Content Detail

• Content META

  Title	Typhoon Blend Wind Field Optimization Using Wave-Height Hindcasts
  ID	6938582
  External ID	7551267891
  Author	Kai-Cheng, Wei-Bo
  Published Date	Feb. 14, 2025, 8:51 a.m.
  Summary	Typhoons cause significant losses and pose substantial threats every year, with an increasing trend observed in recent years. This study evaluates significant wave height (SWH) hindcasts for typhoons
  Tags	None
  URL	https://www.mdpi.com/2077-1312/13/2/354
  Sentiment	Negative: 0.01706    Neutral: 0.83478    Positive: 0.14817
  Relevance	0.00000

• Content Details

  Content ID	4325351
  Key	cbc81cad-c8c8-4f6c-9511-d3e04d669711
  Content	Typhoons cause significant losses and pose substantial threats every year, with an increasing trend observed in recent years. This study evaluates significant wave height (SWH) hindcasts for typhoons affecting Taiwan using optimized wind field configurations within the SCHISM-WWM-III coupled model. To enhance typhoon-induced SWH simulations, the blended wind field integrates ERA5 reanalysis wind data with the modified Rankine vortex wind model. Key parameters, including the parametric wind field start time, best track data, and the radius of maximum wind speed, were carefully selected based on analyses of typhoons Meranti and Megi in 2016. Validation metrics such as the skill core, HH indicator, maximum SWH difference, and peak time difference of the SWH indicate that the optimized setup improves the accuracy of simulation. The findings highlight the effectiveness of the adjusted blended wind field, the high-resolution best track data provided by Taiwan, and the maximum wind speed radius in significantly enhancing the accuracy of typhoon wave modeling for the waters surrounding Taiwan. Department of Hydraulic and Ocean Engineering, National Cheng Kung University, Tainan City 701, Taiwan Tainan Hydraulics Laboratory, National Cheng Kung University, Tainan City 701, Taiwan National Science and Technology Center for Disaster Reduction, New Taipei City 23143, Taiwan Author to whom correspondence should be addressed. 1. Introduction Typhoons significantly impact coastal environments through strong winds, heavy rainfall, storm surges, and severe waves, often leading to substantial economic losses and threats to human safety. Taiwan, located in the western North Pacific, experiences annual sea surface temperatures exceeding 28 °C, creating ideal conditions for typhoon formation. Usually, each year, three to four typhoons strike Taiwan, posing substantial threats to coastal infrastructure. Therefore, designing resilient coastal infrastructure requires a comprehensive consideration of metocean data, including wind, storm surge, and wave statistics. Metocean data can be obtained through field observations, scaled physical experiments, and numerical simulations. While field observations provide precise baseline data, they are costly, time-consuming, and limited in spatial coverage. In contrast, scaled experiments and numerical simulations offer more feasible alternatives. Numerical simulations are particularly suitable for early-stage planning to assess large-scale, long-term typhoon impacts, while scaled physical experiments are better suited for detailed verification under specific conditions. By doing so, the practical application of these two methods provides robust support for infrastructure design works. Moreover, the historical numerical typhoon simulations have become increasingly accurate by validating against buoy station data. Open-source wave–current coupling models, such as SWAN+ROMS (Simulating Waves Nearshore and a Regional Ocean Modeling System), Delft3D (Delft3D-WAVE and FLOW), ADCIRC+SWAN (Advanced Circulation Model and Simulating Waves Nearshore), and SCHISM-WWM-III (Semi-implicit Cross-scale Hydroscience Integrated System Model and Wind Wave Model III), have been extensively utilized in global and regional-scale simulation studies, offering reliable tools for analyzing typhoon-induced waves and storm surges. Notably, wind speed and pressure serve as critical inputs for typhoon wave simulations and can be derived from parametric typhoon wind fields or reanalysis wind fields. The parametric typhoon wind field mathematically describes the spiral structure of typhoons, including formulations such as the Holland model, Rankine vortex model, SLOSH model, Yong and Sobey model, Emanuel model, and Emanuel and Rotunno model. The Holland [ ] uses pressure gradients and the rotational equilibrium to describe the typhoon wind field. The Rankine vortex model [ ] features a two-region structure with rigid body rotation in the inner region and inverse wind speed decay in the outer region. The SLOSH model [ ] is used for storm surge prediction, considering terrain and wind effects. The Young and Sobey model [ ] focuses on wind–wave interaction and typhoon movement