Broadband Synthetic Aperture Matched Field Geoacoustic Inversion

摘要: A typical geoacoustic inversion procedure involves powerful source transmissions received on a large-aperture receiver array. more practical approach is to use moving single source/receiver, broadband, frequency- coherent matched-field strategy that exploits coherently repeated improve estimation of the parameters. The long observation time creates synthetic aperture due relative source- motion. To correlate well with measured field, waveguide Doppler and normal mode theory applied. However, this method uses model constrains source/receiver radial velocity be constant. As result, performance degrades when acceleration exists. Furthermore, processing train pulses all-at-once does not take advantage natural incremental acquisition new along ability assess temporal evolution parameter uncertainty. Therefore, recursive Bayesian developed processes data pulse-by-pulse incrementally updates estimates It also approximates by assuming piecewise constant but linearly changing velocities. When exists, it shown modeling can reduce further biases uncertainties. Finally, above methods depended assumption underlying geophysical change-point detection proposed detect change in parameters using importance samples corresponding weights already are available from inversion. If abruptly, will detected restart pulse measurement after change-point. gradually, (based parameters) may proceed estimate an averaged value until accumulated mismatch significant triggers These form heuristics for controlling integration Examples based either synthetically generated acoustic fields or set low SNR, 100-900 Hz LFM pair Shallow Water 2006 experiment