A COMPARISON OF BROAD-BAND AND NARROW-BAND MODAL INVERSIONS FOR BOTTOM GEOACOUSTIC PROPERTIES AT A SITE NEAR CORPUS-CHRISTI, TEXAS

In September 1985, a series of narrow-band shallow-water acoustics experiments, collectively called project GEMINI, were performed in the Gulf of Mexico near Corpus Christi, Texas. The reasons for the experiments were threefold. The first was to obtain very accurate measurements of the pressure field versus range at a benign (i.e., range-independent, geologically simple) site which had been well studied previously, and then determine how well one could predict the measured fields using state-of-the-art propagation models with high-quality geoacoustic information (both a priori and a posteriori) as input. The second reason was to utilize, extend, and benchmark the narrow-band bottom measurement and inversion techniques previously developed by the authors [Rajan et al., J. Acoust. Soc. Am. 82, 998-1017 (1987)]. The third reason was to compare bottom models generated by narrow-band and broadband techniques. Originally, this was to be restricted to the three-layer model inferred by L. Rubano in a previous broadband study at the same site [L. Rubano, J. Acoust. Soc. Am. 67, 1608-1613 (1980)]. However, by employing Rubano's original modal dispersion data with broadband perturbative inverse techniques also developed by the authors, we could make a more general comparison of narrow-band and broadband inversions for the bottom properties. Detailed comparisons between narrow-band and broadband techniques are presented with emphasis on (a) comparisons of the pressure field versus range predicted, (b) comparisons of the bottom compressional-wave-speed model generated, (c) the variance of the bottom model estimated, and (d) the resolution of the bottom model generated. The range dependence of the bottom medium and the seasonal dependence of the near-surface sediment sound-speed profile are observed and discussed (particularly the former effect-the latter is the subject of a separate paper). We have drawn the following conclusions from this study: (a) The geoacoustic models obtained using narrow-band perturbative inversion on the WHOI data (model I), broadband perturbative inversion on Rubano's data (model II), and J. Matthews' synthesis of archival data are generally consistent with one another for the top 90 m of sediment. Meaningful quantitative comparisons of models I and II with Matthews' model are difficult because of the lack of error estimates for the archival data. (b) The continuous compressional-wave-speed profile inferred using the narrow-band perturbative inversion approach yields the best agreement between the theoretically computed and measured pressure fields. This result is partly due to the apparent inaccuracies of Matthews' model for the deeper sediment depths ( > 90 m), the constraint of three homogeneous layers imposed by Rubano's original model, and some error in fitting the broadband dispersion curve and correcting for seasonal effects in the broadband perturbative inverse calculation based on Rubano's data. (c) The differences between models I and II fall within the error bounds predicted by linear inverse theory everywhere except the top 10 m of sediment. There the difference is attributed to the hypothesis that the compressional wave speed in the uppermost part of the sediment column is influenced by the water column temperature; the Rubano and WHOI experiments were conducted at different times of year with significantly different water temperatures, thereby yielding the differing sediment results.