Estimation of thermocline depths from SAR imagery and a two layer density model

Abstract

Imagery from the ERS-1 synthetic aperture radar (SAR) collected on July 18, 1992 was used to estimate the depth of the seasonal thermocline in the New York Bight. It was assumed that the internal wave packets observed in the SAR imagery were tidally generated at the shelf break and that the temporal spacing between the packets was 12.42 hours i.e. one M/sub 2/ tidal cycle. The wave speed of a packet was estimated by measuring the distance from the leading crest in one packet to the leading crest in the previous one and dividing by the tidal period. For each packet pair, nine separate measurements, evenly spaced along the wave crest, were used to obtain a mean separation between the packets. The mean of the observed internal wave speeds far the southern packet was 0.79 m/s with a deviation about the mean of 11%. The depth of the thermocline was computed from the wave speed derived from a simple two-layer fluid model for these internal wave dynamics. The densities were obtained from a joint Russian-United States internal wave experiment that was concurrent with the SAR overpass. The computed mean depth of the seasonal thermocline is 22 m. The average depth of the seasonal thermocline observed during the summer season from historical data was on the order of 20 m, which agrees with the estimated thermocline depth from the SAR imagery.