1. Introduction
On March 11, 2011, the tsunami caused by the 9.0-magnitude Great East Japan Earthquake struck the Fukushima Daiichi Nuclear Power Plant, highlighting the inundation risk of nuclear power plants (NPPs) worldwide. The approximately 14 m tsunami inundated the NPP site and cut off all alternating current power sources. This caused the loss of the cooling function and core meltdown, leading to a major accident of level 7, which is the highest level based on the International Nuclear and Radiological Event Scale (INES) (Anzai et al., 2011). This accident resulted in the emission of large amounts of radioactive materials (e.g., xenon-133 and cesium-137) into the atmosphere (Stohl et al., 2012), and complete recovery has yet to be achieved despite enormous recovery costs and 14 years of effort. This case shows that thorough risk assessment is required for inundation disasters in NPPs located in coastal areas.
In particular, in the case of coastal NPPs that can be exposed to both tsunamis and storm surges, comprehensive and quantitative risk analysis considering various marine disasters is essential (Mori et al., 2014). South Korea has 24 nuclear reactors in operation. Among them, 18 units are concentrated on the east coast, resulting in high geographical density (Min et al., 2023). The storm surges caused by extreme weather phenomena (e.g., typhoons) may cause serious damage to coastal NPP sites (Liu et al., 2012). The risk is escalating particularly owing to the frequency and intensity of extreme weather events increased by climate change (Vousdoukas et al., 2018). Accordingly, it is necessary to quantitatively assess the inundation risk at NPP sites on the east coast and prepare corresponding response strategies.
Mei and Xie (2016) analyzed typhoons that occurred in the northwest Pacific from 1977 to 2014, and reported that the intensity of typhoons that landed in East Asia and Southeast Asia increased by 12% to 15% and the proportion of strong typhoons in categories 4 to 5 increased by two to three times. They mentioned that this is closely related to the rise in sea surface temperature, and warned that this tendency is likely to continue in the future. Yoon and Shim (2013) precisely simulated flood inundation using the finite volume coastal ocean model (FVCOM). They revealed that additional sea level rises of approximately 0.97 and 0.84 m may occur at the Masan and Kori NPP sites, respectively, during the invasion of Typhoon Maemi (0314) under a virtual scenario that applied extreme tides. These results indicate that storm surge hazards are gradually increasing across the Korean Peninsula, and that they may have a significant influence on NPPs located on the east coast.
To assess the potential risk of storm surges, Lin et al. (2010) calculated the extreme storm surge height in New York, USA based on probabilistic synthetic typhoons. They reported that the surge height is affected by a combination of the intensity, maximum wind speed radius, and path of the typhoon as well as its relative location with respect to the target region. Särkkä et al. (2024) calculated the maximum storm surge height (MSSH) on the Baltic coast through a simulation that utilized a synthetic low-pressure system, and revealed that the surge height may vary depending on the tropical cyclone path and speed despite the same tropical cyclone intensity. Lee et al. (2021) predicted the surge height for three actual hurricanes and 1,031 synthetic tropical cyclones using machine learning, and revealed that the surge height increases as the landing angle is more perpendicular to the coastline and the movement speed decreases. They warned that large surge heights can be induced in a wide area even by a typhoon that does not land when its path is parallel to the coastline, and that the risk can be underestimated when it is judged simply based on landing. Shi et al. (2025) revealed that slight differences in typhoon path significantly affect the arrival location, time, and intensity of storm surges. Gharehtoragh and Johnson (2024) reported that even the same typhoon may have different surge heights and spatial distributions depending on environmental conditions through a synthetic typhoon simulation that reflected climate change and topographic changes. Jin et al. (2024) and Kim et al. (2025a) combined the past typhoons that affected the Korean Peninsula and analyzed the influences of the landing time and path on the storm surge height using the advanced circulation (ADCIRC) model. They also analyzed the volumetric characteristics of time waveforms for identifying the degree of overtopping and overflow of storm surges.
