Different Modes of Wave Response over the Past Four Decades: Coastal vs. Open-Ocean Regions

Abstract

Tropical cyclone-induced waves (TCWs) are projected to intensify under global warming, with recent evidence suggesting that their growth outpaces the increase in surface winds. Yet, how TCWs differ between coastal and open-ocean environments remains poorly understood. Here, we investigate TCW characteristics during two climatic periods (1979–2000 and 2001–2023) using a coupled analysis of buoy observations and ERA5 reanalysis. Our results reveal a striking contrast: while open-ocean TCWs exhibited a pronounced intensification of up to 19% (~74 cm) over the past four decades, coastal TCWs show only a muted increase of 26 cm (~8%). This discrepancy is primarily linked to weaker wind forcing and a contraction of effective fetch in coastal regions. On a broader scale, global wave heights (GWs) demonstrate strong temporal and regional variability. The 1979–2000 period featured widespread increases exceeding 10 cm per decade, whereas 2001–2023 displayed pronounced regional disparities, with declines in the Pacific and Indian Oceans but increases in the North Atlantic, Southern Ocean, and Arctic. Notably, the Arctic exhibits a significant rise in extreme wave heights, consistent with reduced ice cover and enhanced wind-driven fetch, highlighting critical feedback to global warming. These findings underscore the importance of distinguishing coastal from open-ocean wave responses when assessing future hazards. By revealing the divergent trajectories of TCWs and GWs under climate change, our study provides a refined framework for understanding storm-induced risks and for improving projections of wave-driven coastal impacts.

1. Introduction

Tropical cyclones (TCs) are a major hazard for many countries, often bringing heavy rainfall [1], storm surges [2,3,4,5], and swells [6], which are also known as long waves [7]. These long waves are typically caused by the influence of strong low-pressure systems [8]. When they reach the shore, they can suddenly generate larger wave heights due to topographic effects, causing damage to coastal infrastructure, ships at sea, and resulting in loss of life and property [9,10,11,12,13,14,15]. Additionally, long waves can gradually erode the coastline [16], leading to destructive geological effects that cause coastal retreat and land loss [17,18]. Furthermore, they increase the risks of maritime operations and cause significant disruptions. With the impact of climate change, many studies suggest that the intensity and frequency of TCs will change in the future [19]. The average intensity is expected to increase, while the average frequency will decrease [20]. To balance this, the frequency of stronger TCs will increase [21], based on theoretical and high-resolution dynamical model predictions, indicated that the greenhouse effect will lead to a global increase in the average intensity of TCs, with an expected increase of 2–11% by 2100. The global average frequency of TCs is projected to decrease by 6–34% [21]. Therefore, it can be confirmed that the intensity of TC will gradually increase in the future.

Under the strengthening of TC intensity due to global warming, waves are expected to change [7]. Shi et al. [7] used ERA5 reanalysis data to show that the maximum wave height and area of TC wave (TCW) have increased globally by about 3%/decade and 6%/decade, respectively. Globally, the energy of TCWs transmitted from the atmosphere to the ocean has increased by approximately 9%/decade. Extreme wave events are largely associated with TCs [7], and many studies have used models to estimate changes in extreme wind and wave events in the future [22,23]. Belmadani et al. [11] using ERA5 reanalysis and satellite altimetry data, assessed the performance of wave models for historical periods. Future predictions show that due to the weakening of subtropical anticyclones, there will be a slight but widespread decrease in seasonal average wave heights [11]. In large areas extending from the coast of Africa to the North American continent, wind-sea and extreme wave heights associated with TCs will significantly increase [11]. Meucci et al. [9] also predicted that by the end of the 21st century, extreme wave events in the North Atlantic will decrease in low to mid-latitude regions (by approximately 5% to 15%) but increase in high-latitude regions (by about 10%). In the North Pacific, extremely significant wave heights (SWHs) in high-latitude regions will increase by 5% to 10%. Previous studies have used numerous modeling approaches to predict changes in future extreme wave events and average wave heights [9,23]. However, there has yet to be an in-depth investigation into the differences and characteristics of TCWs and wave height variations in coastal and open ocean areas. Most prior studies emphasize basin-scale or global perspectives, overlooking the distinct dynamics near coastlines where human and ecological vulnerabilities are highest. A better understanding of how waves will change in coastal regions due to global warming will certainly help in proposing appropriate disaster mitigation measures before the arrival of future TCs, reducing disasters and losses.

