No abstract

In this study, the predictability of ocean wave fields is considered based on marine radar measurement data. Phase-resolved components obtained by applying 3D FFT-based reconstruction to a sequence of radar images are utilized for wave field prediction, and two different prediction approaches are introduced: (i) snapshot data-based prediction through the adjustment of the frequency and phase of each component, and (ii) spatiotemporal data-based prediction through the data assimilation for reconstructed wave fields. Furthermore, the time evolution of a predictable zone is derived for different shapes of measurement domains including rectangular and ring-shaped domains. To validate the proposed wave propagation modeling method, numerical simulations are conducted on synthetic radar images created by reflecting geometrical shadowing effects, and the prediction accuracy is examined in relation to the derived predictable zone. Lastly, the forecasting performance, which is represented by the predictable time range at a radar location, is discussed with respect to the prediction techniques, specifications of the reconstruction domain, and moving measurements.