Journal of Coastal Research SI 113 ***–*** Charlotte, North Carolina 2024

Extreme Surge Heights Incorporating Long-term Tidal Variation

and Sea-level Rise
Seung-Won Suh†*and Hyeon-Jeong Kim††
††
†
Department of Ocean Science and Engineering Disaster Information Research Division
Kunsan National University National Disaster Management Research Institute
Kunsan, Republic of Korea Ulsan, Republic of Korea www.cerf-jcr.org
Yeongdo, B†usan
ABSTRACT

Suh, S.-W. and Kim, H.-J., 2024. Extreme surge heights incorporating long-term tidal variation and sea-level rise. In:
Phillips, M.R.; Al-Naemi, S., and Duarte, C.M. (eds.), Coastlines under Global Change: Proceedings from the
International Coastal Symposium (ICS) 2024 (Doha, Qatar). Journal of Coastal Research, Special Issue No. 113, pp.
***–***. Charlotte (North Carolina), ISSN 0749-0208.

www.JCRonline.org Global warming is a primary factor for acceleration of sea-level rise (SLR). Most previous studies have dealt
only with mean SLR; however, in terms of storm surge inundation, not only the mean SLR, but also the
summer seasonal SLR, which is locally almost 20 cm higher on the west coast of Korea because of thermal
expansion and riverine runoff, must be considered. In addition, the nodal effects of the 18.6-year tidal variation
should be considered for long-term (through 2100) surge hazard analyses. Tidal analyses were conducted at
the Incheon station, which is located in the middle of the west coast of Korea, using records from 1960 to
2018. Peak M2 amplitudes varying from 2.96 to 3.02 m were obtained in 1979, 1997, and 2015, while the
lowest value of 2.76 m was obtained in 1968, 1985 and 2006. The analysis predicted that the resulting tidal
range, would show peak values this century in 2025, 2043, 2062, and 2099. These extreme SHs are potentially
hazardous to existing coastal infrastructure such as levees and dikes. This study is relevant to the mitigation of
coastal hazards, incorporating not only extreme SH but also wave-induced overtopping hazards on artificial
coasts where anthropogenic impacts were made without considering future sea-level changes.

ADDITIONAL INDEX WORDS: Global warming, sea-level rise, extreme surge height, tidal variation.

INTRODUCTION seasonal SLR, and the inter-decadal nodal factor variations over
In previous research, it is noted that sea-levels (SLs) an 18.6-year period should be included.
continuously vary over time spans as short as a day or as long as The causes of SLR, and information on long-term intra-
a season or year. Moreover, SLR and the acceleration of global decadal variation in SL essential in preparing coastal disaster
warming are pressing issues in the context of coastal prevention plans for low-lying coastal zones has been rigorously
vulnerability. SLR and/or acceleration of SLR have tended to studied by several researchers (Abeysirigunawardena and
increase over the last several decades, and thus, it is anticipated Walker, 2008; Atkinson et al., 2013; Baart et al., 2012;
that future trends will be much higher than in the past (Kim and Houston, 2013; Irish and Resio, 2013; Suh, 2016; Watson, 2019;
Suh, 2019). It is addressed the summer seasonal sea-level in the Zhang, 2011).
yellow sea (YS), especially for the west coast of Korea (WCK), In this study, to prepare coastal areas that are exposed to
which is ~ 20 cm higher than the annual mean sea-level (MSL). extreme surges along the WCK and the south coast of Korea
Incorporating the summer seasonal SLR (S3LR) is essential in (SCK), a simple linear S3LR trend to predict SL and intra-
future storm surge height (SH) analysis because the annual decadal tidal fluctuation due to the nodal factor was
typhoons usually occur in the summer season from July to incorporated. Based on such projected SL fluctuations,
October. synthetically generated typhoons were applied, as recently
The WCK lies in a macro-tidal environment, thus the suggested by Kim and Suh (2018), with randomly triggered
modulation of ± 3.7% of M 2 tidal amplitudes of ~3 m, yields surges without manually fixing the tidal phase to account for
approx. ± 11 cm of sea-level fluctuation due to the nodal factor. realistic non-linear tide-surge interactions in a macro-tidal area.
This fluctuation was demonstrated over a period of 18.6 years, The simulated SHs were analyzed using extreme probability
which relates the regional SLR to the spatial variability (Baart et distribution in the R-package to determine the targeted water
al., 2012). Thus, in the storm SH and extreme value analysis for level on the WCK by the end of this century.
the coastal hazards mitigation plan, possible SLR as a result of As the sea-level increases, the depth-limitation of waves
the steric effect, the inverse atmospheric pressure effect, the relaxes, resulting in waves with larger periods, greater
amplitudes, and higher run-ups. Additionally, depth and
frictional changes affect tide, surge, and wave characteristics,
____________________ altering the relative importance of other risk factors (Arne et al.,
DOI: 10.2112/JCR-SI113-XXX.1 received Day Month Year; accepted 2017). Thus, the inclusion of SLR in planning the mitigation of
in revision Day Month Year.
*Corresponding author: [email protected]
multiple coastal hazards is necessary as emphasized by this
©
Coastal Education and Research Foundation, Inc. 2024 study. However, there can be limitations; insufficient archived

