The deep ocean (200 m or deeper) covers 66% of Earth’s surface, encompasses the largest ecosystem on our planet, provides critical services and resources to humankind, and is the least explored and understood biome on Earth. Due to rapid absorption of electromagnetic waves by seawater, limiting their range to tens or hundreds of meters, acoustics remains the only practical tool for sensing, communication, and exploration in the ocean, efficiently propagating over vast distances (tens to thousands of kilometers).
However, the high inhomogeneity of the marine environment, characterized by strong spatial and temporal variability, results in extremely complex acoustic conditions. This complexity poses significant challenges for both direct problems (predicting sound propagation) and inverse problems (determining environmental properties from sound measurements) in ocean acoustics. This seminar explores these difficulties by addressing one direct and one inverse problem, each set within drastically different acoustic environment.
First, we investigate how internal Kelvin gravity waves affect sound propagation in shallow water. Using an experimental data from Lake Kinneret and advanced sound propagation modeling, we demonstrate that internal Kelvin waves can produce a 12 dB drop in sound intensity (a fourfold pressure decrease) over a 5-km acoustic track, and cause substantial coupling of acoustic normal modes, making conventional adiabatic approximation invalid.
The second part focuses on the Challenger Deep, Earth’s deepest point, where we use a synthetic-aperture autonomous acoustic recording system to address an inverse problem. We demonstrate how wind-driven ambient noise can be used to infer ocean acidity (pH), a key metric for assessing ecosystem health. By examining the depth-dependent attenuation of sound and applying passive absorption spectroscopy, we present a novel, scalable method for estimating the volume-integrated pH of seawater. This approach offers the potential for large-scale acidity measurements without relying on traditional, localized point sampling.
