forecasting Center State Oceanic Administration.
54.
Rogers
,
W. E.
,
Hwang
,
P. A.
, and
Wang
,
D. W.
,
2003
, “
Investigation of Wave Growth and Decay in the SWAN Model: Three Regional-Scale Applications
,”
J. Phys. Oceanogr.
,
33
(
2
), pp.
366
389
. doi:10.1175/1520-0485(2003)033<0366:IOWGAD>2.0.CO;2
55.
Chun
,
H.
, and
Suh
,
K.-D.
,
2019
, “
Empirical Formulas for Estimating Maximum Wave Height and Period in Numerical Wave Hindcasting Model
,”
Ocean Eng.
,
193
, p.
106608
.
56.
Ou
,
S.-H.
,
Liau
,
J.-M.
,
Hsu
,
T.-W.
, and
Tzang
,
S.-Y.
,
2002
, “
Simulating Typhoon Waves by SWAN Wave Model in Coastal Waters of Taiwan
,”
Ocean Eng.
,
29
(
8
), pp.
947
971
.
57.
Wang
,
X.
,
Yao
,
C.
,
Gao
,
G.
,
Jiang
,
H.
,
Xu
,
D.
,
Chen
,
G.
, and
Zhang
,
Z.
,
2020
, “
Simulating Tropical Cyclone Waves in the East China Sea With an Event-Based, Parametric-Adjusted Model
,”
J. Oceanogr.
,
76
(
6
), pp.
439
457
.
58.
Li
,
J.
,
Pan
,
S.
,
Chen
,
Y.
,
Fan
,
Y.-M.
, and
Pan
,
Y.
,
2018
, “
Numerical Estimation of Extreme Waves and Surges Over the Northwest Pacific Ocean
,”
Ocean Eng.
,
153
(
7
), pp.
225
241
.
59.
Imani
,
H.
, and
Kamranzad
,
B.
,
2015
, “
Wave Data Assimilation in Chabahar Bay
,”
6th International Offshore Industries Conference
,
Tehran, Iran
,
May 4
. https://civilica.com/doc/482625/
60.
Wei
,
K.
,
Arwade
,
S. R.
,
Myers
,
A. T.
,
Hallowell
,
S.
,
Hajjar
,
J. F.
,
Hines
,
E. M.
, and
Pang
,
W.
,
2016
, “
Toward Performance-Based Evaluation for Offshore Wind Turbine Jacket Support Structures
,”
Renew. Energy
,
97
(
13
), pp.
709
721
.
61.
Liu
,
T.
,
Yu
,
J.
,
Yu
,
Y.
,
Huang
,
H.
,
Zhang
,
X.
, and
Zhou
,
B.
,
2023
, “
Statistical Characteristics of Sea Surface Waves Derived From Buoy Measurements in the Eastern Guangdong Offshore Water
,”
J. Sea Res.
,
195
(
5
), p.
102440
.
62.
Yang
,
B.
,
Feng
,
W.
, and
Zhang
,
Y.
,
2014
, “
Wave Characteristics at the South Part of the Radial Sand Ridges of the Southern Yellow Sea
,”
China Ocean Eng.
,
28
(
3
), pp.
317
330
.
63.
Yang
,
B.
,
Shi
,
W.
,
Ye
,
Q.
,
Zhang
,
Z.
,
Yang
,
W.
, and
Song
,
Z.
,
2017
, “
Characteristics of Waves in Coastal Waters of Northeast Zhoushan Island During Typhoons
,”
Adv. Water Sci.
,
28
(
1
), pp.
106
115
.
64.
Yang
,
B.
,
Yang
,
Y.
,
Yang
,
Z.
,
Feng
,
X.
,
Ye
,
Q.
,
Yu
,
L.
, and
Ou
,
J.
,
2021
, “
Wave Characteristics in a Semi-Enclosed Offshore Windfarm Influenced by East Asian Monsoon and Extreme Weather: A Study of Central Hangzhou Bay, China
,”
Ocean Dyn.
,
71
(
10
), pp.
1011
1031
.
65.
Zhou
,
Y.
,
Ye
,
Q.
,
Shi
,
W.
,
Yang
,
B.
,
Song
,
Z.
