Ocean acoustic tomography: a scheme for large scale monitoring

We consider the problem of monitoring ocean basins for mesoscale fluctuations, using acoustic inverse techniques. The procedure, which has much in common with conventional seismology, consists of measuring perturbations in travel time between acoustic sources and receivers. Because the number of pieces of information is the product of the number of sources, receivers, and resolvable multipath arrivals, the economics of the system is enhanced over usual spot measurements. The temporal resolution required to distinguish multipath arrivals is estimated at 50 ms; the precision required to measure mesosclae perturbations is estimated at 25 ms. The required resolution and precision can be achieved by existing low-frequency (100- to 200-Hz) broadband (> 20-Hz) sources, but we are ultimately limited to 1000-km ranges by the variable ocean finestructure and associated micropaths. There appear to be no practical range limits imposed by micropaths if such broadband sources could be centered at 30 Hz. Given the travel time measurements and their noise estimates, we show how actually to invert the system for the interior changes in sound speed and, by inference, for density. The method is analogous to the medical procedure called tomography (from the Greek 'slice'). Measures of the spatial resolution and of formal error bars are obtained. We conclude that such a system is achievable now and has potential for development in a number of directions.