Survey of Changes in Colored Dissolved Organic Matter (CDOM) Using SVM Algorithm and Landsat 8 Satellite Images in Surface Water Resources

Colored dissolved organic matter (CDOM), one of the main constituents of DOM in surface water, is an important indicator of water quality, the biochemical status, and the nutritious material content and plays an important role in the carbon cycle of surface water. In this study, we have analyzed colored dissolved organic matter (CDOM) in lakes using Landsat 8 images in the period of 2013 to 2016 in the watershed area of North Siberia. The support vector regression model was used for selecting the most desirable band in determining the CDOM absorption coefficient, and using the support vector machine model to classify and compare the changes in the amount of〖 α〗_CDOM (440), the CDOM scatter plots for the years of 2014 and 2015 were obtained. Based on the results, regarding the correlation coefficient (R2=0/71) and the amount of errors (MSE=1/60 m-1, RMSE=1/0775 m-1, and MAE=0/9464 m-1), it was concluded that the application of green/red band ratio in Landsat 8 satellite was the most desirable choice for measuring colored dissolved organic matter at a wavelength of 440 nm in watershed resources of North Russia. The scatter plot indicates an increase in colored dissolved organic matter in the lakes in the northeast of the area in 2015 compared to 2014.