An Investigation on the Variations of Sea Level Due to Meteorological Disturbances on the Coast of Japanese Islands(II)* Storms Surges on the Coast of the Japan Sea

Journal of the Oceanographical Society of Japan Vol.24, No.4, pp. 178 to 190, August 1968

An Investigation on the Variations of Sea Level due to Meteorological Disturbances on the Coast of Japanese Islands(II)* Storms Surges on the Coast of the Japan Sea

Ichiro ISOZAKI**

Abstract: Storm surges on the eastern coastal line of the Japan Sea are studied. In contrast

with the Pacific coast, extraordinary destructive surges hardly develop there because shallow waters such as bays or a continental shelf are comparatively small. Generally speaking, the

effect of atmospheric pressure is roughly hydrostatic, and northeasterly winds cause the descent

of sea level and southwesterly winds the ascent of it since the shore line runs from southwest to northeast. However, the fluctuations of sea level are different remarkably according to the

course of atmospheric disturbances as well as topography of the coast. Case studies are made in detail for four storms which took different courses. In some cases we can clearly recognize

a typical external surge which follows the storm considerably later at a very low speed of about

3•`4m/sec along the continental shelf from the southern entrance of the sea to Noto Peninsula. Its low speed is explained by assuming a shelf wave of Robinson's type. A curious fact that

the sea level sinks before the arrival of the storm is also discussed.

to northeast, we can find peculiar surges on the 1. Introduction San'in coast, western half of the region in ques- In Japan, severe storm surges occur mainly in tion, which progresses very slowly as a kind of bays on Pacific coast having their mouth south- free wave and arrives at the coast considerably ward, such as Tokyo Bay, Ise Bay and Osaka late after the storm passed away. SHOJI (1961) Bay. Indeed, at the head of these bays, ab- suggested that this wave might be an internal normal damages have frequently been caused by Kelvin wave induced by a progressive typhoon, the storm surges due to typhoons. Therefore but he could not sufficiently verify validity of the storm surges in such bays were studied ex- the assumption because of scantiness of observed tensively and useful results were obtained by data. many investigators. On the other hand, the When a typhoon crosses West Japan from storm surge on the coast facing open sea directly, south to north and then moves to the eastern especially on the coast of the Japan Sea, has been part of the Japan Sea, atmospheric pressure falls scarecely examined on account of its insignificant considerably and strong northerly winds blow height and rather scanty observation. over the San'in coast. Nevertheless sea level UNOKI (1959) pointed out that there were no does not rise but falls (TANIOKA, 1959). The remarkable storm surges in recent several ten cause of this phenomenon is not yet, well years on the coast of the Japan Sea, and that the known. height of surge was less than 100 cm except for In this paper, the author attempts to clarify Iwasaki. The variation of sea level differs re- the features of such storm surges on the Japan markably according to the course of atmospheric Sea coast of Japan. disturbances. When a typhoon moves across the middle part of the Japan Sea from southwest 2. Data and method of analysis

* Received May 29, 1968 Hourly readings of sea level at fourteen tide ** Meteorological Research Institute, Tokyo gage stations are used. Locations of the stations

(32) An Investigation on the Variations of Sea Level due to Meteorological Disturbances 179 on the Coast of Japanese Islands (II)

Table 1. List of tide gage stations.

