Geographical Distribution and Recent Change in the Meteorological Event Causing the Annual Maximum Wave Height and Storm Surge Around Japan

Abstract

The meteorological events that induced the annual maximum wave heights and storm surges were collected from 35 wave observatories and 50 tide observatories in Japan. The geographical distributions of the meteorological events most responsible for the maximum wave height and storm surge were similar everywhere except the southern part of Hokkaido. The proportions of typhoon- and explosive cyclogenesis-induced maximum wave heights and storm surges increased at many observatories when comparing data collected before 1994 and after 1995. The potential contributions of climate change to these increases were subsequently investigated and discussed based on the results of a detailed analysis conducted in 10-year periods.

1 Introduction

In 2020, the basic policy for coastal protection in Japan was revised to adapt to anticipated climate change. According to this policy, coastal managers must revise the basic coastal protection plan for each coastal unit considering the average value of future projections under the RCP2.6 scenario by 2025. The RCP2.6 scenario is one of the representative concentration pathway scenarios considered in the Fifth Assessment Report of Working Group I of the Intergovernmental Panel on Climate Change.

Wave height and storm surge are among the most critical parameters to consider when designing coastal facilities. Recently, various methods have been proposed to predict changes in these parameters considering climate change; these methods generally require the assumption of a typical meteorological event responsible for causing high waves or storm surges [1]. Furthermore, the results of 4 ℃ warming simulations indicate a high potential for an increased frequency of intense category 4–5 tropical cyclones over the waters south of Japan [2]. Thus, the areas that must be considered when evaluating the impact of typhoons may change in the future. Moreover, it is necessary to confirm whether the effects of climate change are already affecting the observed wave height and storm surge.

In this study, the meteorological events that were responsible for the maximum wave heights and storm surges were determined based on long-term observation data collected by wave and tide observatories in Japan. Changes in the contributions of typhoons and explosive cyclogenesis to high wave and high storm surge events were subsequently investigated.

2 Materials and Methods

2.1 Data Collection

Data from the 35 wave observatories and 50 tide observatories in Japan with observation periods longer than 30 years were selected for use in this study. The latest wave height and tide level data were obtained in 2019 and 2020, respectively. The maximum durations of the wave height and tide observations were 50 years and 97 years, respectively. The annual maximum values of wave height and storm surge were selected for each observatory, and the meteorological events that were responsible for those values were determined. Although the meteorological events that caused the annual maximum values were described in the dataset by each observatory, weather diagrams from the time were also collected and revised if necessary. The meteorological events were classified into four event categories: typhoon, low pressure, explosive cyclogenesis, and others. If the same maximum value was observed multiple times within a year, all values were analyzed. Thus, the total number of determined meteorological events was greater than the number of years in the observation period for several observatories.

Note that explosive cyclogenesis is a rapidly strengthening storm generally defined by a pressure drop 24 of hPa within 24 h. However, there are some variations in this benchmark according to the latitude of the storm. Explosive cyclogenesis was detected in this study based on the definitions proposed by Mori et al. [3] and Yoshida and Asuma [4] using weather diagrams collected from 1958 to 2011 and from 2012, respectively, as well as reports published by the Japan Meteorological Agency and the annual report on nationwide ocean wave information network for ports and harbors.

2.2 Data Analysis

The obtained data were separated into two groups by period (from start to 1994 and from 1995 to end) to investigate recent changes in the contribution of meteorological events to the annual maximum wave height and storm surge. The 1994/1995 boundary was determined as the sample sizes of the wave data were nearly the same on either side; the same separation was applied to the tide data. As the tide observation periods were longer than the wave observation periods for most observatories, storm surges observed before 1968 were not used when comparing the two groups.

Fisher’s exact test was used to determine the presence of a significant difference between the two groups. The significances of the deviation from a null hypothesis (p value) were calculated using statistical computing platform R version 3.5.2. Data from observatories exhibiting significant or nearly significant differences were further separated into 10-year periods that were analyzed in detail to assess the cause of the difference.

3 Results

3.1 Distribution of Responsible Meteorological Events

The type of meteorological event most responsible for the annual maximum wave height in all collected data was determined and the resulting distribution is presented in Fig. 1. Typhoons were the most responsible events at 16 of the35 observatories, most of which face the Pacific Ocean. Explosive cyclogenesis was dominant in the northern part of Japan (Hokkaido and Tohoku district), and low pressure was dominant on the coast from Ishikawa prefecture to Fukuoka prefecture. These geographical distribution patterns were nearly the same for the storm surge (Fig. 2) with clearer geographical boundaries than for the wave height.

