pulse occurrence to sequence of events during the detonation; (6) correlation of device
characteristics and pulse characteristics, both close-in and, as far as possible, at
distances; (7) experimentation with prototype surveillance equipment; (8) measurement
of azimuthal errors in direction-finding equipment; and (9) determination of times of
pulse reception to within 1 msec in world time.
In order to achieve these objectives, two fundamental problemsfirst had to be solved:
(1) the discrimination of nuclear-device pulses from nature! atmospherics and (2) the
cetermination of the maximum information on the source tiself and external conditions
at detonation time from the characteristics of this electromagnetic pulse.
7.1.1 Pulse Identification. One means of identifying a nuclear-detonation pulse with
an experimental system (when recording at distances from the detonation point) is by
knowledge of the time of detonation.
To aid pulse identification during Castle, local
. timing signals were referred to world time. Both timing signals and pulse signals were
corrected for propagation, giving an accuracy of 1 msec for world time and less than 1
megec for the pulse. Reception and identification of such pulses when time of detonation
was known to millisecond accuracy was relatively easy; doing the same thing on a 24-hour
basis 1f the detonation time had not been known would have been much moredifficult.
More information was found to be needed on techniques of discrimination, much of which
could be learned by studying naturally occurring atmospherics.
In locating the puise source, azimuthal errors were generally within the error ordi-
narily experienced with the location equipment used: +3 degrees.
7.1.2 Pulse Characteristics. All close-in records showed the characteristic first
negative-going pulse; wherever the effect of the second stage was apparent (except Shot 3)
the first portion of the secondary pulse went positive. Wave forms were recorded at
distances up to 12,000 km; however, beyond about 2,000 to 4,000 km, close-in detail disappeared. The changes in wave form causedbythe filtering effects of the ionosphere
(decreased reflection of the higher-frequency components) and interference between different sky-wave modes was quite apparent as the broad-band pulse was recorded at
greater distances: the pulse lost character and presented a damped-sine-wave appearance. The broad-band wave formsat the far stations, in general, covered about 6 to
100 kc, which encompassed the greatest portion of the energy available.
7.13 Field Strength. Data from Guam, Shemya, and Point Barrow were generally
low.
The reasons were not definitely known, and these anomalies are being investigated.
Contributing causes may have been interference between sky-wave modes, ionospheric
absorption, ground constants, and in the case of Point Barrow, attenuation due to aurora!
absorption. In addition, it was believed that the Shemya results may have been low because of local conditions at the receiving site.
There was apparently considerable variation from day to day and during the day. Dayand-night variation in signal strength was generally more pronounced on the north-south
path than the east-west. The magnitude of diminution in signal from dark-to-daylight
path was apparently greater when the auroral zone was penetrated. Field strengths were
lower during magnetically disturbed periods (i.e. , 24 March 1954) than during relatively
quiet magnetic periods.
7.1.4 Yield Determinations. Field strength, especially at distant points, was only a
very-approximate measure of yield; however, the vagaries of propagation were only imperfectly known—yield is also more properly a function of total energy emitted. For an
93