In this study, buoy data from the National Data Buoy Center (NDBC) were used to compare with wave height data retrieved from ERA5 reanalysis. This comparison aimed to validate the consistency between in situ measurements and ERA5 wave heights. Afterward, with a focus on the potential discrepancies between TCWs in different regions, the characteristics of TCWs were separated into two main areas: coastal regions vs. open ocean. After highlighting the differences in the characteristics of coastal TCWs relative to existing understanding, the characteristics and long-term variations in global wave heights, as well as potential controlling factors of global wave heights in a warming environment, were investigated. Finally, the migration and changing trends of the maximum wave height areas were presented.

2. Data and Methods

2.1. IBTrACS Best Track Data

The International Best Track Archive for Climate Stewardship (IBTrACS) is a global archive of TC best track data maintained by the National Environmental Satellite, Data, and Information Service (NESDIS) of NOAA. IBTrACS provides comprehensive data for TCs globally, including information from the Atlantic, Pacific, Indian, and Southern Oceans. The best track data released by IBTrACS is compiled from multiple sources, including satellite observations, aircraft reports, ship reports, and other observational platforms [24]. The data is available at various time intervals, typically every 6 h, and includes details such as the TC’s center coordinates, maximum sustained wind speed, minimum central pressure, and intensity classification. The IBTrACS best track data can be accessed via https://www.ncei.noaa.gov/products/international-best-track-archive (accessed on 20 July 2025). In this study, for analysis, data from 1979 to 2023 were collected and processed.

2.2. NDBC Buoy Data

The NDBC, a part of NOAA’s National Weather Service, maintains the world’s largest real-time marine observation network dedicated to maritime safety and environmental monitoring. NDBC’s observing system includes more than 100 moored ocean buoys and about 50 Coastal-Marine Automated Network (C-MAN) stations, complemented by additional volunteer vessel reports, mobile platforms, and partnerships with international organizations.

The NDBC buoy and C-MAN stations provide a wide range of meteorological and oceanographic parameters, including barometric pressure, wind speed and direction, wind gusts, air temperature, relative humidity, sea surface temperature, and wave energy spectra. From these measurements, key wave parameters such as SWHs, dominant and average wave periods, and wave propagation direction are derived. All data undergo rigorous automated quality control procedures, including range checks, temporal continuity checks, and consistency tests between wave height and wave period [25].

Because buoy-based wave observations are not assimilated into the ERA5 reanalysis, NDBC buoy data serve as an independent validation dataset. For the analysis period considered in this study, only those stations that provided continuous, high-quality SWHs records were retained. Specifically, we included stations for which the official Keyhole Markup Language (KML) system could directly supply automatically downloadable buoy data (in total 1077, Figure 1). After that, we excluded any observations collected from ships and restricted the dataset to measurements recorded between 1979 and 2023, and the sea area only. Ship-based observations were excluded due to their relatively random sampling, along with extra uncertainties.

