Remote sensing estimation method for colored dissolved organic matters in lake under soil erosion condition

The invention discloses a lake colored dissolved organic matter remote sensing estimation method under a soil erosion condition, which comprises the following steps: collecting a sample of a sampling point, and obtaining a colored dissolved organic matter CDOM absorption coefficient of the sample; acquiring the satellite remote sensing reflectivity of the sampling point according to the sampling time and the sampling place; constructing a data set based on the CDOM absorption coefficient and the satellite remote sensing reflectivity; constructing a CDOM machine learning estimation model by using XGBoost and the data set, and obtaining an estimated CDOM absorption coefficient through the CDOM machine learning estimation model; and carrying out inversion on the estimated CDOM absorption coefficient to obtain the content of the colored dissolved organic matters in the lake under the soil erosion condition. According to the method, the CDOM absorption coefficient can be accurately estimated, and the method has ex