Comments: download file (pdf
)
Ref.: SiPLAB Report 02/02, University of Algarve,
2002.
Abstract: The estimation of range-dependent properties with
a single array-source pair is a problem which solution may not be
unique. Being aware that this dificulty is particularly true when data
are contaminated with noise (i.e., in all real cases), our objective
was to obtain a variability trend rather than very accurate estimates
of the properties. In other words, detection and global tracking of
the filaments were of prime interest rather than detailed mapping
of the sea-temperature (or sound-speed) field. While acoustic
travel-time tomography [1] is now a well-developed technique for large-
scale, deep-ocean regions, it is less adapted for studying
filaments which are mesoscale features that develop in relatively
shallow
(400 m) areas. Using matched-field processing to estimate ocean sound
speed [2, 3] is a more recent approach than tomography. However this
approach has already shown good results and can treat any type of
environment equally. It is therefore the approach adopted in this study.
ACKNOWLEDGMENT: this work was partially supported by ATOMS project (FCT, Portugal)