2.3. ERA5 Wave Reanalysis

Hourly SWHs from ERA5, spanning from 1979 to 2023, with a resolution of 0.5° × 0.5° were used in this study. ERA5 employs the WAve Modeling (WAM) model for ocean wave simulations. The WAM model has a horizontal grid size of 28 km and discretizes the wave spectrum into 36 directions and 36 frequencies. In ERA5, ocean wave data assimilation follows an optimal interpolation (OI) approach with a 12 h window. The assimilation process incorporates wave heights from space-borne altimeters, adjusting the wave fields in WAM. During data assimilation, the two-dimensional spectra of SWHs are corrected based on observations using the OI scheme, and the adjusted field serves as the initial condition for the next stage of model integration [7,26]. Meanwhile, for comparison, 10 m wind forcing retrieved from ERA5 was also downloaded. The ERA5 reanalysis was processed and distributed by the ECMWF [26]. This product can be accessed through https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=download (accessed on 20 May 2025).

2.4. Comparison Between Buoy and ERA5 Data

The ERA5 grid points closest to the buoy station locations were selected for comparison. A linear regression analysis was applied to the time series of wave height to evaluate long-term trends. The regression model is expressed as:

$$Y = {\beta }_0 + {\beta }_{1}X$$

(1)

where Y is the dependent variable (wave height), X is the independent variable (time), ${\beta }_0$ is the intercept, and ${\beta }_1$ is the regression coefficient representing the rate of change. Any missing values in the dataset were excluded from the analysis to avoid bias.

To assess the consistency between ERA5 and buoy observations, the Pearson product–moment correlation coefficient (PPMCC) was calculated as:

$$r ={\displaystyle \frac{\sum (X_i-\overline{X})(Y_i-\overline{Y})}{\sqrt{\sum {\left(X_i-\overline{X}\right)}^2\sum {\left(Y_i-\overline{Y}\right)}^2,}}}$$

(2)

where $X_i$ and $Y_i$ denote buoy and ERA5 wave heights, respectively, and $\overline{X} \mathrm{a}\mathrm{n}\mathrm{d} \overline{Y}$ are their corresponding means. Missing values were omitted prior to calculation to ensure reliable correlation estimates. In addition, the root mean square error (RMSE) was computed to quantify the discrepancy between buoy and ERA5 data.

$$RMSE=\sqrt{ {\displaystyle \frac{1}{n}}{\sum }_{i=1}^n{\left(y_{i-}{\widehat{y}}_i\right)}^2,}$$

(3)

where $y_i$ is the buoy observation, ${\widehat{y}}_i$ is the corresponding ERA5 value, and $n$ is the number of valid paired samples. A smaller RMSE indicates a closer agreement between ERA5 and buoy measurements.

3. Results

3.1. Comparison of Wave Heights from ERA5 and NDBC Buoys (Coastal Region vs. Open Ocean)

We first assessed the reliability of ERA5 reanalysis wave heights by comparison with in situ observations from NDBC buoys, focusing on differences between coastal and open-ocean environments. Here, the 500 m isobath is used to distinguish shelf and deep-ocean wave regimes, following Combes et al. [27], who define the shelf as the region extending from the coast to the 500 m isobath, while the term open ocean refers to the region outside the shelf. Previous studies have suggested that reanalysis products may underestimate or misrepresent wave heights due to limited model resolution and unresolved coastal processes [28]. Here, we evaluate whether ERA5 data can be effectively applied to both nearshore and offshore wave height estimations.

Given the large volume of buoy records and their irregular temporal sampling, we adopted a random monthly sampling strategy to ensure physically consistent comparisons rather than averaging over long composite periods. In addition, the random monthly sampling was performed using Python (version 3.10.12) ’s built-in random module, which allows the sampling process to be fully reproducible by preserving the random seed. Aside from applying this algorithm, we did not impose any predefined threshold on the number of available observations per month. An inspection of the sampled dataset shows that fewer than 15% of the selected monthly records contain fewer than 500 observations within a given month.

Specifically, all historical buoy records from 1979–2023 were compiled into continuous time series for each station. From these, one month of valid observations was randomly selected, and the corresponding ERA5 significant wave heights (SWHs; 0.25° × 0.25° grid) at the nearest grid point were extracted for the same period. The ERA5 hourly time series were temporally aligned with buoy measurements, and statistical diagnostics, including root mean square error (RMSE), bias, correlation coefficients, and scatterplot comparisons, were computed.