XX
 Suh and Kim
_________________________________________________________________________________________________

water level data, the linear extrapolation of future SLR, an magnitude is too small to produce SLR on the order of a meter
uncertain SLR trend in recent years that might alter the trend or greater if projected to 2100. The annual S 3LR is at least 20.7
from linear to non-linear, under resolved future coastal change cm higher than the annual MSL and the deviations increase up
due to morphological change, numerical errors in the applied to 27.1 cm with respect to higher latitudes in the Yellow Sea and
model itself, and other as yet unrevealed issues. Therefore, the East China Sea (Kim and Suh, 2019).
continued study of these issues is required to improve the
relevant results.

METHODS
In this study, to determine the vulnerabilities to coastal storm
surges, archived historical tidal data along the WCK through
2018 was analyzed. Hourly SL records were harmonically
analyzed. In the analysis, not only the annual MSL but also the
summer seasonal annual peak level were determined to see the
inverse barometric effects (Kim and Suh, 2019).

Tidal Data Analysis on the WCK

Annually, three to four typhoons usually make a landfall on
the Korean coast between July and October. Due to the steric
effect, the long inter-annual sea-level fluctuation, Sa, should be
considered for the summer seasonal SLR, which reached at least
the 20 cm (∆ Z 0 i.e., S3LR-annual SLR) on the WCK (Kim and
Suh, 2019). Therefore, incorporation of the S 3LR rather than the
annual SLR for future SH vulnerability analysis would be
preferable. Monthly and annual MSLs were determined by
analyzing the Korea Hydrographic and Oceanographic Agency’s
hourly recorded SL data from 1964 to 2018 along the WCK at
Incheon (ICN), Kunsan (KUN), Mokpo (MOK), and Jeju (JEJ),
through TIRA (TASK-2000 Package; Bell et al., 1999). Watson
(2019) adapted a singular spectrum analysis at 21 sites around
the Korean Peninsula for annual average SL records, available
from the Permanent Service for Mean Sea-Level (≤50 years) for
trend analysis.
The SL can be expressed in harmonics as in Eq. (1) when the
nodal factor is incorporated. Up to 120 harmonic constituents,
plus the mean level, were recognized in TIRA. Figure 1. Location map for the stations used for the analysis of historical
vs. synthesized typhoon tracks.
N
h(t)=Z 0+ ∑ H n f n cos ( V n +u n−gn +σ n t ) ,(1)
n=1

