Estimation of Internal Wave from Sentinel-1A Images by the Combination of 2-D and 1-D Empirical Mode Decomposition

Inhalt:
The ocean internal waves oscillating along the stratification surfaces within the ocean are important meso-scale phenomenon in the ocean. In some cases, they may bring out great damages to shipping and offshore platforms. SAR has been one of the main tools to observe the internal waves. During this process, 1-D EMD or EEMD method is often adopted to decrease noises and extract the internal wave signal. But the estimated parameters of the internal wave are very sensitive to the location of a profile in the original SAR image. In this paper the combination of 2-D and 1-D EMD method is proposed. At first a 2-D EMD is applied to decompose a SAR internal wave image into a number of 2-D IMFs, 1-D EMD is then used on the characteristic IMF which has almost the same streaks as the original image, finally the parameter estimation is carried out. Experiments show the effectiveness of this proposed method.