1. Introduction
The ongoing climate crisis has caused a steady increase in sea level and temperature, contributing to intense and widespread typhoons. These changes in sea level and temperature amplify the impact of typhoons and increase the likelihood of super typhoons striking the Korean Peninsula. The scale and extent of storm surges significantly vary depending on factors such as the minimum atmospheric pressure, maximum wind speed, storm radius, movement speed, and typhoon track. As climate change intensifies, damage resulting from storm surges will become more severe. This highlights the need for further research on predicting typhoon characteristics, coastal inundation, and the extent of the associated damage.
Consequently, numerous studies have been conducted to analyze the relationship between typhoon characteristics and storm surges. Kang et al. (2009) simulated a linearized hypothetical typhoon and adjusted the key typhoon parameters to examine the variations in storm surges along the coast of Gyeongnam. Rego and Li (2009) employed a finite-volume coastal ocean model (FVCOM) to investigate how a hurricane’s movement speed and duration affect storm surges, confirming an inversely proportional relationship between these factors. Park and Park (2021) utilized the Advanced Circulation (ADCIRC) model and found that a typhoon’s fast movement speed led to a sharp increase in storm surges, whereas a slow speed resulted in a more gradual rise.
Several studies have actively explored typhoon tracks and their impacts on storm surges. Kim et al. (2014) developed a probabilistic clustering method to classify typhoons based on the locations and tracks of 197 typhoons that affected the Korean Peninsula between 1951 and 2012. Ku et al. (2019) used the Sea, Lake, and Overland Surges from Hurricanes (SLOSH) model to examine storm surges caused by Typhoons Maemi (0314) and Bolaven (1215) and to determine the applicability of the model by comparing their maximum surge heights. Lee et al. (2016) conducted a numerical simulation incorporating the track of Typhoon Maemi (0314) using the Delft3D model and derived predictions for storm surges affecting the southern and western coastal areas of the Korean Peninsula. Park et al. (2008) examined the movement characteristics of various tropical depressions that affected the Gyeongsangbuk-do region between 1978 and 2006. Lee et al. (1992) conducted a statistical analysis of the central pressure and maximum wind speed of typhoons affecting Korea in the East Asian mid-latitude region between 1960 and 1989. They identified the characteristics of each typhoon by classifying it based on its movement patterns. Park et al. (2006) classified the tracks of typhoons affecting Korea into seven types and found that 24.6% passed through China, and 22.9% passed through Japan before reaching Korea. Moon et al. (2007) developed hypothetical scenarios using past typhoon data, including the central pressure, maximum wind speed radius, track, and speed. They then created a database from the results to provide real-time predictions of the expected inundation areas and scales based on similar typhoon data.
Global research has actively been conducted on the relationship between typhoon tracks and storm surges. Qin et al. (2023) created a hypothetical scenario based on the intensity and track of typhoons to simulate the inundation caused by extreme typhoons in Ningbo, China, and carried out a numerical simulation using the Regional Ocean Modeling System (ROMS). McInnes et al. (2003) simulated a simple hypothetical typhoon by varying its entry angle for the Cairns coast in northeastern Australia. Peng et al. (2006) used the Princeton Ocean model to conduct a numerical simulation of storm surge and inundation responses in Charleston Harbor and nearby coastal regions and found that both storm surge and inundation were highly sensitive to changes in the hurricane’s path. Weisberg and Zheng (2006) employed FVCOM to simulate storm surge responses near Tampa Bay, Florida, and found that storm surges were significantly influenced by factors such as hurricane speed, landfall location, approach direction, and intensity. Du et al. (2020) used the ADCIRC model to simulate 23 newly defined typhoon tracks in Wenzhou, China, and found that the impact of storm surges significantly varied based on the proximity of the typhoon track to the observation station. Consequently, they conducted simulations of 55 typhoon tracks that affected the Zhejiang coast between 1951 and 2017.
