基于EEMD的SAR海洋内波参数反演

Extracting Internal Wave Parameters From SAR Image Based on Ensemble Empirical Mode Decomposition

经验模态分解(EMD)方法对非平稳信号进行分解,容易出现模式混叠和边界效应,从而不能得到有物理意义的特征信息.集成经验模态分解(EEMD)能够有效地克服模式混叠和边界效应问题,可准确地提取信号的本质特征信息.在分析SAR图像反演海洋内波参数机理的基础上,本文提出了一种基于EEMD的海洋内波参数反演方法.实验结果表明:与小波分解和EMD方法相比,该方法能更有效地克服模式混叠现象,所提取的分量更接近内波波动的物理本质,所反演的内波参数与经验数据吻合.

Empirical mode decomposition （EMD） sometimes cannot reveal the non-stationary signal characteristics accurately because of the problem of mode mixing and edge effect. Ensemble empirical mode decomposition （EEMD） was developed recently to alleviate this problem. Utilizing the advantage of EEMD, this paper proposes a new EEMD-based extracting method for internal wave parameters from SAR image. The application results indicate that, the EEMD is superior to the wavelet method and the EMD method; EEMI） can eliminate modemixing, the extracted internal wave component is close to the true physical nature, and the internal wave amplitude retrieved by the method is in agreement with the empirical data.