Asymptotic Accuracy of Geoacoustic Inversions

Michele Zanolin, Ian Ingram, Aaron Thode, Nicholas C. Makris

Research output: Contribution to journalArticlepeer-review

Abstract

Criteria necessary to accurately estimate a set of unknown geoacoustic parameters from remote acoustic measurements are developed in order to aid the design of geoacoustic experiments. The approach is to have estimation error fall within a specified design threshold by adjusting controllable quantities such as experimental sample size or signal-to-noise ratio (SNR). This is done by computing conditions on sample size and SNR necessary for any estimate to have a variance that (1) asymptotically attains the Cramer–Rao lower bound (CRLB) and (2) has a CRLB that falls within the specified design error threshold. Applications to narrow band deterministic signals received with additive noise by vertical and horizontal arrays in typical continental shelf waveguides are explored. For typical low-frequency scenarios, necessary SNRs and samples sizes can often approach prohibitively large values when a few or more important geoacoustic parameters are unknown, making it difficult to attain practical design thresholds for allowable estimation error. © 2004 Acoustical Society of America.
Original languageAmerican English
JournalThe Journal of the Acoustical Society of America
Volume116
DOIs
StatePublished - Oct 2004
Externally publishedYes

Keywords

  • underwater sound
  • inverse problems
  • parameter estimation
  • acoustic waveguides
  • acoustic measurement
  • oceanographic techniques

Disciplines

  • Acoustics, Dynamics, and Controls
  • Oceanography and Atmospheric Sciences and Meteorology
  • Other Physics

Cite this