Relationships Between DOC and CDOM Based on the Total Carbon‐Specific Fluorescence Intensities for River Waters Across China

Abstract

Weak correlations between chromophoric dissolved organic matter (CDOM) absorption coefficient a(440) and dissolved organic carbon (DOC) were observed due to weak absorption for river waters especially in the Qinghai‐Tibetan Plateau, and thus, it is impossible to estimate DOC concentrations by CDOM absorption across China. Therefore, it is necessary to estimate DOC concentrations through the fluorescent fraction of CDOM (i.e., FDOM). FDOM was proposed as a new method for estimation of large‐scale DOC concentrations in river waters across China. A total of 301 water samples from eight river basins across China were selected to assess FDOM by excitation‐emission matrix (EEM) fluorescence. Five fluorescent regions were obtained by EEM coupled with fluorescence regional integration (FRI) (EEM‐FRI) method. However, weak correlations between the fluorescence intensities FR(1–5), FSUM, and DOC concentrations were observed for all water samples, respectively. Therefore, the total carbon‐specific fluorescence intensities FSUM/DOC were used to establish relationships between DOC and FDOM. All 301 water samples excluding five outliers with FSUM/DOC higher than 8,000 nm/(mg/L)) were divided into four groups based on the threshold values for FSUM/DOC, and strong positive correlations between FSUM and DOC were observed in each group (R2 = 0.843, FSUM/DOC < 1,000 nm/(mg/L); R2 = 0.928, 1,000 < FSUM/DOC < 2,000 nm/(mg/L); R2 = 0.964, 2,000 < FSUM/DOC < 3,000 nm/(mg/L); and R2 = 0.953, 3,000 < FSUM/DOC < 8,000 nm/(mg/L)), respectively. This result indicated that the riverine DOC concentrations on large‐scale across China can be estimated directly by FDOM properties.

Optical properties of turbid coastal water and their remote sensing retrievals based on radiative transfer approaches

“…Empirical models have also been extensively employed to retrieve the concentrations of CDOM and suspended sediments in coastal waters (D'Sa et al, 2007;Doerffer and Fischer, 1994;Liu et al, 2014;Loisel et al, 2014;Siswanto et al, 2011;Zhang et al, 2010). Other biogeochemical parameters, such as dissolved organic carbon (DOC), through its relationship with CDOM absorption properties, can also be empirically retrieved from earth observation data (Cao et al, 2018;Zhao and Song, 2018). K d , as a QIOP, has also been empirically related to reflectance measurements for both clear waters and optically complex waters (Austin and Petzold, 1981;Mueller, 2000;Werdell and Bailey, 2005;Zhang and Fell, 2007).…”

“…For turbid waters, one of the challenges for ocean color remote sensing is the scarcity of "ground-true" data for model validation, which is because current commercial instruments for IOPs measurements, such as Wet Labs ac-s meter and bb9 meter, are prone to saturate in high turbidity environments. The other challenges is that due to the large variation of the IOPs and concentrations of in-water constituents, most semianalytical models have been less successful in retrieving IOPs or K d (IOCCG, 2000(IOCCG, , 2006Salama et al, 2011;Wang et al, 2017;Zibordi et al, 2009), which is mainly attributed to the strong backscattering by the suspended particles. Therefore, developed models for the retrieval of SPM and Chl-a concentrations in the Yangtze estuary are all empirical or semi-analytical based Shen et al, 2010c;Shen et al, 2013;Sokoletsky et al, 2014).…”

“…Algorithms for the remote retrieval of K d have been developed for clear waters (Austin and Petzold, 1981;Mueller, 2000), but they always generate huge errors in turbid waters due to the complexity in optical properties and limitations in the algorithms (IOCCG, 2000(IOCCG, , 2006Salama et al, 2011;Zibordi et al, 2009). To retrieve K d in optically complex waters, several methods were adopted.…”