Jl. Galloway et H. Melling, TRACKING THE MOTION OF SEA-ICE BY CORRELATION SONAR, Journal of atmospheric and oceanic technology, 14(3), 1997, pp. 616-629
A prototype correlation sonar has been developed for the measurement o
f ice motion in polar seas. It operates in the very shallow-water mode
as a two-pulse spatial correlation log. The design was guided by nume
rical signal simulations, which justified the implementation of first-
difference filtering of the signal prior to cross correlation. Four co
mponents of velocity are measured separately and combined in optimal l
east squares fashion. Controlled trials in air and water have clearly
demonstrated the acquisition of useable interference patterns from sim
ulated ice targets. The prototype has been deployed under moving sea i
ce during three winters in the Beaufort Sea, using four pulse interval
s between 0.08 and 10 s that provided a speed range of 50 cm s(-1). Th
e observations have been evaluated in relation to independent ice spee
d and topography measurements by Doppler and ice-profiling sonars inst
alled nearby. Useable data were obtained about 80% of the time over wi
de ranges in the speed and character of the ice target. Except for ver
y low speeds, the estimates by correlation were noisy relative to Dopp
ler determinations. This characteristic was traceable to the nature of
the operating algorithm, to the design of the receiving antenna, and,
at times, to the highly specular character of the echo from the ice.
Only about 1% of data loss was attributable to reasonable failures of
the operating algorithm. Within the remaining fraction, the incidence
of loss increased with increasing displacement of the interference pat
tern across the antenna between pulses. This is consistent with a decr
easing ability to track the interference pattern using a linear array
of hydrophones when pattern displacement transverse to the array excee
ds the pattern decorrelation scale. In future development of the corre
lation sonar for this application, it is recommended that the design o
f the transmitting and receiving arrays be modified to reduce the inci
dence of tracking failures, that all hydrophones in the receiving ante
nna be operated simultaneously, that the operating mode be converted f
rom a spatial to a temporal correlation concept, and that the dynamic
range be extended. With these enhancements, the correlation sonar will
be an effective tool for ice observation in polar seas.