* Authority G: Geographical Survey Institute Fig. 1. Location of tide gage stations and maxi- H: Hydrographic Office mum storm surge observed at each station M: Japan Meteorological Agency during1953•`1962. are shown in Fig. 1 and Table 1. Eleven of corded at each station is presentied in Fig. 1. them are situated along the Japan Sea coast, At all stations except Iwasaki, the surge height while Tomie, Izuhara and Shimonoseki face the was less than 1 m. Only at Iw4saki, we can East China Sea, the Korean Channel and the find a height of surge higher then 1m. The Straits of Kammon, respectively. The period surge with the peak value of 122cm, was caused for the used data is ten years from 1953 to by Typhoon Marie, Sept. 27, 1954. Nine of the 1962. fourteen surges mentioned above were generated The height of meteorological tide or storm by extratropical cyclones which passed through surge was obtained by subtracting predicted tides the Japan Sea in winter or spring, and five were from the observed ones. In predicting the astro- due to typhoons which crossed the Japan Sea. nomical tide, fifteen tidal constituents were taken At five stations, Wakkanai, Iwasaki, Sakai, in most cases and the monthly mean sea level Tomie and Shimonoseki, the biggest surge was was assumed to coincide with the mean of ob- brought about by typhoons and at other stations served levels. The astronomical tide is incons- it was attributed to extratropical cyclones. This picuous in the Japan Sea and the estimated is in contrast with the circumstances on the surge height is of sufficient accuracy. In the Pacific coast, where remarkable storm surges neiboring seas to which Izuhara and Tomie face, with height exceeding three meters or so, are however, the astronomical tide is rather great, always caused by typhoons. so thirty constituents of tide were used (See The tracks of the fourteen atmospheric dis- ISOZAKI, 1968). turbances in question are given in Fig. 2. From The number of atmospheric disturbances which the figure we can clearly recognize that the severe caused a storm surge with height exceeding 40 surges on the Japan Sea coast were never caused cm at any of the tide gage stations on the coast of by the storms which progressed eastward on the the Japan Sea was fourteen for ten years. Table Pacific side of Japan. It should be also kept in 2 gives the maximum height in each case at the mind that northerly winter monsoon from the respective stations. The maximum height re- Asian Continent does not generate heavy storm

(33) 180 Jour. Oceanogr. Soc. Japan, Vol.24, No.4 (1968)

(34) An Investigation on the Variations of Sea Level due to Meteorological Disturbances 181 on the Coast of Japanese Islands (II)

Fig. 2. The tracks of typhoons and extratropical cyclones which caused severe storm surges Fig. 3. Track of Typhoon Sarah, Sept. 17-18, on the Japan Sea coast of Japan. 1959. Thin lines indicate the typhoon position when the maximum storm surge appeared at surges in spite of its great strength, long dura- each station. tion and direction normal to the coast, because of narrowness of shallow water in the Japan Sea. The storms which bring about surges on the 18th of September, 1959. The central pressure Japan Sea coast may be roughly divided into of the typhoon kept about 970mb during its three groups according to their tracks since the passage over the Japan Sea, and considerable surges generated by them are considerably dif- damages were caused on the Japan Sea coast by ferent from each other. They are (A) storms accompanied wind waves and storm surges. that enter the Japan Sea after crossing Korea or The track of the typnoon is represented by a Korean Channel and moves away northeast heavy line in Fig. 3. The successive position of through the middle of the Japan Sea, (B) storms the center of typhoon is connected by a straight that travel northward over the western part of line with the station at which highest surge is the Japan Sea after crossing Kyushu, and (C) observed simultaneously. It is noticeable that the storms that advance from West Japan toward sea level becomes highest considerably later than the eastern part of the Japan Sea. Atmospheric the time of the nearest approach of the strom, disturbances belonging to Group A are most im- except in the southwestern and northeastern part portant as the source of storm surges. In the of the Sea where the elevation attains its maxi- following sections typical examples of each group mum at the nearest approach of the typhoon. will be examined somewhat in detail. The time lag of the surge from the arrival of the storm is especially large on the coast west of Noto Peninsula. At Shimonoseki and Saigo 3. Case studies we can find two peak levels, the first of which 3.1 Case A: An example where a storm appears at the time of nearest approach of the passes through the middle of the Japan typhoon. Sea from southwest to northeast Time variations of surge height at various Typhoon Sarah entered the Japan Sea through station for five days are presented in Fig. 4. Korean Channel and progressed northeast at a The maximum height is 97 cm at Iwasaki, 56 mean speed of about 45km/hr on the 17th•` cm at Tonoura and 52 cm at Sakai. The dotted

