Biases and caustics in long-range acoustic tomography

Ray chaos and eigenrays

“…The penalty, however, is the introduction of a ''mesoscale bias'' whose magnitude ͑and even sign͒ is unknown. 18,19 Estimates of this ''mesoscale bias'' give at least 50 ms at 1 Mm, which is comparable to the prediction uncertainty introduced by ray chaos. Another way out of this dilemma is to abandon ray methods altogether and develop more costly and less sensitive full-wave range-dependent acoustic models.…”

Parameter-Based Rules for the Definition of Detectable Ducts for an RFC System

“…This hypothesis is relevant in S-band with a sea state not too high. Indeed, with sea state 3 (wave height ) at 3 GHz, and considering grazing angles , the Rayleigh criterion expressed as (26) characterizing the smooth surfaces (e.g., [29]) is barely fulfilled. For higher frequencies or rougher sea states, the surface roughness should be considered.…”

“…However, rays are not sufficient to extensively describe the wave behavior. In the trapping layer, caustics appear and reach its upper and lower parts [26], [27]. The energy is focused at the vicinity of the caustics (a phenomenon known as "edge focusing" [28]).…”

Tomographic observations of deep convection and the thermal evolution of the Greenland Sea Gyre, 1988-1989

“…[Mercer and Booker, 1983;Munk and Wunsch, 1985;Munk and Wunsch, 1987]. This is of greatest importance during the February /March, where (as we shall show in Chapter 7) the tomographic result is an average though a large number of plumes.…”