in which, h is the tidal height;Z 0 is the value of the MSL; H n Extreme Surge Height Distribution (GEV) Analysis
and gnare the amplitude and phase, respectively; f n and un are In extreme SH analysis along the Korean coast, the empirical
simulation technique, annual maximum series, peaks over
the nodal factor and equilibrium argument, respectively; and V n threshold, and generalized Pareto distributions were applied
is the tide generating potential at Greenwich at t=0 ; σ n is the indirectly to the storm characteristics not to SH itself (Suh et al.,
2009). Despite the extreme value approach for evaluating
angular speed of constituent; and N is the number of constituents.
forcing or response, the result could be almost identical SH
values. However, to understand the extreme distribution of
Monthly Maximum Sea-level Trend
results from SH analysis over the target area, a GEV frequency
Extreme SHs usually happen in the summer season because
analysis was performed and the maximum likelihood method
cyclone vortex development relies on atmospheric and heated
was chosen for parameter estimation for the resulting SH. Storm
sea surface conditions; therefore, the annual peak summer
surge simulations were performed based on synthetic typhoons
seasonal sea-level and its trend should be considered. In a
of 174,689 set (Kim and Suh, 2018) to define extreme values
previous approach, Kim and Suh (2019) applied a simple linear
with additional consideration given to SLR, S 3LR, and the nodal
regression analysis to fit the annual MSL for selected tidal
factor effect in long-term sea-level trends. In simulations,
stations on the WCK. In future variation analysis, the magnitude
ADvanced CIRCulation (ADCIRC) was considered as the basic
of the acceleration should be applied, even though the
surge modeling under an unstructured grid named NWP-G258k,

Journal of Coastal Research, Special Issue No. 113, 2024

XX
 Extreme Surge Heights Incorporating Long-term Tidal Variation and Sea-level Rise
_________________________________________________________________________________________________
which has 258,046 vertices covering the north-western Pacific Therefore, the simulated tidal SH is expected to be randomly
(NWP) region that can efficiently simulate tide and surge allocated at a target location.
interactions (Suh et al., 2015). To identify the storm surge phase with the tide, 12
RESULTS synthetically generated typhoons, which could yield extreme
S3LR Trend vs Mean SLR SHs at ICN, were arbitrarily selected. Then, the phase of the
According to the analysis of annual mean and summer season storm was shifted relative to high tide to match the beginning of
SL trends, as seen in Figure. 2, when a linear trend by the end of the flood tide, maximum flood tide, high tide, maximum ebb
the century is assumed, increasing trends of SL appear in all tide, and ending ebb tide by adjusting the appearance time -5, -3,
stations of the WCK. However, S3LR that occurred mainly in 0, +3, and +5 h respectively. The simulated SH results were
August (55 to 70%), showed a ~20 cm rise over the annual compared with total water level (TWL) as shown in Fig 3. The
MSL. The highest recorded value during this period was 23 cm results revealed that higher SHs of 1.2 to 2.4 m can appear as the
at MOK but decreased to 19 cm at ICN and KUN, and finally to tidal phase approaches low tide, i.e., the beginning of the flood
17 cm at JEJ. tide and ending of the ebb tide because of non-linear tide-surge
The importance of the non-linear interaction between the interactions in macro-tidal environments. Meanwhile, the lower
surge and tide in macro-tidal environments was addressed by SHs of 0.6 to 1.4 m can be seen when the tidal phase is closely
Abeysirigunawardena and Walker (2008) for Northern British aligned with high tide. However, in terms of TWL, only
Columbia and by Suh and Lee (2018) for WCK in YS. As accounting for extreme SH may lead to an under-estimation of
Watson (2019) demonstrated, the highest rates of relative mean coastal vulnerability. Some historical typhoons have hit the ICN
SLR were observed at MOK and JEJ (5.8 ±0.8 mm/year and 5.4 area and showed relatively low TWL even when the typhoon
± 0.8 mm/year, respectively). strength was significant, as was the case with typhoon Bolaven
in 2012. At that time, the SH was marked at 1.37 m; however,
landfall corresponded to the beginning of the flood tide and the
TWL was marked as 0.46 m below the MSL. Nonetheless, it
should be pointed out that probable maximum TWL will be
realized when extreme SH matches S3LR in a special year when
the nodal factor is maximized, such as in 2090.

Figure 2. Linearly regressed annual mean SLR (black) and S 3LR (red)
trend along the WCK.
Figure 3. Effects of tide-surge interaction by comparison of TWL and
SH for synthetic and observed typhoons.

