Quantitative evaluation of a passively leveled ocean bottom seismometer

A problem in the use of ocean bottom seismometers is the difficulty in leveling the sensors while ensuring good coupling to the seafloor. We have investigated the coupling characteristics of the seismic sensors in the new ONR ocean bottom seismometer. In the deployable sensor package for that instrument, a three-component seismometer set is suspended on a 2-axis passive leveling gimbal and is immersed in a viscous fluid. We report tests, conducted in a seismic vault, comparing the output of a gimbaled seismometer set to that of a set rigidly coupled to the ground. Our results show that the degree to which the gimbaled set is coupled to ground motion is a function of the viscosity of the coupling fluid. The coherence between the two sensor sets is poor (<0.4) at some frequencies within the band of interest (0.15 to 20 Hz) and on some components when the viscosity of the coupling fluid is comparatively low (14 Pa-s or 0.16 kSt kinematic viscosity). In addition, the outputs of some components over portions of this frequency band are attenuated and are phase-shifted relative to the outputs of the set rigidly coupled to the ground. Coherence and phase response similarity improve as the viscosity of the coupling fluid is increased. With a coupling fluid viscosity of 980 Pa-s (10 kSt), coherence and phase agreement between the two sensor sets is good (>0.9) across nearly the entire band of interest on all three components. A simple analytical model of the gimbaled seismometer set as a damped, driven, compound-pendulum provides a basis for understanding the test results.