Kim et al. (2025b) analyzed storm surge heights that may occur at the Hanbit NPP site according to the key parameters of typhoons (central pressure, maximum wind speed, and the radius of maximum wind) and the closest approach distance through a virtual typhoon simulation that utilized the tropical cyclone risk model (TCRM). The analysis results showed that the storm surge height is affected by a combination of the closest approach distance of the typhoon and the topographic characteristics around the NPP site as well as the intensity of the typhoon.
This study aimed to analyze the effects of typhoon path characteristics on storm surges for NPP sites located on the Korea’s eastern coast. To this end, probabilistic synthetic typhoons were generated using the TCRM by applying the path and intensity information of the past typhoons, and a storm surge simulation was performed using the ADCIRC model (Luettich et al., 1992). The storm surge characteristics according to the key parameters of typhoons (central pressure, maximum wind speed, the radius of maximum wind, and closest approach distance) were then analyzed using indicators for evaluating the sustainability of storm surges beyond the conventional method of evaluating only the MSSH to discuss potential risks, including the degree of overflow and inundation.
2. Numerical Method
2.1 Overview of Synthetic Typhoon Scenarios
The synthetic typhoons used in this study were based on the International Best Track Archive for Climate Stewardship (IBTrACS) data provided by NOAA. These data are the records of typhoons that occurred in the northwest Pacific from 1848 to 2020, and the information used includes the locations, movement paths (latitude and longitude), travel speeds, and formation and dissipation times of typhoons. The area of the synthetic typhoon simulation in the TCRM was set to latitude 0° to 55° and longitude 100° to 170°. Approximately 3,000 virtual typhoons that entered this area were generated as shown in Fig. 1. Based on northern latitude 30° and longitude 120° to 135°, 235 typhoons that may affect the Korean Peninsula were selected. Among them, 37 virtual typhoons that enter the East Sea through the Korea Strait were determined as the final scenarios. For the selected scenarios, a numerical simulation of storm surges was performed using the ADCIRC model. The synthetic typhoon catalog constructed in this study assumed a return period of approximately 1,000 years. Accordingly, the occurrence characteristics for a 100-year frequency were calculated. Further details on the algorithm and probabilistic synthesis process of the TCRM can be found in Kim et al. (2025b).
The 37 virtual typhoons exhibit various characteristics, including the central pressure ranging from 885 to 986 hPa, the maximum wind speed from 26.24 to 77.17 m/s, and the radius of maximum wind from 9.26 to 51.86 km under the maximum intensity. In addition, their closest approach distance to major NPP sites on the east coast was 0.95 to 550.78 km for the Kori NPP, 0.12 to 551.61 km for the Saeul NPP, 2.16 to 561.39 km for the Weolseong NPP, and 6.07 to 579.71 km for the Hanul NPP.
To examine the validity and reproducibility of the ADCIRC model used in this study, a simulation was performed for Typhoon Maemi (0314), one of the most powerful typhoons in history. The typhoon information was acquired from the Joint Typhoon Warning Center (JTWC, n.d.), and its path is shown in Fig. 1.
2.2 Storm Surge Simulation Using the ADCIRC Model
ADCIRC is a finite-element-method-based numerical model suitable for analyzing the changes in water levels, tides, and storm surges in coastal and continental shelf areas, and it has been widely used for predicting coastal disasters. In this study, the sea level rise effect was calculated by entering the meteorological field of each virtual typhoon, and the maximum height and time-series data of storm surges were obtained for each NPP site on the Korea’s eastern coast.
For precise numerical simulation, the computational domain and grid system were constructed based on the latest nautical chart and bathymetric data as shown in Fig. 2. The entire computational domain includes the major waters of the northwest Pacific as well as the entire waters of the Korean Peninsula (including the East Sea, South Sea, and West Sea) as it is expanded to Japan (mainland excluding Hokkaido), Taiwan, the east coast of China, and the area surrounding the Ogasawara Islands.