(35) 182 Jour. Oceanogr. Soc. Japan, Vol.24, No.4 (1968)

indicated by dash-dotted line. It should be noticed that it splits into two parts at Wajima located at the top of Noto Peninsula. Thick arrows represent the time interval in which the surge height exceeds half of the maximum height at each station. The values of maximum height are shown by numerals written close to the arrows, but two values are given when double peaks appeared. From Figs. 3, 4 and 5 it is suggested that the mechanism of the generation of storm surges is different for the following four coastal waters: from Tomie to Shimonoseki, from Shimonoseki to Wajima, from Wajima to Iwasaki, and from Oshoro to Wakkanai. First, in the shallow water just to the west of the Japan Sea in which Tomie and Shimono- seki are located, the peak of storm surge appears about the time of nearest approach of the typhoon. The anomaly of sea level is nearly equal to the value estimated hydrostatically from the fall of atmospheric pressure, and the effect of wind seems to be small. Meanwhile, two peaks of sea level are recorded at Shimonseki, facing to Straits of Kammon through which the shallow water in question is connected with the

Fig. 4. Storm surge elevations at various tide gage stations during Typhoon Sarah, Sept. 16- 20, 1959. Ordinate : elevation above predicted tide (cm), Absissa : time JST, Dash line : connection of minimum pressure time of each station.

line in the figure connects the time when the atmospheric pressure becomes minimum at respective stations, or roughly the time when the storm is nearest. Relation between the occurrence times of the maximum surge height and the minimum atmospheric pressure is graphically re- Fig. 5. Storm surge propagation diagram for presented in Fig. 5, in which the abscissa is the Typhoon Sarah. Ordinate: time JST, Absissa: distance measured from southwest to northeast distance along the coast of the Japan Sea, Thick along the coast and the ordinate the time in arrows : duration of storm surges larger than JST. In the figure the fine curves are isobars the half of maximum surge height, Dash-dotted and the heavy broken curve passing through line : surge peak propagation line with numerals the pressure trough shows the propagation of of maximum surge height for each station (cm), pressure minimum. On the other hand, the Thin full line : isobars, Thick dash line : atmos- propagation curve of maximum surge height is pheric pressure minimum propagation line.

(36) An Investigation on the Variations of Sea Level due to Meteorological Disturbances 183 on the Coast of Japanese Islands (II)

Seto Inland Sea. The second peak may be due Finally, on the west coast of Hokkaido repre-

to the storm surge in the Seto Inland Sea. sented by Oshoro and Wakkanai, the peak of

Second, on the coast extending from the the surge height appears about the time when

southwestern entrance of the Sea to Noto Penin- the storm center approaches nearest. The maxi-

sula, which includes San'in coast and west half mum height roughly accords with what is eva- of Hokuriku coast, the surge appears consider- luated from the fall of atmospheric pressure

ably late after the arrival of the storm center, based on the assumption of hydrostatic equi-

and the time lag increases with distance from librium. It is noticeable that the surge continues

the entrance of the Sea. Namely, it is 9 hours long after the storm has passed far away and

at Tonoura, 15.5 hours at Sakai 20 hours at the strong wind has dropped, but the reason for

Miyazu and Maizuru. Moreover, the faint trace this is not clear at present. Possibly the arrival

of this surge is recognized as far north as at of a shelf wave with a low speed may be also

Wajima with a time lag of about 30 hours as suggested.

seen in Fig. 4. The mean velocity of the surge The Typhoon Emma, Sept. 10-11, 1956, took

is estimated at 3.5m/sec approximately, while the similar course as the Typhoon Sarah and that of the atmospheric disturbance is 13m/sec. caused a similar surge. Moreover, as will be