Jin et al. (2024) developed several scenarios representing different landfall locations of major typhoons from the past based on a single track of Typhoon Maemi (0314). They examined the maximum storm surge height, temporal waveform, and volume at 67 locations across Jeju-do and the southwest and southeast coasts. However, there are limitations in fully assessing the characteristics of storm surges across all coastal regions of Korea. Thus, this study aimed to analyze the impact of typhoon tracks through numerical simulations using hypothetical typhoons in the western, southern, and eastern coastal regions. Accordingly, we created hypothetical typhoon scenarios based on past typhoons that made landfall on the Korean Peninsula and explored the characteristics of regional storm surges. The results of this study can help evaluate the risk and vulnerability of coastal regions to storm surges in relation to typhoon tracks and provide essential data for developing effective coastal disaster prevention strategies.
2. Hypothetical Typhoon Scenarios
2.1 Selection of the Target Typhoons
This study selected six representative typhoons with tracks that reached the west, south, and east coasts of the Korean Peninsula and maintained their intensities, as shown in Fig. 1. Typhoons Kompasu (1007) and Bolaven (1215) traveled north along the west coast, bypassing China and Taiwan. Typhoons Rusa (0215) and Tembin (1214) made landfall on the southern coast after passing west and east of Jeju-do, respectively. Typhoons Bolaven (0006) and Haishen (2010) affected the east coast after passing through the Korea Strait.
2.2 Synthetic Typhoon Combinations
Hypothetical typhoon scenarios were created by combining various typhoons based on S1, S2, S3, SM1, and SM2 (Table 1). S1 represents the scenario where the target typhoon made landfall at the same time as it actually did, whereas S2 depicts the scenario where the target typhoon made landfall when it reached its minimum central atmospheric pressure. S3 shows a scenario in which the target typhoon made landfall when it reached its maximum wind speed. In SM1, the target typhoon makes landfall based on its track at the actual landfall time of Typhoon Maemi. In contrast, in SM2, the target typhoon followed its track and made landfall at the maximum intensity (minimum central pressure and maximum wind speed) of Maemi. Here, S2/S3 represents the lowest central pressure and the highest wind speed, respectively, with all past typhoons analyzed in this study falling into these categories.
The landfall times of the typhoons were modified and adjusted according to typhoon data provided by the U.S. Joint Typhoon Warning Center. A hypothetical typhoon applied to the SM2 scenario is shown in Fig. 2. Furthermore, Table 2 provides a summary of the characteristics of each typhoon when it was closest to or made landfall on the Korean Peninsula, emphasizing key variables such as central pressure, maximum wind speed, and wind radius.
3. Numerical Method
3.1 Numerical Model
The ADCIRC ver. 55 (Luettich et al., 1992) utilized in this study is a numerical model developed using the finite element method, which offers high grid flexibility through an unstructured grid system. Accordingly, intricate coastlines and diverse water-depth conditions can be accurately simulated. Moreover, high-resolution grids can be strategically concentrated in the target region to accurately capture detailed geographical and physical features. A parallel computation function was implemented to efficiently simulate spatiotemporal marine phenomena across large-scale regions.
3.2 Reference Coastal Points
Storm surge characteristics were analyzed by comparing the key locations along the track of the target typhoon that were significantly affected by storm surges (Fig. 3). Specifically, the analysis points were chosen based on the following geographical conditions: A total of 10 points on the west coast (Nos. 1–10), 16 points on the south coast (Nos. 11–26), and nine points on the east coast (Nos. 27–35) were chosen. This arrangement of analysis points facilitated the mathematical examination of storm surges across the coastal regions of the Korean Peninsula. In this study, these reference points were used to analyze the influence of the geographical features of the west, south, and east coasts and typhoon intensity on storm surge impacts.
4. Numerical Results
4.1 Spatial Characteristics of Maximum Storm Surge Heights
Figs. 4,–6 illustrate the spatial distribution of the maximum storm surge heights observed in hypothetical typhoon scenarios that impacted the west, south, and east coasts of Korea in the past. In this study, storm surge simulation results are presented for the west, south, and east coasts based on hypothetical scenarios reflecting the tracks of Typhoon Bolaven (1215), Typhoon Rusa (0215), and Typhoon Haishen (2010).
