Mg. Brown et al., ON THE MEASUREMENT OF MODAL GROUP TIME DELAYS IN THE DEEP-OCEAN, The Journal of the Acoustical Society of America, 100(4), 1996, pp. 2093-2102
The modal description of sound propagation in deep ocean environments
is considered. Recently published inversion algorithms have assumed th
at modal group time delays can be measured. Such a measurement is stra
ightforward to make if either: (1) the modal group arrival of interest
is resolved in time; or (2) the wave field is sampled on a dense vert
ical array which spans much of the water column, thereby enabling the
orthogonality of the modes to be exploited. In order to temporally res
olve modal group arrivals using measurements made on a single hydropho
ne, the frequency bandwidth must: (a) be sufficiently broad that the t
emporal separation between successive modal arrivals exceeds the recip
rocal bandwidth; and (b) be sufficiently narrow that, across the band,
the group slownesses of neighboring mode numbers do not overlap. To s
atisfy both conditions [(a) and (b)] the ratio of the range to the cen
ter frequency must be large. Unlike ray arrivals, modal group arrivals
broaden as range increases due to dispersive spreading. To minimize d
ispersive spreading so that accurate group delay time estimates can be
obtained, the ratio of range to center frequency should be kept small
. Thus the requirements for temporally resolving modal group arrivals
conflict with the conditions that minimize dispersive spreading. Numer
ical results are presented which give quantitative estimates of which
combinations of range, center frequency, bandwidth, and mode number pr
oduce conditions which are favorable for temporally resolving modal gr
oup arrivals in six regions of the deep ocean at mid and low latitudes
. These results suggest that it is extremely difficult to find conditi
ons which allow modal group arrivals to be temporally resolved and sim
ultaneously allow group delay times to be measured with sufficient acc
uracy to be useful for tomography. The situation is further complicate
d by internal waves which appear to cause mode coupling and significan
t broadening of modal group arrivals at frequencies above approximatel
y 50 Hz. The combined effects of dispersive spreading and internal wav
e-induced mode coupling suggest that, without employing mode filtering
techniques, modal group time delay-based inversion schemes in the dee
p ocean do not appear to be promising. (C) 1996 Acoustical Society of
America.