When the surge arrived in the neighborhood of seen in Fig. 2, most of the surges in the Japan

the station, the atmospheric pressure had returned Sea are raised by both extratropical and tropical to nearly normal and the wind had dropped, since cyclones which take the course from southwest to the storm was very distant. Therefore, it is northeast in the Japan Sea. Accordingly, fluctua-

suggested that this surge is not a forced wave tions of sea level stated here will be considered

following the typhoon directly but an external to be typical on the eastern boundary water of

wave in essence. We may regard it as a kind the Japan Sea. In case of extratropical cyclones,

of boundary wave which travels along. the con- however, variations of sea level are rather com-

tinental shelf of San'in coast after generated in plicated and the features of surge described

the west shallow water adjacent to the Japan above becomes somewhat vague compared with

Sea. The low velocity of surge can be explained the surges due to typhoons, because the former

when we assume the shelf wave of Robinson's storms are weaker in intensity, distributions of

type as discussed later (Section 4). atmospheric pressure and wind are more com-

On the other hand, in the coastal water north- plicated as expected from existence of fronts, and the movement, growth and decay of the east of Noto Peninsula where Wajima, Kashiwa- storms differ considerably case by case. zaki and Nezugaseki are located, the time interval

from the occurence of minimum pressure to that 3.2 Case B: An example when a storm of maximum surge height is about 7•`8 hours

everywhere, suggesting that the surge is not progresses northward through the western Japan Sea progressive. Moreover, the mature stage of the surge is almost coincident with the period of Typhoon June which landed at the southern

strong southwesterly wind which prevails behind end of Kyushu with the central pressure of 956

the center of typhoon. Since this wind blows mb at 15h of Sept. 13, 1954, entered the Japan

nearly parallel to the coast, we may say that the Sea at the midnight from 13th to 14th after it

surge is not attributed to direct effect of wind set- had crossed Kyushu from south to north. The

up but to mass transport which is directed to the track of this storm is presented in Fig. 6. The

right of the wind in the northern hemisphere. typhoon moved straight northward across the

The station Iwasaki is, however, located on the western Japan Sea and landed again on the op- head of a small bay whose mouth opens wide posite coast of Asian Continent. In the figure,

southwestward, so the effect of wind set-up is pre- the position of typhoon at the time when the

dominant. In fact the peak height of the surge surge height reached maximum at respective

at Iwasaki is larger than at any of the other tidal stations is shown by thin line.

stations. Fig. 7 indicates the time variation of accom-

(37) 184 Jour. Oceanogr. Soc. Japan, Vol.24, No.4 (1968)

Fig. 6. Track of Typhoon June, Sept. 13-14, 1954. Thin lines indicate tne typhoon position when the maximum storm surge appeared at each station.

Fig. 7. Storm surge elevations at various gaging panied storm surges at each station during Sept. stations during Typhoon June, Sept. 13-17, 13-17, 1954, with peak values. The surge is 1954. See legend in Fig. 4 for explanation. not high because the storm is distant. The recorded highest surge is 74 cm at Shimonoseki : In Fig. 7, although a very faint secondary at other stations the height is less than 50cm. peak of the elevation can be traced near the It is especially noteworthy that the sea level occurrence time of minimum pressure at each falls before the arrival of the storm all over the station from Tonoura to Maizuru, the highest eastern coastal water of the Japan Sea, though elevation appears at all stations considerably late this is not seen at Tomie and Shimonoseki just after the passage of pressure minimum. At outside of the Japan Sea. The decent of most of them it occurred after the storm landed sea level begins when the typhoon lies about on the opposite coast (See Fig. 6). On the 300km south off Kyushu, and at that time the other hand, the surge on the eastern coast of typhoon center is about 600km distant from the Noto Peninsula is different from that on the southern end of the Japan Sea and more than western coast. 1500km from the northern part of it. As the This is clearly confirmed in Fig. 8. It is sur- storm approaches the Japan Sea, the atmospheric prising that Fig. 5 resembles to Fig. 8 in essen- pressure on the sea falls gradually but the sea tial features despite the tracks of storm differ surface continues to descend. This will be attri- considerably. And as in the former case, we buted to the seaward transport of water due to can also piont out an external surge which the northeasterly gale in front of the typhoon travels at a low speed of 2.8m/sec along the center that prevails in the southern part of the continental shelf from the southern entrance to Japan Sea with a direction parallel to its coast. Noto Peninsula, while the pressure minimum The fall of sea level continues till the storm moves at a mean speed of about 20m/sec. On center invades the Japan Sea in the early morn- the other hand, on the coast northeast of Noto ing of the 14th. Then as soon as the wind Peninsula the peak of the elevation occurs almost shifts to southwest bringing about shoreward simultaneously everywhere and it corresponds transport of water the sea level turns to rising. to the time when the southwesterly wind accom-

