In situ acoustic calibration for a large aperture array

During September 1987, a large aperture acoustic array was deployed vertically in the Northeast Pacific to study low-frequency noise in the ocean. Coherent combination of the 120-channel outputs requires knowledge of individual element amplitude and phase response for accurate results. Two in situ methods of array calibration are described and results from the September experiment are presented. The first method used transmissions from a low-frequency source of known location and power level. Simulating the conditions encountered during the transmission, the power arriving at the array was predicted by several acoustic propagation models. By comparing the array response at specific frequencies to the response predicted by the models, an absolute calibration was obtained. An error curve for the phase data was generated by unwrapping the phase, accounting for a sampling offset in the array, and subtracting a multiple linear regression curve. The second method determines relative amplitude levels by examining the average ambient noise power output of a specified frequency band across the array. Using spectral, coherence, and directionality plots, the level of self-noise in the array was shown to be below that of the ambient noise being measured. These two independent methods provide a consistent set of element calibration values used for array beamforming. [Work supported by ONT.]