基于希尔伯特-黄变换,发展了合成孔径雷达内波参数提取的新方法,解决了传统的傅里叶变换不能提取非线性内波群各个孤立子波信息的问题。利用该方法对巴士海峡附近的一幅ERS-1 SAR内波图像进行了内波参数提取。希尔伯特-黄变换对内波剖面数据进行了尺度分解,根据归一化方差最大来提取内波分量。该方法具有自适应性。结果表明,该方法获取的内波平均波长与傅里叶变换和小波分析具有较好的一致性,提取得到的反映各个孤立子波信息的波形比小波分析的要清晰,提高了数据的质量。通过提取的内波信号,利用极值序列计算了剖面线前后段平均波长大小,消除了谱分析当中的180°模糊问题。计算了明暗点的间隔,反演得到各个孤立子波的半振幅宽度。利用CTD数据获得的跃层深度,反演前导波振幅,与验证算例结果在量级上一致。

A newmethod to extract internalwave parameters from SAR mi agery has been developed based on\ntheHilbert-HuangTransform(HHT). Internalwave parametershave been retrieved from theERS-1 SAR mi age\nin the Luzou strait. HHT is used to decompose the internalwave profile data. The internalwave component is\nextracted according to themaxmi um normalized deflection. The results show that the newmethod can not only\nextract thewavelength of internalwaves, but also extractwave shapesmodulated by the internalwaves. The\nmeanwavelengths of the internalwaves are calculated in order to solve the 180°ambiguity of the internalwave\ndirection. The halfwidth of each soliton is calculated. The internal wave amplitudes retrieved by this new\nmethod are about104m which is in agreementwith Liu’s report.