Evaluation of the Effectiveness of Multiple Machine Learning Methods in Remote Sensing Quantitative Retrieval of Suspended Matter Concentrations: A Case Study of Nansi Lake in North China

Total suspended matter (TSM) is a core parameter in the quantitative retrieval of ocean color remote sensing and an important indicator for evaluating the quality of the aquatic environment. This study selects part of Nansi Lake in North China as the study area. Researchers used Hyperion remote sensing data and field-measured TSM concentration as data sources. Firstly, the characteristic variables with high correlation were selected based on spectral analysis. Then, seven methods such as linear regression, BP neural network (BP), KNN, random forest (RF), and random forest based on genetic algorithm optimization (GA_RF) are used to construct the inversion model of TSM concentration. The retrieval accuracy of each model shows that the machine learning models are much more accurate than the linear model. Among them, the GA_RF model retrieves the suspended solids concentration with the best performance and the highest prediction accuracy, with a determination coefficient R2 of 0.98, a root mean square error (RMSE) of 1.715 mg/L, and an average relative error (ARE) of 6.83%. Additionally, the spatial distribution of TSM concentration was inversed by Hyperion remote sensing image. The results showed that the concentration of TSM was lower in the northwest and higher in the southeast, and the concentration distribution was uneven, showing the characteristics of a typical shallow macrophytic lake. This study provides an effective method for monitoring TSM concentration and other water quality parameters in the shallow macrophytic lake and further proves the advantages of machine learning in ocean color inversion. All in all, this research provides some useful methods and suggestions for quantitative inversion of TSM concentration in shallow macrophytic lakes.

Characterizing variations of dissolved organic matter (DOM) properties in Nansi Lake: an important reservoir along the eastern route of China’s South-to-North Water Diversion Project

Variations in dissolved organic matter (DOM) quality has far-reaching implications that affect, e.g., aquatic productivity, food web structures, trace element and pollutant transport. In this study, a total of 186 water samples were collected at 62 sites (three points in time within one year) in Nansi Lake. UV-Vis spectra, Synchronous fluorescence (SF) spectra, the excitation-emission matrix and parallel factor analysis (EEM-PARAFAC) were applied to indicate the source and quality of DOM. Water transferring of the eastern route of China’s South-to-North Water Diversion Project had a great influence on the water level of Nansi Lake. Results of SF spectra, EEM-PARAFAC and principal component analysis (PCA) suggested that protein-like substances played a more important role in DOM properties in April and July than October. This result is related to a high fluorescence intensity occurred in April (Fmax=0.72±0.03 in the upper lake and 1.84±0.13 in the lower lake) and July (Fmax=1.10±0.05 in the upper lake and 1.49±0.04 in the lower lake), which might be caused by water transferring from other lakes to Nansi Lake, death and decomposition of submerged plants. At the same time, relatively good correlations were found between humic-like substances, DOC and a254 in April, July and October, which indicated the important contribution of humic-like substances to Nansi Lake. With the completion of the water diversion, the ratio of the fluorescence intensity of component to the total fluorescence intensity (%Fmax) suggested that the proportion of humic-like substances started to increase. And when it came to October, humic-like substances become the main substance in DOM collected from Nansi Lake (%Fmax=66.56%±0.58% in the upper lake and 61.98%±0.99% in the lower lake). Moreover, among the two areas in Nansi Lake, the upper lake always had a higher degree of humification (HIX=2.23±0.06, 2.38±0.11 and 3.10±0.05 in April, July and October, respectively) than the lower lake (HIX=1.06±0.05, 1.68±0.05 and 2.62±0.08 in April, July and October, respectively), which implied extraneous contaminants might have a more important impact on DOM properties in the upper lake.