The Ma’an Archipelago, located in the northeastern region of Shengsi County, including Huaniao Island, Bixiashan Island, Lvhua Island, Gouqi Island and Shengshan Island ( Figure 2B ). The majority of existing suspended mussel farms are distributed along the island coastline, primarily situated in the northwest, southwest, and southeast waters. ( Figure 2C ). This location serves three purposes: first, the water depth is appropriate for aquaculture; second, it is handy for fisherman to salvage work; and third, the island will also act as a shelter during typhoons. In order to comprehensively assess the risk of typhoon-induced waves in this area, the locations are classified shown in Figure 2C as follows: current mussel farms (F1 to F5), undeveloped farming areas (S6 to S8) and the marginal points of the farms for assessing risk (R9 to R11). It is also vital to classify the typhoons impacting Gouqi Island. Because the rotating wind fields caused by the typhoons on different tracks are different, resulting in distinct local wind directions.

The data of typhoon tracks in the northwest Pacific Ocean from 2000 to 2022 is downloaded from the China Meteorological Administration Tropical Cyclone Data Center (https://tcdata.typhoon.org.cn/zjljsjj.html) (Ying et al., 2014; Lu et al., 2021), and the Tropical Cyclone Best Track Dataset of China Meteorological Administration on Central Meteorological Observatory Typhoon Network (http://typhoon.nmc.cn/web.html).

In this study, typhoons that pass within 300 km of Gouqi Island between 2000 and 2022 are categorized into three groups: (1) West sweep type (TPW), (2) Intermediate sweep type (TPM) and (3) East sweep type (TPE). The first type of typhoon track enters within 300 km of the center of the island, with the center of the typhoon situated on its western flank. The second is the typhoon center entering within a 100 km range of Gouqi Island and subsequently passed over Gouqi Island, before moving directly northward. The third type is within 300 km of Gouqi Island, and the center of the typhoon is located on the east side of the island.

Between 2000 and 2022, there were recorded instances of 16 typhoons that approached within a distance of 100 km, and 44 that entered at a range of 300 km without nearing a distance of 100 km. For each classification track, three typhoons were selected. The following are the primary selection criteria for these typhoons: (1) the typhoon center entered within 300 km. (2) The typhoon is moving across Gouqi Island or its surrounding, greatly affecting the island. (3) Given that TPW may make landfall more readily than the other two types, the wind speed of the typhoon center has to be greater than 17.20 m/s. (4) For TPM and TPE, the wind speed of typhoon center has to be greater than 32.70 m/s. Figure 3 illustrates the classification outcomes and the chosen typhoon tracks.

In the present study, wave measurements are observed by the Acoustic Doppler Current Profiler (ADCP, RDI whs-300k) installed on buoys. The wave buoy MF05003 is located at W1 (122°36′53.28″, 31°44′7.8″), near the Yangtze River estuary (see Figure 2A ). The observation time of the buoy MF05003 is from 8 June 2022 to 14 June 2023, output once every 30 minutes, receiving rate reached 99.5%.

The wind parameters of the data source were validated using wave buoy observations. The SWAN model runed from 1 June 2022 to the end of the observation period. Model results were evaluated by means of the skill score (SS) (Warner et al., 2005) and correlation coefficient (CC) (Allen et al., 2007), calculated by the following formula ( Equations 3, 4):

where N represents the number of samples of time series, X is the variable and X ¯ is the time average. X_model and X_obs are the predicted and observed value respectively, x m ¯ and x o ¯ are the average of the variables. Model performance levels evaluated by SS and CC are categorized as: >0.65 excellent; 0.65-0.5 very good; 0.5–0.2 good;< 0.2 poor (Maréchal, 2004; Allen et al., 2007).

The validation results of the wind data used to drive the wave model are presented in Figure 4 and Table 1 . The verification period covers from June to October, which corresponds to the cycle of frequent typhoons. With a few differences in wind direction, the wind speed of ERA5 dataset compares the observed local wind data well, and the overall pattern remains consistent. The comparison results of SS and CC at field station W1 were all greater than 0.65 ( Table 1 ). It is evidence that the ERA5 dataset compares very well with observed local wind speed and direction, indicating that it is can be used as the driving force dataset of the SWAN model.

The model results of the significant wave height and the average wave period were validated using the observed data of wave buoy. Comparison results of the observed and simulated data of the average wave period and significant wave height at station W1 was presented in Figure 5 . Assessment results of model performance were shown in Table 1 . Because three typhoons affected the East China Sea in September—Typhoon Hinnamnor (2022) on September 5, Typhoon Muifa (2022) on September 14, and Typhoon Nanmadol (2022) on September 18—there were three peaks in the wave height ( Figure 5A ).

