We have developed a productive and innovative research  program based on the development and application of ocean acoustic inverse methods,  and the study of seismo-acoustic propagation in the ocean and sea bed. To date,  the research projects have supported thirteen graduate students, eight of whom have  completed their thesis programs, and over 35 journal publications.

     Much of the research on inversion is based on a powerful  new technique known as matched field processing (MFP). In its simplest terms, MFP  is a model-based, full wave inversion method in which measurements of the acoustic  field are compared to calculated fields. Matched field (MF) inversion seeks the set  of environmental model parameters of the ocean that optimize the match between the  measured and modelled fields. The technique is applied for inversion of geometrical  parameters of the experiment (source localization), and for the estimation of  geoacoustic properties of the ocean bottom. The inversion methods have also been  applied to the study of methane gas hydrates in a collaboration with the Geophysics  Group at SEOS and Dr. R. Hyndman at PGC.

     Although the program is strongly linked to research at  DND, it is designed to complement and not duplicate the research activities at DREA  and EDRD. For example, the OA group has investigated new inversion methods such as  phase demodulation and interpretation of head wave data, and studied background issues  in MFP such as cross relation MF processors, advanced global search algorithms and range  bias errors in source localization in range dependent environments that are not necessarily  within the mandate of research at the DND laboratories.

     The OA group has interacted effectively with industry to  develop collaborative research projects. A three-year project to investigate the use  of radiated noise from passing ships as sound sources for geoacoustic inversion was  carried out with MacDonald-Dettwiler (MDA). This research was supported by the  NSERC/DND/Industry partnership program. A continuing research program was established  with Quester Tangent on extending their in-house methods for acoustic sea bed classification  to make use of sub-bottom features.

     We have also made use of other funding opportunities to  broaden the scope of our research program. Funding from the US Office of Naval Research  (ONR) partially supported the group's participation in the Haro Strait experiment in  June 96, and subsequently funded the analysis and interpretation of the data. ONR  funding also supported the Geoacoustic Inversion Workshop that was in co-operation  with MacDonald Dettwiler in June 97. Stan Dosso received a CFI new investigator  award for development of a seismic survey system that has enabled new 3-D seismic  experiments on the gas hydrates offshore Vancouver Island. He also received funding  from the US Navy to process and interpret the array element localization data from the  joint Canadian/US field trials in the Arctic; this work lead to further contract  work from the US, and has supported several Co-op students.

The highlights of the research program are listed below.

• Geoacoustic inversion using broadband ship noise :
  This project investigated the use of radiated noise from passing ships as  sound sources for geoacoustic inversion. A new cross-relation MF processor was  developed for broadband random signals such as ship noise that was superior  to conventional processors. Its performance was demonstrated for source localization  and geoacoustic inversion in shallow water.
  
  
• Haro Strait Matched Field Geoacoustic Tomography Experiment :
  This experiment was the first to use multiple hydrophone arrays and multiple sound  sources to estimate the range variation of geoacoustic parameters in an area of  the sea bed. A unique feature of the design was the use of light bulbs as sound  sources in waters where conventional sound sources are restricted due to the presence  of marine mammals. This work has resulted in the development of vertical slice  tomography with multiple hydrophone arrays, and a statistical freeze bath inversion  method for appraisal of the estimated parameters.
  
  
• Advanced techniques in geoacoustic inversion :
  New efficient hybrid search techniques that combine simulated annealing,  genetic algorithms and downhill simplex methods were developed for MF inversion.   Matched mode processing was investigated as an alternative means for inversion of  acoustic field data.
  
  
• Array element localization algorithms :
  Linear inversion algorithms based on ray theory were developed to determine  the positions of array elements. The method treats source and receivers as unknowns,  and accepts a priori information consisting of initial position estimates and/or  the assumption of a smooth array shape.
  
  
• The COAMS array project :
  The COAMS multichannel hydrophone array was modified for use as a marine  seismic survey system. This project was funded by the Canadian Foundation for  Innovation, and has provided an effective experimental system for marine seismic  research that is unique on the west coast of Canada.
  
  
• Submarine gas hydrates :
  Marine seismic experiments with the US Navy advanced deep towed array  (DTAGS) and new 3-D surveys with COAMS have been carried out to investigate the  vertical distribution of methane hydrate offshore Vancouver Island, and to  determine the relationship of hydrate with the geological structure in the  sediments of the accretionary margin.
  
  
• Head waves and vertical hydrophone arrays :
  The advantages of a vertical array geometry were demonstrated for extracting  geoacoustic parameters from head wave data.
  
  
• ONR Geoacoustic Inversion Workshop :
  A workshop that attracted participation of over 40 scientists from Europe  and Asia as well as Canada and the US was held to benchmark geoacoustic inversion  methods. The proceedings that were edited by Dr. Chapman and published in the  Journal of Computational Acoustics represent the state of the art in geoacoustic  inversion research.