平行因子分析法在太湖水体三维荧光峰比值分析中的应用

以太湖水样三维荧光光谱数据为例,提出在采用平行因子分析法(PARAFAC)处理后的荧光数据中提取荧光峰强度计算荧光峰比值进行水环境分析评价的方法,较直接在水样原始荧光谱图中获取的荧光峰强度更加准确客观。天然水体中各水样间由于受荧光团复杂多样性等因素的影响,某类荧光物质荧光峰的激发发射波长位置并不是固定不变的,就同一水样而言各类荧光峰之间的相互重叠干扰也将影响到荧光峰强度和位置的准确判断。而在PARAFAC模型各因子中提取相应荧光峰值可以保证各水样间同类荧光物质荧光峰在同一位置又有效减弱同一水样中各类荧光物质荧光峰之间的相互干扰,更加高效准确的利用荧光峰比值进行水环境分析。区域差异性分析时水样因子得分比值的区域变化与原始荧光峰比值变化趋势是一致的。

Parallel factor analysis as an analysis technique for the ratio of three-dimensional fluorescence peak in Taihu Lake

The present paper proposes a new method to find the ratio of three-dimensional fluorescence peak. At first, the excitation-emission fluorescence matrix of water samples was treated with parallel factor analysis (PARAFAC) and then fluorescence peaks intensity and ratio of fluorescence peak were obtained from the parallel factor analysis model. From the parallel factor analysis model, the same fluorescence peaks of different water samples lie at the same excitation-emission wavelength and the overlap of different fluorescence peaks of the same water sample is reduced. Analysing regional characteristic in Taihu Lake, the ratio of factor score and the ratio of fluorescence peak showed strong correlation.