SS of significant wave height at station W1 in September 2022 and October 2022 were 0.96 and 0.96, respectively, indicating model performance is good. The CC were 0.96 and 0.93, respectively, indicating model performs excellent on the simulation of significant wave height. The model performs well in simulating the impact of typhoons on significant wave height. The maximum difference between the simulated and observed wave period was about two seconds ( Figures 5B, D ). The evaluation results of wave period at station W1 was greater than 0.5, ranking “good”. In general, the SWAN model reproduced the temporal variation of significant wave height and period well during typhoon period, and better captured the wave eruption and dissipation processes, indicating that it could be used to study the characteristics of typhoon-induced wave around Gouqi Island (Jiang et al., 2023).

In order to understand the relationships between local wind direction and wave field during typhoon, this study simulated the wave field caused by typhoons entering the 300 km range of Gouqi Island from 2000 to 2022, a total of 60 typhoons ( Figure 3B ). The statistical results of local wind and wave during 60 typhoons in different stations of mussel farms are shown in Figure 6 .

Results showed that local wind greater than 10 m/s during typhoons mainly occurred in the NE, E, SE, and S, more than 15 m/s in the NE, E and SE, and wind speeds of 15 to 22 m/s occur most frequently ( Figure 6 ). The majority of wind speeds over 23 m/s are located in the NE-E-SE direction, while the northwest is also home to the greatest wind speed of 27 m/s. It is evidence that waves generated by winds are easier influence the southwest and southeast farming area of Gouqi Island.

It should be noted that most of the reanalysis data products are generated after a series of algorithmic corrections using satellite monitoring data and measured data. It is often difficult for satellites to obtain accurate data for typhoon centers that move at high speeds. Therefore, the ERA5 wind field data used as an input file does not reflect the maximum instantaneous wind speed during typhoons, this results in an underestimation of the typhoon disaster by the simulation process (Zhang et al., 2003; Hersbach et al., 2020).

During typhoons, F1 and F2, which are to the north of the island, both have southwest waves ( Figure 6 ). For F1, the south waves are also higher, and it can reach to about 6 m. F3, F4, and F5 are mainly affected by waves from the southeast, among which F4 is the riskiest farm, and the wave height in the southeast direction can reach more than 8m. The wave heights of R9, R10 and R11 at the edge of the mussel farm are much higher than those of F1, and the occurrence frequency of 5~6 m waves increase to more than 10%. This suggests that the waves will bypass the island and that further west or north of the farming area will be at greater risk. For S6, S7 and S8, the wave height of S8 can reach up to 6~7 m, and the wave comes from the southeast. Under the protection of the islands on both sides of S6 and S7, the waves mainly come from the southwest during the typhoon, and the frequency of 4~5 m wave height is less than 10%.

Wind fetch, wind speed, and duration of winds are the primary factors influencing the properties of the wave field. Under the constant wind fetch, wave heights increase with longer time of wind durations and higher wind speeds. Most of the local winds affecting Gouqi Island come from the northwest, north, northeast, east, southeast and south, but the frequency of wave direction occurred distribution in the aquaculture area is different from the wind direction, which is mainly due to the temporal and spatial differences of the local wind in the island and mussel farms.

Among the three typhoon track types mentioned above (TPW, TPM and TPE), three typical typhoons were selected from each category to simulate the wave field. A typical typhoon is selected from each category for analysis here, including Matsa, Chan-hom and Muifa. The spatial distributions of global and local wave fields before, during, and after the passage of three typical typhoons are shown in Figure 7 . The time series of significant wave heights at different locations (F1 to F5 and S6 to S8) during typhoons in mussel farming areas are shown in Figure 8 .

In terms of type TPW (Matsa), the center of the typhoon is always located to the west of Gouqi Island ( Figure 7A1-A3 ), and the local wind direction in the study area changes from east to southeast ( Figure 7A4-A6 ). The wave heights at station F1 and F2 are much smaller due to the blockage of the island. The center of TPW typhoon is still moving northward and is getting weaker. Eventually, the winds will gradually turn southward.

Wind direction of TPM (Chan-hom) changes from NE to ESE when the center of the typhoon is still at low latitudes. The local wind direction stayed to the southeast with the center of the TPM reaching Gouqi Island, and at this time the wave height in the sea had reached 8m. The center of the typhoon passed through island and continued to move northward, under the influence of the rotating wind field, the local wind direction changed to WSW, due to the lagging nature of the wave field, the waves are still coming from the southeast. F3, F4 and F5 are easily damaged by waves come from southeastern ( Figure 7B ).