asymmetry. The Emanuel model [ ] incorporates thermodynamic principles to account for environmental influences on typhoon intensity, while the Emanuel and Rotunno model [ ] refines this by adding self-organization characteristics to the wind field. These models differ in complexity and application, with the Holland and Rankine models being simpler and the Emanuel and SLOSH models accounting for more physical mechanisms. Reanalysis wind fields, as part of reanalysis dataset products, are reconstructed from multi-source observational data (e.g., satellites, weather stations, and buoys) using data assimilation techniques to represent historical climate states. Studies indicate that reanalysis wind fields generally align with observed trends for significant wave heights (SWHs) and mean periods but tend to underestimate wind-induced wave heights within the typhoon's impact zone. Meanwhile, the parametric wind field models yield better results for wave simulations within the typhoon's core but perform poorly at greater distances from the center [ ]. Therefore, improvement works are needed to address underestimations of wind speeds and SWHs near typhoon centers. Two primary approaches have been proposed for improving wind fields of typhoon centers: direct modification methods [ ] and blended wind fields [ ]. The former adjusts the reanalysis wind field near the typhoon center by calibrating the ratio between observed and reanalysis-derived maximum wind speeds, while the latter combines parametric and reanalysis wind fields through a superposition process, enhancing performance in the typhoon-affected region. Recent studies suggest that the blended wind field formula by Shao et al. [ ] is particularly suited to the waters surrounding Taiwan [ ]. Recent articles focus on specific typhoons or localized stations to analyze wind fields for typhoon simulations. This study first improves the hybrid wind field by adjusting the blending interval and applies the SCHISM-WWM-III (V5.3.1) model to evaluate ERA5 and hybrid wind fields for different typhoons under various trajectories, ultimately proposing an optimal wind field scheme for Taiwanese waters. 2. Methodology 2.1. Wave–Current Model, SCHISM-WWM-III SCHISM-WWM-III (V5.3.1) couples the SCHISM-2D, two-dimensional SCHISM [ ] ocean circulation model with the WWM-III [ ] spectral wave model. SCHISM is an open-source, unstructured grid-based model that employs a semi-implicit finite element/volume method with hydrostatic and Boussinesq approximations. WWM-III, a third-generation spectral wave model, solves the wave action balance equation using a fractional step method on unstructured grids, making the coupled system a robust tool for simulating wave-current interactions. SCHISM-2D has been widely adopted in Taiwan, particularly for typhoon storm surge hindcasting and coastal flooding simulations in vulnerable regions in recent decades [ ]. In this paper, the coefficient Manning for bottom shear stress is commonly set to 0.025 in SCHISM-2D. For WWM-III, the discrete wave frequency ranges from 0.03 to 1.0 Hz, while the full directional range of 360° is represented. Both wave frequency and direction are discretized into 36 bins. The JONSWAP spectrum is applied with a peak enhancement factor of 3.3. The wave breaking criterion is set at 0.78, and the bottom friction coefficient is specified as 0.067. The SCHISM-2D and WWM-III models share the same unstructured grid. Data exchange between the models occurs every five SCHISM-2D time steps, with SCHISM-2D operating at a 120 s time step and WWM-III running at a 600 s time step. The computational grid covers the region from longitudes 111° E to 135° E and latitudes 18° N to 30° N, ensuring sufficient coverage to capture typhoon-driven dynamics. Bathymetric data were sourced from two datasets: the global General Bathymetric Chart of the Oceans (GEBCO-2014) with a resolution of 30 arc seconds, and a higher-resolution dataset with 200 m provided by the Ocean Data Bank, National Science and Technology Council, Taiwan. Open boundary conditions for SCHISM-WWM-III were specified using tidal elevations and horizontal velocities derived from eight major tidal constituents (M2, S2, N2, K2, K1, O1, P1, and Q1) of the 1/30° global tidal model TPXO8 ( Figure 1 2.2. Wind Fields 2.2.1. Reanalysis Wind Field Reanalysis wind fields are commonly used as input conditions for typhoon hindcasting [ ]. In this study, the ERA5 dataset is adopted as the reanalysis wind field for hindcasting simulations. ERA5 is the fifth-generation global climate and weather reanalysis product by the ECMWF (European Centre for Medium-Range Weather Forecasts). Its single-level dataset provides hourly 0.25°-resolution data of various oceanic, terrestrial, and atmospheric variables from 1940 to the present. 