Fig. 1

Distribution of meteorological events responsible for the annual maximum wave height at each wave observatory. The names of the observatories are listed below, and the number of collected data points is described in the following parentheses. 1: Tomakomai (50), 2: Kojohama (39), 3: Mutsu Ogawara (46), 4: Hachinohe (49), 5: Miyako (30), 6: Kamaishi (40), 7: Sendai New Port (41), 8: Souma (37), 9: Onahama (41), 10: Hitachinaka (41), 11: Kashima (46), 12: Jonan (35), 13: Shiono Misaki (48), 14: Kobe (49), 15: Eigashima (40), 16: Shibushi (40), 17: Ioujima (46), 18: Tower Sougou (40), 19: Genkainada (38), 20: Ainoshima (46), 21: Hamada (46), 22: Hiedu (39), 23: Tottori (41), 24: Fukui (36), 25: Kanazawa (49), 26: Wajima (41), 27: Tanaka (40), 28: Sakata (49), 29: Akita (38), 30: Fukaura (36), 31: Setana (37), 32: Rumoi (50), 33: Nase (43), 34: Nakagusuku-wan (47), 35: Naha (47)

Fig. 2

Distribution of meteorological events responsible for the annual maximum storm surge at each tide observatory. The names of the observatories are listed below, and the number of collected data is described in the following parentheses. 1: Wakkanai (71), 2: Abashiri (71), 3: Hanasaki (71), 4: Kushiro (83), 5: Hakodate (74), 6: Miyako (89), 7: Ofunato (61), 8: Ayukawa (79); 9: Onahama (72), 10: Mera (91), 11: Tokyo (70), 12: Okada (80), 13: Chichijima (51), 14: Uchiura (87), 15: Shimizu Minato (74), 16: Maisaka (63), 17: Nagoya (73), 18: Toba (81), 19: Owase (66), 20: Uragami (83), 21: Kushimoto (94), 22: Shirahama (54), 23: Wakayama (80); 24: Tanwa (82), 25: Osaka (87), 26: Kobe (103), 27: Sumoto (85), 28: Uno (72), 29: Matsuyama (72), 30: Takamatsu (74), 31: Komatsushima (82), 32: Muroto Misaki (53), 33: Kochi (72), 34: Tosa Shimizu (93), 35: Uwajima (78), 36: Aburatsu (95), 37: Kagoshima (76), 38: Makurazaki (70), 39: Naha (55), 40: Ishigaki (60), 41: Oura (55), 42: Kuchinotsu (57), 43: Nagasaki (58), 44: Fukue (58), 45: Hamada (38), 46: Sakai (98), 47: Saigo (71), 48: Maizuru (59), 49: Toyama (59), 50: Fukaura (52)

3.2 Changes in Annual Maximum Wave Height

The proportion of typhoon-induced annual maximum wave heights increased between the two groups at 25 wave observatories (Fig. 3). Although statistically significant differences (p < 0.05) were limited to the Naha observatory (no. 35, p = 0.020), the Sendai New Port (no. 7, p = 0.050) and Shiono Misaki (no. 13, p = 0.065) observatories exhibited nearly significant p-values. The number of observatories where typhoons were the dominant cause of annual maximum wave height slightly increased from 15 (up to 1994) to 16 (1995 onward).

Fig. 3

Change in the proportion of typhoon-induced annual maximum wave heights at 35 wave observatories. Each observatory identification number is the same as that described in Fig. 1

The proportions of typhoon-induced annual maximum wave height steadily increased every decade at Naha (Fig. 4). Although the Sendai New Port observatory exhibited a similar change, there was no difference between the previous (2000–2009) and most recent (2010–2019) decade. In contrast, the increase in typhoon-induced annual maximum wave heights was unclear at Shiono Misaki, except following the oldest decade considered (1970–1979).

Fig. 4

Decadal change in the proportion of typhoon-induced annual maximum wave heights (typhoon-induced/observed)

The proportion of explosive cyclogenesis-induced annual maximum wave heights increased between two groups at 13 observatories (Fig. 5). However, only the Tottori observatory (no. 23) indicated a statistically significant difference (p = 0.040). The number of observatories where explosive cyclogenesis was dominant the cause of the annual maximum wave height slightly increased from 12 (up to 1994) to 13 (1995 onward).

Fig. 5

Change in the proportion of explosive cyclogenesis-induced annual maximum wave heights at 35 wave observatories. Each observatory identification number is the same as that described in Fig. 1

Interestingly, the proportions of explosive cyclogenesis-induced annual maximum wave heights gradually decreased from 0.3 to 0 over the last two decades (2000–2019) at Tottori (Fig. 6).

