Directional Shift of Trade Winds at Honolulu 
Chester K. Wentworth 1 
For MANY years it has been known to those 
persons especially interested in weather con- 
ditions that a gradual shift of the direction of 
prevailing surface wind at Honolulu has been 
in progress. The data on which such a con- 
clusion can be reached are now sufficient to 
give a more complete picture of the shift, 
though they have not been made available in 
a uniform published form. The present paper 
has made use of data, partly published and 
partly unpublished, furnished by the Weather 
Bureau office at Honolulu. From 1905 to 1922 
the weather vane was on the Young Hotel 
Building, 121 feet above the ground and 26 
feet above the roof. Since 1922 the weather 
vane has been on the Federal Building, 99 feet 
above the ground and 17 feet above the roof. 
The former location is about 700 feet north- 
westerly from the latter and about 300 feet 
nearer the crest of the Koolau Range. The 
records show no indication of change due to 
the difference in location. 
The Hawaiian Islands are in a latitude zone 
dominated by northeasterly trade winds, as 
shown by the fact that over the past 40 years 
of record the winds in the northeast and east 
sectors at Honolulu have averaged about 81 
per cent of the total time in hours. Similar con- 
ditions are shown at various other stations but 
it is only at the Honolulu station that the record 
is of sufficient length and that the data are tab- 
ulated in a form which permits detailed presen- 
tation. 
The shift of the prevailing wind direction 
was first noted in published form by E. A. Beals 
in 1927 (The northeast trade winds of the 
North Pacific. U. S. Monthly Weather Rev. 
1 Geologist, Honolulu Board of Water Supply. 
Manuscript received May 23, 1948. 
55: 211-221). He showed that a shift from 
northeast toward east was then in progress. He 
discussed the possible influence of changes in 
value and position of high pressure areas north 
and east of Honolulu, as well as changes in 
ocean temperature. He concluded that the "true 
explanation is not at once apparent.” 
In 1938, in an unpublished manuscript 
(Wentworth, C. K. Geology and ground water 
resources of the Palolo-Waialae district (Hono- 
lulu) . Board of Water Supply, Manuscript Re- 
port, 274 pages; p. 29), the writer extended the 
tabulation by 5 -year intervals and showed that 
the swing from northeast toward east continued 
until 1930 and appeared to slow down from 
1930 to 1934. Since there is no possibility of 
such a shift being continuous in one direction, 
it then appeared that a reversal might be im- 
minent. Only recently, in the course of further 
discussion of climatic data, the percentages of 
hours of wind in the northeast and east sectors 
for the years 1935 to 1946 were secured through 
the courtesy of the Weather Bureau office at 
Honolulu and the analysis was extended to in- 
clude the whole period from 1905 to 1946. It 
is now clearly shown that the anticipated re- 
versal has taken place. The 5 -year moving 
average used as an index had, by 1944, returned 
from the extreme easterly position more than 
halfway to the apparent extreme northeasterly 
position suggested by the data for the period 
1907 to 1910 (Fig. 1). 
In order that the reader may evaluate the 
general picture presented here, some explana- 
tion of the method of analysis is warranted. 
The fundamental data consist of a summarized 
record of wind in each of the eight sectors cen- 
tered at the four positions of north, west, south, 
and east, and the four intermediate (45°) posi- 
I86'J 
