Environmental Acoustic Feature Vectors For Identification Of Similar Ocean Acoustic Environments

Ocean analyses, forecasts and ensembles characterizing oceanographic uncertainty are becoming widely available. Acousticians are interested in the impacts of oceanographic variability (temporal and spatial) and uncertainty on acoustic propagation, however, it is very time consuming to compute acoustic propagation along many potential paths for multiple sensors, over a geographic area for many analysis/forecast times and/or ensembles.To facilitate faster prediction over many model configurations of interest, a method to identify similar acoustic waveguides prior to executing acoustic models has been developed. Environmental acoustic feature vectors (EAFVs) are developed to include characteristics of the waveguide that are important to acoustic transmission. The EAFVs are very quick to compute and are used with various techniques, such as mathematical clustering, to identify environmental-acoustic similarities over a geographic area, forecast time and/or ensemble prior to running an acoustic model. It is shown that clustering on EAFVs produces reasonable estimates of TL clusters. A representative environment is then identified as the full waveguide with EAFVs closest to the mathematical cluster centroid and can be used in place of each of the members of a cluster, thus requiring less computation time over an area/time of interest.