Results of the comparison are shown in Figure 2. Both ERA5 and buoy observations consistently reveal that SWHs in coastal regions are systematically lower than those in the open ocean. ERA5 estimates yield mean values of 1.591 m in the open ocean and 1.110 m in coastal regions, closely matching buoy observations of 1.564 m and 1.024 m, respectively.

The comparison between the diagonal and regression lines in the scatter plots indicates that, regardless of whether it is for open ocean or coastal regions, ERA5 reanalysis data shows a good ability to reflect SWHs. Generally speaking, the reanalysis product demonstrates a better ability to reflect SWHs in open ocean regions than in nearshore areas, given its smaller RMSE and bias. This result is attributed to that, in nearshore areas, due to the presence of more influencing factors, such as topographic effects and wave-current interaction effects, the conditions and required quality control are more complex [7,29], and thus the WAM model may not be as effective in accurately reflecting wave heights. Bathymetry and wave–current interactions substantially alter nearshore wave climates, producing responses that differ from offshore conditions. Studies show that limited fetch, shallow depths, and enhanced dissipation in systems such as Lake Michigan [30] and the Bohai Sea [31], as well as wave–current coupling on continental shelves [32], can suppress and spatially modulate coastal wave heights. In addition, open ocean SWHs benefit from more sufficient high-quality satellite data for ERA5 WAM model wave height data assimilation [26].

3.2. Manifestation of the Height and the Area of TC Waves (Coastal Region vs. Open Ocean)

We next examine the characteristics of TCWs and compare their differences between open-ocean and coastal regions. To ensure sufficient sample size, the analysis is restricted to the Northern Hemisphere (NH), as extremely reduced land coverage in the Southern Hemisphere severely limits the availability of coastal observations for robust statistics. Figure 3a,b present the composites of SWHs at the stage of maximum TC intensity for all TCs that occurred in the open ocean in two distinct periods: (a) 1979–2000 and (b) 2001–2023. Accordingly, following Shi et al. [7] we divided the satellite era into two equal-length periods to maintain a uniform comparison basis. This approach ensures that any identified differences reflect genuine regime contrasts rather than artifacts of period selection. Figure 3c illustrates the difference between the two periods, with statistically significant regions (95% confidence level) indicated by black dots. The statistical significance shown in Figure 3c was determined using a two-sample Student’s t-test (through Python SciPy library [33]. Specifically, for each grid point, we applied an independent t-test (Welch’s t-test) between the two composited groups, without assuming equal variances. Only grid points with at least three valid observations in both groups were included in the test. Regions with p < 0.05 were marked as statistically significant. In addition, no additional correction for temporal autocorrelation was applied.

Overall, the distribution of TCWs exhibits a pronounced asymmetry, with larger wave heights located to the right (first quadrant) of the TC center. This asymmetry shows great consistency with the pattern of corresponding wind forcing, and the asymmetric wave height, along with the biased footprint, are attributed to the influence of the so-called dangerous semicircle [34,35]. In the NH, the right-hand side of the TC track, particularly the right-front quadrant, is commonly referred to as the dangerous semicircle. On this side of the storm, the rotational winds combine with the translational motion of the cyclone, producing stronger earth-relative winds and consequently higher sea surface waves [34,35,36].

Compared to the period of 1979–2000, the maximum composited TCW derived from 2001–2023 has increased by up to 74 cm, representing an 19% increase during the two climatic periods. Figure 3c shows the spatial pattern of the TCW enhancement between the two climatic periods. The most pronounced increase reaches about 34 cm per decade, occurring predominantly to the right of the TC center. Since larger values also appear in the first quadrant of the storm center, the significant enhancement of wave heights is likely associated with wind forcing and its variability [35].