The center of TPE (Muifa) is situated to the east of the island, and the local wind direction changed from NNE to NNW and finally to NW ( Figure 7C ). Wave direction changes from SE to NE due to swell domination. With changes of the wave direction, the significant wave height at F4 is the largest initially, while growing progressively greater at F1 and F2.

The time series of wave height changes during typhoons in different location of the aquaculture area are also different ( Figure 8 ). TPW (Matsa) has moved to 25.60°N on August 5, with a maximum wind speed of 45 m/s, and has developed into a severe typhoon. At this time, Gouqi Island was already affected by the outer edge of the Matsa, and the swell of farming area was less than 3m, and the wind direction was east. Then Matsa made landfall in Zhejiang and weakened into a typhoon, with a maximum wind speed of 35 m/s. When it moved to the same latitude with Gouqi Island, the intensity of Matsa have weakened into a tropical storm, with the maximum wind speed of 23 m/s and local wind direction is southeast ( Figure 8A ). Meanwhile, station F4 has the biggest significant wave height, about 7 m, due to there is no sheltering of the island and waves coming from the southeast. Slightly lower than F4 and similar to F5, F3 is situated southwest of Gouqi Island. Nevertheless, diffracted waves have the ability to increase its wave height to about 7 m ( Figure 8B ). Similar patterns and wave heights in F1, S6, S7, and S8 suggest that new mussel farms can be considered. Since F2 is shielded by the island, it always has the lowest wave height.

For TPM (Chan-hom), wave heights of F1 to S8 continue to increase, due to the impact of northeast wind. The wave height of F4 is always greater than that of other stations during Chan-hom, with the maximum value can reach 7.5 m/s. As the typhoon eye approaches Gouqi Island, the wind is coming from the east and the waves are propagating from the southeast, reaching a maximum wave height of F3, F4 and F5 in the absence of island protection ( Figure 8D ). The wave heights of F1, S6, S7 and S8 are comparable, whereas F2 remains the smallest.

There is a noticeable temporal variation in the wave height fluctuation during the passage of TPE type (Muifa). The wave heights of F3 and F4 increase first when the typhoon center is lower than the latitude of Gouqi Island. As the typhoon center moves northward, the wave direction shifts from southeast to northeast, the wave height of the northern culture area rises, and F3 and F4 start to decrease ( Figure 8F ). In conclusion, the TPW and TPM exhibit their maximum wave heights when the center of the typhoon is situated at a low latitude. While the peak wave height for TPE type occurs when the center is located at an identical latitude.

In view of the complexity of the typhoon path, in addition to analyzing the characteristics of the actual typhoon waves, in order to more comprehensively evaluate the spatial distribution characteristics of typhoon waves in mussel farms, we chosen the typical wind speed of 25 m/s and simulated wave fields under the driving of 16 kinds of wind directions. Overall, the wave field surrounding Gouqi Island is dominated by the N, NE, ESE, and S wave directions without considering the blocking effect of islands ( Figure 9 ). In the extreme situation, the wave heights of F1 and F2 are smaller, while the farms located in the southwest and southeast of the island are higher, even experiencing huge waves of 14m. The highest significant wave height will reach its maximum value in 48 hours, based on 15 days of winds at 25 m/s. However, not many typhoons are able to stay in one place for longer than 12 hours in reality, therefore it is important to be aware of typhoons with unusual paths and extended lives.

Modeled results of significant wave height under extreme wind conditions at different stations of mussel farms are shown in Figure 10 . There is a maximum wave height about 9 m of F1 and F2 under the wind of NNE direction ( Figure 10 ). F1 and F2 have the minimum wave height when the wind direction is southeast, due to the island can block the wind and waves. In particular, because F1 is situated to the west of F2, waves from the southeast are avoided, resulting in the wave height of F1 being higher than F2 overall. As a result, the F1 wave height minimum happens when the wind direction is ESE, while the F2 wave height minimum is occurred during SSE wind.

F3 is completely protected by Gouqi Island and has the lowest wave height values when there is a northerly wind. However, waves will diffract from the middle of Gouqi and ShengShan islands, resulting in even higher wave height values for F5, with the minimum value occurring in the direction of the wind as NE. F4 has the highest wave height about 12 m for wind direction SE. Westerly wind experiences the smallest wave height, due to the sheltering of islands. The highest wave height for the undeveloped farming area (S6, S7, and S8) occurs at S6 in the case of northeast wind and at S8 in the case of east wind. S7 is situated between S6 and S8, and it has the smallest wave height due to the protection of the islands.

