Pseudoseisms from Military Exercises — Krivoy, Johnson, and Koyanagi 
125 
in which the disturbances were recorded and felt 
is approximately 190 km from the Kahoolawe 
source. Kaula Rock is 400 km to the west of 
Kahoolawe. An approximate schedule supplied 
by the Navy did not specify the exact time of 
any individual explosion or the total number of 
bombs dropped. By virtue of overlap in attack 
periods on the two islands, however, there is 
weak evidence, based on schedules of exercises 
for 19 and 20 December 1961, that all dis- 
turbances recorded and felt on the island of 
Hawaii originated at the closer target, Kahoo- 
lawe. On 13 February 1962, both targets were 
bombed simultaneously, but because the bomb- 
ing log is skeletal, no conclusions can be drawn. 
No log was provided for the exercise of 29 
October 1963; but it was confirmed that bomb- 
ing had taken place and it was implied that 
only Kahoolawe was involved. 
Bombing exercises on these targets during 
1961 only, and qualitative results from them 
obtained 
by searching the seismograms follow: 
DATE 
HOURS 
RESULTS 
24 Feb. 
06:40-14:30 
No recordings, eruption 
tremor in progress. No 
sensations reported. 
16 May 
05:30-13:00 
Neither recordings nor 
sensations. 
22 Aug. 
05:00-12:00 
Same as above. 
15:30-20:00 
Same as above. 
23 Aug. 
09:35-13:30 
Same as above. 
18 Nov. 
05:40-13:10 
Weakly recorded on Ahua 
only, no sensations re- 
ported. 
20 Nov. 
10:30-13:40 
Neither recordings nor sen- 
sations. 
17:45-19:20 
Weakly recorded on Ahua 
only, no sensations. 
21 Nov. 
09:30-13:35 
Very weak on Mauna Loa 
and Ahua only, no "felt” 
reports. 
19 Dec. 
06:30-12:30 
Weakly recorded, no sen- 
sations reported. 
17:45-20:00 
Many reports and good 
records (Table l). 
20 Dec. 
09:30-13:00 
Some reports and good 
records (Table 2). 
GRAPHICAL ANALYSIS: FIGURE 1 
Because of the almost simultaneous arrival of 
the sonic waves at Uwekahuna and Desert, it 
was convenient to use a graphical technique to 
estimate the azimuth to the source on the basis 
of data summarized in Table 3. On Figure 1, 
therefore, an assumed wave front has been 
drawn at the instant it passes Desert seis- 
mometer. The seismic stations which time such 
wave fronts are arrayed along a chord which 
is approximately 20 km long and about 20 km 
from the source. With this geometry, the chord 
is within 0.3 km of the circular segment it inter- 
sects; therefore, straight-line wave fronts have 
been assumed. 
Seismograph recordings of sonic disturbances 
differ considerably from those of local earth- 
quakes. There are no definite phase identifica- 
tions for sonic arrivals as recorded by short- 
period seismometers; instead, the maxima have 
been read on each record. Some events were 
recorded as featureless bursts barely resolvable 
above the normal background noise. The Desert 
record as illustrated in Figure 2 usually gave the 
clearest and largest arrival. We can only specu- 
late on the possible interference of wave groups 
following slightly different paths from the 
source to the receivers and on the effect such 
interference would have on the times of maxima 
at different receivers. However, the relatively 
low velocity of sound in air reduces the degree 
of precision necessary in timing sonic arrivals 
compared with that for seismic waves, for 
example. 
If we assume that the correlation of the sonic 
arrival between receivers was in error by 5 or 
10 seconds, and if we apply all of this error at 
either extreme of the hypothetical 20-km record- 
ing chord, errors of only 1 0 of arc would result. 
These errors are so small that confusing a source 
on Kahoolawe with one on Kaula Rock seems 
unlikely. 
If we make these simplifications and allow- 
ances for error, and if we assume further that 
constant velocity prevailed in the seismic record- 
ing zone and over the 200-km propagation path, 
a direction of N 44° -46° W may be read from 
the diagram on Figure 1. This solution is excel- 
lent for a Kahoolawe source. (The Kaula Rock 
target is about N 64° W of the seismic pickup 
location.) 
VARIATIONS IN APPARENT VELOCITY 
The traveltimes (averaged from Table 3) 
for each leg in the wave front’s passage are 
shown in Figure 1. The spacing provided by 
