Optical properties and compositions of dissolved organic matter (DOM) along trophic gradients: Implications for a COD proxy in urban lakes

Aiming to control lake eutrophication, proposed methods for convenient and faithworthy lake water quality evaluation are warranted. Optical measurement of dissolved organic matter (DOM) demonstrates great potential for estimating organic matter (OM) composition, and can thus serve as a proxy for conventional chemical oxygen demand (COD_Mn) measurements, which are considered as imprecise and environmentally unfriendly. Hence, we conducted a field campaign across 30 lakes in Wuhan's metropolitan area, collecting 255 samples from varying trophic states to evaluate the predictive capability of COD_Mn using DOM optical measurements combined with parallel factor (PARAFAC) analysis. The DOM optical properties and chemical composition exhibited considerable variability across varying trophic state levels (TSLs). Fluorescence components C1-C3 and C5, fluorescence index (FI), and absorption at 254 nm (α₂₅₄), increased as TSL increased, while the DOM spectral slope (S_R) decreased. COD_Mn was positively and significantly correlated with fluorescence components C1–C3 and C5, freshness index (β/α), autochthonous index (BIX), humification index (HIX), α₂₅₄, the ratio of α₂₅₀ to α₃₆₅ (E2/E3) while being negatively correlated with S_R. Parameters α₂₅₄, C1, C3, C4, FI, β/α, and HIX were identified as key predictors of COD_Mn. The multiple linear regression model successfully predicted COD_Mn (r² = 0.63, p < 0.01, n = 1113) and demonstrated superior performance in mesotrophic lakes. These findings highlight the potential for establishing high-frequency, continuous, and multi-regional COD monitoring programs.