2.2.2. Parametric Typhoon Wind Field The parametric typhoon wind field formula used in this study is a modified Rankine vortex (MRV) proposed by Phadke et al. [ ], where V V m a x represents the maximum wind speed, r r is the distance from any point to the typhoon center, R R m a x the radius of maximum wind speed, and shape parameter X X = 0.5 , according to the recommendations by Bastidas et al. [ ]. The value of V − P V m a x = 3.44 1010 − P c 0.644 is calculated as proposed by Atkinson and Holliday [ ]. P P c represents the central pressure of the typhoon. R R m a x is a key parameter in the parametric typhoon wind field. This study will later evaluate the following six empirical formulas listed in Table 1 to determine which is suitable for the study area. ϕ V f The pressure field of the parametric typhoon wind field is based on the exponential form proposed by Schloemer [ ] and Holland [ ], where P P r is the pressure at a distance r r from the center and P P n the unaffected atmospheric pressure (1013.3 hPa). The peak parameter B − P − B = 2 − P c − 900 / 160 , as proposed by Harper and Holland [ ], which typically ranges from 1 to 2.5. To match the 10 m wind speed ( V V 10 ) for the wave model, the parametric wind field is adjusted using a 0.8 conversion factor [ ] and a 0.92 gust factor to convert sustained winds to hourly mean wind speeds [ ]. Additionally, the tangential deviation angle is corrected based on Bretschneider [ ], and the asymmetry caused by V V f is adjusted as proposed by Jelesnianski [ ]. The synthetic wind speed V V p at any point is as follows: 2.2.3. Blended Wind Field Shao et al. [ ] published a blended wind field V V b by using R 2 R m a x and R 7 R m a x as boundaries to correct the underestimation of the reanalysis wind field V V a in the typhoon center region by incorporating adjustments from the parametric wind field. For r R r < 2 R m a x V V p replaces the reanalysis wind speed V V a . For r R r > 7 R m a x , the reanalysis wind speed V V a is retained. In the intermediate region R ≤ r ≤ R 2 R m a x ≤ r ≤ 7 R m a x , the wind speed is blended by the following formula: where γ − r R γ = 7 − r / R m a x / 5 is an adjustment factor. We noted that Shao et al. [ ] utilized ERA-Interim and the Holland model for their blended wind field. In this study, the higher-temporal-and-spatial-resolution ERA5 and the previously mentioned parametric typhoon wind field were incorporated instead. 2.3. Best Track and Observation Data After a typhoon dissipates, meteorological agencies from various countries compile and revise key typhoon information, such as tracks and central pressure, into best track data. These data serve as an essential source for typhoon wave predictions. However, discrepancies exist among countries due to differences in measurement instruments. In this study, best track data from the Japan Meteorological Agency (JMA) and Central Weather Administration (CWA) of Taiwan were utilized. The temporal resolution of these datasets ranges from 6 h before landfall to 3 h after landfall. In recent years, the CWA's best track data have achieved a temporal resolution of up to 1 h. Observation data were measured by buoys stations of Taiwan's CWA. The positions and information on the latitude, longitude, water depth, and data coverage period of buoy stations are given in Figure 2 and Table 2 2.4. Selected Typhoons The subsequent model tests and validations in this study primarily focus on two typhoons that affected Taiwan in September 2016: Meranti and Megi. Meranti, a super typhoon, and Megi, a moderate typhoon, both had significant impacts on wave conditions. Figure 3 and Figure 4 illustrate the tracks of Typhoon Meranti and Typhoon Megi, respectively. 2.5. Validation Metrics Willmott [ ] introduced the s k i l l s k i l l (skill index), which is more sensitive to the differences and proportional changes between observed and simulated values. According to Mentaschi et al. [ ], when the mean value of the data is relatively small, commonly used error metrics may be dominated by extreme values, leading to an inaccurate assessment of model performance. Therefore, the skill index provides a more reasonable measure for evaluating the accuracy of numerical simulations, especially in cases where data values are relatively low. Therefore, the indicator H H H H introduced by Hanna and Heinold [ ] provides a more accurate and reliable assessment of numerical simulation performance as it is not affected by low mean values. Since both the s k i l l s k i l l and H H H H are calculated based on overall error, they cannot independently account for the time when the typhoon is closest to the observation station. To evaluate whether the blended wind field effectively corrects the reanalysis wind field around peak conditions, the indices D D and D D p are used to quantify the difference between simulated and observed SWHs at peak conditions. Additionally, the time difference between the maximum