It is constructed using unstructured grids, and low-resolution grids were used in the open sea area. High-resolution grids at the intervals of approximately 10 m were applied to the coastal areas near the Kori, Saeul, Weolseong, and Hanul NPP sites, which are the areas of interest, to precisely simulate the coastal reflection and local amplification effects caused by the propagation of storm surges. The constructed computational grids had 398,909 and 216,241 elements, respectively, which allowed stable storm surge simulation.
2.3 Validation of the ADCIRC Model
To examine the reliability and reproducibility of the ADCIRC model, a storm surge simulation was performed for Typhoon Maemi (0314), which caused significant damage to the Korean Peninsula in 2003. Typhoon Maemi (0314) landed on the south coast of Korea and moved toward the East Sea, resulting in storm surge damage across the south and east coasts. For the comparison and verification of storm surges, the observation data from 11 tidal stations (5 stations on the south coast, 3 stations on the east coast, 2 stations in Jeju, and 1 station in Ulleungdo) shown in Fig. 3 were utilized.
Fig. 4 compares the observed MSSH with the numerical simulation results. A high correlation was observed, as the coefficient of determination (R2) was 0.864. Underestimated or overestimated simulation results were observed from some points, but the overall consistent trend between the observed and simulated values was well reproduced. This provides a sufficient foundation for ensuring the reliability of the virtual-typhoon-based storm surge simulation performed in this study.
3. Numerical Results
3.1 Spatial Distribution of Maximum Storm Surge Height
Fig. 5 shows the spatial distribution of the MSSH on the Korea’s eastern coast. Three representative typhoon cases were presented: (a) typhoon No. 29, which moved in close proximity to the Kori, Saeul, and Weolseong NPPs, (b) typhoon No. 23, which landed on the south coast and moved toward the East Sea, and (c) typhoon No. 22, which passed through the Korea Strait while maintaining a certain distance and moved northeast. Table 1 presents the meteorological parameters (central pressure (pc), maximum wind speed (Vmax), and the radius of maximum wind (Rmax)) and the closest approach distance (Dmin) when these typhoons are closest to each NPP site.
Typhoon No. 29 shown in Fig. 5(a) traveled along a path very close to the Kori, Saeul, and Weolseong NPP sites and caused large storm surge heights in nearby waters. It had a significant impact on the Weolseong NPP as well. Typhoon No. 23 shown in Fig. 5(b) had a similar intensity, but it rapidly entered the East Sea after landing on the south coast. It exhibited a large storm surge height near the landing point. Although this typhoon included the Kori, Saeul, and Weolseong NPPs within the dangerous semicircle, the impacts of storm surges did not directly reach these sites owing to topographic effects. This typhoon rather traveled to the East Sea through a path closer to the Hanul NPP, causing significant storm surges near the site. In the case of typhoon No. 22 with the highest intensity (Fig. 5(c)), it affected storm surges at the Kori and Saeul NPP sites while moving northeast along the Korea Strait. However, as its path changed toward the northeast, its impact on the Weolseong NPP relatively decreased and the Hanul NPP was minimally affected. This indicates that the path and closest approach distance of a typhoon as well as its key meteorological parameters can be the key indicators for the risk of storm surges.
This analysis confirms that the storm surge height distribution and storm surge hazards are significantly affected by the travel path of the typhoon and its closest approach distance to NPP sites as well as its meteorological parameters, such as the central pressure, maximum wind speed, and the radius of maximum wind.
3.2 Storm Surge Waveform
Fig. 6 shows the storm surge waveforms during the invasion of virtual typhoons No. 6, No. 22, and No. 29 for the Kori NPP site as a representative case. As evident from Table 1, typhoon No. 29 had a path closest to the Kori NPP, and Dmin is 12.49 km. Typhoons No. 22 and No. 6 followed gradually separated paths with the Dmin values of 47.83 and 95.12 km, respectively. Among the three typhoons, No. 22 had the highest intensity, whereas No. 6 was the weakest typhoon. To compare the surge response characteristics of the storm surge waveforms, the time axis was reset and the waveforms were arranged based on the time point of the MSSH (tp). Accordingly, the effects of the intensity and path of each typhoon on storm surge waveforms could be quantitatively compared and analyzed.
