基于多极化后向散射系数的海洋悬浮物反演研究

Inverse Model of the Ocean Suspended Solids Based on Multi-polar Backscattering Coefficient

以珠江口东岸香港海域为研究对象,准同步获取实测悬浮物浓度和Radarsat-2影像数据。对影像进行滤波处理和掩模处理后,利用Radarsat-2四种极化下的后向散射系数建立悬浮物浓度单极化回归模型;通过多极化后向散射系数构造多个遥感参数,运用相关性分析得到4个敏感因子,建立悬浮物浓度多极化回归模型。最终得到研究区域悬浮物浓度的反演模型为:SSC=11.08+0.06(HH+VV)-0.002(HH+VV)2,R2=0.84,其中SSC为悬浮物浓度,HH和VV为该极化下的后向散射系数,R2为决定系数。研究表明:HH和VV极化的后向散射系数之和对研究区域悬浮物反演最为敏感,得到的反演模型能较好预测海洋悬浮物的分布情况。

Taking East coast of the Pearl River Estuary-Hong Kong Sea as research object,the suspended solids concentration（SSC） and Radarsat-2 image data are quasi-synchronized acquired to be filtered and masked.Then,the SSC single-polar regression model is established using Radarsat-2 four polarized backscattering coefficient.Multiple remote sensing parameters are constructed using multi-polarization backscatter coefficient.Four sensitive factors are obtained using the correlation analysis.Finally,the multi-polar regression model of SSC is established.The regression model is： SSC=11.08＋0.06（HH＋VV）-0.002（HH＋VV）2,R2=0.84,which HH and VV is backscatter coefficient of HH and VV polarization,R2 is determination coefficient.The results show that： The sum of HH and VV polarized backscattering coefficient is most sensitive to the study area SSC and the inverse model can predict the distribution of SSC better.