Travel Times of Seismic Sea Waves to Honolulu 
Bernard D. Zetler 1 
The seismic sea wave which struck the 
Hawaiian Islands on April 1, 1946, has 
again focused attention on the necessity for 
adequate protective measures against similar 
disasters in the future. The problem is obvi¬ 
ously complex, involving rapid location of 
the epicenter, the detection of the sea wave 
as it moves toward the Hawaiian Islands, a 
quick method of determining the time the 
wave will reach the islands, and finally an 
adequate means of providing security for 
people and property. The purpose of this 
study was the preparation of a chart of the 
Pacific Ocean which would show the travel 
time to Honolulu of a seismic sea wave from 
the plotted position of an earthquake epi¬ 
center (see Fig. 1, insert sheet). Given the 
time of the disturbance, the arrival time of 
the wave at Honolulu becomes immediately 
available. 
Oceanographers have long accepted the 
concept that the velocity of a seismic sea 
wave is a function of the depth of water and 
they have expressed it mathematically as 
v = \/gd, where v is the velocity of the 
wave, g the acceleration of gravity, and d 
the depth of the water. However, this 
formula for velocity has been considered by 
some authorities to be a rough approxima¬ 
tion; it was believed that the actual velocity 
would always be somewhat slower. 
The results of the computations made in 
the course of the study by Green (1946) 
created more confidence in the accuracy of 
travel times computed by means of this 
formula. These computations were not in¬ 
fluenced by the recorded arrival times; the 
times to several of the more distant places 
1 Mathematician, U. S. Coast and Geodetic Sur¬ 
vey, Washington, D. C. Manuscript received 
March 17, 1947. 
[ 1 
were computed before it was known that the 
sea wave had been recorded on the gages. 
Table 1 of that report lists 12 places whose 
distances from the epicenter vary from 1,610 
to 8,066 statute miles. A comparison of 
observed with computed travel times to these 
places shows an average variation of 1.2 per 
cent, which is not consistently in one direc¬ 
tion. 
It was decided that the procedures used 
in the above project could be adopted in the 
preparation of a chart which would show 
the travel time of any seismic sea wave to 
Honolulu. A series of ^-hour curves would 
be drawn on the chart such that each would 
represent the length of time a sea wave 
would take to reach Honolulu from an epi¬ 
center at any point on the curve. 
Points to be used as epicenters of sea 
waves were selected in various directions 
from Honolulu, and travel times were com¬ 
puted, using soundings from large-scale 
charts, along arcs of great circles between 
these points and Honolulu. Half-hour inter¬ 
vals were plotted along each arc with numer¬ 
ical time values increasing with distance 
from Honolulu. The time curves were drawn 
by connecting the respective l^-hour points. 
Although observed travel times were 
available for a number of sea waves which 
had previously been recorded on the Hono¬ 
lulu tide gage, it was considered desirable to 
use computed rather than observed data in 
the construction of the chart. However, the 
positions of the epicenters of recorded waves 
were included among the points from which 
travel times were computed in order to make 
available a comparison of observed and com¬ 
puted travel times (see Table 1). 
In preparing the time data along any par¬ 
ticular path, a great circle course between 
5 ] 