simulated and observed values (denoted as ∆ t ∆ t p ) is also considered. Therefore, the indices s i l l s k i l l H H D D D D p , and ∆ t ∆ t p are used as the validation metrics for the SWH simulations. The s k i l l ranges from 1 (perfect agreement) to below 0.2 (poor performance), with thresholds of 0.65, 0.5, and 0.2 indicating very good, fair, and acceptable performance, respectively [ ]. For H H D D p , and ∆ t p , lower values indicate a better simulation performance. The formulas for these metrics are as follows: where n n is the number of data points, O i O i and i H i are the observed and simulated values, O i O i ¯ is the mean observed value, D D is the difference between the maximum simulated and observed values, and D p D p is D D as a proportion of the maximum observed value. It noted that the metrics are calculated only during the period of the maritime typhoon warning issued by the Taiwan CWA to exclude unaffected times. 3. Results and Discussion 3.1. Effects of Parametric Typhoon Wind Field Duration The blended wind field enhances the ERA5 reanalysis wind field by correcting the underestimation of wind speeds in the central region using the parametric typhoon wind field. However, the articles did not provide guidance on the optimal timing for incorporating the parametric typhoon wind field. Therefore, this section tests and discusses the generation times of different blended wind fields on wave height simulations. To create various blended wind fields, a time delay parameter δ is introduced for control, as illustrated in Figure 5 . The parametric typhoon wind field in this section is configured using the Japan Meteorological Agency (JMA) best track data, with R m a x R m a x calculated based on the empirical formula proposed by Graham and Nunn [ ]. For Typhoon Meranti, the parametric typhoon wind field start time is set at 18:00 UTC on 9 September 2016, while for Typhoon Megi, it is set at 00:00 UTC on 23 September 2016. The delay intervals are set at 12 h, but in cases with significant variations, they are refined to 6 h intervals. The comparisons of SWH of Typhoon Meranti under different parametric wind field start times and hourly observation data from the CWA are shown in Figure 6 . The Hualien, Guishan Island, and Taitung buoys, located on Taiwan's eastern coast, are directly affected by the two typhoons. The SWH comparisons for these three stations are shown in Figure 6 ; the times are different when the three stations are significantly affected by the typhoon. During the period when the typhoon approaches the stations (low delay), the SWHs show an early increase, resulting in overestimated simulation results. As the start time of the parametric typhoon wind field is further delayed, this premature increase and overestimation of SWH gradually improve. The likely reason is that the blended wind speeds in the transition region are greater than the actual wind speeds, causing abnormal waves to arrive at the stations before the typhoon itself. Conversely, excessive delays mean the typhoon center is too close to the stations, resulting in an insufficient time for wave growth and leading to underestimations, as seen with the 96 h delay, which is significantly lower than the observed values. This highlights the importance of accurately determining the start time for introducing the parametric typhoon wind field to ensure the reliability of the blended wind field and the final simulation results. In Figure 6 , it can be observed that delaying the start time by 78 or 84 h results in wave height simulations at the three stations closely matching the observed values. To determine the optimal start time for incorporating the parametric typhoon wind field into the blended wind field during Typhoon Meranti around Taiwan, comparisons were performed for all 16 buoy stations listed in Table 1 with delays of 78 and 84 h. Then, the five average validation metrics in Section 2.5 were calculated and are summarized in Table 3 . The results show that delaying by 84 h yielded slightly better overall validation metrics than delaying by 78 h. Therefore, the start time for the parametric typhoon wind field in the simulation of Typhoon Meranti is set to 0600 UTC on 23 September (84 h delay). At stations that are directly affected by typhoon-induced waves in the eastern region, significant differences are observed at each time point, particularly during the period when the typhoon approaches the stations. This results in early wave height increases, leading to overestimations. As the start time of the parametric typhoon wind field is delayed, these overestimations gradually improve. This phenomenon is likely due to the blended formula's inability to accurately represent strong typhoon wind fields or the overestimation of wind speeds in transitional regions, causing abnormal waves to reach the stations before the typhoon itself. Conversely, the final delay of 96 h results in underestimation because the start time is too close to the stations, leaving insufficient time for wave growth. The SWHs for Typhoon Megi are shown in Figure 7 . For all stations, except the final delay of 108 h (black line) and the Taitung station delay of 96 h (red line), which caused significant underestimations due to an insufficient time for wind-driven wave growth near the buoy, there are no notable differences in SWHs for other time points. Therefore, the simulation of Typhoon Megi uses 00:00 UTC 23 September 2016 as the start time for the parametric typhoon wind field. 