Skip Navigation Links
NOAA NWS United States Department of Commerce

Tropical Cyclones During the Years 1900 - 1952


The next storm of record was tracked from December 23, 1902 to January 2, 1903 by Rev. Louis Froc, S.J., Director of the Ziwakei Observatory, Shanghai. The dates involved indicate that possibly it was a "Kona",storm. However, the parabolic track which headed to 4ON 178W before recurvature and final dissipation in the Bering Sea, as well as respect for Froc's integrity, suggest a cyclone of tropical origin. The storm track was published as a typhoon track on the U.S.W.B. January 1913 issue of the Meteorological Chart of the North Pacific.

This storm had its origin on December 23 about 40 miles southwest of Lanai and on December 24 was located in Kaulakahi Channel between Kauai and Niihau. No records of unusual weather over the State of Hawaii on these dates have been discovered, so it is likely that on the 23rd and 24th of December the storm was in its formative stages.


This storm was first located near 17.5N 160W - about 250 miles southwest of Honolulu. It was tracked from records at the Zikawei Observatory northwestward to 150 miles west-southwest of Barking Sands, Kauai. Shortly thereafter it began to trace a broad parabolic curve whose westernmost point was 35N 171W. The last estimated position of the remnants of the storm was south of Sitka, Alaska.

The initial track of this storm, as well as the dates involved, are very reminiscent of hurricane NINA -- another late season storm which inflicted considerable damage on Oahu and Kauai on November 30 and December 1 of 1957. The Zikawei track of the 1904 cyclone placed it within 20 miles of that of NINA when both were southwest of Barking Sands. Lack of definite reports of the storm of 1904's effects on the Hawaiian Islands indicate it must have been either much smaller or less intense than NINA.


This is another example of a tropical cyclone which occurred much later in the year than those of more recent history. The track, another obtained by Hurd from Zikawei Observatory records. and published on the United States Weather Bureau Metqorolooical Chart of the North Pacific, was first located at 15.5N 156W - 230 mi south-southwest of South Point, Hawaii. The cyclone was tracked to about 150 miles north of Johnston Island and then to 270 miles south of Midway Island. The storm subsequently crossed the International Date Line and continued to 39N 170E before recurving to the northeast.


The early May dates of this storm seem as out of season to present day meteorologists as the late season December dates previously mentioned. The month of May has long been recognized by tropical cyclone historians as a month without such storms in the Central North Pacific. This was the first such occurrence recorded in this area.

The track of the 1906 cyclone was compiled by Pere H. Gauthier of the Zikawei Observatory and was plotted on the May 1913 United States Weather Bureau Meteorological Chart of the North Pacific from Gauthier's records. The storm was 0 first located far south- southeast of the Big Island at 12.5N 153W. From that point it was followed to latitude 25N where it crossed the International Date Line. The storm continued in a familiar parabolic track to latitude 4N longitude 173E before turning to the northeast and finally dissipating in the Bering Sea.

The track of this early season storm was very similar to those followed by the late season storms of December 1832 and 1904.


This storm was first recorded 120 miles southeast of Hilo. It moved west southwestward to about 60 miles south of South Point, Hawaii. From there it progressed northwestward along a sharp parabolic curve and skirted the southwest tip of Niihau before moving nearly straight northward to latitude 27N. At that point it began a somewhat unusual curve to the northwest to 35 N on the International Date Line before heading northnortheast, finally dying in the Bering Sea.

The complete track of the-storm was included in the Typhoon Tracks for October on the January 1913 United States Weather Bureau Meteorological Chart of the North Pacific as compiled by the Zikawei Observatory.

The Hawaiian Section of the October 1906 Climatological Service of the U.S. Weather Bureau stated that Makawao, Maui on October 4 had a "maulu," or mountain shower, consisting of a steady downpour lasting 4-1/2 hours which was "the heaviest in years" and which totaled 12.70". That same publication added that the barometer was "relatively low" at the Weather Bureau observatories at Honolulu and Ewa Beach during the period from October 2 to the 8th.


The first known position on of this cyclone was 90 miles south of Honolulu at 20N 157.9W. It was tracked north-northwestward by the Zikawei observatory midway through the Kauai channel. it appears that this cyclone was either extremely small or was in a weak tropical depression stage at that location since examination of climatological records indicates no unusual rainfall, wind, surf, or pressure at any of the Hawaiian Islands during that pebiod. Shipping reports allowed it to be tracked northward to 35N 16OW whence it recurved to the northeast and dissipated just west of British Columbia.