(38) An Investigation on the Variations of Sea Level due to Meteorological Disturbances 185 on the Coast of Japanese Islands (II)

over dikes, flooded the area of about 31km2. As seen in Fig. 9 the elevation for every station becomes the highest at about the time when the typhoon approaches nearest in contrast with the previous two cases. Fig. 10 indicates the time variations of surge height during Sept. 14-18, 1961. The largest height attained is only 40cm at Kashiwazaki, and storm surges were remarkably weak on the coast of the Japan Sea in spite of great size of the typhoon itself. It is especially noteworthy that sea levels segin to fall all over the coast of the Japan Sea in the middle of 15th when the typhoon approaches about 300km south-south-west of Kyushu, same as the case of Typhoon June. The descent of sea level continued till the typhoon reached Osaka Bay after crossing through eastern part of Shikoku in the morning of 16th, and then it turned to rise. Over the coast of San'in, at Fig. 8. Storm surge propagation diagram for that time, atmospheric pressure fell to about 980 Typhoon June. See legend in Fig. 5 for mb and strong northesterly winds were blowing, explanation. nevertheless the sea levels continued to fall by reason of that the lowering effect by strong panied to the rear of the typhoon center prevails northeasterly winds overcompensated the effect in the northern part of the Japan Sea. None of of pressure fall. typhoons took this kind of course for the ten At the stations north of Miyazu, the maxi- years in question except for the one mentioned mum height appears near the time when the here.

3.3 Case C: An example when a storm advances northeastward through the eastern Japan Sea Typhoon Nancy was one of the most violent tropical cyclones which had attacked the Japanese Islands: at the stage of its maximum intensity (15h, Sept. 13th, 1961) the central pressure was 885mb and the maximum wind speed 150kts. The track of the typhoon is shown in Fig. 9. The typhoon first landed on Shikoku Island with the central pressure of 930 mb at 9h 30min of Sept. 16, 1961, and entered the Japan Sea at 1711 of Sept. 16 after having the passed the western part of Osaka Bay and western Honshu. Then it progressed north-north-east along the Japan Sea coast and reached to southern end of Saghalin at 9h of Sept. 17. The typhoon caused severe storm surges on Fig. 9. Track of the Typhoon Nancy, Sept. 16- the coast around the Osaka Bay. Its height 17, 1961. Thin lines indicate the typhoon posi- was especially large at Osaka harbor reaching tion when the maximum storm surge appeared 245cm, and sea water invaded the land, flowing at each station. (39) 186 Jour. Oceanogr. Soc. Japan, Vol.24, No.4 (1968)

Fig. 11. Storm surge propagation diagram for Typhoon Nancy. See legend in Fig. 5 far explanation.

surge height and the minimum atmospheric pressure, the duration of the surge is very short for every station. The surge peak propagates with a mean speed of about 15m/sec lagging a few hours behind the propagation of the minimum pressure. The typhoons which took the similar course as Typhoon Nancy are the Typhoons Marie, Sept. 26-27, 1954 and Vera, Sept. 17-18, 1959 in the ten years in question. These typhoons caused also qualitatively the similar surges on the Japan Sea coast as those described above.

