Denoising Ocean Turbulence Microstructure Signals for Application in Estimating Turbulence Kinetic Energy Dissipation Rates Based on EMD-PCA

When ocean turbulence signals are collected using turbulence observation instruments in real marine environments, the effective signals in the acquired data set are often polluted by noise. In order to eliminate the noise component contained in the non-stationary and nonlinear ocean turbulence signals, a new multi-scale turbulence signal denoising method is proposed by combining the empirical mode decomposition (EMD) and principle component analysis (PCA). First, the time series of turbulence signals are decomposed into a couple of components by EMD algorithm and approximately calculate the noise energy in each intrinsic mode function (IMF). Then, PCA is implemented on each IMF. The appropriate principal components are selected according to the decomposition characteristics of PCA and the noise energy proportion in IMF. Each IMF is reconstructed by the selected principle components. At last, the effective ocean turbulence signals are reconstructed by the corrected IMFs and the residue. Ocean turbulence signals collected in the South China Sea (SCS) are used to evaluate the effectiveness of the proposed method. The results show that the proposed method can effectively eliminate the noise and maintain the characteristics of the effective turbulence signals under high noise. Turbulence kinetic energy (TKE) is also estimated from the denoised signals, which provide a reliable data basis for the analysis of the turbulent characteristics in later stage.