Combination of self-organizing map and parallel factor analysis to characterize the evolution of fluorescent dissolved organic matter in a full-scale landfill leachate treatment plant.

The dissolved organic matter (DOM) characterization in a full-scale landfill leachate treatment plant is of great importance for the design and operation of treatment processes. In this study, the long-term removal behaviors of DOM during landfill leachate treatment were explored using excitation emission matrix fluorescence spectroscopy (EEMs) coupled with parallel factor analysis (PARAFAC) and self-organizing map (SOM). Results indicated that the application of combining PARAFAC and SOM on EEMs analysis effectively characterized long-term removal behaviors of DOM during leachate treatment. The DOM in raw leachate was dominated by humic substances, while its composition exhibited significant seasonal differences. A large proportion of protein-like fluorescent dissolved organic matter (FDOM) and bulk DOM were removed within membrane bioreactor (MBR) system. Meanwhile the humic-like FDOM removal capacity in nanofiltration (NF) process was well comparable with those in the MBR system owing to the bio-recalcitrant nature of humic substances. The protein-like FDOM and bulk DOM were removed synchronously in both the process of MBR and NF. Moreover, samples distribution exhibited obvious differences among NF concentrate samples. In general, the performance of MBR-NF treatment for landfill leachate displayed reasonable stability in DOM removal irrespective of seasonal variations. This study enhanced our understanding of EEMs application in characterizing leachate-derived DOM composition and has potential implications for the associated monitoring investigations in engineered systems. (C) 2018 Elsevier B.V. All rights reserved.