3.4 Case D: An example when a storm Fig. 10. Storm surge elevations at various gaging moves away along the Pacific coast of stations during Typhoon Nancy, Sept. 14-18, Japan 1961. See legend in Fig. 4 for explanation. The effect of wind set-up does not prevail on pressure becomes minimum. This is ob- the coast of the Japan Sea because there is not viously the surge which progresses with the so broad shallow water region along the coast typhoon. At the stations south of Saigo, the (HISHIDA and WAKABAYASHI, 1950; SHIOMI, sea level does not rise remarkably but we can 1959). On the other hand, strong southwesterly find a very faint secondary peak of the ele- winds blowing parallel to the coast raise the sea vation at about the occurrence time of minimum level and strong north-easterly winds reduce it, pressure. as seen in previous exmples. To examine the At Izuhara and Tomie there arises a surge latter effect in more detail, the sea level varia- with a height more than 20cm but it does not tions caused by Typhoon Ellen, Aug. 6-10, 1959 invade the Japan Sea as an external surge, and are shown. we cannot see the surge which propagates to the The typhoon approached to Kyushu from the coast of San'in as a free wave. southern East China Sea and progressed east- In Fig. 11, which represents the relation north-east along the Pacific coast of Japan. The between the occurrence time of the maximum path of the typhoon is shown in Fig. 12. (40) An Investigation on the Variations of Sea Level due to Meteorological Disturbances 187 on the Coast of Japanese Islands (II)

Fig. 12. Track of Typhoon Ellen, Aug. 6-10, 1959.

In Fig. 13, which indicates the time variation of sea level during Aug. 6-10, 1959, notable surge is not seen except for Shimonoseki and Tomie where the maximum of surge of 33 cm and 39cm respectively appeared a few hours before the occurrence of the minimum pressure. The sea level falls slightly before the arrival of the storm over the coast to the north of Noto Peninsula. Namely the sea level begins to sink in the evening of Aug. 7th when the typhoon approaches about 150km west-south-west off Kyushu, continues to sink till the typhoon Fig. 13. Sea level at various gaging stations during arrives at the coast of Shikoku at the early Typhoon Ellen, Aug. 6-10, 1959. See legend morning of 9th, and then turns to rising. The in Fig. 4 for explanation. sea level descent is not seen on the coast southwest of Noto Peninsula. mass transport by the strong northeasterly winds At Miyazu the rise of sea level reached to 14 on the sea level will overcome the piling up cm at 4h of Aug. 9th when the storm approached effect by the atmospheric pressure fall. nearest. This is too small in view of that the atmospheric pressure for Miyazu was at this time 983mb. If the hydrostatic pressure is the 4. A shelf wave which progresses very slowly along the continental shelf off only effect on the sea, the level should rise about San'in 30 cm. At other stations on the San'in coast sea level did not rise in spite of the falling of It was shown in Section 3 that the typhoons the pressure over the coast. or extratropical cyclones which took the course In spite of the minimum pressure of 986.2 mb of the Case A or B induced storm surge which (1h 52min of Aug. 9th) and the maximum wind propagated slowly as a free wave along the coast of NE 17.4m/sec (21h 30min of Aug. 8th) ob- of San'in. Remarkable examples are listed in served at Sakai, sea level did not rise as seen Table 3. in Fig. 13. This may be explained by reason- SHOJI (1961) presumed that this might be a ing that the lowering effect of the off-shore kind of internal Kelvin wave induced by an

(41) 188 Jour. Oceanogr. Soc. Japan, Vol.24, No.4 (1968)

Table 3. Examples of slowly moving storm Let x, y be respectively the distance normal

surge along San'in coast. and parallel to the coast and z the distance ver-

tically upward from the undisturbed sea surface .