Typhoon No. 29 was closest to the Kori NPP and caused a sharp increase in water level within a short period of time while recording the largest MSSH as well. This shows that the wave height rapidly increases and the time to reach its peak decreases as the closest approach distance decreases. In contrast, the strongest typhoon No. 22 had a somewhat separated path, causing a relatively gradual increase in water level and relatively long duration of the storm surge. This indicates that even a strong typhoon can have a limited increase in storm surge height if it has no direct impact on the site. Meanwhile, the weakest typhoon No. 6 traveled along the farthest path from the NPP site, resulting in the small amplitude of the overall waveform and an insignificant increase in storm surge height. These results highlight that the storm surge waveform characteristics depend on the closest approach distance to the NPP site rather than being simply determined by the typhoon intensity alone.
Therefore, a comprehensive approach that considers the relative location to the site and the approach path as well as meteorological parameters (e.g., the central pressure, maximum wind speed, and the radius of the typhoon) is required to quantitatively assess the effects of storm surges. As such, in this study, the time-integrated storm surge height (TISSH), which can reflect both the sustainability and scale of storm surges, was calculated using Eq. (1).
where η(t) is the storm surge height at time t, and t1 and t2 are the start and end times of a significant increase in storm surge height, respectively.
3.3 Effects of Typhoon Intensity
Fig. 7 and Fig. 8 show the (a) central pressure (pc), (b) maximum wind speed (Vmax), and (c) radius of maximum wind (Rmax) for the 37 virtual typhoons and their relationships with the MSSH and TISSH calculated at each NPP site. Accordingly, the effects of the intensities and paths of the typhoons on the storm surge response characteristics were analyzed.
In Fig. 7(a), there is a negative correlation between pc and MSSH. This reflects the typical physical mechanism whereby a lower pc amplifies the sea level rise by increasing the pressure gradient. In Fig. 7(b), there is a positive correlation between Vmax and MSSH. This is because the long-term water level rise is further amplified as the wind speed increases owing to the increase in the wind stress acting on the sea level. In Fig. 7(c), the MSSH shows a tendency to increase as Rmax decreases. This is because the local pressure gradient increases as the radius of maximum wind decreases, as the pressure change occurs within a short distance based on the same central pressure. In addition, as the radius decreases, the centripetal force increases because the turning radius decreases based on the same wind speed. This causes the concentration of water level rise around the center of the typhoon. In other words, if the radius of maximum wind is small, large storm surge heights are highly likely to occur on the coast owing to the concentration of the influence of air pressure and rotational force in a small area.
In Fig. 8, the relationships between typhoon intensity and TISSH at each NPP site are generally similar to those in Fig. 7, but wider dispersion and somewhat weak correlations are observed. This is because the TISSH reflects the sustainability and overall volume of the storm surge rather than the simple instantaneous MSSH, indicating that the TISSH is significantly affected by the passage time and path of the typhoon.
Meanwhile, each NPP site shows different storm surge response characteristics. The Kori and Saeul NPP sites exhibit a relatively high sensitivity to the MSSH. This appears to be because the southern coast of the East Sea has a relatively gentle seabed slope and is directly exposed to the northward typhoon path. On the other hand, the Hanul NPP is the most sensitive to the TISSH. This results from the storm surge distribution characteristics in which the water level, which is not high, continues to rise despite no direct approach of the typhoons.
Overall, the meteorological parameters of typhoons are closely related to the MSSH and TISSH, which is consistent with physical mechanisms. However, in some cases, distributions that deviate from this tendency are also observed. For example, a large MSSH occurs at the Kori and Saeul NPP sites when even a weak typhoon approaches them, whereas the storm surge height is small despite a strong typhoon at the Hanul and Weolseong NPP sites when its path is set to be far from them. In particular, the TISSH is significantly large in some cases at the Hanul NPP site. This is because the storm surge waveform distribution is wide and gentle owing to the long typhoon approach distance.