Fig. 6

Decadal change in the proportion of explosive cyclogenesis-induced annual maximum wave heights (explosive cyclogenesis -induced/observed) at Tottori

3.3 Changes in Annual Maximum Storm Surge

The proportion of typhoon-induced annual maximum storm surges increased between the two groups at 37 tide observatories (Fig. 7). No observatory indicated statistically significant difference (p < 0.05), though the Mera (no. 10, p = 0.086) and Okada (no. 12, p = 0.064) observatories exhibited nearly significant p-values. The number of observatories where typhoons were the dominant cause of the annual maximum wave height increased from 35 (up to 1994) to 38 (1995 onward).

Fig. 7

Change in the proportion of typhoon-induced annual maximum storm surges at 50 tide observatories. Each observatory identification number is the same as that described in Fig. 2

At the Mera observatory, the proportion of typhoon-induced annual maximum storm surges gradually decreased during 1951–1970, increased during 1981–1990, obviously decreased during 1991–2010, and increased again during 2011–2020 (Fig. 8).

Fig. 8

Decadal change in the proportion of typhoon-induced annual maximum storm surge (typhoon-induced/observed)

The proportion of explosive cyclogenesis-induced annual maximum storm surges increased between the two groups at 30 tide observatories (Fig. 9). No observatory exhibited a statistically significant difference (p < 0.05) between the two data groups, though the number of observatories where explosive cyclogenesis was the dominant cause of the annual maximum storm surge decreased slightly from 11 (up to 1994) to 9 (1995 onward).

Fig. 9

Change in the proportion of explosive cyclogenesis-induced annual maximum storm surges at 50 tide observatories. Each observatory identification number is the same as that described in Fig. 2

4 Discussion

4.1 Geographical Distribution of Meteorological Events

The geographical distributions of meteorological events were similar for wave height (Fig. 1) and storm surge (Fig. 2), though there were several exceptions for wave height. While explosive cyclogenesis was dominant in the Hokkaido and Tohoku districts, low pressures were dominant at Tomakomai (no. 1) and Kojohama (no. 2), which are located in the southern part of Hokkaido.

This difference between the dominant causes of storm surge and wave height is likely a result of the differences between their respective generation mechanisms. Storm surge tends to be large near the center of a typhoon because it is strongly affected by local decreases in air pressure and coastward wind. In contrast, wave height depends on the strength of the wind and the distance over which it blows (known as fetch). Low pressure-induced winds can reach as far north as the southern coast of Hokkaido with long fetches; thus, low pressure could be dominant cause of wave height in these areas.

4.2 Recent Increase in Typhoon-Induced Annual Maximum Wave Heights

The typhoon-induced annual maximum wave heights (Fig. 3) and storm surges (Fig. 7) increased at many observatories. The observed increase in the Hokkaido and Tohoku districts may coincide with the Japan Meteorological Agency’s report of a northward shift in the latitude of maximum typhoon intensity [5]. However, we must be cautious of concluding that all results were caused by climate change as few observatories indicated statistically significant differences.

One potential limitation on these differences is the selection of the boundary year (1994/1995), which was simply determined to evenly divide the wave height data points available in this study. This boundary might not be appropriate for determining whether the observed changes were caused by climate change. This possibility is supported by the results showing different time series trends among observatories according to decade (Figs. 4, 6, and 8). Moreover, the time when the effects of climate change appear may vary depending on the region.

The second potential source of limitation is the shortage of wave observation periods. The decadal time series analyses of storm surge data indicated that the proportion of typhoon-induced annual maxima was also high prior to 1970 (Fig. 8), suggesting the existence of periodic fluctuations in wave height and storm surge. The Japan Meteorological Agency reported that no long-term trend has been observed in the number of strong typhoons or in their percentage among all typhoons [5]. However, the existence of decadal oscillation in typhoon behavior is known in the North Pacific [6]. The recent increase in typhoon-induced annual maximum wave heights may represent only one aspect of these fluctuations. Longer wave observation periods are necessary to assess this issue.

4.3 Recent Increase in Explosive Cyclogenesis-Induced Annual Maximum Wave Heights

The proportion of explosive cyclogenesis-induced annual maximum wave heights clearly increased at many observatories (Figs. 5 and 9). Koike et al. [7] projected that the number of intense explosive cyclogenesis tends to increase in the future. Although the specific mechanisms through which climate change affects the occurrence of explosive cyclogenesis are unknown, these changes are clearly a critical issue for disaster prevention departments such as coastal managers.

5 Conclusions

This study determined the meteorological events that induced annual maximum wave height and storm surge at each coastal observatory in Japan. The geographical distributions of the meteorological events most responsible for maximum wave heights and storm surges were similar everywhere except the southern part of Hokkaido. The proportions of typhoon-induced annual maximum wave heights and storm surges have increased at many observatories, as have the proportions of explosive cyclogenesis-induced annual maximum wave heights and storm surges. However, as few observatories reported statistically significant differences between data groups, we must be careful when concluding that these results were caused by climate change.