Spectra Adaptive Correction Based on the Empirical Mode Decomposition and Thresholding Algorithms: Application to the Ocean Turbulence Time Series.

Measurements and observations of turbulence in the open sea have contributed significantly to improving our understanding of ocean mixing processes. However, the measured microstructure shear time series are strongly nonstationary and easily contaminated by various types of noise. Extended observation data sets with higher accuracy are needed to resolve significant issues in ocean circulation systems. In this article, an adaptive and multiscale correction algorithm is developed to eliminate vortex-induced vibration contamination from the calculated turbulent kinetic energy. First, an adaptive empirical mode decomposition method is implemented to decompose three-axis vibration accelerations into a series of oscillatory intrinsic mode functions (IMFs) and a residual mode. Then, a statistical noise model is developed from the energy density distribution function of the extracted IMF components. The key modes that are noise dominated are minimized, and the remaining shear IMFs are used to reconstruct the true signals. Finally, a technique is developed to automatically set a threshold for severe vortexinduced turbulent conditions. The developed method is verified using a turbulence data set recorded in the South China Sea, and the results show that this adaptive noise correction algorithm is fairly effective in removing vortex-induced vibration noise in comparison with the classical cross-spectrum method. [ABSTRACT FROM AUTHOR]