3.2. Effects of Best Track Data Sources The key information in best track data includes the latitude, longitude, and central pressure of the typhoon, which are essential for constructing the parametric typhoon wind field. However, discrepancies may arise between best track data from different sources due to variations in measurement instruments across countries. In this section, the best track data from the JMA and Taiwan CWA are compared. The temporal resolution for both datasets is generally 6 h. However, the CWA increases the resolution to 3 h or even 1 h when a typhoon affects Taiwan, whereas the JMA only increases the resolution to 3 h for major impacts or landfall events in Taiwan. Table 4 and Table 5 list the JMA and CWA best track data for typhoons Meranti and Megi at various time points. The differences in typhoon center location between the two datasets range from 0.1° to 0.2° in latitude and longitude, and the central pressure differences can reach 15 hPa. The setup in this section includes the ERA5 reanalysis wind field, the MRV parametric typhoon wind field formula from Phadke et al. [ ], the blended wind field formula from Shao et al. [ ], the parametric wind field start times determined in Section 3.1 , and R m a x R m a x calculated using the empirical formula of Graham and Nunn [ ]. Figure 8 and Figure 9 illustrate the SWH comparison results obtained using the best track data from the JMA (red line) and CWA (black line). The comparison shows that the SWHs at Hualien, Guishan Island, and Taitung stations based on the two sets of best track data are very close. During Typhoon Meranti, slight differences in peak SWHs were observed at the Taitung station, while during Typhoon Megi, such differences occurred at the Hualien, Guishan Island, and Taitung stations. For Typhoon Meranti, the differences between the best track data from the JMA and CWA were approximately 0.1° in latitude and longitude and 5 to 10 hPa in central pressure. During Typhoon Megi, the differences were about 0.1° to 0.3° in latitude and longitude and 10 to 15 hPa in central pressure. The differences between the best track data from the JMA and CWA for Typhoon Meranti were smaller than those for Typhoon Megi. This likely explains why the discrepancies at the three stations were smaller for Typhoon Meranti. Similarly, the five mean indices for all 16 buoy stations listed in Table 1 were quantified and are summarized in Table 6 . for Typhoon Meranti, the s i l l s k i l l and H H indices were better when using the best track data from the CWA, while for Typhoon Megi, the indices were slightly better when using the JMA's data. However, the differences in skill and H H between the two datasets were not significant. In contrast, the CWA best track data consistently showed smaller errors in terms of D D and D p D p , indicating that the higher temporal resolution of the CWA's data enables more accurate simulations of significant wave height peaks. When typhoons affect Taiwan, the CWA increases the temporal resolution of its best track data from 6 h to 3 h or even to 1 h in recent years. Meanwhile, the JMA's best track data often remain at 6 h intervals if the typhoon does not make landfall in Taiwan. The best track data from the CWA are more suitable and are therefore selected for subsequent simulations in this study. 3.3. Effects of the Radius of Maximum Wind The radius of maximum wind R m a x R m a x is a critical parameter in both parametric typhoon wind fields and blended wind fields. Selecting the proper R m a x R m a x for the study area is essential for accurate typhoon wave simulation works. However, R m a x R m a x may have regional and temporal applicability limitations, necessitating testing and optimization to identify the most appropriate empirical formula for the study area. Therefore, simulations and wave height analyses were conducted using six R m a x R m a x empirical formulas (Equations (2)–(7)) mentioned in Section 2.2.2 Table 1 Figure 10 and Figure 11 illustrate SWHs with different R m a x R m a x formulas for typhoons Meranti and Megi, respectively. Wave simulation with different R m a x R m a x formulas has a noticeable impact on SWH estimations. Based on recent research, R m a x R m a x generally