Spectral Characteristics Analysis Of Dissolved Organic Matter In Aquatic Systems Of The Middle Reaches Of Huai River

Dissolved organic matter（DOM）,as the most active component in the aquatic environment,has an extremely important influence and indication on the aquatic ecosystem.An in-depth understanding of its content,composition,change,and environmental impact is of great significance for the field of ecological environment.Spectral technology is the most convenient,rapid,and economic tool to track the source,composition,and environmental behavior of DOM.In this study,the water body in the middle reaches of Huai River was taken as the research object（309 water samples collected）,and UV-Vis absorption spectrum（UV-Vis）and excitation and emission matrix（EEMs）were used as the technical means for the spectral characterization of DOM,supplemented by water quality indicators to explore the content,composition,migration and transformation characteristics of DOM in the basin water.Through the combination of multivariate statistics and machine learning data analysis methods,the potential correlation between the source,content,composition,and transformation characteristics of DOM and the water quality indexes of water body was further explored,which provided a theoretical basis for water quality improvement and ecological restoration in the middle reaches of Huai River.The main conclusions are:（1）The analysis results of UV-Vis parameters and fluorescence parameters showed that the relative molecular weight,hydrophobicity,aromaticity,and humification degree of DOM in lake water in the middle reaches of Huai River were significantly lower than those in the ditch and river water.The humic-like substances in DOM of lakes and rivers were mainly fulvic-like acids,while the humic-like acids were mainly humic-like acids in ditches.The hydrophobicity,relative molecular mass,humification degree,and aromaticity of DOM in PJ and SX regions were higher than those in other regions,and there were more exogenous inputs.The humic-like components of DOM in HY,PJ,YS,TJA,and DT areas were mainly fulvic-like acids.（2）The results of three-dimensional fluorescence regional integration（FRI）analysis showed that the DOM content of lake water in the middle reaches of Huai River was significantly higher than that of ditches and rivers,and the DOM content of water in the PJ region was significantly higher than that of other regions.DOM in lakes and ditches contains more components generated by microbial activities.The proportion of humic-like acid components in the DOM of water in the PJ region（21%±6%）was significantly higher than that in other regions.The DOM in the middle reaches of Huai River were mainly tryptophan-like（region II）and fulvic-like（region III）substances,while tyrosine-like（region I）,humic-like（region V）,and soluble microbial products（region IV）accounted for a relatively small proportion.（3）Four significant chemical components（UVA-like fulvic acid C1,tryptophan-like C2,humic-like acid C3,and UVC-like fulvic acid C4）were isolated from DOM in the middle reaches of Huai River by parallel factor analysis（PARAFAC）.DOM content in lake water（1.26±0.62 R.U.）was the highest;DOM in all types of water was dominated by the C2 component（accounting for more than 30%）.In the middle reaches of Huai River,the terrestrial input of DOM in the dry season and water season is less,and the water body is mainly composed of autogenetic substances.Humus-like substances produced by microbial activities in the DOM of the water body in the SX region account for a relatively high proportion（29%±5%）.The humic component of water DOM in the PJ region（29%±17%）was higher than that in other regions,and the photochemical oxidation process of water DOM in the PJ region was inhibited by surface photovoltaic panels,and the DOM content accumulated to a high level（2.01±0.82R.U.）.（4）The results of t-distributed stochastic neighbor embedding（t-SNE）show that spectral parameters can better distinguish the differences between different types of water bodies than water quality indicators.The results of redundancy analysis（RDA）showed that the chemical oxygen demand（COD）in the water body of the middle reaches of Huai River in the dry season and water period mainly came from the self-generating process of the water body.DOM in lake water is mainly composed of autogenic components,with small relative molecular mass,low humification degree,low hydrophobicity,and low carbon skeleton polymerization degree.The water body in the TJA area is mainly urban water,and the level of self-origin is strong.The DOM humification level in the PJ area is high,and the molecular weight is large.The results of the gradient boosting machine（GBM）model show that PARAFAC and FRI are better than UV-Vis parameters and fluorescence parameters in analyzing DOM composition and distinguishing various water characteristics.The results of the random forest regression（RF）model showed that NH₄⁺was the most important predictor of DOM composition in water,and the change of NH₄⁺content in the water of the Huai River during the dry season and the water period was significantly driven by the change of DOM composition.The C4 component is UVC fulvic acid,which is not sensitive to photochemical oxidation and biodegradation and is not easily driven by environmental factors.（5）The self-organizing map（SOM）model was similar to the results obtained by the traditional spectral analysis method.The DOM in the Yaogao Reservoir was mainly characterized by protein-like components,while the DOM in the Gaotang Lake was characterized by humus-like components.The principal component analysis（PCA）showed that DOM in the Gaotang Lake had a higher degree of humification,aromaticity,the proportion of humic-like components,and hydrophobicity than that in the Yaogao Reservoir.SOM model is based on the original data input of water EEMs,which can realize the preliminary analysis of water DOM composition.After verification,the analysis results are similar to those of other spectral analysis results and can be used as a powerful tool for rapid analysis,classification,and qualitative analysis of EEMs.Fig[50]Table[7]Reference[194]...