OCTOBER 17-18, 1908 (VORTEX )

An unusual storm track, designated as a typhoon on the January 1913 United States Weather Bureau Meteorological Chart of the North Pacific, begins at 27N 166W and ends at 38N 160W after proceeding along a fairly sharp parabolic curve. It is the only cyclone charted by Visher, Zikawei Observatory, Redfield, or Hurd whose origin is at such a northerly latitude. This track as plotted is reminiscent of that of the remnants of hurricane GILMA in July 1978.

That coincidence suggests that possibly this storm too had its origin as the decaying remnant of a cyclone which entered the Central Pacific near latitude 24 N, since other storms entering the area at more northerly latitudes have A tendency to track westward near or just south of latitude 25N to 155 or 160W before curving northwestward.

The Period from 1910-1926

A study-(Hurd, 1929) of all the vessel weather reports received by the United States Weather Bureau from the Eastern and Central North Pacific for the period 1910 to 1928, inclusive, together with information furnished by the Mexican meteorological service, indicates that at least 95 tropical cyclones occurred in those waters during that 19-year period -- an average of 5 per year.

Deutsche Seewarte records of cyclones in the same waters during the 61-year period from 1832 to 1892 indicate a total of 45 storms -- a yearly average of 1.36.

The apparent increase in activity during 1910-1928 is a result of the greater number of observations made possible by the opening of the Panama Canal in 1914, and especially by the rehabilitation of ocean commerce in 1921 following the depression caused by World War I.

Thirteen (13) known cyclones in 1925 mark it as a year of maximum frequency during this period.


A single ship report on this date located a cyclone with winds of Beaufort force 10 and barometric pressure of 29.49" (998.7 mb) near 20N 147W (Monthly Weather Review, 1929). No other reports concerning this storm are presently available. A newspaper account of very rough seas which capsized a boat off Waikiki and which were observed shortly after dark on September 30 may or may not have been related to this storm.

THE YEARS 1920-1923

According to Visher (1925) Messrs. Hurd and Tingley of the U.S. Weather Bureau reported 15 cases of tropical cyclones of depression or greater intensity passing to the north or northeast of the Hawaiian Islands during the years 1920-1923.

Analysis of the list of those 15 cyclones indicates that more than likely 7 of those listed were not true tropical cyclones but were winter "Kona"-type storms or were cold frontal passages. Dates of occurrence of those 7 storms ranged from January through March. This particular observation should not disparage the very studious work of Visher. Rather it should alert the casual reader that more recent studies in meteorology have clarified the distinct differences between true tropical, "Kona," and higher latitude storms. Synopses of 3 of the 8 remaining cyclones reported by Hurd and Tingley are presented in this chronology. No corroborating data are presently available concerning the other 5.


This occurrence was noted as being of tropical origin but no track has been discovered which indicates it as ever having achieved more than tropical disturbance intensity. The following excerpt is from a summary of October 1921 weather over the North Pacific Ocean

(MWR, 1921):

"The second disturbance formed near the Hawaiian Islands about the 5th and during the next several days moved slowly northward toward the Gulf of Alaska, there merging with a depression which had advanced eastward over the Aleutian Islands on the 7th. From the 8th to the 10th the combined depression was ill defined, but on the llth it increased in energy and by the 12th covered the whole of the Gulf of Alaska, as well as nearby portions of the ocean. Between the 12th and 16th it remained nearly stationary as a fairly vigorous disturbance, causing moderate to fresh westerly gales along the eastern portion of the northern steamer lane. On the 17th it moved inland on the British Columbian coast and a fresh depression formed in the ocean north of the Hawaiian Islands."


The track of this particular cyclone was extracted from a map constructed by Visher from data supplied him by Messrs. Tingley and Hurd, as well as U.S. Pilot Charts. The only present day references to the cyclone that have been discovered as of this date consist of the reference quoted from F. G. Tingley in the last sentence of the description of the October 5, 1921 cyclone and the following quote from the same article:

"In connection with the latter depression an interesting development was reported by Capt. H. J. Satterly of the British S. S. ADNA. At noon on the 17th the ADNA, San Francisco to Kobe, was in latitude 36 40'N, longitude 147 29'W, barometer 29.32 inches, wind SSW. At 2:30 PM the barometer had fallen to 29.10 inches and the wind had backed to southerly, increasing to force 10-11. There was a very heavy, breaking southerly sea. At 3 PM the wind started to veer to westward, and at 4 PM, according to Capt. Satterly, the disturbance suddenly blew out, the barometer standing at 29.04 inches. There was a very confused sea. After 4 PM the barometer rose quickly, and at 6:15 PM read 29.42 inches. After the passing of the disturbance the wind came down from the NW in a heavy blast, later settling down to a moderate gale. Later on the previous day, states Capt. Satterly, the wind had twice in quick succession veered round the compass, and formations that looked like small whirlwinds were seen in various places traveling in a northeasterly direction. Capt. Satterly expressed the opinion, which subsequent developments appear to confirm, that these signs pointed to the formation of a small circular storm."

