Structure of Koolau Volcano — Furumoto, Thompson, and Woollard 
309 
Fig. 4. Travel time plots of refraction traverses in Kawainui Swamp. 
and critically examined, corrections for water 
depth at shot points and elevation at recording 
sites were made to the raw data. 
The corrected data were fed into an IBM 7040 
computer for processing. The computer pro- 
gram gave the slopes, velocities, and intercepts 
of the various branches of the travel-time plots 
and the variances of the slopes and intercepts. 
Calculations of depth to the top of the various 
layers were done by using desk calculators. 
DATA AND INTERPRETATION 
After reviewing the gravity and magnetic 
data the authors decided to run seismic refrac- 
tion surveys over the caldera area, the north- 
west rift zone, and one of the flanks not dis- 
turbed by a rift zone. For the flank investiga- 
tion the western flank was chosen first, but 
inclement weather forced the transfer of the 
operations to the southern flank, which is on 
the leeward side of the island. The short tra- 
verses and the GASHOUSE series were designed 
to outline the volcanic plug in the caldera area; 
the K-K and ABLE series were planned to paral- 
lel as closely as possible the northwest rift zone 
as outlined by the ridge of high gravity anoma- 
lies. The BRAVO series was designed to run 
along the strike of the south flank. The DELTA 
series was intended to extend the investigation 
to the island of Molokai, but data were inade- 
quate to attempt any analysis. 
Caldera Area 
Two consecutive traverses were carried out 
in the drainage canal which cuts across the 
northern section of Kawainui Swamp located in 
the caldera area. For the first traverse recording 
units were located at C-l and C-2 of Figure 3 
and shots were detonated in the canal at 200-ft 
or 400-ft intervals. For the second traverse the 
recording units were placed farther apart at H-l 
and H-2, and the shots were again detonated in 
the canal at 400-ft intervals. The recording units 
consisted of a 250-ft geophone spread with geo- 
phones spaced 50 ft apart. The spread was in 
line with the shot line. 
The travel-time plots for both traverses are 
shown in Figure 4. The abscissa is given in 
terms of feet, but the velocities have been con- 
verted into units of km/sec for easy compari- 
son with other plots to follow. The larger dots 
represent the arrival times of the first geophone 
in the spread. From the figure it can be seen 
that the apparent velocity across the spread 
often does not agree with the step-out velocity. 
This indicates complicated bedding beneath the 
spread. 
A few unusual features are conspicuous on 
the travel-time plots. For both traverses the 
plots indicate very high velocities at the ends 
of the profile. Because the distances involved 
preclude penetration to great depths, these ap- 
parent high velocities have been interpreted in 
