During the Seabed Characterization Experiment, a multi-institutional field effort held at the New England mud patch, the autonomous passive acoustic lander Deep Sound made a series of ambient noise measurements from the seafloor. The instrument platform carried four hydrophones, arranged in an inverted ‘T’ shape with three spaced in the horizontal and two in the vertical, and landed on the seafloor with the bottom phones 30 cm above the interface. Pressure time series, vertical and horizontal noise coherence (directionality), were recorded continuously for periods of 9 hours over the acoustic bandwidth of 5 Hz to 30 kHz, along with the local temperature, conductivity, and depth. An analytical Pekeris waveguide noise model was fitted to the data in order to determine the bulk sound speed, sheer speed, density, and frequency dependent attenuation in the bottom fluid half-space. Acoustic properties of the mud were determined by comparing the data to the output of a range independent noise model, featuring a realistic multi-layered seabed. [Research supported by ONR.]During the Seabed Characterization Experiment, a multi-institutional field effort held at the New England mud patch, the autonomous passive acoustic lander Deep Sound made a series of ambient noise measurements from the seafloor. The instrument platform carried four hydrophones, arranged in an inverted ‘T’ shape with three spaced in the horizontal and two in the vertical, and landed on the seafloor with the bottom phones 30 cm above the interface. Pressure time series, vertical and horizontal noise coherence (directionality), were recorded continuously for periods of 9 hours over the acoustic bandwidth of 5 Hz to 30 kHz, along with the local temperature, conductivity, and depth. An analytical Pekeris waveguide noise model was fitted to the data in order to determine the bulk sound speed, sheer speed, density, and frequency dependent attenuation in the bottom fluid half-space. Acoustic properties of the mud were determined by comparing the data to the output of a range independent noise model, featuring a ...