Overall, because these historical typhoons did not make landfall at their peak intensities, scenario S2 exhibited a higher storm surge height than scenario S1. As shown in Table 2, because the maximum intensity of the historical typhoons (S2) was stronger than the actual landfall time of Typhoon Maemi (SM1), the storm surge height was greater in S2 than in SM1. However, when only comparing the intensity of the typhoons, past typhoons were relatively weaker than Typhoon Maemi, which led to the highest storm surge height in Scenario SM2.
This tendency was evident on the west and south coasts, where the coastline is intricate and the bay terrain is prominent. On the west coast, as shown in Fig. 4, the storm surge height increased significantly in Haeju Bay, Hwanghae-do Province, with high storm surges extending deep inland along the Hangang River Estuary and riverbed. Specifically, given the absence of an estuary bank along the Hangang River, the metropolitan area would face severe impacts if a super-typhoon made landfall along a track similar to Typhoon Bolaven (1215).
In the scenario based on Typhoon Rusa (0215), which has caused significant damage in the past, substantial increases in storm surge height were observed in the bay terrain on the southwest coast (Fig. 5). In particular, storm surge heights were exceptionally high in Suncheon Bay and Gwangyang Bay, which were significantly impacted by the scenario. This suggests that bay areas are the most vulnerable when powerful typhoons make landfall or strike the southern coast.
In contrast, the east coast features a relatively straightforward coastline and typhoons typically move from south to north along their tracks. Thus, storm surge heights were generally low, except in areas near the typhoon path. However, a relatively high storm surge height was observed near Ulsan because of a stalled northward-propagating surge. However, the increase in storm surge height observed in each scenario was not as pronounced as that on the west and south coasts (Fig. 6). Given the impact of storm surges, the east coast, with its simple coastline and deep waters, is less affected by geographical characteristics.
4.2 Regional Variability of Storm Surge Magnitudes
To analyze the regional characteristics of storm surges for each (a) Bolaven-S1 (b) Bolaven-S2/S3 (c) Bolaven-SM1 (d) Bolaven-SM2 typhoon scenario, the maximum storm surge heights at 10 reference points on the west coast, 16 points on the south coast, and 9 points on the east coast (Fig. 3) were non-dimensionalized, as presented in Figs. 7,–9. In this analysis, the storm surge heights for each hypothetical scenario (S2, S3, SM1, and SM2) were nondimensionalized by dividing them by the storm surge height of the reference typhoon.
Fig. 7(a) shows that, on average, scenarios S2/S3 exhibited a maximum storm surge height 2.37 times higher, SM1 was 1.45 times higher, and SM2 was 2.8 times higher compared to Typhoon Bolaven (1215). Notably, the highest storm surge height was recorded in scenario SM2, with Incheon (No. 1), Pyeongtaek Bay (No. 2), and Seosan (No. 3) experiencing increases of 4.6 times, 5.04 times, and 4.37 times, respectively, compared to scenario S1. Fig. 7(b) illustrates the case of Typhoon Kompasu (1007), where the maximum storm surge heights were, on average, 1.81 times higher in scenarios S2/S3, 1.05 times higher in SM1, and 2.34 times higher in SM2. In scenario SM2, Seosan (No. 3), Taean (No. 4), and Gunsan (No. 5) recorded maximum storm surge heights that were 2.53 times, 2.9 times, and 2.88 times higher, respectively, compared to scenario S1. Areas with high storm surge heights generally feature narrow channels and bay terrains. Specifically, bays that opened southward experienced higher storm surge heights because of the greater impact of the typhoon’s moving track.
Similar topographical patterns were observed in storm surge simulations based on Typhoons Rusa (0215) and Tembin (1214) (Fig. 8). In the scenarios based on the track of Typhoon Rusa (0215), the maximum storm surge heights were, on average, 2.36 times higher in S2/S3, 1.33 times higher in SM1, and 3.23 times higher in SM2 (Fig. 8(a)). Suncheon Bay (No. 16), Yeoja Bay (No. 17), and Gamak Bay (No. 19) recorded storm surge heights that increased by 5.47 times, 5.67 times, and 5.17 times, respectively, under scenario SM2. In the scenarios based on the track of Typhoon Tembin (1214), the maximum storm surge heights were, on average, 8.3 times higher in S2/S3, 5.45 times higher in SM1, and reached the highest increase of 13.47 times in SM2 (Fig. 8(b)). In scenario SM2, Gangjin Bay (No. 12), Boseong Bay (No. 13), and Deukryang Bay (No. 14) experienced substantial increases in maximum storm surge heights of 21.81 times, 21.14 times, and 21.62 times, respectively. These results suggest that the intensity of Typhoon Tembin during actual landfall was weaker than that in the hypothetical scenarios. Storm surges are influenced by meteorological factors, such as the typhoon’s central pressure, maximum wind speed, and storm radius, and these effects are particularly amplified in bay terrains.