decreases as the typhoon intensity increases. When R m a x R m a x is smaller, the distance from the minimum wind speed near the typhoon center to the maximum wind speed at R m a x R m a x becomes shorter, leading to more abrupt wind speed changes and resulting in higher estimated SWH values [ ]. From the SWH comparison at different stations, simulations using various R m a x R m a x formulas for the same typhoon event exhibit a similar trend. For instance, in Figure 10 , the overall SWH for Typhoon Meranti at Hualien, Guishan Island, and Taitung stations is largest with Knaff 2007 (Equation (5)) and smallest with Silva 2002 (Equation (3)). Similarly, in Figure 11 , the overall SWH for Typhoon Megi at the same stations is largest with Knaff 2007 and smallest with Graham 1959 (Equation (2)). Although the formula Knaff 2007 yields peak SWH simulations close to the observed peaks at Guishan Island and Taitung stations for Typhoon Megi, the simulations at Hualien station for Typhoon Megi, as well as at Hualien, Guishan Island, and Taitung stations for Typhoon Meranti, show that the overall SWH values from Knaff 2007 significantly overestimate the observed values and the results from other R m a x R m a x formulas. Based on the validation indices across all stations ( Table 7 and Table 8 ), it can be observed that the SWH estimations using R m a x R m a x values from Knaff 2007 and Knaff 2015 (Equation (6)) show significant discrepancies in s i l l s k i l l H H , and ∆ t p ∆ t p compared to the other R m a x R m a x formulas. Table 7 indicates that for Typhoon Meranti, the overall best-performing formulas are Silva 2002 and Willoughby 2006 (Equation (4)). While Silva 2002 slightly outperforms Willoughby 2006 in s i l l s k i l l H H D D , and D p D p , Willoughby 2006 achieves the smallest ∆ t p ∆ t p . Although Graham 1959 performs comparably to Willoughby 2006, it still lags behind the top-performing Silva 2002. For Typhoon Megi, Table 8 shows that Graham 1959, Willoughby 2006, and Nederhoff 2019 yield nearly identical validation indices. Silva 2002 excels in D D and D p D p but is noticeably outperformed by Graham 1959, Willoughby 2006, and Nederhoff 2019 (Equation (7) in s i l l s k i l l H H , and ∆ t p ∆ t p . Considering the consistent performance of Willoughby 2006 for both Typhoon Meranti and Typhoon Megi, it is selected as the optimal formula for determining R m a x R m a x in subsequent typhoon wave simulations surrounding Taiwan. 3.4. Optimized Simulation From Section 3.1 Section 3.2 and Section 3.3 , the optimal configuration for the waters surrounding Taiwan includes the selected parametric typhoon wind field start time, the CWA's best track data, and the empirical formula for R m a x R m a x proposed by Willoughby et al. [ ]. These optimized parameters were used to simulate the 2016 typhoons Meranti and Megi, with the results compared against hourly observational data from buoy stations and the simulations of Hsiao et al. [ ]. Figure 12 and Figure 13 illustrate the SWH distributions under typhoons Meranti and Megi, represented by the optimized simulation (red line), Hsiao et al. [ ] (green line), and observational data (blue circles). Evidently, when compared with observational data, the SWH simulations using the optimized parameters from this study exhibit a better overall consistency. For Meranti, the results at the Hualien and Guishan Island stations are closer to the observed values than those of Hsiao et al. [ ]. While peak values at Guishan Island and Taitung stations during Megi are still underestimated, the optimized simulation shows improved agreement with observations compared to Hsiao et al. [ ]. Table 9 presents the mean metrics for the Hualien, Guishan Island, and Taitung buoy stations, highlighting that the present model significantly improves peak SWH conditions and slightly enhances the skill score. These findings suggest that the parameter optimization in this study effectively enhances the accuracy of typhoon-induced SWH predictions. 4. Conclusions This study focuses on improving the accuracy of SWH simulations induced by typhoons in the waters surrounding Taiwan and enhancing the precision of blended wind fields by optimizing the introduction timing of the parametric typhoon wind field. Using typhoons Meranti and Megi in 2016 as simulation cases, this study optimizes the selection of the parametric wind field start time, best track wind field, and R m a x R m a x formula. Simulations based on the optimized configurations demonstrate that the refined wind field setup captures the spatiotemporal characteristics of SWH more accurately in regions significantly impacted by typhoons and shows greater reliability in peak SWH simulations. The optimization process proposed in this study proves its potential application in simulating extreme weather events, particularly in scenarios where typhoons pose significant threats to infrastructure and coastal environments. This research not only provides a reference for typhoon wave studies but also offers robust support for future disaster risk assessments and infrastructural design. Author Contributions Conceptualization, S.