累积性发射光谱结合多元统计研究蘑菇湖水体DOM的特征

三维荧光光谱技术（3DEEM）因其方便迅速、灵敏度高等一系列优点，广泛应用于表征DOM。在实际应用中，3DEEM步骤较为繁琐，PARAFAC等方法相较于寻峰法虽然更加直观可靠，但是往往需要借助MATLAB等数学软件，因此该研究希望借助一种新的方法能够更加简便迅速地表征DOM。以蘑菇湖水体DOM为例，基于累积性发射光谱（AFEs），结合多元统计及二阶导数等方法，对DOM各类荧光组分及含量进行表征。利用主成分分析（PCA）对AFEs进行因子载荷分析, 并确定荧光峰的类型及其含量的差异; 通过二阶导数转换，得到二阶导数AFEs；通过对所有采样点二阶导数AFEs各荧光峰进行绝对面积积分，分析DOM中各组分的含量及变化；通过聚类分析, 分析不同点位组分的差异性或相似性。研究表明，通过AFEs得到5类荧光峰，分别为类蛋白峰、类富里酸峰、陆源类腐殖酸峰以及腐殖酸峰。基于AFEs及其对荧光强度之和的分析，可以看出蘑菇湖水体中的DOM主要以不稳定、易被降解的、相对分子质量较小的类蛋白及富里酸为主，腐殖化程度由滨湖区向深湖区递减。基于AFEs得分图，得到5类荧光峰，且荧光峰中类蛋白及类富里酸峰占主导；基于点位得分矩阵，可以说明各点位之间的荧光组分存在差异。二阶导数AFEs被分为5个荧光波段, DOM以相对质量较小的有机质为主，腐殖化程度、芳香度较小，空间差异上不显著。通过对荧光峰面积以及采样点进行聚类分析，荧光峰被分为3类，其中富里酸含量占比较大，且岸边采样点和位于湖心区域的采样点之间存在差异。综上研究显示，AFEs相对简单迅速，能够代替3DEEM对DOM进行表征。蘑菇湖水体DOM是以相对分子质量较小、不稳定、易被降解的类蛋白、类富里酸物质为主，总体上腐殖化程度及相对分子质量具有由滨湖区向深湖区递减的趋势，但空间上差异性较小。

Accumulative Fluorescence Emission Spectra Combing Multivariate Statistics to Study the Characteristics of DOM in Moguhu Lake

This study takes the DOM of the Moguhu Lake water as an example, based on accumulative emission spectra (AFEs), combined with multivariate statistics and second derivative, to characterize the various fluorescent components and content of DOM. Principal Component Analysis (PCA) was used to analyze the factor loading of AFEs and to determine the difference between the types of fluorescent peaks and their contents. The second derivative AFEs are obtained by AFEs through second derivative conversion. Analysis of the content and variation of each component in DOM by absolute area integration of the fluorescence peaks of the second derivative AFEs of all sampling points. Cluster analysis was used to analyze the difference or similarity of DOM components at different points. Studies have shown that five types of fluorescent peaks are obtained by AFEs, namely protein-like peaks, fulvic acid-like peaks, marine or terrestrial humic acid peaks, and humic acid peaks. Based on the analysis of the AFEs and sum of fluorescence intensity, it can be seen that the DOM in the water of the Moguhu Lake is mainly unstable, easily degraded, and the protein with less molecular weight and fulvic acid. The degree of humification of its DOM is reduced from the lakeside area to the deep lake area. According to the AFEs score map, five types of fluorescent peaks are obtained, and the protein-like and fulvic acid-like peaks in the fluorescent peak are the main ones. Based on the point-score matrix, it can be explained that there is a difference in the fluorescence components between the points. The second derivative AFEs are divided into five fluorescent bands. The DOM is mainly composed of organic matter with small molecular mass, and the degree of humification and aromaticity are small, and the spatial difference is not significant. By clustering the fluorescence peak area and sampling points, the fluorescence peaks are divided into three categories, of which fulvic acid content is relatively large, and there is a difference between the shore sampling point and the sampling point located in the lake center area. In summary, AFEs are relatively simple and rapid, and can be used to characterize DOM instead of 3DEEM. The DOM in the water of the Moguhu Lake is mainly composed of protein-like and fulvic acid-like substances with relatively low molecular weight, instability and easy degradation. In general, the degree of humification and relative molecular mass have a tendency to decrease from the lakeside area to the deep lake area, but the spatial difference is small.