The generation of N-layered geoacoustic descriptions of the subbottom using NAVOCEANO-collected environmental data

A technique has been developed and implemented by the Naval Oceanographic Office for the generation of N-layer low-frequency bottom-loss (LFBL) databases. This technique employs a data collection methodology, a geoacoustic inversion technique, and fine-scale seismic data interpretation to generate a regional layered geoacoustic (compressional velocity, density, attenuation) definition of the subbottom. These geoacoustic descriptions are optimized to characterize the attenuation of waterborne acoustic energy by the bottom, in the range of 20-1000 Hz. The process begins with the collection of multiple types of environmental data on two types of survey operations. On the first type of survey operation, collocated trackline measurements of broadband transmission loss seismic, chirp, and bathymetry are made. Trackline lengths are generally on the order of 15-20 nmi. In addition, in situ conductivity-temperature-depth (CTD) and expendable bathythermograph (XBT) measurements are made. On the second type of survey operation, seismic and chirp data are collected along regional tracklines. The spatial separation of the tracklines is a function of the complexity of the geologic layering and is chosen so that accurate regional layer thickness maps can be generated. From the data collected on the first type of survey, subbottom geoacoustic parameters are inverted for using a global optimization method genetic algorithm (GA). Seismic and chirp data at each transmission-loss location are interpreted for range-dependent layer depths; for each layer a starting field of geoacoustic parameters is generated using sequence stratigraphy, along with empirical relationships between lithology, burial depth, and geoacoustic values. For each geoacoustic parameter, a bound is imposed, representing realistic expected values. The layered starting field of geoacoustics is input to the PE5.0 model and is allowed to iterate through the imposed bounds, using the GA to minimize the error between the transmission-loss model output and the measured transmission-loss data. Regional layered descriptions are generated by extrapolating the inverted for layered descriptions from each transmission-loss measurement location throughout the area. Layer thickness maps are first generated using the- regional seismic data. A provinced set of velocities, densities, and attenuations, ground-truthed at the locations at which transmission-loss data were collected, are applied to each layer throughout the region.