-C.H.; Methodology, T.-C.C., K.-C.H. and W.-B.C.; Software, H.-L.W.; Validation, H.-L.W.; Formal analysis, W.-S.L., W.-B.C. and W.-S.C.; Resources, W.-B.C.; Data curation, W.-S.L.; Writing—original draft, T.-C.C. and K.-C.H.; Writing—review & editing, K.-C.H., W.-S.C. and S.-C.H.; Project administration, S.-C.H. All authors have read and agreed to the published version of the manuscript. Funding This research was financially supported by the National Science and Technology Council, Taiwan (NSTC 112-2221-E-006-143-MY3). Institutional Review Board Statement Not applicable. Informed Consent Statement Not applicable. Data Availability Statement The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author. Conflicts of Interest The authors declare no conflict of interest. References Holland, G.J. An analytic model of the wind and pressure profiles in hurricanes. Mon. Weather Rev. , 1212–1218. [ Google Scholar ] [ CrossRef ] Deppermann, S.J.; Charles, E. Notes on the Origin and Structure of Philippine Typhoons. Bull. Am. Meteorol. Soc. , 399–404. [ Google Scholar ] [ CrossRef ] Phadke, A.C.; Martino, C.D.; Cheung, K.F.; Houston, S.H. Modeling of tropical cyclone winds and waves for emergency management. Ocean Eng. , 553–578. [ Google Scholar ] [ CrossRef ] Young, I.; Sobey, R. The Numerical Prediction of Tropical Cyclone Wind-Waves ; Department of Civil & Systems Engineering, James Cook University of North Queensland: Douglas, QLD, Australia, 1981. [ Google Scholar ] Emanuel, K. Tropical cyclone energetics and structure. Atmos. Turbul. Mesoscale Meteorol. , 165–191. [ Google Scholar ] Emanuel, K.; Rotunno, R. Self-stratification of tropical cyclone outflow. Part I: Implications for storm structure. J. Atmos. Sci. , 2236–2249. [ Google Scholar ] Hersbach, H.; Bell, B.; Berrisford, P.; Hirahara, S.; Horányi, A.; Muñoz-Sabater, J.; Nicolas, J.; Peubey, C.; Radu, R.; Schepers, D.; et al. The ERA5 global reanalysis. Q. J. R. Meteorol. Soc. , 1999–2049. [ Google Scholar ] [ CrossRef ] Li, D.; Staneva, J.; Grayek, S.; Behrens, A.; Feng, J.; Yin, B. Skill assessment of an atmosphere-wave regional coupled model over the east china sea with a focus on typhoons. Atmosphere , 252. [ Google Scholar ] [ CrossRef ] Pan, Y.; Chen, Y.P.; Li, J.X.; Ding, X.L. Improvement of wind field hindcasts for tropical cyclones. Water Sci. Eng. , 58–66. [ Google Scholar ] [ CrossRef ] Li, J.; Pan, S.; Chen, Y.; Fan, Y.M.; Pan, Y. Numerical estimation of extreme waves and surges over the northwest Pacific Ocean. Ocean Eng. , 225–241. [ Google Scholar ] [ CrossRef ] Shao, Z.; Liang, B.; Li, H.; Wu, G.; Wu, Z. Blended wind fields for wave modeling of tropical cyclones in the South China Sea and East China Sea. Appl. Ocean Res. , 20–33. [ Google Scholar ] [ CrossRef ] Chen, W.B.; Chen, H.; Hsiao, S.C.; Chang, C.H.; Lin, L.Y. Wind forcing effect on hindcasting of typhoon-driven extreme waves. Ocean. Eng. , 106260. [ Google Scholar ] [ CrossRef ] Hsiao, S.C.; Chen, H.; Wu, H.L.; Chen, W.B.; Chang, C.H.; Guo, W.D.; Chen, Y.M.; Lin, L.Y. Numerical simulation of large wave heights from Super Typhoon Nepartak (2016) in the eastern waters of Taiwan. J. Mar. Sci. Eng. , 217. [ Google Scholar ] [ CrossRef ] Zhang, Y.J.; Ye, F.; Stanev, E.V.; Grashorn, S. Seamless cross-scale modeling with SCHISM. Ocean Model. , 64–81. [ Google Scholar ] [ CrossRef ] Hsu, T.W.; Ou, S.H.; Liau, J.M. Hindcasting nearshore wind waves using a FEM code for SWAN. Coast. Eng. , 177–195. [ Google Scholar ] [ CrossRef ] Roland, A. Development of the WWM II (Wind Wave Model II)—Spectral Wave Modelling on Unstructured Meshes. Ph.D. Thesis, Technische Universität Darmstadt, Darmstadt, Germany, 2009; p. 212. [ Google Scholar ] Chang, T.-Y.; Chen, H.; Fu, H.-S.; Chen, W.-B.; Yu, Y.-C.; Su, W.-R.; Lin, L.-Y. An operational high-performance forecasting system for city-scale pluvial flash floods in the southwestern plain areas of Taiwan. Water , 405. [ Google Scholar ] [ CrossRef ] Hsiao, S.-C.; Chiang, W.-S.; Jang, J.-H.; Wu, H.-L.; Lu, W.-S.; Chen, W.-B.; Wu, Y.-T. Flood risk influenced by the compound effect of storm surge and rainfall under climate change for low-lying coastal areas. Sci. Total Environ. , 144439. [ Google Scholar ] [ CrossRef ] [ PubMed ] Hsiao, S.-C.; Fu, H.-S.; Chen, W.-B.; Chang, T.-Y.; Wu, H.-L.; Liang, T.-Y. Assessment of future possible maximum flooding extent in the midwestern coastal region of Taiwan resulting from sea-level rise and land subsidence. Environ. Res. Commun. , 095007. [ Google Scholar ] [ CrossRef ] Hsiao, S.-C.; Fu, H.-S.; Wu, H.-L.; Liang, T.-Y.; Chang, C.-H.; Chen, Y.-M.; Lin, L.