Search for: All records

Abstract. Barium is widely used as a proxy for dissolved silicon and particulateorganic carbon fluxes in seawater. However, these proxy applications arelimited by insufficient knowledge of the dissolved distribution of Ba([Ba]). For example, there is significant spatial variability in thebarium–silicon relationship, and ocean chemistry may influence sedimentaryBa preservation. To help address these issues, we developed 4095 models forpredicting [Ba] using Gaussian process regression machine learning. Thesemodels were trained to predict [Ba] from standard oceanographic observationsusing GEOTRACES data from the Arctic, Atlantic, Pacific, and Southernoceans. Trained models were then validated by comparing predictions againstwithheld [Ba] data from the Indian Ocean. We find that a model trained usingdepth, temperature, and salinity, as well as dissolved dioxygen, phosphate,nitrate, and silicate, can accurately predict [Ba] in the Indian Ocean with amean absolute percentage deviation of 6.0 %. We use this model tosimulate [Ba] on a global basis using these same seven predictors in theWorld Ocean Atlas. The resulting [Ba] distribution constrains the Ba budgetof the ocean to 122(±7) × 1012 mol and revealsoceanographically consistent variability in the barium–silicon relationship. We then calculate the saturation state of seawater with respect to barite. This calculation reveals systematic spatial and vertical variations in marine barite saturation and shows that the ocean below 1000 m is at equilibrium with respect tobarite. We describe a number of possible applications for our model outputs, ranging from use in mechanistic biogeochemical models to paleoproxy calibration. Ourapproach demonstrates the utility of machine learning in accurately simulatingthe distributions of tracers in the sea and provides a framework that couldbe extended to other trace elements. Our model, the data used in training and validation, and global outputs are available in Horner and Mete (2023, https://doi.org/10.26008/1912/bco-dmo.885506.2). 

Abstract Processes controlling dissolved barium (dBa) were investigated along the GEOTRACES GA03 North Atlantic and GP16 Eastern Tropical Pacific transects, which traversed similar physical and biogeochemical provinces. Dissolved Ba concentrations are lowest in surface waters (∼35–50 nmol kg⁻¹) and increase to 70–80 and 140–150 nmol kg⁻¹in deep waters of the Atlantic and Pacific transects, respectively. Using water mass mixing models, we estimate conservative mixing that accounts for most of dBa variability in both transects. To examine nonconservative processes, particulate excess Ba (pBa_xs) formation and dissolution rates were tracked by normalizing particulate excess²³⁰Th activities. Th‐normalized pBa_xsfluxes, with barite as the likely phase, have subsurface maxima in the top 1,000 m (∼100–200 μmol m⁻² year⁻¹average) in both basins. Barite precipitation depletes dBa within oxygen minimum zones from concentrations predicted by water mass mixing, whereas inputs from continental margins, particle dissolution in the water column, and benthic diffusive flux raise dBa above predications. Average pBa_xsburial efficiencies along GA03 and GP16 are ∼37% and 17%–100%, respectively, and do not seem to be predicated on barite saturation indices in the overlying water column. Using published values, we reevaluate the global freshwater dBa river input as 6.6 ± 3.9 Gmol year⁻¹. Estuarine mixing processes may add another 3–13 Gmol year⁻¹. Dissolved Ba inputs from broad shallow continental margins, previously unaccounted for in global marine summaries, are substantial (∼17 Gmol year⁻¹), exceeding terrestrial freshwater inputs. Revising river and shelf dBa inputs may help bring the marine Ba isotope budget more into balance.