High-resolution geoacoustic inversion in shallow water: a joint time- and frequency-domain technique

High-resolution geoacoustic data are required for accurate predictions of acoustic propagation and scattering in shallow water. Since direct measurement of geoacoustic data is difficult, time-consuming, and expensive, inversion of acoustic data is a promising alternative. However, the main problem encountered in geoacoustic inversion is the problem of uniqueness, i.e., many diverse geoacoustic models can be made to fit the same data set. A key, and perhaps unique, aspect of this approach is the combination of data analysis in both the space-time and the space-frequency domains. This combination attempts to ameliorate the uniqueness problem by exploiting as much independent data as possible. In order to meet the stringent requirements of high spatial resolution and uniqueness, an entire method has been developed including a new measurement technique, processing/analysis technique, and inversion strategy. These techniques are described and then illustrated with a shallow-water data set. Sound-speed gradients in the upper few meters of the sub-bottom appear to be much higher (one order of magnitude) than generally assumed. And, although often ignored, a large density gradient was observed in the top layer that played an acoustically significant role.