Water Column Seismic Images as Maps of Temperature Gradient

Abstract

Multichannel seismic imaging of ocean water column features is anew interdisciplinary study that may become an accepted oceanographic toot in coming years. We now know that reflectors are associated with water column thermohaline fine structures such as internal waves and intrusions (on a scale of similar to 10-50 m) associated with ocean mixing, and also that the images outline larger-scale oceanographic features such as currents, water-mass boundaries, eddies, meddies, and fronts. The synopticity and detail showing the relationships between mesoscale and fine-scale features promises improved insight into the processes that cascade energy from mesoscales to mixing scales. In order to trust a new tool, oceanographers require a quantitative understanding of how the new tool acts upon physical properties to yield a final result. We explain the basic principles of multichannel seismics, and show that the imaging process can be viewed as a filtering operation acting on the acoustic impedance field, which, on the scales that matter, is primarily (but not completely) associated with temperature variations. Synthetic seismic images show the derivative of acoustic impedance, averaged over the resolution scale of the acoustic source wavelet-they are, aside from side-lobe effects, essentially smoothed maps of temperature gradient. We use a conductivity-temperature-depth (CTD) trace from the periphery of a meddy to estimate the contribution of thermal (83%) and saline (17%) anomalies to a synthetic seismic trace, and then use multiple CTD traces from the same data set to construct a synthetic seismic image. This synthetic image compares favorably to a real seismic image of a different meddy with important differences that can be ascribed to the higher lateral resolution of the seismic technique.