ConvLSTM-Based Wave Forecasts in the South and East China Seas

Numerical wave models have been developed for the forecast in last two decades; however, it faces challenges terms of requirement large computing resources and improvement accuracy. Based on a convolutional long short-term memory (ConvLSTM) algorithm, this paper establishes two-dimensional (2D) significant height (SWH) prediction model South East China Seas trained by WaveWatch III (WW3) reanalysis data. We conduct 24-h predictions under normal extreme conditions, respectively. Under condition, 6-, 12-, forecasting, their correlation coefficients are 0.98, 0.93, 0.83, mean absolute percentage errors 15, 29, 61%. condition (typhoon), 6 12 h, 0.98 0.94, 19 40%, which is better than all It concluded that ConvLSTM can be applied to 2D with high accuracy efficiency.