基于光谱混合模型的河口表层悬浮泥沙浓度卫星遥感反演研究

选取长江口及周边海区为研究区域，利用长江口取样泥沙的不同配比浓度的实验室光谱测量数据，以2002年6月1日的MODIS Terra卫星遥感数据为研究案例，考虑高、低悬浮泥沙浓度和叶绿素a 3个端元组分，采用线性光谱混合模型反演长江口及周边海区表层悬浮泥沙浓度。反演结果很好地反映了长江口及周边海区的悬浮泥沙空间分布精细结构，空间分布特征和数值量级比较分析表明结果真实可信，初步说明该算法在我国近岸河口海区的应用效果较好。与传统的经验算法相比，基于线性光谱混合模型的悬浮泥沙反演算法更少地依赖现场实测数据，可以更加充分地利用多光谱波段信息，对于近岸河口高浑浊水体的悬浮泥沙浓度反演是一种可行的方法。 

Estimating Surface Suspended Sediment Concentration in the Estuary Based on the Spectral Mixture Model with Remotely Sensed Data

A technique is presented in this paper for estimating the Suspended Sediment Concentration (SSC) in the high turbid water of the Yangtze River estuary and its adjacent sea area based on the Linear Spectral Mixture Model (LSMM) with the MODIS imagery.The spectral data of different SSC measured in the lab is used to get the SSC~f highnonlinear calibration formula which is then adopted to adjust the SSC retrieved from the atmospheric-corrected MODIS imagery.The highest,lowest SSC end-members from the experimental spectral data and the Chl-a end-member selected from the atmospheric-corrected MODIS imagery are chosen in the LSMM.It is shown that the result is believable and convincible by analyzing the SSC spatial distribution and validation using several historical statistical SSC data in the study sea area.Compared to the empirical method,the presented method needs little the real-time in-suit data and may produce more accurate result in high turbid estuarine water since it incorporates analyses of multiple wavelength bands and a nonlinear calibration.
