Converting Seismic Reflection Sections into Oceanic Temperature and Salinity: A New Approach

Oceanic submesoscale dynamics are associated with horizontal scales between tens of meters and tens of kilometers, and timescales of hours to weeks. Through impacting the transfer of energy and other fundamental ocean properties such as heat, salt, carbon and nutrients, submesoscale processes are believed to play an important role in the climate system and marine biosphere. However, direct observations of these processes, especially in the ocean interior, remain limited due to their transient nature. Marine seismic reflection surveys offer a solution, resolving thermohaline structures on scales of order 10 m vertically, and 100 m horizontally, and capturing 100 km swathes in hours. While seismic data provides vertical temperature/salinity gradients, legacy datasets are often hindered by sparse hydrographic validation and uncertain inversions. Here, we present an improved inversion method combining root mean square sound velocity analysis and iterative Markov Chain Monte Carlo techniques to extract thermohaline fields with quantified uncertainties. The method is validated using Gulf of Cadiz seismic data with coincident hydrographic measurements and applied to a new Mozambique Channel dataset capturing mesoscale and submesoscale activities. Uncertainties for inverted temperature and salinity are 2.5°C (1.65°C) and 0.5 psu (0.08 psu) in the Gulf of Cadiz (Mozambique Channel), with Dix equation-derived velocity conversion identified as the primary error source. This novel approach expands the use of legacy seismic reflection data as a tool for ocean finescale to submesoscale analyses, and will aid new, global insights into previously difficult-to-observe ocean dynamics.