Dynamic mode decomposition for estimating vortices and lee waves in a stratified wake

Dynamic mode decomposition (DMD) is an analysis technique for extracting flow patterns and their dynamics from experimental or simulated velocity fields. Here, DMD is applied to experimental data in the vertical center-plane of wakes generated by a towed grid in a stably stratified background, at varying values of the dimensionless Froude and Reynolds Number. The primary goal was to identify dynamically important patterns and reveal the influence of stratification on their initiation and evolution. It is demonstrated that DMD captures lee wave and vortical modes with different length scales successfully. Further, one can construct a mode energy spectrum which shows a clear dependence on Froude Number, with energy transfer to larger scales in the near wake, as the initial shear-triggered Kelvin–Helmholtz roll-ups diffuse and pair with neighbors. Finally, this paper serves as a detailed example of the application of DMD to time-resolved particle imaging velocimetry data for a stratified flow. The results confirm its utility in objective identification of dynamics at different scales of complex fluid flows.