There is no invariable wind field and symmetrical wave field during the movement of typhoon center, the relationship between winds and waves is related to wind duration, wind fetch, topography, and the strength of the subtropical high pressure. Typhoons are characterized by an unequal distribution of air pressure, and powerful wind fetch are naturally distributed unevenly. In general, wave and wind directions are basically the same in the right front quadrant of the typhoon forward direction (from east to west or northwest), and the relative wind speed is higher. This is due to the proximity to the western Pacific subtropical high pressure. The wind field constantly affects waves in this quadrant, where energy is continuously accumulated, wave height greatly increases. However, there is a time lag in the response of waves to wind in the left quadrant of the typhoon path. The area here have short wind fetch and are dominated by swells, the wave significant height decreases. The asymmetry of wave field is more obvious than wind field (Moon et al., 2004; Liu et al., 2007; Zhou et al., 2008).

In addition to the influence of wind field, the intensity and energy of typhoon are also important factors that determine the asymmetry of wave field. The TPW type typhoon is not discussed here because its characteristics on the sea surface after landfall are atypical. Right Front (RF), Right Behind (RB), Left Behind (LB) and Left Front (LF) represent the NE, SE, SW and NW of the typhoon center, respectively. Taking Chan-hom as an example, the effects of three different typhoon intensities: super typhoon, strong typhoon and typhoon are discussed. While Muifa is used to illustrate the impact of typhoon energy on the same typhoon intensity. The decrease in energy due to the loss of continuous heat supply as the center of the typhoon moves further north.

It is evident that the wave height initially increases, then decreases as the distance from the typhoon center increases ( Figure 11 ). Additionally, there is a gradual decrease in wave height with the reduction of typhoon intensity in both RF and LF. Typhoon-induced waves are always obstructed and diffracted by islands in the LF quadrant, leading to the manifestation of the lowest values (blue pentagram in Figure 11 ). The asymmetry of the typhoon and the short wind fetch on the left side result in complex variations in RB and LB. The significant wave height in the RB quadrant exhibits a tendency of first reducing and then growing when the intensity of the typhoon is relatively high ( Figures 11A, D ). When the intensity of typhoon is relatively low, the wave height typically increases before declining ( Figures 11C, F ). The unstructured grid points used for calculation are scattered at both ends of the selected line section, which is located at the edge of the typhoon wind circle, resulting in a zigzag-shaped line section ( Figure 11E ).

The variation in wave height within the LB quadrant closely parallels that observed in the RB. There is a notable trend in wave height changes, this is characterized by RF>LF>LB>RB when the intensity of the typhoon is high. Due to the northward shift of the typhoon center and the impact of topography, the wave heights follow the pattern: RF>LF>RB>LB. As the center of the typhoon moved northward, the energy is no longer able to consistently maintain heat supply, leading to the wave height of RB exceeding that of RF. The findings indicate a significant correlation between the asymmetry of the wave field and both the intensity and energy of typhoons.

The asymmetry and hysteresis of the waves result in various types of waves dominating Gouqi Island at distinct times. The correlation coefficients, calculated for wind and wave parameters across the entire typhoon transit time series, do not adequately elucidate the changing processes of dominant wind waves and swells. Divide the time series into three distinct periods: T1(Before), T2(During) and T3(After) ( Figure 12B ). The correlation coefficients between the winds and waves for the three time periods are shown in Table 2 .

For TPW type, the wind and wave direction on the right side of the track basically coincides with the movement direction of the typhoon ( Figures 12A, B ). During the T1 period, the latitude of the typhoon center was lower than that of Gouqi Island. The dominated wave at this time was primarily swells. The correlation between wind direction and wave direction within the farms is suboptimal, minimum is 0.07. Wind waves are predominant when winds persist in the same direction during T2 and T3. It is worth noting that F2 is protected by the Gouqi island and impacted by a mixed waves dominated by swells.

In the case of TPM type typhoons, swells predominantly govern the wave structure on Gouqi Island. As the typhoon center traverses Gouqi Island, the local wind direction on the island undergoes a sudden change due to the influence of the rotating wind field ( Figures 12C, D ). Variations in wind fields lead to a decrease in the duration of wind hours and a reduction in the size of wind fetch. Similarly, the Gouqi Island area is predominantly influenced by swells during the passage of TPE type typhoons. The correlation coefficient of farm F3 is only -0.36 during T2. As the wave direction alters, the angle between the winds and waves directions progressively reduces ( Figures 12E, F ). Concurrently, the correlation coefficient exhibits a corresponding increase at T3. The results underscore the pronounced lag characteristics observed in waves.