Storm Surge Phase Interaction with Tide Extreme Surge Height Distribution
In this study, the time of occurrence of the storm surge and After considering the nodal factor, the resulting SH over a
the phase of the tidal current were randomly seeded. 100 year period would vary from 1.18 to 1.31 m on the WCK,
Consequently, the peak surge time and high tide have no while it would vary from 0.68 m to 0.71 m on the SCK. These
relation. These results show that actual storm surges may vary extreme SHs are potential vulnerabilities for existing coastal
depending on the tidal situations, such as low or high tides. infrastructure, such as levees and dikes.
Journal of Coastal Research, Special Issue No. 113, 2024
XX
 Suh and Kim
_________________________________________________________________________________________________

the SCK. They found the same pattern occurred for the non-
DISCUSSION linear effect: the non-linear term opposed the tide by enhancing
SLR Incorporation storm tide heights when the tide was near its minimum. Feng et
Even though there might be uncertainties in future climate al. (2016) demonstrated the delay and damping of the SH, which
conditions, incorporating SLR in coastal flood analysis is is the non-linear interaction between the tide and surge, resulting
necessary for assessing future flood probabilities; reliable in a semidiurnal fluctuation influencing the M2 signal.
hurricane flood probability estimates are essential for effective
management and engineering in coastal environments (Irish and Synthetic Storms and Tide-surge Modeling
Resio, 2013). To overcome the uncertainties due to the latest Hoeke et al. (2015) attempted efficient analysis by using a
information, Houston (2013) demonstrated that SLR projections total of 30 synthetic ensemble members. This “reduced
can sometimes mislead recent global climate variations when ensemble” approach allows the range of variability in the wave
based on the static projection of SLR based on past records. setup to be explored for a given storm tide level. They
Considerations for target SLR scenario applications (Clara et al., concluded that these local changes in momentum balances will
2015; Pelling et al., 2013; Suh, 2016) were applied by adding result in changed patterns of inundation and sediment transport,
1.0 or 2.0 m of SLR for future coastal vulnerability analyses. particularly under progressively changing SL. Thus, it would be
Applying long-term SLR, controlled by climatic change, in valuable to account any intra- or inter-annual sea surface
macro-tidal environments and the linear regression model based fluctuation, especially S3LR. Because the SLR report by the
on the highest SLs. i.e., max SL due to the surges of each IPCC encompasses reasonable aspects, a scenario based SLR
calendar year, may yield a trend that exceeds twice that of MSL projection based on IPCC reports could be an alternative
as demonstrated by Abeysirigunawardena and Walker (2008). approach, such as the addition of 1 or 2 m to the SL by the end
Thus, all types of probable SLR, should be incorporated in of this century (Suh, 2016). Whereas, S3LR was applied here,
coastal storm surge hazard analysis, regardless of uncertainties which is at least 20 cm higher than the annual MSL; thus, the
in climate change and assumptions about SLR. combination of extreme SH and GEV analysis could give more
conservative possible TWL but can yield a slightly higher value.
Nodal Factor
In terms of including periodic fluctuation over a longer Counter Measure Approach
duration such as the 18.6-year nodal cycle, data obtained from a Storm surge elevations were compared with the results from
longer time would be a plausible choice in SL analysis. Even simulating the operation of three storm surge barriers at the
though the S3LR accounted for the seasonal vulnerability, entrances to the NY-NJ Harbor for several storm surge events
particularly focusing on the summer seasons, there was not under three projected SLRs representing conditions in the 2020s,
sufficient analysis of extreme SLs because of the limited amount 2050s, and 2080s (Kim et al., 2013).
of archived data. However, due to numerous uncertainties in In the meantime, wave overtopping (WOT) vulnerability for
SLR and trend analysis, it is highly difficult to define an exact complex types of coastal dikes, which could more intensively
future for SLR. The possible trends and summations of the affect coastal inundation during storms, has been successfully
additional effects described in this paper, such as the summer reproduced (Suh and Kim, 2018). As noted by them, the TWL is
seasonal SLR in response to the inverse barometric effect and a result of the summation of SLs, the inclusion of extreme SH in
thermal expansion, should be incorporated in future related accordance with the S3LR would be regarded as a conservative
studies. Baart et al., (2012) has demonstrated intra-annual SL parameter for the purpose of the prevention of future coastal
fluctuations showing the importance of including the 18.6-year damage.
cycle in regional sea-level estimates. When existing appropriate offensive coastal disaster
prevention structures, such as storm surge barriers or similar
Tide-Surge Interaction infrastructure, incorporate extreme SH with S 3LR based on
In previous studies (Kim and Suh, 2018; Suh and Lee, 2018), possible projections or scenario based SLR trends, they would
researchers allocated synthetic typhoon storm surge modeling by be applied as a conservative approach.
randomly setting the tidal phase to acquire non-linear tidal-surge
interactions. Ascending or descending tidal phase interactions CONCLUSIONS
with the storm surge propagation direction could constructively In the analysis of coastal vulnerability for surge inundation
or destructively interfere with surge heights; however, the total mapping, bath-tub approaches are widely applied because of
water surface elevation can behave non-linearly. It is especially their simple applications in comparing TWL and topographic
important in macro-tidal environments because non-linear tidal data. However, in the hindcasting method of storm surge
current interactions with storm waves enhance or reduce surface inundation due to WOT (Suh and Kim, 2018) the analysis
elevation greatly, as described in a previous paper (Suh and Lee, showed possible threats caused by wave overtopping prior to
2018). meeting the bath-tub condition (under the sufficient positive
Rego and Li (2010) computed the non-linear interaction term free-board condition). That is, the TWL incorporation of high-
in surge simulations and found a negative interaction term near water levels and significant wave heights does not meet the crest
the peak surge, while non-linearity decreased the peak storm level of coastal dikes, and the storm wave-induced inundation
tide. To confirm if this pattern was maintained with a different occurred and spread into landward inundation.
tide, Suh and Lee (2018) used the same wind forcing, with Though the Korean coast has been artificially (i.e., hardly
landfall occurring at low tide and a doubled tidal amplitude on with concrete materials) treated with dikes or levees, some of