Figure 4 extends the composite analysis to TCWs in coastal regions. Although an asymmetric pattern is still evident, the spatial signature diverges substantially from that observed over the open oceans (Figure 3). In particular, both the footprint and amplitude of coastal TCWs remain muted across the two climatic periods. The interdecadal comparison (1979–2000 vs. 2001–2023) reveals only a modest increase of 26 cm, equivalent to ~8%. This muted response stands in stark contrast to the pronounced open-ocean intensification observed in the open ocean (Figure 3c).

Collectively, these results demonstrate that the characteristics and long-term variability of coastal TCWs diverge fundamentally from those of the open ocean. This divergence has important implications: storm-induced wave hazards near coastlines cannot be directly inferred from global or basin-scale TCW behavior (e.g., Shi et al. [7]). The striking contrast observed under comparable warming conditions over the past four decades raises a critical unresolved question: what mechanisms govern the suppressed response of coastal TCWs compared to their open-ocean counterparts?

3.3. Possible Mechanism Driving the Contrasting TCW Responses (Coastal vs. Open Ocean)

The observed enhancement of TCWs between the two climatic periods is broadly consistent with the theoretical framework of the dangerous semicircle [34,35,36]. In addition, processes such as wave–current interaction [37], bathymetry and bottom stress [38,39], water level changes [40], and wave interference, while potentially important, are not expected to exert a basin-wide influence and are therefore not considered central to our analysis. Accordingly, we infer that the observed long-term variability in wave heights (shown in Figure 3 and Figure 4) is primarily attributable to changes in wind forcing. Previous studies have emphasized three fundamental drivers of wave growth: wind speed, wind duration, and the size of the wind fetch [41,42]. Guided by this framework, we next examine the interannual variations and long-term trends of these parameters in conjunction with wave height changes.

Figure 5 presents the interannual variations and linear trends of four key TC metrics over the open ocean from 1979 to 2023, including annual mean TCWs at peak intensity (blue line), maximum wind speed (yellow line), wind fetch (green line, defined as the area with wind speed exceeding 10 m/s), and TC translation speed (red line, derived from IBTrACS records). The composited maximum TCW exhibits a robust upward trend of ~22.8 cm per decade, broadly consistent with the concurrent intensification of maximum wind speed (yellow dashed line). Correlation analyses further reveal that interannual TCW variability is strongly associated with maximum TC wind speed (r ≈ 0.97) and wind fetch (r ≈ 0.61), but negatively correlated with TC translation speed (r ≈ –0.26). Importantly, even when substituting reanalysis-derived winds with IBTrACS maximum sustained winds, the correlation with SWH remains high (r ≈ 0.67), reinforcing the robustness of the wind–wave relationship.

These findings are physically intuitive in the context of classical wave generation theory, in which wave growth is fundamentally controlled by wind intensity, the fetch or wind-affected area, and the effective wind duration, which is closely tied to TC translation speed (e.g., figure 3 in Lin et al. [43]). The long-term trends (dashed lines in Figure 5) further suggest that the significant increase in open-ocean TCWs over the past four decades is most likely driven by the intensification of wind speed. While an expansion of fetch would also be expected to promote wave growth, the fetch over the open ocean (green dashed line) has remained nearly unchanged during this period. By contrast, the slight increase in TC translation speed (red dashed line) may have exerted a counteracting influence, limiting further enhancement of wave heights.

Figure 6 depicts the long-term variations in TCWs and wind-related parameters in coastal regions from 1979 to 2023. In coastal environments, TCWs show only a modest increase (blue dashed line), consistent with the two-period comparison presented earlier (Figure 4c). Compared with the open ocean, the magnitudes of changes in TCWs, wind speed, fetch, and TC translation speed are substantially weaker. From a long-term perspective, the slight increase in coastal TCWs appears primarily attributable to a modest rise in wind speed (~0.24 m/s per decade). Meanwhile, the composited TC translation speed has remained nearly unchanged, and fetch has exhibited a weak contraction, which may have partially offset the contribution of wind-speed intensification to TCWs growth.