These results indicate that storm surge characteristics are not simply determined by the intensity parameters of the typhoon alone, and the travel path of the typhoon has a decisive influence on them (Kim et al., 2025a; 2025b). Even a typhoon with similar intensity cannot induce large storm surge heights when it is far from the site. Conversely, even a relatively weak typhoon can significantly amplify storm surges when it passes in close proximity to the NPP site.
This difference also results from the geographic locations of the NPP sites on the east coast. While the Hanul and Weolseong NPPs are higher in latitude and somewhat deviate from the main travel path of typhoons, the Kori and Saeul NPPs have conditions that facilitate the occurrence of relatively large storm surge heights as they are close to the south coast and adjacent to the typhoon path.
3.4 Effects of Closest Approach Distance
Fig. 9 shows the influence of Dmin on the (a) MSSH and (b) TISSH at the four NPP sites on the east coast. Overall, the MSSH and TISSH increased as Dmin decreased. This appears to be because, as typhoons approach the NPP sites, their impact on the coastline increases and the loss of storm surge propagation energy decreases. Hence, significant MSSH and TISSH did not occur for most typhoons at Dmin > 200 km.
In Fig. 9(a), the MSSH at the Kori and Saeul NPP sites was generally higher compared with that at the other sites, and the difference was more significant for Dmin ≤ 100 km. This appears to be because the two sites are under significant direct and indirect influence owing to closer proximity to the main path of typhoons, and the typhoon energy is concentrated on the coast owing to the relatively gentle seabed slope and shallow depths. Meanwhile, the decreasing tendency of the TISSH in Fig. 9(b) with the increase in Dmin was relatively gentle compared with that of the MSSH. As Dmin decreased, a high and narrow storm surge waveform was generated (see Fig. 6). As Dmin increased, the width of the waveform increased, but the water level gradually decreased. Consequently, the TISSH also tended to decrease beyond a certain distance. Accordingly, a relatively large TISSH was frequently observed at the Hanul NPP site. This appears to be due to the characteristics of the waters where the water level rise in a wide area continued over an extended period of time despite no direct typhoon landing.
In addition, there are some cases in which the MSSH and TISSH are not large when the typhoon intensity is weak despite a short Dmin. For example, the water level rise was not apparent when the central pressure was high and the wind speed was low even under typhoon proximity conditions at the Kori and Saeul NPP sites, and narrow and low storm surge waveforms were generated.
In summary, the closest approach distance serves as a key factor that determines the storm surge height along with the typhoon intensity. Its influence is particularly more sensitive at sites on the east coast. Therefore, the closest approach distance information should be considered for accurate storm surge assessment at NPP sites on the east coast, and simple intensity parameters may underestimate or overestimate storm surge hazards.
4. Discussion
In this study, the results of the storm surge simulation performed for four NPP sites located on the Korea’s eastern coast were comprehensively analyzed. The results showed that the storm surge scale significantly varies depending on the closest approach distance and geographic factors as well as simple typhoon intensity. In particular, the local MSSH rapidly increased during the direct typhoon approach, but a relatively large TISSH occurred even when typhoons passed along a path farther than a certain distance in several cases. This indicates that the typhoon energy concentrated near the site causes short and strong impacts, and that the accumulation of relatively low water levels over an extended period of time may rather increase overflow and inundation effects when typhoons pass at a certain distance.
The correlation between the MSSH and TISSH is not explained by simple intensity factors. The TISSH, which reflects sustainability and volumetric effects, is useful in explaining the actual disaster scale of storm surges more clearly. As the sustainability and damage scale of inundation cannot be fully assessed using the MSSH alone, it is necessary to utilize additional volume-based indicators, such as the TISSH. In particular, when the ground height of the site is smaller than the MSSH, the inundation risk may rapidly increase as the duration of seawater inflow increases, which should be considered during the establishment of disaster prevention plans for NPP sites in the future.
