Publication | Open Access
Deep Velocity Profiling with Self-contained ADCPs
178
Citations
0
References
1993
Year
EngineeringMachine LearningAcoustical OceanographyUnderwater AcousticOceanographyMarine EngineeringMarine Geophysical DataEarth ScienceGeophysicsOcean AcousticsGlobal Positioning SystemSparse Neural NetworkEmbedded Machine LearningCtd-rosette SamplerOcean InstrumentationMarine GeologyMachine VisionProfiling ToolDeep VelocityComputer ScienceDeep LearningAdcp DepthOcean EngineeringAerospace EngineeringPhysical OceanographyOcean Acoustic
Ocean deep velocity profiles were obtained by lowering a self-contained 153.6-kHz acoustic Doppler current profiler (ADCP) attached to a CTD-rosette sampler. The data were sampled during two Meteor cruises in the western tropical Atlantic. The ADCP depth was determined by integration of the vertical velocity measurements, and the maximum depth of the cast was in good agreement with the CTD depth. Vertical shears were calculated for individual ADCP velocity profiles of 140–300-m range to eliminate the unknown horizontal motion of the instrument package. Subsequent raw shear profiles were then averaged with respect to depth to obtain a mean shear profile and its statistics. Typically, the shear standard deviations were about 10−3 s−1 when using up and down traces simultaneously. The shear profiles were then vertically integrated to get relative velocity profiles. Different methods were tested to transform the relative velocities into absolute velocity profiles, and the results were compared with Pegasus dropsonde measurements. The best results were obtained by integrating the raw velocities and relative velocities over the duration of the cast and correcting for the ship drift determined from the Global Positioning System. Below 1000-m depth a reduction of the measurement range was observed, which results either from a lack of scatterers or instrumental problems at higher pressures.