黄河吴忠段水体叶绿素a与悬浮物质量浓度的光谱反演模型OA

In response to the complex optical characteristics of water bodies in the Wuzhong section of the Yellow River and the poor adaptability of existing models,30 sets of water samples and synchronous spectral data were collected.The original spectra were subjected to natural logarithm transformation and first-order differential transformation to enhance spectral features.The optimal sensitive band combinations were then selected using the normalized difference spectral index(NDSI)and band ratio indices to construct a multiple linear regression inversion model for chlorophyll-a(Chl-a)and total suspended solids(TSS)concentrations.The results demonstrate that the inversion model combining first-order differential transformation and NDSI yields the best performance,achieving R2 values of 0.98 and 0.92 for Chl-a and TSS,respectively,significantly outperforming traditional methods.This approach effectively improves the utilization of hyperspectral information and provides a reference for high-precision,localized remote sensing monitoring of water quality in the Wuzhong section of the Yellow River.