Close inspection of the dates and locations set forth by Tingley and Capt. Satterly, as well as the track plotted by Visher, reveals discrepancies which at this time cannot be accounted for.

Since Capt. Satterly was actually "in the storm" it seems only logical to place first priority on his report. So doing immediately makes the reader question whether or not it was a tropical cyclone. Perhaps it was the vortical remnant of a more vigorous circulation or it may have been a tropical development similar to that which occurred some 1500 km northeast of Honolulu in early September 1975 and which is explained in detail at that point in time of this chronology.

This particular cyclone and that of September 1975 are among those that might fit into the "rare" category due to formation at such northerly latitudes. They bring to mind similar instances in the Atlantic Ocean which include hurricane ANNA in 1965 which developed into a hurricane at 38N after being tracked from 32N as a lesser development; hurricane BETTY in 1972 which became a hurricane at 39N after subtropical low beginnings at 36N; and hurricane CANDICE in 1976 which became a hurricane at 41N after initiation as a tropical depression at 32N.


A singular, but extremely interesting report of a tropical storm west of Hawaii was written by Thomas A. Blair, meteorologist at the U.S. Weather Bureau on August 31, 1923.

"The U.S.S. VEGA at noon August 20, 1923, encountered a well-developed cyclone in latitude 210 OO'N, longitude 165 OO'W, about 450 miles west of Honolulu. Such storms are extremely rare in these waters, and indeed appeal to have been previously unrecorded in the summer season. The excellent report of the storm made by Lieut. P. J. Gundlach, U.S. Navy, follows:

'Left Guam for Pearl Harbor on 9th August, 1923. Up to 20th August very little change in weather conditions, about 50 percent overcast and cloudy with some rain, wind from east and northeast, force 3 to 4.

'About midnight 20th August wind increased to force 5, barometer normal. Wind increasing from that time, barometer falling slowly, 0.01 to 0.02 per hour. At 5 A.M. wind . increased to force 7, same direction (45 true), barometer falling 0.04 in one hour. Heavy swells commenced from the southeast, wind veering (backing) to north. Heavy cross sea commenced. From 5 A.M. to noon wind increased from force 7 to 11 and barometer dropped from 29.84 to 29.32, largest drop 0.14 between 9 and 10 A.M. About 12:30 barometer commenced rising and wind veered (backed) to left to about northwest. From 1 P.M. wind veering (backing) to left, to west, and then south, decreasing in force, weather moderating. At 5 P.M. wind south, force 6, barometer 29.71; barometer rising to 29.96 at midnight; no change in force and direction of wind. Then gradual change in wind to left, south to east, and barometer steadily rising to normal. Next day, conditions normal, sky clear.

'At noon, 20th, at period of lowest barometer, sky cleared and sun shone brightly about 20 minutes, apparently "eye of storm center."

Blair continues:

"This was evidently a typical cyclone of small dimensions. On the same date high seas from the southwest occurred on the southwest coast of Kauai. These subsided on the 21st. The only effect on the weather at Honolulu was the interruption of the trade winds on a portion of the 21st and 22nd by light variable breezes from a southerly quadrant, accompanied by increased cloudiness, and on the 23rd by the highest temperature recorded for two years."

"No further report of the storm has been received, and its origin and path remain unknown. Pressure at Midway island, Honolulu, and Hilo remained normal. Although there was simultaneously an area of low pressure in Alaskan waters, pressure records from Midway Island and Honolulu and from ship reports indicate that there was no break in the belt of high pressure in middle latitudes. It does not seem probable, therefore, that this storm was in any way influenced in its origin by the northern LOW, as frequently appears to be the case with winter storms known as "Konas." The storm may, however, have moved northward and merged with the northern LOW on the 22nd, since the observations on the afternoon of the 21st and the morning of the 22nd indicate that the center of high pressure had moved eastward, and that the Alaskan LOW had extended southward a considerable distance between the 150th and 160th meridians."


