固体介质中的重金属源解析及其在中国近海沉积物中的应用

沉积物中重金属源解析对甄别人为活动与自然变化对近海生态系统演化的影响有重要作用.本文总结了近年来污染物源解析常用的多元统计分析、地球化学方法和地质统计分析3种主要研究方法,剖析了不同方法的优劣及适用性,提出正定矩阵因子分析、Pb同位素示踪在重金属来源定量化研究中具有良好应用前景.梳理了中国近海沉积物中重金属来源的主要研究结果,发现河口和海湾是沉积物重金属受人为来源影响剧烈的典型近海区域,不同定量解析方法(多元统计分析、背景值估算、Pb同位素分析)均表明中国近海沉积物重金属的人为来源贡献率接近或超过50%.当前中国近海沉积物中重金属源解析研究还存在源识别端元模糊、解析结果缺乏相应的可靠性评价等问题.据此提出近海沉积物重金属源解析研究应使用多种解析技术手段综合、集成与优化,提高源解析的准确性;建立完善指标体系,筛选代表特定人为活动和自然过程的指标;甄别人为源重金属入海方式及过程,为沉积物数据信息的解译提供理论基础.

Source apportionment of heavy metals in solid and its application in coastal sediment of China.

Source apportionment of heavy metals in sediments is significant to distinguish the influence of human activities from nature changes on offshore eco-environment evolution. This overview focused on three widely used source apportionment methods: Multivariate statistical technique, geochemical method and geostatistical analysis and highlighted that the positive matrix factorization and lead isotopic trace were two relatively preferable methods to quantitatively assess the anthropogenic source of heavy metals. Based on the heavy metal concentrations in coastal sediments of China, heavy metals were primarily of anthropogenic origin with their contribution close to or higher than 50% in coastal sediment of China evaluated by combination of different source identification me-thods (multivariate analysis, estimation based on background value and lead isotopic trace), and estuaries and bays were typical polluted regions. The current researches on source apportionment had the problems that the source endmembers were still vague and the interpretation of data lacks corresponding reliability evaluation. The future research should be focused on the integration and optimization of multi-methods to improve the accuracy of source apportionment, extracting informative indicators to represent specific anthropogenic activities or natural source and recognizing the geoche-mical process and mechanism of anthropogenic emissions of heavy metals when entering into coastal environment to provide referential fingerprints for source interpretation.