In addition, the NPP sites on the east coast are mostly located on open coasts unlike the west coast, thereby lacking topographic buffering effects. For example, waves and storm surges are dispersed by the archipelago topography at the Hanbit NPP site on the west coast, but the NPP sites on the east coast are directly exposed to storm surges owing to the absence of such blocking effects (Kim et al., 2025b). Moreover, the east coast was reported as an area where the damage by storm waves was more significant than that by storm surges in the past damage cases caused by major typhoons (Seo et al., 2023). This is also closely related to the topography of the east coast where waves are amplified according to the direction of waves as well as depth and seabed terrain characteristics. Therefore, it is important to consider the overtopping and inundation effects by storm waves as well as storm surges in an integrated manner, and a multi-indicator approach that reflects the interactions of such complex marine disasters is required. In particular, as the simultaneous effects of storm surges and storm waves are one of the key factors that determine the inundation sustainability and damage patterns of sites on the east coast, they need to be reflected in an integrated manner in future numerical modeling.
In conclusion, a multi-indicator approach that considers storm wave effects as well as the intensity, closest approach distance, and path of the typhoon is required to precisely assess potential marine hazards at NPP sites on the east coast. The analysis of both the MSSH and TISSH can complement sustained risk and volumetric effects, which were missed by conventional assessment approaches focused on simple maximum values. In future studies, the design heights and safety freeboards of each NPP site will be compared to examine the possibility of exceeding them, and a risk rating and quantitative risk assessment system will be prepared by reflecting the duration of inundation. The follow-up study is expected to improve the applicability by linking the inundation risk for each site by storm surges to practical disaster prevention standards.
5. Conclusions
In this study, storm surge characteristics were quantitatively analyzed for four NPP sites located on the Korea’s eastern coast using TCRM-based probabilistic synthetic typhoon scenarios and the ADCIRC numerical model. The effects of key meteorological parameters (e.g., the central pressure of the typhoon, maximum wind speed, and the radius of maximum wind) as well as the geographic characteristics of the sites and the closest approach distance on the intensity and sustainability of storm surges were evaluated. The main conclusions are as follows.
Among approximately 3,000 synthetic typhoons based on TCRM, 37 scenarios that entered the east coast through the Korea Strait were selected for the numerical simulation of storm surges.
The MSSH was sensitive to typhoon intensity and the closest approach distance to the site. The Kori and Saeul sites exhibited relatively high sensitivity.
The TISSH exhibited significant results at the Hanul NPP site by reflecting the sustainability and volumetric characteristics of storm surges, which resulted from the long-lasting effects of the water level increase in a wide area.
The storm surge height and risk significantly varied depending on the path and relative location to the site (closest approach distance) despite similar typhoon intensity.
As the TISSH can assess overflow sustainability and the possibility of severe inundation, which are difficult to evaluate using the MSSH alone, more accurately, it can be used to calculate the overflow discharge, inundation range, and depth.
The NPP sites on the east coast lacked storm surge buffering effects by natural terrain unlike the west coast. As the effects of storm waves as well as storm surges can be significant, an integrated analysis of complex disasters is required.
The results of this study showed that volumetric indicators, such as the TISSH, can play an important role in evaluating marine hazards at NPP sites on the Korea’s eastern coast. They also presented the need to consider complex factors, such as storm waves (Seo et al., 2023; Hwang et al., 2024), tidal overlapping effects (Park et al., 2010; Kang et al., 2014), and sea level rise (IPCC, 2021), as well as storm surges. Combining such multirisk assessment and high-resolution three-dimensional numerical models (Hur et al., 2019; Hwang et al., 2022) is expected to ensure the long-term safety of NPP sites more conservatively and contribute to the establishment of comprehensive disaster prevention strategies that reflect climate change conditions.
PDF Links
PubReader
ePub Link
Full text via DOI
Download Citation
Print