Generally, the sea area of Gouqi Island is predominantly impacted by swells originating from the northeast (NE), east (E), and southeast (SE) due to the hysteresis of the wave field. The swells of shallow water concentrate the energy, leading to an increase in wave height. Therefore, greater emphasis should be placed on the farm F1 effected by diffracted waves and the farm F4 devoid of island protection. However, this study merely provides a qualitative analysis. For a more quantitative assessment of wind waves and swells, it is imperative to employ specific methodologies in the further study (Hwang et al., 2012; Zheng et al., 2019).

In according with the technical directives for risk assessment and zoning of marine disaster—Part 2: Ocean wave (Ministry of Natural Resources of the People’s Republic of China, 2022), the ocean wave intensity levels in offshore waters are categorized into four levels. This classification is based on the significant wave height at each grid point: extremely strong, strong, relatively strong and general, corresponding to levels I (H_s ≥ 14.0 m), II (9.0 m ≤ H_s < 14.0 m), III (6.0 m ≤ H_s < 9.0 m) and IV (4.0 m ≤ H_s < 6.0 m) respectively.

The variability in wave field characteristics across individual typhoons introduces uncertainty and contingency to the assessment of typhoon-induced wave risk levels within the Gouqi Island area. Consequently, the evaluation will be conducted with respect to both extreme conditions and actual typhoons. Under extreme situations, the 16 wind directions manifest with varying weights during typhoons. Therefore, the weights of the 16 wind directions were determined based on the frequency of wind direction with a wind speed exceeding 20 m/s in section 4.1 during typhoons from 2000 to 2022, the results of weights are 0.05, 0.07, 0.12, 0.14, 0.12, 0.14, 0.01, 0.00, 0.00, 0.00, 0.00, 0.03, 0.00, 0.12, 0.10, 0.10, respectively (from 0° to 337.5°). The risk levels are assessed by significant wave heights of each grid point, are determined through the follow formula ( Equation 5 ):

where HS represents the significant wave height for each grid point; N is the total number of the wind directions; MHS_n is the maximum wave height of the nth wind direction and W_n is the weight of the nth direction for each grid. The wave risk map in extreme cases with different wind direction frequencies of study area is constructed using a 0.005°×0.005° matrix diagram ( Figure 13A ).

Figures 13B–D display the average value of maximum significant wave height caused by TPW (Matsa, LEKIMA and Hagupit), TPM (Prapiroon, Chan-hom and MITAG) and TPE (Muifa, LINGLING and Hinnamnor) types typhoon, respectively. The method of average calculation entails computing the mean of the maximum significant wave height for each typhoon within the designated track categories (Tian and Zhang, 2021; Jiang et al., 2023). This methodology mitigates the influence of individual variations stemming from the unpredictable paths of typhoons. Simultaneously, it elucidates the overarching distribution and interrelation of various typhoon paths, emphasizing their shared spatial characteristics. The matrix enables a more visual depiction of the typhoon-induced wave distribution and risk level at large-scale suspended mussel farms. ( Equation 6 ) shows how to calculate the significant wave height

for each category, HS_i is the maximum significant wave height of matrix; h_i is the maximum value of each typhoon; k is the total number of the typhoons; i represent the number of mesh points in matrix.

In extreme cases ( Figure 13A ), the risk level is high and the most severe wave height can approach more than 9 m, defined as Level II. The risk levels in the western section of the F1 farm, as well as F3, F4, and F5 located in the southern of islands, are significantly high. This lack of island blocking increases both the risk and wave heights. The topography of the island protects the F2, resulting in a lower risk level. For TPW, TPM and TPE types, the area surrounding Gouqi Island controlled by swells from southeastern. F4 presents the highest risk similarly. The landfall of TPW type attenuates wind speed and reduces risk compared to other types. In contrast, TPM and TPE do not make landfall, often maintaining high wind speeds that significantly impact the farms. Consequently, it is crucial to prioritize typhoon forecasting and early warning systems for TPM and TPE.

In summary, under the influence of extreme conditions and three types of typhoon paths, the risk level will reach Level II. Given that the data source utilized is an average dataset for reanalysis, it is important to note that the measurements of wind speed and wave height may be underestimated. Therefore, in the event of a super typhoon or with long circling time, the risk level may reach Level I. The risk of proposed development area is comparable to the current farms F1 and F2. Furthermore, the depth of water in this region is conducive to large-scale cultivation. In the western of the proposed marine area, water depths approach 40 m, presenting opportunities for deep-water net cage culture. In the event of a typhoon, it is imperative to focus on safeguarding the farms situated in the southern and northwestern parts of the islands.