The composition and photochemical transformations of dissolved organic matter (DOM) in the northern North Pacific Ocean were investigated at the molecular level using ultrahigh resolution mass spectrometry and geochemical tracers. Analyses included vertical profiles and experiments in which deep sea DOM was exposed to sunlight and incubated in the dark. The composition of the deep sea DOM was found to be approximately uniform and enriched with highly unsaturated compounds, with highly aromatic compounds, and with polycyclic aromatics. Surface DOM had a significantly different composition, being enriched with both highly unsaturated and with unsaturated aliphatic compounds potentially due to the addition of photodegradation products and phytoplankton inputs. Deep sea DOM composition is transformed by photoreactions, becoming more similar to surface DOM. The influence of photochemistry extends beyond the photic zone, presumably because of vertical export of DOM previously modified at the surface.

Hydrogen peroxide (H₂O₂) and its precursor, superoxide (O₂(-)), are well-studied photochemical products that are pivotal in regulating redox transformations of trace metals and organic matter in the surface ocean. In attempts to understand the magnitude of both H₂O₂ and O₂(-) photoproduction on a global scale, we implemented a model to calculate photochemical fluxes of these products from remotely sensed ocean color and modeled solar irradiances. We generated monthly climatologies for open ocean H₂O₂ photoproduction rates using an average apparent quantum yield (AQY) spectrum determined from laboratory irradiations of oligotrophic water collected in the Gulf of Alaska. Because the formation of H₂O₂ depends on secondary thermal reactions involving O₂(-), we also implemented a temperature correction for the H₂O₂ AQY using remotely sensed sea surface temperature and an Arrhenius relationship for H₂O₂ photoproduction. Daily photoproduction rates of H₂O₂ ranged from <1 to over 100 nM per day, amounting to ∼30 μM per year in highly productive regions. When production rates were calculated without the temperature correction, maximum daily rates were underestimated by 15-25%, highlighting the importance of including the temperature modification for H₂O₂ in these models. By making assumptions about the relationship between H₂O₂ and O₂(-) photoproduction rates and O₂(-) decay kinetics, we present a method for calculating midday O₂(-) steady-state concentrations ([O₂(-)]ss) in the open ocean. Estimated [O₂(-)]ss ranged from 0.1-5 nM assuming biomolecular dismutation was the only sink for O₂(-), but were reduced to 0.1-290 pM when catalytic pathways were included. While the approach presented here provides the first global scale estimates of marine [O₂(-)]ss from remote sensing, the potential of this model to quantify O₂(-) photoproduction rates and [O₂(-)]ss will not be fully realized until the mechanisms controlling O₂(-) photoproduction and decay are better understood.

Dissolved organic matter (DOM) in aquatic environments forms a vast reservoir of carbon present as a complex supermixture of compounds. An efficient approach to tracking the production and removal of specific DOM fractions is needed across disciplines, for purposes that range from improving global carbon budgets to optimizing water treatment in engineered systems. Although widely used to study DOM, fluorescence spectroscopy has yet to deliver specific fractions with known spectral properties and predictable distributions. Here, we mathematically isolate four visible-wavelength fluorescent fractions in samples from contrasting lake, river, and ocean environments. Using parallel factor analysis (PARAFAC), we show that most measured fluorescence in environmental samples can be explained by ubiquitous spectra with nearly stable optical properties and photodegradation behaviors over environmental pH gradients. Sample extraction changed bulk fluorescence spectra but not the number or shape of underlying PARAFAC components, while photobleaching preferentially removed the two longest-wavelength components. New approaches to analyzing fluorescence data sets incorporating these findings should improve the interpretation of DOM fluorescence and increase its utility for tracing organic matter biogeochemistry in aquatic systems.

No abstract

Vast stores of arctic permafrost carbon that have remained frozen for millennia are thawing, releasing ancient dissolved organic carbon (DOC) to arctic inland waters. Once in arctic waters, DOC can be converted to CO2 and emitted to the atmosphere, accelerating climate change. Sunlight‐driven photoreactions oxidize DOC, converting a portion to CO2 and leaving behind a photomodified pool of dissolved organic matter (DOM). Samples from the Kolyma River, its tributaries, and streams draining thawing yedoma permafrost were collected. Irradiation experiments and radiocarbon dating were employed to assess the photolability of ancient permafrost‐DOC in natural and laboratory generated samples containing a mix of modern and ancient DOC. Photolabile DOC was always modern, with no measurable photochemical loss of ancient permafrost‐DOC. However, optical and ultrahigh resolution mass spectrometric measurements revealed that both modern river DOM and ancient permafrost‐DOM were photomodified during the irradiations, converting aromatic compounds to less conjugated compounds. These findings suggest that although sunlight‐driven photoreactions do not directly mineralize permafrost‐DOC, photomodification of permafrost‐DOM chemistry may influence its fate and ecological functions in aquatic systems.