Together, these findings suggest that the limited growth of coastal TCWs is primarily attributable to the weaker enhancement of wind forcing relative to the open ocean, combined with a contraction of storm wind fetches. An interesting paradox arises: why have nearshore regions experienced wind intensification over the past four decades while simultaneously exhibiting a reduction in fetch? Liao and Kaihatu [38] and Chen et al. [39] reported that nearshore bathymetric friction can substantially suppress the efficiency of wind-wave generation. Furthermore, coastal infrastructure such as harbors, breakwaters, and artificial port structures constructed over the past four decades may have obscured or blocked portions of the wind-exposed area, further reducing effective fetch (e.g., Saengsupavanich et al. [44]; Foti et al. [45]). In coastal regions, where wind intensification over the past four decades has been relatively weak, these additional suppressive effects likely contributed to the negligible or even slightly negative trends in fetch. Nonetheless, more comprehensive evidence is required before drawing definitive conclusions.

4. Characteristic of Global Wave Height in a Warming World (1979~2023)

Beyond TCWs, we extended our analysis to global wave heights (GWs), which integrate both TCW contributions and background oceanic variability. Unlike TCWs that focus on storm centers, GWs provide a basin-scale perspective, offering insights into large-scale changes over the past four decades. Figure 7a presents the averaged GWs for the climatic period from 1979 to 2000. Figure 7b displays the averaged GWs from 2001 to 2023. Comparing this with Figure 7a, we find that the GWs characteristics for the two periods are consistent, primarily influenced by the strong westerlies in the mid-latitudes. Nevertheless, the difference between the two periods (Figure 7c) reveals clear regional disparities. Overall, most oceanic regions exhibit an increasing trend in wave heights; however, notable exceptions include a slight decline in the northeast Pacific and localized decreases in parts of the Southern Ocean near continental boundaries.

The temporal evolution of GWs is further illustrated in Figure 8. During 1979–2000, global wave heights experienced a widespread increase, with maximum trends exceeding 10 cm per decade. By contrast, 2001–2023 does not exhibit a uniform global rise; instead, strong regional contrasts emerge. Declines dominate the Pacific and Indian Oceans, whereas increases are concentrated in the North Atlantic and the Southern Ocean south of 60° S. A comparison of Figure 8a,b underscores that the trajectory of global wave variability depends strongly on the climatic period considered.

To assess potential drivers, we examined correlations between GWs and 10 m wind speeds (ERA5) for 1979–2023. The results (Figure 9) show robust positive correlations across most ocean basins, confirming that wind strength is the dominant driver of long-term GW variability. Exceptions occur in subtropical regions, where correlations weaken (blue patches in Figure 9), likely reflecting the influence of multiple atmospheric systems [46]. These findings are consistent with previous studies reporting heterogeneous regional wave responses despite a global intensification of surface winds over the past four decades [29,47,48,49]. From a global perspective, however, our analysis indicates that changes in wind strength remain the primary control on GW evolution.

Finally, we also evaluated extreme SWHs, which often drive the most damaging wave-related disasters [5]. Figure 10a,b display maximum wave heights for 1979–2000 and 2001–2023, with their differences shown in Figure 10c. The results highlight not only higher values in mid-latitudes influenced by westerlies but also anomalous wave patterns tracing the storm tracks of the North Atlantic and Pacific. This finding reinforces the critical role of TCs in shaping extreme wave climatology, consistent with Shi et al. [7]. Although maximum SWHs in the Pacific are relatively lower than those in the North Atlantic, the spatial distribution of maximum SWHs (Figure 10) still highlights the need to remain cautious about the extreme SWHs associated with TCs.

