Acoustic pressure and particle velocity for spatial filtering of bottom arrivals

This paper discusses the advantages of using a combination of acoustic pressure and particle velocitymotion for filtering bottom arrivals. A possible area of application is reflection seismology where, traditionally, the seismic image is extracted from the bottom-reflected broadband acoustic signals received on hydrophones. Since hydrophones are omnidirectional in nature, the received bottom returns are often contaminated by waterborne signals, sea surface reflections, and noise. A substantial part of the processing of the data is dedicated to filtering out these unwanted signals. Today, vector sensors allow us to measure both acoustic pressure and particle velocity motion in a single and compact sensor. The combination of pressure and particle velocity measured at a single location or particle velocity and particle velocity gradient at closely spaced locations allows for spatial beam steering to predetermined directions and filter out unwanted replicas from other directions. Moreover, this can be done at the sensor level, dramatically decreasing the offline processing. The spatial filtering capabilities of various pressure-pressure, particle velocity-particle velocity, and pressure-particle velocity combinations are analyzed in view of filtering the bottom arrivals. It is shown that the combination of pressure and vertical particle velocity and, particularly, the combination of vertical particle velocity and particle velocity gradient enhance bottom arrivals. Moreover, a simple steering procedure combining pressure and particle velocity components of a triaxial sensor allows us to determine the tridimensional structure of the acoustic field and the separation of the bottom reflections. The spatial selectivity of the various sensor combinations is shown with simulations and verified with experimental data acquired with 10 cm separated vector sensors in the 800-1250-Hz band, during the Makai 2005 sea trial, off Kauai Island, HI, USA.This work was supported by the European Union H2020 Research Program under WiMUST Project (Contract 645141).