The steamer, the WEST CALERA, encountered a tropical cyclone on July 31 and August 1 near 15N-152W reporting northeast gales. No further ship reports were received on this particular storm but subsequent observations in the Hawaiian Islands are reported in the Honolulu Advertiser (August 2, 3, and 4, 1925) and by Hurd in the Monthly Weather Review (1929) indicate that it continued west-northwestward remaining well south of and, quite likely, in the lee of the Islands. The following quotes are from Hurd and the Honolulu Advertiser, respectively:

"Easterly gales occurred at Honolulu from August 1 to August 4 but without appreciable barometric depression.

"Late in the afternoon of the lst exceptionally high tide accompanied by strong winds flooded the wharf warehouse Honuapo on the Big Island. All beach houses there and at Punaluu were flooded. The unusually high winds caused a long line of flurtes at Hutchinson plantation to collapse.

Heavy seas and high surf which continued along the southeast coast of the Big Island causing the damage noted above are reminiscent of high surf damage which accompanied hurricane FICO along the same coastline in 1978.

The Advertiser also reported on August 3 that

"Honolulu experienced the highest surf for a long time for this time of year. Fort Kam was flooded; lawns were awash and damaged by huge breakers on the beaches at Diamond Head and Kahala; beach houses and property on 'the other side' of the island were flooded and damaged. Tide was high through late last night. Strong winds blew over a Ford in Pacific Heights by the near cyclone which raged over the island during the day. Rainfall reported at the U. S. Weather Bureau station was only a trace. This storm was believed related to a storm at sea or some earthquake."


The same steamer, the WEST CALERA, that encountered the July 3-August 4 cyclone met another more serious hurricane near 22N, 137 30'W on September 27 and 28. During that passage the barometer fell to 28.53" but no other account of the storm has been located as of this date. It can only be surmised that with no significant synoptic change it may have entered the

Central North Pacific as was the case with the August 1925 cyclone.

The Period from 1926-1950

Other than the report of the possible tropical cyclone described in the section which immediately follows this paragraph, no other observations of tropical cyclones in the Central North Pacific have yet been discovered.

Readers who know of tropical cyclone activity in the Central North Pacific during the 1926-1950 period will be doing a most appreciated favor by passing that information to the CPHC at Honolulu for inclusion in future studies.


The following are extracts from an article which appeared in the August 19, 1938 issue of the Honolulu Star Bulletin:

"A half century of August rainfall record went on the Honolulu Weather Bureau books today, the result of a booming midnight wind and rain storm which accompanied the lowest barometer reading for the month since 1904.

"Rainfall totals for the entire month of August back to 1888 were bettered by last night's downpour.

"Between midnight and 3 a.m. a gale tore over the island, reaching the proportions of a 'whole' gale with a velocity of 60-3/4 miles an hour at the Pan American radio station at Mokapu.

"The FAA station reported winds of 54 m.p.h. at midnight, 52-3/4 m.p.h. at 1:30 and 2:45 a.m. and 56-1/4 at 3 a.m.

"The Weather Bureau's downtown office, never reflecting precipitation in upland and windward areas of the island recorded 1.26 inches of rain for the 24-hour period to 8 a.m. today, while elsewhere the total amounted substantially.

"At midnight the barometer started falling rapidly, and at 12:30 a.m. touched 29.77 inches, the lowest figure recorded since the Weather Bureau office was established here 34 years ago.

"With equal suddenness the wind leaped up, tearing across the island in sudden gusts which seemed intensified by momentary lulls. "Kailua apparently received the full strength of the wind. Although electric power lines to a number of homes were out of commission the general system withstood the battering.

"This was the first August thunder and lightning display officially noted on Oahu in 20 years.

"Waimanalo Plantation reported some damage to trees and cane fields, but nothing serious."

Climatological records for the Islands during this storm reveal many reports of 4-inch totals on both Oahu and Maui; 3 inch totals on Molokai; 2-inch totals on Kauai and generally less than 1 inch on the Big Island.


Hurricane HIKI was first detected on charts at Honolulu Airport during the night of August 12. Inasmuch as this was the first hurricane to affect the Hawaiian Islands in many years, it was designated alphabetically as ABLE, or in the Hawaiian language HIKI. Thus it became widely known as hurricane HIKI.

This storm was first plotted near 14.5N, 144.5W and moved northwestward to arrive east of Hilo, Hawaii on August 14 as an immature storm of very small diameter. Aircraft reconnaissance estimated surface winds of 45-50 knots near the center.

During the next 24 hours the storm intensified, reach full hurricane force; however, it remained small. Gale force winds extended outward less than 150 miles. HIKI continued northwestward on a course nearly paralleling the windward coasts of the Big island, Maui, Molokai, Oahu, and Kauai, 100 to 200 miles offshore.

