Ensemble Empirical Mode Decomposition on Storm Surge Separation from Sea Level Data

This paper concerns the storm surge calculation based on the algorithm of ensemble empirical mode decomposition (EEMD). An accurate storm surge result is key information for coastal disaster warning and prevention. Separation of storm surge magnitude from sea level data has typically been done by specifying tidal input from main tidal harmonics. Obtaining accurate storm surge magnitude with harmonic analysis (HA) requires at least one month. This study discusses possible storm surge separation from short-term sea level time series using EEMD. The current work reveals that EEMD is predominant for short-term sea level data analysis shorter than thirty-five days. Due to different residues obtained from EEMD, this work proposes a method to determine most ideal residue for representing the storm surge.