Frequency–time analysis, low-rank reconstruction and denoising of turbulent flows using SPOD

Four different applications of spectral proper orthogonal decomposition (SPOD): low-rank reconstruction, denoising, frequency-time analysis, and prewhitening are demonstrated on large-eddy simulation data a turbulent jet. SPOD-based reconstruction can be performed by direct inversion truncated SPOD. This problem, however, is ambiguous since SPOD relies estimation. We demonstrate flow field using the algorithm (frequency-domain approach) propose an alternative approach based projection time series onto modes (time-domain approach). While optimally represents in statistical sense, time-domain seeks optimal each instantaneous field. further denoising strategy that hard-thresholding eigenvalues. The proposed achieves significant noise reduction while facilitating drastic compression. In contrast to standard methods analysis such as wavelet transform, yields spectrogram characterizes temporal evolution spatially coherent structures. frequency-domain, time-varying expansion coefficients obtained basing sliding window. approach, computationally intractable, convolution presented. When applied jet data, reveals intermittent occurrence large-scale structures directly associated with high-energy events. work suggests preferable for individual snapshots, frequency-domain analysis.