Characterization of dissolved organic matter at urban and industrial rainwater of Bangladesh by fluorescence spectroscopy and EEM-PARAFAC modeling

Abstract Fluorescent dissolved organic matter (fDOM) was characterized seasonally in urban and industrial rainwater in one of the most polluted countries in the world, Bangladesh, using excitation-emission matrix (EEM) fluorescence spectroscopy and Parallel factor analysis (PARAFAC). EEM and PARAFAC modeling enabled identification of six fluorophores (peaks: A, C, M, T, Tuv, Wuv) and five fDOM components (humic-like, protein-like, fulvic acid (C)-like, humic (M)-like, and detergent-like), in both urban and industrial rain. The concentration of dissolved organic carbon (DOC) varied between 0.2–37.9 mg/L and 0.3–39.4 mg/L in urban and industrial rainwater, respectively. DOC concentration and fDOM component intensity were higher in the dry than in the wet season. DOC concentration was higher in urban than industrial rain in maximum months of the year. However, the higher intensity of fDOM components in industrial rainwater than in urban indicated that the industrial atmosphere might contain more nitrogenous organic compounds besides carbonaceous. The anthropogenic detergent-like component was present throughout the year with high intensity both at urban and industrial rain. fDOM indices and principal component analysis (PCA) described that organic components were not fresh, and originated from both allochthonous and autochthonous sources, and affected by microbial decomposition. fDOM components had strong positive correlations with gases (SO2, NOX, CO, and O3) and particulate matters (PM2.5 and PM10), indicating that they originated from similar local and distant sources. The Hybrid Single-Particle Lagrangian Integrated (HYSPLIT) back trajectory model described the local and distant air mass flow direction around sampling stations and interpreted their role in fDOM intensity fluctuations.