As mentioned above, the east coast exhibits distinct storm surge characteristics compared to the west and south coasts owing to its simpler coastline. Fig. 9(a) shows the scenario based on Typhoon Bolaven (0006), where, on average, storm surge heights were 1.61 times higher in scenarios S2/S3, 2.52 times higher in SM1, and 5.97 times higher in SM2. However, because the storm surge height of the reference typhoon (S1) was relatively low, it was important to consider the absolute height magnitude along with a simple increase in the increase ratio. Fig. 9(b) illustrates the case of Typhoon Haishen (2010), where the maximum storm surge heights were, on average, 2.12 times higher in scenarios S2/S3, 1.15 times higher in SM1, and 2.64 times higher in SM2.
Although the typhoon track moved northward along the east coast, higher storm surges were observed along the southeast coast. This can be attributed to the fact that storm surges do not stall but instead move steadily as the coastline runs nearly parallel to the typhoon’s direction of movement. Therefore, except for certain areas, including Ulsan Bay, which opens to the south, the storm surge height was generally lower along the east coast. In the scenario based on Typhoon Haishen (2010), Yeongil Bay in Pohang, which opens to the north (No. 28), did not show a significant increase in storm surge height despite the intensified typhoon (Fig. 9(b)).
4.3 Temporal Waveforms of Storm Surge
Fig. 10 has been provided to examine the impact of the typhoon track and geographical features on the temporal waveform formation. For example, temporal waveforms were analyzed for each scenario based on the tracks of typhoons Kompasu (1007), Tembin (1214), and Haishen (2010), which affected the west, south, and east coasts of Korea, respectively.
Cheonsu Bay (No. 4) on the west coast has a narrow channel that opens to the south, resulting in significant variations in storm surges based on typhoon intensity because of direct exposure to typhoons (Fig. 3). Consequently, the scenario based on the track of Typhoon Kompasu displayed higher storm surge waveforms with increased typhoon intensity, as shown in Fig. 10(a). Fig. 10(b) presents a scenario based on Typhoon Tembin’s track, where Deukryang Bay (No. 14) on the south coast displayed a broader and higher temporal waveform of storm surge as the typhoon intensity increased. In Yeongdeok (No. 29), on the east coast, the scenario based on the track of Typhoon Haishen displayed high storm surge heights and broad waveforms as the typhoon intensity increased, as shown in Fig. 10(c).
The storm surge height is a key factor in determining whether overtopping/overflow occurs. The distribution of temporal waveforms indicates the duration of a storm surge, and can significantly influence the height and extent of coastal inundation when overtopping/overflow occurs. Therefore, we conducted a comparative analysis of the characteristics of each scenario by applying temporal waveform volume calculation methods (Lee et al., 2022; Lee et al., 2023; Jin et al., 2024).
4.4 Storm Surge Volume
The volume of the storm surge temporal waveform is calculated by taking the integral of the water level above the mean sea level and can be defined as the time-cumulative flow per unit width (unit: m2·s/m). To analyze the characteristics of the maximum storm surge heights in each hypothetical scenario and explore the volume characteristics directly linked to overtopping/overflow, the ratio of the storm surge volume for each scenario compared to the reference typhoon is shown in Figs. 11,–13.
