Anoxic zones, regions of the water column completely devoid of dissolved oxygen, occur in open oceans and coastal zones worldwide. The Baltic Sea is characterized by strong salinity-driven stratification, maintained by occasional water inflows from the Danish Straights and freshwater input from rivers. Between inflow events, the stratification interface between surface and deep waters hinders mixing and ventilation of deep water; consequently, the bottom waters of large regions of the Baltic are anoxic. The onset of the anoxic zone is closely coincident with the depth of the halocline and, as a result, the interface between oxic and anoxic waters corresponds to a strong impedance contrast. Here, we track acoustic scattering from the impedance contrast utilizing a broadband split-beam echosounder in the Western Gotland Basin and link it to a dissolved oxygen level of 2ml/l using ground truth stations. The broadband acoustic dataset provides the means to remotely observe the spatiotemporal variations in the oxic-anoxic interface, map out the extent of the anoxic zone with high resolution, and identify several mechanisms influencing the vertical distribution of oxygen in the water column. The method described here can be used to study other systems with applications in ongoing oceanographic monitoring programs.

Stable fluid bodies, such as the ocean and atmosphere, are composed of a series of increasingly dense layers, defined by density stratification interfaces in which the medium properties (e.g., temperature, salinity) change. The intensity of the stratification between the layers influences the internal mixing dynamics and entrainment, facilitating the transport of dissolved constituents within the fluid medium. Acoustic systems offer the means for high resolution observations of these interfaces, which allow for continuous data collection over broad spatial scales. Here, a one-dimensional acoustic scattering model is presented for predicting acoustic backscatter from stratification interfaces, which is widely applicable to the acoustic water column data collected with ship-mounted sonars. Model predictions based on hydrographic profiles suggest that in many oceanic cases, the density gradient perturbations can be disregarded, and sound speed perturbations alone drive the majority of the acoustic scattering. A frequency-dependent scattering intensity based on the sharpness of the stratification interface is predicted by the model, suggesting a path to remote estimations of the physical medium properties through broadband acoustic inversion.