Identification of pollution properties in typical tropical island watershed and their significant for source apportionment: A case study of Sanya, China

Abstract

Tropical coastal watersheds face complex pollution challenges due to heterogeneous land-use patterns and seasonal anthropogenic activities. This study developed a multi-dimensional pollution fingerprinting framework to resolve source contributions in Central Fishing Port, Sanya, China. We integrated four analytical tiers: (1) multivariate statistical screening of water quality parameters represented pollution sources, (2) seasonal analysis of dissolved organic matter (DOM) with three-dimensional excitation-emission matrix fluorescence spectroscopy combined with parallel factor analysis (EEM-PARAFAC), (3) pollution apportionment using receptor model based on multi-source variables, and (4) validation of pollution sources at molecular-level with Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS). Comparative modeling demonstrated that PCA-APCS-MLR outperformed PMF in traceability accuracy (R² > 0.82 vs. >0.65), particularly in resolving overlapping fishery and domestic sewage contributions. Key findings demonstrated that there was strong spatiotemporal heterogeneity in pollution source contributions due to the fishing activities, and complementary capabilities of EEM-PARAFAC and FT-ICR MS for qualitative source identification. Based on the integration of physicochemical parameters of water quality, fluorescence spectra and mass spectrometry, the traceability and apportionment accuracy of complex and overlapping pollution sources were effectively improved. This integrated approach highlighted the synergistic value of multi-parameter integration (physicochemical-optical-mass spectrometric) for coastal receptor modeling, providing a transferable template for tropical estuary pollution management.