Journal of Coastal Research, Special Issue No. 113, 2024

XX
 Extreme Surge Heights Incorporating Long-term Tidal Variation and Sea-level Rise
_________________________________________________________________________________________________
the risks are arising beyond these general viewpoints. In Houston, J.R., 2013. Global sea level projections to 2100 using
particular, the WOT over coastal dikes is continuously methodology of the intergovernmental panel on climate
increasing (Suh and Kim, 2018). As anticipated, vulnerabilities change. Journal of Waterway, Port, Coastal, and Ocean
of the existing coastal infrastructure on WCK and SCK are Engineering, 139, 82‒87.
exposed to SLR, S3LR, and extreme SH as demonstrated. Irish, J.L. and Resio, D.T., 2013. Method for estimating future
Because uncertainties in SLR would result in a diverse range of hurricane flood probabilities and associated uncertainty.
projections and yield different SLR values, further studies on Journal of Waterway, Port, Coastal, and Ocean
long-term non-linear trend should be provided to enhance the Engineering, 139, 126‒134.
reliability. Thus, it would be valuable in coastal disaster Kim, H.J. and Suh, S.W., 2018. Improved hypothetical typhoon
mitigation plans to incorporate any possible combination generation technique for storm surge frequency analyses on
yielding a maximum TWL in an extreme case. the Southwest Korean Coast. Journal of Coastal Research,
Special Issue No. 85, pp. 516‒520.
ACKNOWLEDGMENTS Kim, H.J. and Suh, S.W., 2019. Summer seasonal sea-level rise
This work was supported by the National Research and associated potential storm surge vulnerability of the
Foundation of Korea (NRF) grant funded by the Ministry of West Coast of Korea. Journal of Coastal Research, Special
Science and ICT (MSICT) (No. 2017R1A2B4008613). Issue No. 91, pp. 21‒25.
Kim, N.B.; George, B.S., and Simmons, P.W., 2013.
LITERATURE CITED Hydrodynamic/GIS simulation of storm surge flooding in
Abeysirigunawardena, D.S. and Walker, I.J., 2008. Sea level the NY/NJ harbor system. Storm Surge Barriers to Protect
responses to climatic variability and change in northern New York City: Against the Deluge, ASCE, 37‒57.
British Columbia. Atmosphere-Ocean, 46(3), 277‒296. Pelling, H.E.; Uehara, K., and Green, J.A.M., 2013. The impact
Arns, A.; Dangendorf, S.; Jensen, J.; Talke, S.; Bender, J., and of rapid coastline changes and sea level rise on the tides in
Pattiaratchi, C., 2017. Sea-level rise induced amplification the Bohai Sea, China. Journal of Geophysical Research:
of coastal protection design heights. Scientific reports, 7, Oceans, 118, 3462‒3472.
40171. Rego, J.L. and Li, C., 2010. Nonlinear terms in storm surge
Atkinson, J.; Smith, J.M., and Bender, C., 2013. Sea-level rise predictions: Effect of tide and shelf geometry with case
