nic disturbance in the equatorial Pacific that affects global weather patterns.

　　Unfortunately, it is not feasible to use conventional techniques to measure mesoscale fields. To measure them properly, monitoring equipment would have to be laid out on a grid at intervals of at most 50 kilometers, with sensors at each grid point lowered deep in the ocean and kept there for many months. Because using these techniques would be prohibitively expensive and time�瞔onsuming, it was proposed in 1979 that tomography be adapted to measuring the physical properties of the ocean. In medical tomography X-rays map the human body’s density variations (and hence internal organs); the information from the X-rays, transmitted through the body along many different paths, is recombined to form three-dimensional images of the body’s interior. It is primarily this multiplicative increase in data obtained from the multipath transmission of signals that accounts for oceanographers’ attraction to tomography: it allows the measurement of vast areas with relatively few instruments. Researchers reasoned that low-frequency sound waves, because they are so well described mathematically and because even small perturbations in emitted sound waves can be detected, could be transmitted through the ocean over many different paths and that the properties of the ocean’s interior—its temperature, salinity, density, and speed of currents—could be deduced on the basis of how the ocean altered the signals. The