Yilmaz (2001), at page 1809, maintains that such reflections are due to density contrasts. Holbrook et al. (2003) use a towed submersible to directly measure the salinity and temperature of water while performing a seismic survey. They conclude that thermohaline anomalies result chiefly in acoustic velocity deviations of less than 20 m/s from the average. Velocities are found by calibrating reflectivity values with local direct measurements, in the context of the aquatic medium, which lacks illumination and focusing problems.
Images of the thermohaline fine structure of the seas can be a useful tool for oceanographers. We describe a processing flow designed to extract seawater reflections from under the shot noise using prediction-error filters (PEFs). We also show an image of thermohaline reflection, as well as evidence that velocity varies with depth and midpoint in the ocean. We propose performing wave-equation migration velocity analysis (WEMVA) to find the corresponding velocity anomalies in the seawater.