Fig. 11(a) shows the west coast case, where scenario SM2 had the highest average volume ratio of 4.98 compared to the reference typhoon (Bolaven (1215)). The areas where the volume ratio exceeds 6.0 are Incheon (No. 1), Seosan (No. 3), and Taean (No. 4), with ratios of 10.2, 6.88, and 8.65, respectively. The average volume ratio of scenarios S2/S3 was 2.06, and Incheon (No. 1) recorded the highest value of 9.22. In contrast, the average volume ratio for scenario SM1 was 0.92, which was lower than that of the reference typhoon, although Incheon (No. 1) exhibited a relatively high value of 5.91. In the scenarios based on Typhoon Kompasu (1007), the average volume ratios were highest at 1.34, 1.09, and 0.79 for SM2, S2/S3, and SM1, respectively. Although the values are generally smaller, the patterns are similar to those shown in Fig. 11(a). However, Cheonsu Bay (No. 4) and Hampyeong (No. 7) recorded an average volume ratio of 1.84 and 2.56, respectively, in scenario SM2. These differences are attributed to the fact that Typhoon Bolaven (1215) and Typhoon Kompasu (1007) followed different tracks, with each making landfall at Gangryong, Hwanghaenam-do Province, and southeastern Ganghwado Island, respectively.
Similarly, on the south coast, compared to the reference typhoons (Rusa (0215) and Tembin (1214)), the storm surge volume ratio showed an increasing trend in the order of SM1, S2/S3, and SM2 (Fig. 12). In the scenarios based on Typhoon Rusa, the average volume ratios were highest at 2.27, 1.43, and 0.63 for SM2, S2/S3, and SM1, respectively (Fig. 12(a)). Conversely, average volume ratios of 7.79, 4.04, and 2.62 were recorded in the scenario based on the track of Typhoon Tembin (Fig. 12(b)). As shown in Table 2, relatively high volume ratios were observed because of the extremely weak intensity of Typhoon Tembin during landfall (Fig. 10(b)). In the case of Typhoon Rusa, Suncheon Bay (No. 16), Yeoja Bay (No. 17), Gwangyang Bay (No. 18), Gamak Bay (No. 19), and Geoje (No. 25) experienced increasing storm surge volume ratios as the typhoon intensity increased. In contrast, during Typhoon Tembin, Gangjin Bay (No. 12), Boseong Bay (No. 13), Deukryang Bay (No. 14), Nogdong Port (No. 15), and Suncheon Bay (No. 16) exhibited high storm surge volume ratios.
On the east coast, the average volume ratio was highest in the order of SM2, S2/S3, and SM1 in the scenarios based on Typhoon Bolaven (0006) and Typhoon Haishen (2010). However, compared to the west and south coasts, the variations in the volume ratio in response to changes in typhoon intensity remained relatively minor, as depicted in Fig. 13. The average volume ratio of scenario SM2 was 2.3 for Typhoon Bolaven and 1.53 for Typhoon Haishen, indicating minimal regional differences owing to the simple coastline. However, Yeongdeok (No. 29) and Jukbyeon Ports (No. 30) in Fig. 13(a), as well as Ulsan (No. 27) in Fig. 13(b), exhibited relatively high volume ratios of 3.04, 3.02, and 2.41, respectively, in scenario SM2.
5. Discussion
In summary, the storm surge simulation results of this study indicate that evaluating disaster risks in coastal areas based solely on maximum storm surge height has limitations. Although storm surge height is undoubtedly a key indicator of a typhoon’s impact, it is also essential to consider the persistence of storm surges and the volume of overtopping/overflow, as these factors determine the actual extent of overtopping and overflow. Thus, even if multiple cases have the same maximum storm surge height, a longer duration, and larger storm surge volume can result in severe coastal inundation damage.
The results of this study indicate that when a typhoon’s track interacts with the regional characteristics of the coast, the height and volume of the storm surge tend to increase along the west and south coasts. Specifically, regions such as the Hangang River estuary, Seosan, and Taean on the west coast, and Suncheon Bay, Gwangyang Bay, and Gamak Bay on the south coast, have complex coastlines and shallow waters. Consequently, these areas are prone to high storm surges, even when typhoon intensity is relatively low. As noted in previous studies, storm surges in bay terrain tend to converge inward, resulting in an amplification effect. This suggests that the impact of storm surges can differ greatly depending on the coastal terrain, even under the same typhoon intensity. This also indicates the possibility that a weak typhoon can generate high storm surge heights and significant overtopping/overflow in certain terrains.
