Removing wave bias from velocity measurements for tracer transport: The harmonic analysis approach

Abstract

Estimates of turbulence properties with Acoustic Doppler Current Profiler (ADCP) measurements can be muddled by the influence of wave orbital velocities. Previous methods-Variance Fit, Vertical Adaptive Filtering (VAF), and Cospectra Fit (CF)-have tried to eliminate wave-induced contamination. However, those methods may not perform well in relatively energetic surface gravity wave or internal wave conditions. The Harmonic Analysis (HA) method proposed here uses power spectral density to identify waves and least squares fits to reconstruct the identified wave signals in current velocity measurements. Then, those reconstructed wave signals are eliminated from the original measurements. Datasets from the northeastern Gulf of Mexico and Cape Canaveral, Florida, are used to test this approach and compare it with the VAF method. Reynolds stress estimates from the HA method agree with the VAF method in the lower half of the water column because wave energy decays with depth. The HA method performs better than the VAF method near the surface during pulses of increased surface gravity wave energy.