Furthermore, it is worth noting that, aside from TCs causing extreme wave heights in tropical and subtropical regions, the Arctic Ocean shows a significant increase in maximum wave heights (Figure 10c). Under warming scenarios, the reduction in ice cover expands the wind fetch areas in the Arctic, making the effects of waves more pronounced. The enhancement of waves also provides positive feedback for the retreat of ice sheet edges [50]. This process, coupled with glacier melt and reduced albedo, creates adverse conditions that accelerate global warming. This positive feedback highlights the Arctic as an emerging hotspot for extreme wave changes under climate change. Collectively, these results demonstrate that focusing solely on global mean wave variability risks obscuring critical regional contrasts. The distinct signatures of coastal versus open-ocean environments, as highlighted in our earlier analyses, remain essential for accurately assessing the hazards posed by waves in a warming climate.

5. Conclusions

This study systematically investigates the evolution and variation in TCWs and global SWHs in both nearshore and open ocean regions from 1979 to 2023. We observed that the relationship between TCs and wave heights exhibits significant regional disparities, particularly between nearshore and open ocean environments. While TCWs in open oceans showed a marked increase in magnitude and footprint over the past decades, nearshore regions demonstrated less pronounced growth. This divergence underscores the need to reassess the hazards posed by storm-induced waves in coastal zones, which cannot be inferred directly from global or basin-scale trends (e.g., Shi et al. [7]). Long-term composite analyses indicate that open-ocean TCW intensification is primarily driven by stronger winds, with little contribution from fetch expansion and partially offset by slightly faster storm translation. By contrast, the muted coastal response is attributable to weaker wind forcing combined with a contraction of effective fetch, likely further suppressed by bathymetric friction and anthropogenic structures. Together, these findings highlight the necessity of distinguishing between coastal and open-ocean wave responses to accurately evaluate future risks to vulnerable coastal communities.

Despite the robust signals identified here, several limitations must be acknowledged. ERA5 and its associated WAM system (≈28 km; Hersbach et al. [51]; Janssen [52]; Komen et al. [53]) do not resolve nearshore wave-transformation processes, including refraction, diffraction, shoaling, and wave breaking, and therefore our coastal wave heights should be interpreted as lower-bound estimates. Additionally, ERA5 does not assimilate in situ buoy observations, such as those from NDBC, which may introduce regional wind–wave biases (Bidlot et al. [54]). Finally, the relative importance of the identified mechanisms may vary across coastlines with distinct bathymetry and geo-morphology, and caution is warranted when generalizing these findings beyond the study region (Masselink & Hughes [55]).

On a broader scale, global SWHs have demonstrated a pronounced and widespread increase during 1979–2000, with the most significant trends exceeding 10 cm per decade in the southern hemisphere’s mid-latitudes. By contrast, the period 2001–2023 no longer shows a uniform global rise but instead exhibits distinct regional disparities: decreasing trends dominate the Pacific and Indian Oceans, while increases are primarily observed in the North Atlantic and Southern Ocean. This pronounced temporal and spatial variability underscores the heterogeneous nature of global wave responses to climate forcing. Another crucial observation is the ongoing increase in extreme wave heights, particularly in the Arctic, where retreating ice cover has expanded wind-driven areas (fetch), allowing for more pronounced wave action. This phenomenon amplifies the feedback loop of global warming, as it accelerates ice sheet retreat and glacier melt, thus further contributing to global climate change. The result highlights the need for heightened attention to extreme wave heights in polar regions.

In conclusion, this study emphasizes the importance of distinguishing between nearshore and open ocean wave height dynamics when assessing the impacts of TCs, particularly in the context of global warming. The findings underline the necessity of improving wave forecasting models for coastal areas and continuing the examination of extreme wave events, especially in vulnerable regions such as the Arctic. With coastal populations and infrastructure facing disproportionate risks, future research and policy must target region-specific wave responses to ensure more effective adaptation and mitigation strategies in a warming world.