Further, let ƒÅ(x, y, t) be the sea level distortion . Following ROBINSON (1964), we assume the

existence of free wave of the form:

and (1)

over the sea which has the bottom configuration of atmospheric disturbance. The wave is caused and not only by typhoons in autumn but extratropi- (2) cal cyclones in winter and spring. In winter, the water of the Japan Sea is homogeneous from the surface to the bottom because of the active where D is the depth in the deep sea region, d vertical mixing due to cooling from the surface is the depth at the edge of shelf and l is the and disturbance by strong northwesterly mon- shelf width. Then, we get the relation soon, and the layer of thermocline is not formed (FUKUOKA, 1962). Accordingly it seems diffi- cult to assume the existence of internal Kelvin (3) wave as SHOJI did. under the assumption of k2l2•á1 and f2l2•ágd On the other hand, in case of typhoon and , where f is the Coriolis parameter, g is the extratropical cyclone which take the course of acceleration of gravity and ƒÈ=lkf/ƒÖ. Using ƒÈ the Case A or B and generate rather conspicuous determined from Eq. (3), the wave velocity is surge in the shallow waters southwest of the Japan Sea, strong southwesterly winds prevail (4) over the East China Sea and the southwestern entrance of the Japan Sea, on the contrary, in This depends only on the bottom configuration case of storm which takes the course of the and Coriolis parameter. The wave speeds cor- Case C or D strong northeasterly winds prevail responding to various bottom configurations are over that area. Moreover, the wave is more shown in Fig. 14, where the latitude is assumed predominant on the coast than on off-shore to be 35•K. The wave can travel along the shelf islands. Then it is plausible to assume that the in one direction only, that is, the coast is on wave will be an external surge which is caused the right when we look in the direction of wave over the East China Sea or southwestern end of motion in northern hemisphere. the Japan Sea and propagates along the coast Fig. 15 a shows the depth contours for 200m with a mode of a kind of boundary wave. and 300m over the south-western Japan Sea,

ROBINSON (1964) showed theoretically the ex- and the latter contour corresponds to shelf edge. istence of a kind of continental shelf wave which As an example, the cross section of depth along propagates slowly only in clockwise direction the line AA' in Fig. 15 a is shown in Fig. 15 b, along a coast line in the northern hemisphere. in which shelf width l is 50km, the depth at HAMON (1962, 1964) reported the existecne of the shelf edge d is 300m and the depth of deep such waves along the east and west coasts of sea D is 1400m. If Robinson's shelf wave Australia. MYSAK (1967 a, 1967 b) modified exists under these configurations, its velocity is

Robinson's results and found the coincidence of 250cm/sec as can be read off from Fig. 14. If the theory with Hamon's observation. we change the shelf width to l=70km, the Now, consider a straight infinite coastline. velocity becomes 340cm/sec. These values agree

(42) An Investigation on the Variations of Sea Level due to Meteorological Disturbances 189 on the Coast of Japanese Islands (II)

5. Conclusions In this paper, the variations of sea level on the coast of the Japan Sea when a typhoon progresses on various tracks are investigated. Con- clusions obtained are summarized as follows: (1) The effect of wind set-up is relatively small. (2) The pressure effect is nearly hydrostatic. (3) Strong southwesterly winds blowing parallel to the coast induce shoreward mass- transport of water, and sea level rises on Fig. 14. Velocity of Robinson's shelf wave. the coast. At Iwasaki, southwesterly winds also have the effect of wind set-up, and the surge becomes very large compared with the other stations. On the other hand, strong northeasterly winds induce seaward mass- transport of water, and sea level is loweiled down on the coast. (4) When a storm passes along the course of Case A or B in Section 3, there exists slowly moving surge on the coast of San'in. We consider that this is a free wave of the mode of Robinson's shelf wave. (5) When a typhoon approaches about 300km Fig. 15a. Depth contours for 200m and 300m south off Kyushu, sea level on the coast of over tne south-western Japan Sea. the Japan Sea begins to descent. At that time, the distance from the typhoon center to the southern end of the Japan Sea is about 600 km and more than 1500km to the northern part of it, and it is not likely that the sea level is under the direct influence of the typhoon. We cannot understand at present why the sinking of sea level is caused.

