Inversion of heavy metals in water by remote sensing is a difficult problem in remote sensing of the water environment. There are still quite a few fundamental problems to be solved. The contribution law of the bottom sediment in the shallow water area to the water-leaving reflectance is one of the important factors affecting the accuracy of the remote sensing inversion model in the future. Especially in the special background of heavy metal pollution, revealing its contribution rule plays an important role in improving the accuracy of the water heavy metal remote sensing model. Meanwhile, the measurement results have reference significance for studying the reflectance spectral characteristics of heavy metal tailings sediments and distinguishing common bottom sediments. Firstly, the reflectance of the Dabaoshan Mountain tailings bottom in Guangdong was obtained by spectrometer measurement. It has reflection peaks at wavelengths of 755, 1 280, 1 620 and 2 200 nm, and has obvious spectral characteristics. The reflectance of three types of sediments coarse sand, silt and stones, which are common in riverbeds, were compared and analyzed. The results show that, on the one hand, the reflectance of coarse sand and silt shows a slowly rising curve, which is different from the sediment in the mining area with multiple characteristic reflection peaks. On the other hand, the reflectance of the stone shows a broad but flat reflection peak in the wavelength range of 550~650 nm, then a trough at the wavelength of 675 nm, and then increases to a wavelength of 750 nm and then tends to be flat. Its characteristic wavelength is different from that of the sediment in the mining area. Therefore, the above characteristic wavelengths can be used as important distinguishing bands of sediment spectra in the background of heavy metal pollution. Then, measure the water-leaving reflectance water depths of 1, 10 cm and deep water. The water-scattered light and water-bottom reflected light were calculated at a water depth of 1 cm, and the contribution of the water-leaving reflectance was discussed. The measurement results of the water-leaving reflectance show that the sediment greatly influences the shallow water area. The contribution law of the water-scattered light and water-bottom reflected light to the water-leaving reflectance takes the wavelength of 515 nm as the dividing line: in the short-wave direction, the water-scattered light is the main contribution, while in the long-wave direction, change to the water bottom reflected light. The contribution is determined by the sediments reflection ability and the water bodys scattering ability. Finally, the accuracy of the remote sensing model after considering the water-bottom reflected light is evaluated. The comparison results of the water-leaving reflectance in the wavelength range of 350~950 nm calculated by the model and the measured value in the field show a significant linear correlation (R2=0.964 2). The relative error is less than 10% in the wavelength range of 560~830 nm and even less than 5% in some of them. The simulation accuracy of the model is good, which is far higher than that when the influence of the sediment is not considered. The model satisfies the requirements of remote sensing inversion of heavy metals in water in the future. The above research results provide important reference data and theoretical basis for processing sediment effects in remote sensing of water heavy metals inversion in the future, which will help promote further development in this field.