Statistical models for ambient noise imaging in temperate and tropical waters

Abstract

The data obtained from the Acoustic Daylight Ocean Noise Imaging System (ADONIS) in San Diego, California, is dominated by ambient noise from snapping shrimp. This is expected in all warm and shallow waters, except with very high winds or precipitation. It is shown that these data exhibit a log-normal intensity distribution. This exactly determines the number of independent statistical parameters that can be used to create images. Further, ambient noise data taken near Singapore are also shown to be log-normal in intensity. This distribution is common for unsaturated, multiply-scattered acoustic fields, but it is shown that it is a property of the sources and not the propagation. It is also demonstrated that good ambient noise images can be created from cross correlations of pixel intensities, independently of mean or other statistical measures. Image examples will be presented and some implications for the next-generation ambient noise imaging system will be discussed. [This work was funded by NUS and DSO, Singapore. ADONIS data provided by Scripps Institution of Oceanography.]