抄録
We present a technique which allows determination of spatial modes oscillating at a single frequency. It is Koopman Mode Decomposition (KMD) which is based on spectral analysis of the Koopman operator, an infinite-dimensional linear operator. We performed two algorithms, Fourier-based and Arnoldi-based KMD, on 30-day beam velocities measured by an Horizontal Acoustic Doppler Current Profiler (HADCP) on the Hudson River Estuary. The flow is shown to be dominated by semidiurnal tide, supplemented by sub-harmonic (diurnal tide) and harmonics. Harmonic-average profiles (Fourier-based Koopman modes) correspond to a sinusoidal variation at a tidal frequency, whence a precise temporal rate of change. They also quantify the phase lags at different distances from the river bank. Results show that both algorithms provide quite similar phase lags at these dominant tidal modes.