The northwest movement of the hurricane became blocked temporarily on the 17th north of Kauai as a high pressure ridge intensified west of the islands. During the following 24 hours the hurricane was forced slowly southwestward, affected both by lower tropospheric currents which would have carried it further westward and by high level westerlies which might have nudged it northeastward.

At this stage aircraft reconnaissance reported winds up to 90 miles per hour just south of the storm center, and it became necessary to alert the population of Kauai and Oahu against possible loop movement which would have carried it east and northeastward through the Kauai channel. Fortunately, the blocking action to the west eased and HIKI resumed its westward movement, passing south of French Frigate Shoals and Midway Island before it began its recurvature near 17 east longitude.

The eye of the storm was entered by two reconnaissance aircraft and several U.S. Navy ships. It was described as being 20-25 miles in diameter and overcast with lowest pressure 29.02 inches (982.7mb).

Because all of the territory remained in the southern or less dangerous semicircle of the hurricane, HIKI caused little wind damage. Highest sustained winds reported in the islands were: Kilauea Lighthouse, Kauai - 68 m.p.h.; Niihau - 50 m.p.h.; and Lanai Airport - 48 m.p.h. The roofs of several houses were blown off, and unbraced masonry walls of one home under construction were blown over. One life was lost at Kohala, Hawaii when a farmer contacted a live wire blown down by strong winds.

The greatest property damage attributed to HIKI resulted from flooding of the Waimea River on Kauai. More than 200 residents of Waimea Valley were evacuated. While no structures were swept away by the flood, water damage was considerable in a few areas.

At the Kanalohuluhulu Ranger Station on Kauai more than 52 inches of rain was measured between noon on the 14th and noon of the 18th. The gage, capacity 24 inches, after having been emptied at noon of the 15th, was found to be overflowing at 8 a.m. of the 16th. The amount recorded at nearby Mount Waialeale, elevation 5075 feet, usually a much wetter station, was only 13.05 inches on the 16th.

The streets of Waimea were flooded not long before midnight of the 15th, and again early in the morning of the 17th. The Kekaha powerhouse was severely damaged by flooding on the 17th and hundreds of acres of sugar cane land were covered by water.

On Hawaii, at the opposite end of the island chain from Kauai, rainfall totals at low and moderate elevations in the usually wet windward district were substantially below normal on account of downslope southwesterly winds. Maui recorded 11.95 inches of rainfall on August 15, a greater 24-hour amount than during any previous August of record.

The track taken by HIKI -- off the windward coasts of the Islands -- was the first such known in Hawaiian tropical cyclone history. Twenty-two years later tropical storm (ex- hurricane) DIANA took a similar course in August 1972. DIANA's track off the windward coasts not only closely followed HIKI's but was 20 to 30 miles closer to land. Then, 11 days later, as if to prove that the HIKI and DIANA tracks weren't really unusual, tropical storm (ex- hurricane) FERNANDA followed a comparable path somewhat further northeastward away from the Islands. The only other known tropical cyclone of modern history which had a similar track was hurricane KATE whose center passed 240 miles east northeast of Hilo during September 1976.


Surface map analysis of 08/0600Z data indicated a weak tropical disturbance near 12N 169W which was analyzed as a wave in the easterlies. The system moved west-northwestward at 7-8 knots to 12.5N 179W where surface observational data indicated a closed LOW center. Turning to the northwest at 8.9 knots it headed for Wake Island.

At 14/1200Z (midnight Wake Island time) it was classified a tropical disturbance as it intensified while moving at right angles to the 700 millibar contour flow -- heading directly- towards the center of the High as analyzed on the 700 millibar chart.

By 15/1200Z at 18.6N 171.5E it was designated a tropical storm and named OLIVE -- still heading northwest towards the vicinity of Wake at about 9 knots. From that point the storm turned to a slightly more westerly track which headed it directly at Wake.

B-29 aircraft reconnaissance at 15/2248Z placed the center at 19 42'N 166 56'E, approximately 30 nautical miles northeast of Wake, with a minimum sea level pressure of 27.90" (945 mb). Forty-two minutes later, along with serious flooding, the Wake Island observatory reported NNW winds of 105 knots with gusts to 123 knots.

OLIVE was the second typhoon to strike Wake since weather records were begun at that station in 1935. The first, an unnamed typhoon, struck on October 14, 1940 (Tomita, 1968), 12 years earlier. Fifteen years later to the day, on September 16, 1967, Wake was to be devasted by typhoon SARAH.

After striking Wake, OLIVE continued to move northwestward around the 700 mb High as that center moved eastward.