effects on storm surge and nearshore waves on the Texas study from hurricane Rita. Journal of Geophysical
coast: Influence of landscape and storm characteristics. Research, 115, C06020.
Journal of Waterway, Port, Coastal, and Ocean Suh, K.D.; Yang, Y.C.; Jun, K.C., and Lee, D.Y., 2009. Extreme
Engineering, 139, 98‒117. sea level analysis in coastal waters around Korean
Baart, F.; van Koningsveld, M., and Stive, M.J.F., 2012. Trends peninsula using empirical simulation technique. Journal of
in sea-level trend analysis. Journal of Coastal Research, Korean Society of Coastal and Ocean Engineers, 21(3),
28(2), 311‒315. 254‒265. (in Korean)
Baart, F.; van Gelder, P.H.A.J.M.; de Ronde, J.; van Suh, S.W.; Lee, H.Y.; Kim, H.J., and Fleming, J.G., 2015. An
Koningsveld, M., and Wouters, B., 2012. The effect of the efficient early warning system for typhoon storm surge
18.6-year lunar nodal cycle on regional sea-level rise based on time-varying advisories by coupled ADCIRC and
estimates. Journal of Coastal Research, 28(2), 511‒516. SWAN. Ocean Dynamics, 65(5), 617‒646.
Bell, C.; Vassie, J.M., and Woodworth, P.L., 1999. POL/PSMSL Suh, S.W., 2016 Tidal asymmetry and energy variations due to
Tidal analysis software kit 2000(TASK-2000). Permanent sea-level rise in a macro tidal bay. Journal of Coastal
Service for Mean Sea Level, CCMS Proudman Research, Special Issue No. 75, pp. 765‒769.
Oceanographic Laboratory, Bidston Observatory, Suh, S.W. and Kim, H.J., 2018. Simulation of wave overtopping
Birkenhead, Merseyside, 20p. and inundation over a dike caused typhoon Chaba at
Clara, L.M.; Simionato, C.G.; D’Onofrio, E., and Moreira, D., Marine city, Busan, Korea. Journal of Coastal Research,
2015. Future sea level rise and changes on tides in the Special Issue No. 85, pp. 711‒715.
patagonian continental shelf. Journal of Coastal Research, Suh, S.W. and Lee, H.Y., 2018. Forerunner storm surge under
31(3), 519‒535. macro-tidal environmental conditions in shallow coastal
Feng, X.; Olabarrieta, M., and Valle-Levinson, A., 2016. Storm- zones of the Yellow Sea. Continental Shelf Research, 169,
induced semidiurnal perturbations to surges on the US 1‒16.
Eastern Seaboard. Continental Shelf Research, 114, 54‒71. Watson, P. J., 2019. Updated mean sea-level analysis: South
Hoeke, R.K.; McInnes, K.L., and O’Grady, J.G., 2015. Wind Korea. Journal of Coastal Research, 35(2), 241‒250.
and wave setup contributions to extreme sea levels at a Zhang, K., 2011. Catastrophic inundation from sea level rise and
tropical high island: a stochastic cyclone simulation study its policy implication. Solutions to Coastal Disasters 2011 -
for Apia, Samoa. Journal of Marine Science and Proceedings of the 2011 Solutions to Coastal Disasters
Engineering, 3(3), 1117‒1135. Conference, pp. 502‒510.

Journal of Coastal Research, Special Issue No. 113, 2024

XX