Characterization of Patches Along Transects Using High-Resolution 70-kHz Integrated Acoustic Data
A patch recognition algorithm was applied to high-resolution (1 m vertical and 25 m horizontal) daytime sonar date collected from a 20-km-length transect to a depth of 200 m. The transect was oriented perpendicular to the Gulf Stream frontal zone, 105 km east–northeast ENE of Cape Hatteras, North Carolina, on August 8 1985. An adaptive high-pass filter was used to identify patches of high-intensity echo strengths. For a broad based averaging "window size" of 13 m deep by 1.4 km long and an echo strength threshold of 1.4 × integrated echo units patches resemble fine-scale features of the original echogram. A discrimination of patches using sonar statistics from within the patches gave good separation of slope water patches from patches belonging to four other water masses Slope water patches were characteristically small and of low mean scattering. Large but infrequent targets were present In the Gulf Stream, by contrast, patches contained more uniformly distributed targets with a higher mean scattering The observed correlation between echo patches, biological structures, and oceanographic features suggests that the measurement of echo statistics and our patch recognition techniques produce biologically meaningful parameters.