Acknowledgments The author wishes to thank Prof. S. UNOKI and Prof. K. KAJIURA for the helpfull advice Fig. 15b. Depth profile along the line AA' and encouragement in connection with this work. in Fig. 15a. The aid of Dr. K. SUDA, Chief of the Typhoon Res. Lab., Meteorol. Res. Inst., is also greatly roughly with the velocities mentioned in Table appreciated in preparing the paper. 3. Accordingly, we can assume that the surge on San'in coast is a free wave of the mode of Robinson's shelf wave rather than an internal References Kelvin wave. FUKUOKA, 3. (1962): Characteristics of hydrography of the Japan Sea-In comparison with hydrography of the North Pacific-. Jour. Oceanogr.

(43) 190 Jour. Oceanogr. Soc. Japan, Vol.24, No.4 (1968)

Soc. Japan, 20th Anniversary Volume, 180-188. and the response of sea level to weather system. HAMON, B. V. (1962): The spectrums of mean sea Jour. Geo. Res., 69, 367-368. level at Sydney, Coif's Harbaur and Lord Howe SHIOMI, N. (1961): Storm surge in Wakasa Bay. Island. Jour. Geo. Res., 67, 5147-5155. Technical Report of the Japan Meteorological HAMON, B. V. (1966): Continental shelf waves and Agency, No.7, Report of the Ise Bay Typhoon the effects of atmospheric pressure and wind stress (No.5914) in September 1959, 458-465. on sea level. Jour. Geo. Res., 71, 2883-2893. SHOJI, D. (1961): On the variation of the daily HISHIDA, K. and T. WAKABAYASHI (1950): On mean sea levels along the Japanese Islands. Jour. the storm tide in the Wakasa Bay on September Ocearnogr. Soc. Japan, 17, 21-32. 3, 1950. Oceanogr. Report, Centr. Meteorol. TANIOKA, K. (1961): Storm surge on the San'in Obs., 1, 185-191. Coast. Technical Report of the Japan Meteoro- ISOZAKI, I. (1968): An investigation on the varia- logical Agency, No.7, Report of the Ise Bay tions of sea level due to meteorological distur- Typhoon (No. 5915) in September 1959, 243- bances on the coast of Japanese Islands (1): On 426. the accuracy of tide predictions. Papers in UNOKI, S. (1959): An investigation on meteorological Meteorol. and Geophys., Tokyo, 19. tides in the neighbouring sea of Japan (4): Ge- MYSAK, L. A. (1967a): On the very low frequency neral features of regional distribution and seasonal spectrum of the sea level on a continental shelf. variation. Oceanogr. Mag., 2, 51-63. Jour. Geo. Res., 72, 3043-3047. UNOKI, S. and I. ISOZAKI (1965): Mean sea level MYSAK, L. A. (1967b): On the theory of conti- in bays, with special reference to the mean slope nental shelf waves. Jour. Mar. Res., 25, 205- of sea surface due to the standing oscillation of 227. tide. Oceanogr. Mag., 17, 11-35. ROBINSON, A. R. (1964): Continental shelf waves

気象じよう乱によって起る日本沿岸の水位変動の研究(II) ― 日本海沿岸の高潮―

磯崎一郎

要旨 1953年より1962年までの10年間の潮汐記録を用 staticと考えてよい. いて日本海沿岸の高潮の一般的特性を明らかにした. 東支那海方面から日本海に侵入する高潮があると,こ日本海沿岸では風の吹き寄せ効果は小さく,したがっれは山陰沿岸を3～4m/secの非常に遅い速度で伝播すて海岸に直角方向に吹く北西風では顕著な高潮は起つてる.これはROBINSON (1964) のshelf waveのモードいない.海岸に平行に吹く強い南西風は岸に向かう水ので進む自由波であると考えるとよく説明される. 輸送を強制し海岸で潮位が上昇する.反対に,強い北東台風が九州の南方約300km位に接近すると日本海沿風では沖に向って水の輸送が起り,海岸での潮位は低下岸全般に潮位の下降が始まる.台風からの距離が非常にする.このような性質から,目本海岸の顕著な高潮は台遠いので,台風の直接の効果とは考えにくいが,その機風や低気圧が日本海中部を南西から北東に通過する時に構はまだよくわからない. 起っている.気圧低下による潮位の上昇はほぼhydro-

(44)