When terrigenous dissolved organic carbon (tDOC) rich in chromophoric dissolved organic matter (tCDOM) enters the ocean, solar radiation mineralizes it partially into dissolved inorganic carbon (DIC). This study addresses the amount and the rates of DIC photoproduction from tDOC and the area of ocean required to photomineralize tDOC. We collected water samples from 10 major rivers, mixed them with artificial seawater, and irradiated them with simulated solar radiation to measure DIC photoproduction and the photobleaching of tCDOM. The linear relationship between DIC photoproduction and tCDOM photobleaching was used to estimate the amount of photoproduced DIC from the tCDOM fluxes of the study rivers. Solar radiation was estimated to mineralize 12.5 ± 3.7 Tg C yr−1 (10 rivers)−1 or 18 ± 8% of tDOC flux. The irradiation experiments also approximated typical apparent spectral quantum yields for DIC photoproduction (ϕλ) over the entire lifetime of the tCDOM. Based on ϕλs and the local solar irradiances in river plumes, the annual areal DIC photoproduction rates from tDOC were calculated to range from 52 ± 4 (Lena River) to 157 ± 2 mmol C m−2 yr−1 (Mississippi River). When the amount of photoproduced DIC was divided by the areal rate, 9.6 ± 2.5 × 106 km2 of ocean was required for the photomineralization of tDOC from the study rivers. Extrapolation to the global tDOC flux yields 45 (31–58) Tg of photoproduced DIC per year in the river plumes that cover 34 (25–43) × 106 km2 of the ocean.

Solar radiation mineralizes dissolved organic matter (DOM) to dissolved inorganic carbon through photochemical reactions (DIC photoproduction) that are influenced by iron (Fe) and pH. This study addressed as to what extent Fe contributes to the optical properties of the chromophoric DOM (CDOM) and DIC photoproduction at different pH values. We created the associations of Fe and DOM (Fe-DOM) that cover the range of loadings of Fe on DOM and pH values found in freshwaters. The introduced Fe enhanced the light absorption by CDOM independent of pH. Simulated solar irradiation decreased the light absorption by CDOM (i.e., caused photobleaching). Fe raised the rate of photobleaching and steepened the spectral slopes of CDOM in low pH but resisted the slope steepening in neutral to alkaline pH. The combination of a low pH (down to pH 4) and high Fe loading on DOM (up to 3.5 μmol mg DOM –1 ) increased the DIC photoproduction rate and the apparent quantum yields for DIC photoproduction up to 7-fold compared to the corresponding experiments at pH >6 or without Fe. The action spectrum for DIC photoproduction shifted toward the visible spectrum range at low pH in the presence of Fe. Our results demonstrated that Fe can contribute to DIC photoproduction by up to 86% and produce DIC even at the visible spectrum range in acidic waters. However, the stimulatory effect of Fe is negligible at pH >7.

Solar radiation initiates photochemical oxidation of dissolved organic carbon (DOC) to dissolved inorganic carbon (DIC) in inland waters, contributing to their carbon dioxide emissions to the atmosphere. Models can determine photochemical DIC production over large spatiotemporal scales and assess its role in aquatic C cycling. The apparent quantum yield (AQY) spectrum for photochemical DIC production, defined as mol DIC produced per mol chromophoric dissolved organic matter‐absorbed photons, is a critical model parameter. In previous studies, the principle for the determination of AQY spectra is the same but methodological specifics differ, and the extent to which these differences influence AQY spectra and simulated aquatic DIC photoproduction is unclear. Here, four laboratories determined AQY spectra from water samples of eight inland waters that are situated in Alaska, Finland, and Sweden and span a nearly 10‐fold range in DOM absorption coefficients. All AQY values fell within the range previously reported for inland waters. The inter‐laboratory coefficient of variation (CV) for wavelength‐integrated AQY spectra (300–450 nm) averaged 38% ± 3% SE, and the inter‐water CV averaged 63% ± 1%. The inter‐laboratory CV for simulated photochemical DIC production (conducted for the five Swedish lakes) averaged 49% ± 12%, and the inter‐water CV averaged 77% ± 10%. This uncertainty is not surprising given the complexities and methodological choices involved in determining DIC AQY spectra and needs to be considered when applying photochemical rate modeling. Thus, we also highlight current methodological limitations and suggest future improvements for DIC AQY determination to reduce inter‐laboratory uncertainty.