Matched-field localization of explosive sources in the Barents Sea using a horizontal array

Matched-field processing (MFP) techniques are used to estimate the range and depth of underwater explosive sources. Data from an experiment conducted by FFI in the Barents Sea 1999 is used. The experiment was conducted in a relatively flat area, using explosive (SUS) sources and a 31-element receiver array. Low-frequency data from shots endfire to the ten-element bottom mounted horizontal section of the array (length 820 m) is used in this report. A two-step method is employed. The first step (environment focusing) makes use of data from known sources in matched-field inversion for the estimation of seabed geoacoustic model parameters. The second step (source localization) employs use of the incoherent broadband Bartlett processor and a fast normal mode acoustic propagation model (C-SNAP) in an exhaustive search for source range and depth in the model environment obtained in the first step. The second step is fast, requiring less than a minute of computer time on a 1.1 GHz processor. Excellent results are obtained when combining data from two 10 Hz wide frequency bands centred at 40 Hz and within 80-110 Hz, with source positions determined to within 200 m in range and 6 m in depth for ten sources at ranges 3-27 km from the array.