An improved adjoint-based ocean wave reconstruction and prediction method

We propose an improved adjoint-based method for the reconstruction and prediction of nonlinear wave field from coarse-resolution measurement data. adopt data assimilation framework using adjoint equation to search optimal initial match simulation result at later times with given Compared conventional approach where optimised surface elevation velocity potential are independent each other, our features additional constraint dynamically connect these two control variables based on dispersion relation waves. The performance new is assessed generated phase-resolved simulations high-order spectral method. consider a variety steepness noise levels irregular It found that tends overestimate in high-frequency region underestimate potential. In comparison, shows significantly instantaneous elevation, statistics, including skewness kurtosis.