A Digital Filter Technique for Estimating Small-Scale Structure of Salinity,Temperature,and Sound Speed.

摘要

A series of vertical profiles of temperature, salinity, and sound speed was made from a surface vessel using an STD/SV environmental profiling system in water depths of 350 m off Southern California. The data were recorded digitally on magnetic tape at a 0.2-sec sample rate. The spatial resolution of the sensors was degraded by noise introduced into the data by vertical speed variations of the instrumentation package which resulted from wave-induced motions of the ship, variations in suspension cable winch speed, and velocity of underwater currents. The problem discussed is the removal of the noise from the data without removal of the small-scale structure. To accomplish this, each variable, e.g., salinity, temperature, and sound speed, is treated as a time series and the techniques of spectral analysis are employed to establish a cutoff frequency, c. Using this as a criterion, a digital filter was designed to remove the noise from the data. Criteria are presented for the design and use of digital filters to remove the noise and improve resolution of STD/SV digital cases.