-Y.; Chen, W.-B. Impact assessment of sea level rise-induced high tide flooding and socioeconomic losses in a highly vulnerable coastal region. J. Hydrol. Reg. Stud. , 101921. [ Google Scholar ] [ CrossRef ] Hsiao, S.C.; Chen, H.; Chen, W.B.; Chang, C.H.; Lin, L.Y. Quantifying the contribution of nonlinear interactions to storm tide simulations during a super typhoon event. Ocean. Eng. , 106661. [ Google Scholar ] [ CrossRef ] Hsiao, S.-C.; Wu, H.-L.; Chen, W.-B. Study of the optimal grid resolution and effect of wave–wave interaction during simulation of extreme waves induced by three ensuing typhoons. J. Mar. Sci. Eng. , 653. [ Google Scholar ] [ CrossRef ] Bastidas, L.A.; Knighton, J.; Kline, S.W. Parameter sensitivity and uncertainty analysis for a storm surge and wave model. Nat. Hazards Earth Syst. Sci. , 2195–2210. [ Google Scholar ] [ CrossRef ] Atkinson, G.D.; Holliday, C.R. Tropical cyclone minimum sea level pressure/maximum sustained wind relationship for the western North Pacific. Mon. Weather Rev. , 421–427. [ Google Scholar ] [ CrossRef ] Graham, H.; Nunn, D. Meteorological Considerations Pertinent to Standard Project Hurricane. Atlantic and Gulf Coasts ; US National Hurricane Research Project Report No 33; Weather Bureau: Washington, DC, USA, 1959. [ Google Scholar ] Silva, R.; Govaere, G.; Salles, P.; Bautista, G.; Díaz, G. Oceanographic Vulnerability to Hurricanes on the Mexican Coast. Coast. Eng. Proc. , 39–51. [ Google Scholar ] [ CrossRef ] Willoughby, H.E.; Darling, R.W.R.; Rahn, M.E. Parametric representation of the primary hurricane vortex. Part II: A new family of sectionally continuous profiles. Mon. Weather Rev. , 1102–1120. [ Google Scholar ] [ CrossRef ] Knaff, J.A.; Sampson, C.R.; DeMaria, M.; Marchok, T.P.; Gross, J.M.; McAdie, C.J. Statistical tropical cyclone wind radii prediction using climatology and persistence. Weather Forecast. , 781–791. [ Google Scholar ] [ CrossRef ] Knaff, J.A.; Longmore, S.P.; DeMaria, R.T.; Molenar, D.A. Improved tropical-cyclone flight-level wind estimates using routine infrared satellite reconnaissance. J. Appl. Meteorol. Climatol. , 463–478. [ Google Scholar ] [ CrossRef ] Nederhoff, K.; Giardino, A.; Van Ormondt, M.; Vatvani, D. Estimates of tropical cyclone geometry parameters based on best-track data. Nat. Hazards Earth Syst. Sci. , 2359–2370. [ Google Scholar ] [ CrossRef ] Schloemer, R.W. Analysis and Synthesis of Hurricane Wind Patterns over Lake Okeechobee, Florida ; Weather Bureau: Washington, DC, USA, 1954; p. 49. [ Google Scholar ] Harper, B.; Holland, G. An updated parametric model of the tropical cyclone. In Proceedings of the 23rd Conferences Hurricanes and Tropical Meteorology, Dallas, TX, USA, 10–15 January 1999; pp. 10–15. [ Google Scholar ] Powell, M.D.; Black, P.G. The relationship of hurricane reconnaissance flight-level wind measurements to winds measured by NOAA's oceanic platforms. J. Wind Eng. Ind. Aerodyn. , 381–392. [ Google Scholar ] [ CrossRef ] Krayer, W.R.; Marshall, R.D. Gust factors applied to hurricane winds. Bull. Am. Meteorol. Soc. , 613–618. [ Google Scholar ] [ CrossRef ] Bretschneider, C.L. A non-dimensional stationary hurricane wave model. In Proceedings of the Offshore Technology Conference, Houston, TX, USA, 30 April–2 May 1972; p. OTC-1517-MS. [ Google Scholar ] Jelesnianski, C.P. Numerical Computations of Storm Surges Without Bottom Stress. Mon. Weather Rev. , 379–394. [ Google Scholar ] [ CrossRef ] Willmott, C.J. On the validation of models. Phys. Geogr. , 184–194. [ Google Scholar ] [ CrossRef ] Mentaschi, L.; Besio, G.; Cassola, F.; Mazzino, A. Problems in RMSE-based wave model validations. Ocean Model. , 53–58. [ Google Scholar ] [ CrossRef ] Hanna, S.R.; Heinold, D.W. Development and Application of a Simple Method for Evaluating Air Quality ; American Petroleum Institute: Washington, DC, USA, 1985. [ Google Scholar ] Chen, W.; Chen, K.; Kuang, C.; Zhu, D.Z.; He, L.; Mao, X.; Liang, H.; Song, H. Influence of sea level rise on saline water intrusion in the Yangtze River Estuary, China. Appl. Ocean Res. , 12–25. [ Google Scholar ] [ CrossRef ] Kimball, S.K.; Mulekar, M.S. A 15-year climatology of North Atlantic tropical cyclones. Part I: Size parameters. J. Clim. , 3555–3575. [ Google Scholar ] [ CrossRef ] © 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license ( https://creativecommons.org/licenses/by/4.0/
  Content Reference URL	None

• Publisher

  Publisher Name	International Journal of Molecular Sciences (#6058)
  Publisher URL	https://www.mdpi.com/
  Description	None
  Trust Flow Score	67
  Citation Flow Score	53

• Data Source

  Data Source Name	Quantexa News (#5)
  Data Source Type	news
  Data Source URL	https://app.aylien.com/dashboard
  Description	Welcome to Quantexa News API — the most powerful way to aggregate, filter, and integrate global news data into your apps and models.
  Weight	10