基于植被指数的Sentinel-1失相干评估

Sentinel-1 decorrelation assessment based on vegetation index

Sentinel-1数据自全球免费开放以来已被广泛应用于合成孔径雷达干涉测量及相关应用领域,但是其短波长的特性使其容易在部分低相干区域,特别是在植被覆盖区产生失相干现象,因而目前主流的干涉后失相干评估策略可能会造成Sentinel-1数据的浪费和研究成本的提升。针对此情况,本文在顾及极化方式的前提下,分别建立了光学遥感植被指数NDVI和VV/VH极化下Sentinel-1相干系数的定量模型,并在邻近的验证区对模型精度进行了验证,证明了模型的可靠性。基于建立的模型,可以在Sentinel-1数据干涉前对失相干进行定量评估,以克服上述干涉后评估策略的缺点,提高干涉测量的效率并降低研究成本。

Sentinel-1 data has been widely used in synthetic aperture radar interferometry technology and related application fields since it is opened for free in the world,but its short wavelength characteristics make it easy to generate decorrelation in some low coherence areas,especially in vegetation coverage areas.However,the current mainstream post-interference incoherence evaluation strategy may cause data waste and increase research costs,taking into account the polarization mode,this paper establishes the quantitative models of Sentinel-1 coherence and optical remote sensing vegetation index NDVI at VV and VH polarization,and verifies the accuracy of the models in the adjacent validation area,indicating the reliability of the model.Based on the established models,the decorrelation can be quantitatively evaluated before the interference of Sentinel-1 data,in order to overcome the shortcomings of the above post-interference evaluation strategy,improve the efficiency of interferometry and reduce the research costs.