Compared with the west and south coasts, the east coast showed a lower maximum storm surge height. The east coast is characterized by a relatively simple coastline and the absence of a bay terrain open to the south. Consequently, typhoons tend to dissipate quickly without lingering in coastal areas when moving northward. However, as evidenced by recent cases such as Typhoons Maysak (2009) and Haishen (2010), the eastern coast still experienced significant damage. The damage resulted not only from storm surges but also from the substantial impact of storm waves that accompanied typhoons. Therefore, relying solely on storm surge analysis has limitations for fully assessing the disaster vulnerability of coastal areas. In the future, a more detailed coastal disaster analysis is required, incorporating additional meteorological and oceanographic factors, such as storm waves (Son and Do, 2022; Seo et al., 2023; Hwang et al., 2024) and tidal effects (Park et al., 2011; Park et al., 2013; Kim et al., 2018).
In the hypothetical typhoon scenarios analyzed in this study, the lower storm surge height and volume in scenario SM1 compared with scenario S1 have significant implications. Even in the case of Typhoon Maemi (0314), which caused severe damage to the Korean Peninsula, the level of damage could vary if the typhoon followed a different path. In other words, while the intensity and strength of a typhoon are key factors in determining the scale of storm surges, their impact can vary significantly depending on the coastal terrain and typhoon track. Consequently, the uncertainty of a typhoon’s track is a crucial factor in coastal disaster prevention, given that the track can vary significantly owing to changing weather conditions. Despite advancements in prediction technology, accurate forecasting tracks in advance remain a challenging task.
In conclusion, the analysis of hypothetical typhoon scenarios in this study provides a quantitative assessment of how typhoon tracks and coastal terrain influence the storm surge extent. Therefore, disaster prevention measures must be tailored to the specific characteristics of vulnerable coastal areas. As a result, more accurate and effective coastal disaster predictions can be made, along with the development of practical disaster prevention strategies.
6. Conclusion
This study examines the characteristics of storm surges in the coastal areas of the Korean Peninsula by applying hypothetical typhoon scenarios based on past typhoon tracks. The maximum height, temporal waveform, and volume of storm surges were analyzed and compared across key locations on the western, southern, and eastern coasts, and differences in regional storm surge characteristics were quantitatively evaluated.
On the west coast, high storm surges were observed in areas such as Haeju Bay, Hwanghae-do, the Hangang River estuary, Seosan, and Taean owing to the interaction between the typhoon’s path and coastal terrain. In particular, a notable trend was observed in which the storm surge height increased as the surge concentrated inward when passing through the bay terrains. This suggests that there is a high likelihood of significant localized storm surge amplification effects on the West Coast owing to the complex interaction between the typhoon’s approach and the coastline.
On the south coast, areas such as Suncheon Bay, Gwangyang Bay, and Gamak Bay experienced high storm surge heights and volumes, with surge characteristics varying significantly based on the typhoon track and landfall location. Storm surges were notably amplified in the bay terrain along the south coast. This indicates that if a typhoon makes landfall on the southern coast while retaining its high intensity, the coastal areas will experience more severe inundation and damage.
On the east coast, storm surge heights exceeding 1 m were observed in some areas near the typhoon track; however, the surges were relatively lower than those on the west and south coasts. These results can be attributed to the east coast being less affected by storm surges than the west and south coasts because of their geographical and topographic characteristics.
On the west and south coasts, stronger typhoon intensities resulted in longer storm surge durations, which in turn caused an increase in surge volumes. In particular, when typhoons approach bay terrains on the west and south coasts, storm surges tend to persist longer, and their volumes increase owing to the accumulation of seawater. In contrast, typhoons move quickly on the east coast, leading to shorter storm surge durations and smaller volume increases.
This study quantitatively examined regional variations in storm surge characteristics along typhoon tracks on the Korean Peninsula coast, confirming that typhoon intensity can have a substantial impact, combined with the interaction between storm surge characteristics and coastal terrain, in specific coastal areas. A more thorough coastal disaster assessment is required to consider the interactions among storm surges, tidal effects, and storm waves.
PDF Links
PubReader
ePub Link
Full text via DOI
Download Citation
Print