, and
Yan
,
D.
,
2020
, “
Wave Characteristics in the Nearshore Waters of Sanmen Bay
,”
Appl. Ocean Res.
,
101
(
8
), p.
102236
.
66.
Munk
,
W. H.
,
1944
, “
Proposed Uniform Procedure for Observing Waves and Interpreting Instrument Records
,”
Scripps Inst. Oceanogr., Wave Project Rep.
,
26
(
1
), p.
22
. https://www.google.com/books/edition/Proposed_Uniform_Procedure_for_Observing/ExQjNQEACAAJ?hl=en
67.
Putz
,
R. R.
,
1950
,
Wave-height Variability: Prediction of Distribution Function
, Vol.
318
,
University of California, Institute of Engineering Research, Fluid Mechanics Laboratory
.
68.
Seiwell
,
H.
,
1948
,
Results of Research on Surface Waves of the Western North Atlantic
, Vol.
X, No. 4
,
MIT and Woods Hole Oceanographic Institution
, pp.
1
56
.
69.
Wiegel
,
R. L.
,
1949
, “
An Analysis of Data From Wave Recorders on the Pacific Coast of the United States
,”
Eos, Trans. Am. Geophys. Union
,
30
(
5
), pp.
700
704
.
70.
Rattanapitikon
,
W.
, and
Shibayama
,
T.
,
2007
, “
Estimation of Shallow Water Representative Wave Heights
,”
Coast. Eng. J.
,
49
(
3
), pp.
291
310
.
71.
Ko
,
D.-H.
,
Lee
,
U.-J.
,
Cho
,
H.
, and
Lee
,
B. W.
,
2021
, “
Estimation of Relations Between Representative Wave Heights Using Measured Wave Data for the West Sea of Korea
,”
J. Coast. Res.
,
37
(
8
), pp.
51
55
.
72.
Yan
,
J.
,
Tao
,
A.
,
Lin
,
Y.
,
Pei
,
Y.
,
Liu
,
Y.
, and
Su
,
J.
,
2020
, “
Swell Source Analysis of East China Sea Under the Influence of Typical Typhoon Scenarios
,”
China Ocean Eng.
,
34
(
2
), pp.
210
222
.
73.
Takahashi
,
T.
,
Hirose
,
M.
, and
Sasaki
,
T.
,
1979
, “Relationship Between Spectral Width Parameters and Wave Statistics,” Technical Note of the Port and Harbour Research Institute, Ministry of Transport, Japan, (No.315), pp.
3
65
.
74.
Huang
,
W.
, and
Dong
,
S.
,
2019
, “
Probability Distribution of Wave Periods in Combined Sea States With Finite Mixture Models
,”
Appl. Ocean Res.
,
92
(
11
), p.
101938
.
75.
Lucas
,
C.
, and
Soares
,
C. G.
,
2015
, “
Bivariate Distributions of Significant Wave Height and Mean Wave Period of Combined Sea States
,”
Ocean Eng.
,
106
(
14
), pp.
341
353
.
76.
Drago
,
M.
,
Ciuffardi
,
T.
, and
Giovanetti
,
G.
,
2011
, “
Estimators of Wave Peak Period for Large Return Period Seastates
,”
Proceedings of the ASME 2011 30th International Conference on Ocean, Offshore and Arctic Engineering, Volume 6: Ocean Engineering
,
Rotterdam, The Netherlands
,
June 19–24
, pp.
899
907
.
77.
Haselsteiner
,
A. F.
,
Sander
,
A.
,
Ohlendorf
,
J.-H.
, and
Thoben
,
K.-D.
,
2020
, “
Global Hierarchical Models for Wind and Wave Contours: Physical Interpretations of the Dependence Functions
,”
Proceedings of the ASME 2020 39th International Conference on Ocean, Offshore and Arctic Engineering, Volume 2A: Structures, Safety, and Reliability
,
Virtual, Online
,
Aug. 3–7
.
78.
Wei
,
K.
,
Myers
,
A. T.
, and
Arwade
,
S. R.
,
2017
, “
Dynamic Effects in the Response of Offshore Wind Turbines Supported by Jackets Under Wave Loading
,”
Eng. Struct.
,
142
(
13
), pp.
36
45
.