Global trends in marine nitrate N isotopes from observations and a neural network-based climatology

Rafter, Patrick A.; Bagnell, Aaron; Marconi, Dario; DeVries, Tim

doi:10.5194/bg-16-2617-2019

Cited by 34 publications

(46 citation statements)

References 94 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For this step, we designed and trained an artificial neural network (ANN) that uses a suite of well‐sampled predictors to interpolate gaps in the dissolved Cu data set. ANNs are versatile models inspired by biological neural systems (Hassoun, 1995) and have proven to perform well for interpolation of sparse data in the oceanographic environment (Bowles et al, 2014; Landschützer et al, 2014; Rafter et al, 2019; Roshan & DeVries, 2017; Roshan et al, 2018). Our procedure is similar to that used by Roshan and DeVries (2017) to map DOC concentrations and by Roshan et al (2018) to map dissolved zinc (Zn) concentrations.…”

Section: Methodsmentioning

confidence: 99%

Constraining the Global Ocean Cu Cycle With a Data‐Assimilated Diagnostic Model

Roshan

DeVries

2020

Global Biogeochemical Cycles

View full text Add to dashboard Cite

Copper (Cu) is a biologically important trace metal for marine plankton, but it is also toxic at high concentrations. Understanding the global distribution of Cu and the processes controlling its cycling in the ocean is important for understanding how the distribution of this important element can respond to climate change. Here, we use available observations of dissolved copper, an artificial neural network, and an ocean circulation inverse model, to derive a global estimate of the three-dimensional distribution and cycling of dissolved Cu in the ocean. We find that there is net removal by bio-assimilation and/or scavenging of dissolved Cu in the surface ocean at a rate of~1.7 Gmol yr −1 and that both the concentration and export of dissolved Cu are highest in the Southern Ocean. In the subsurface above the near-sediment layer, dissolved Cu is removed at a net rate of~2.4 Gmol yr −1 , consistent with scavenging onto sinking particles, contributing to an increase in the flux of particulate Cu with depth. This removal of Cu by scavenging in the interior ocean is balanced by a net near-sediment source of dissolved Cu, which sustains a gradual increase in the concentration of dissolved Cu with depth. Globally, this net near-sediment source is estimated at~2.6 Gmol yr −1 in the deep ocean and~0.8 Gmol yr −1 along continental shelves and slopes. Our results suggest an active oceanic dissolved Cu cycle with a mean internal ocean residence time of~530 years, highlighting the potential for climate-driven changes in the marine Cu cycle.

show abstract

Section: Methodsmentioning

confidence: 99%

Constraining the Global Ocean Cu Cycle With a Data‐Assimilated Diagnostic Model

Roshan

DeVries

2020

Global Biogeochemical Cycles

View full text Add to dashboard Cite

show abstract

“…Remaining parameters were either experimentally manipulated in different model runs or calculated via an optimization procedure under a variety of model parameterizations; the optimized parameters are the rate constants for NO 3 − reduction ( k NAR ), NO 2 − reduction ( k NIR ), NO 2 − oxidation ( k NXR ), and DON remineralization ( k remin ; Table S1). The optimization procedure, as described by Martin et al (), compares the model output 14 N and 15 N concentrations to a large database of NO 3 − and NO 2 − concentration and isotope data (Martin et al, ; Rafter et al, ). The parameters were simultaneously modified over the course of the optimization to find the combination with which the 14 N and 15 N output best fit the observations and minimized a cost function consisting of a weighted least squares algorithm between modeled NO 3 − and NO 2 − concentrations and isotopes and their observed distributions interpolated to the model grid (Martin et al, ).…”

Section: Methodsmentioning

confidence: 99%

Assessing Marine Nitrogen Cycle Rates and Process Sensitivities With a Global 3‐D Inverse Model

Martin

Primeau

Casciotti

2019

Global Biogeochemical Cycles

View full text Add to dashboard Cite

We present results from a global inverse marine nitrogen (N) cycle model that include nitrate (NO3−) and nitrite (NO2−) concentrations and their N isotopic compositions as constraints on N cycle process rates in marine oxygen deficient zones (ODZs). NO2− is an important intermediate in the N cycle, particularly in ODZs where it is a substrate in the N loss processes, denitrification, and anammox. Similar to earlier work, our model yields a total water column N loss rate of 61 ± 10 Tg N/year. However, by including NO2− and its N isotopic composition, we are able to assess the relative contributions of denitrification and anammox to N loss and examine some of the potential drivers of that balance. We find that anammox contributes 60% of global water column N loss, dominating N loss along the edges of ODZs, while denitrification is more important in the anoxic ODZ cores. The decoupling of anammox and denitrification is supported by NO2− oxidation, which co‐occurs with NO3− reduction and anammox in ODZs. High rates of NO2− oxidation (up to 400 nM/day), which are tightly coupled to heterotrophic NO3− reduction, are required to match NO3− and NO2− concentration and isotope observations in marine ODZs. Lowering the rate of NO2− oxidation in ODZs by adjusting O2‐sensitive parameters results in higher rates of water column N loss, highlighting the role of NO2− oxidation in maintaining the marine fixed N inventory.

show abstract

“…For example, Roshan and DeVries (2017) applied an artificial neural network (ANN) to extrapolate observed dissolved organic carbon (DOC) to the global ocean. Rafter et al (2019) used an ensemble of neural networks to study oceanic δ 15 N distribution. ANNs have also been used to study DMS on regional scales (e.g., Humphries et al, 2012).…”

Section: Introductionmentioning

confidence: 99%

Global ocean dimethyl sulfide climatology estimated from observations and an artificial neural network

et al. 2020

View full text Add to dashboard Cite

Abstract. Marine dimethyl sulfide (DMS) is important to climate due to the ability of DMS to alter Earth's radiation budget. Knowledge of the global-scale distribution, seasonal variability, and sea-to-air flux of DMS is needed in order to improve understanding of atmospheric sulfur, aerosol/cloud dynamics, and albedo. Here we examine the use of an artificial neural network (ANN) to extrapolate available DMS measurements to the global ocean and produce a global climatology with monthly temporal resolution. A global database of 82 996 ship-based DMS measurements in surface waters was used along with a suite of environmental parameters consisting of latitude–longitude coordinates, time of day, time of year, solar radiation, mixed layer depth, sea surface temperature, salinity, nitrate, phosphate, and silicate. Linear regressions of DMS against the environmental parameters show that on a global-scale mixed layer depth and solar radiation are the strongest predictors of DMS. These parameters capture ∼9 % and ∼7 % of the raw DMS data variance, respectively. Multilinear regression can capture more of the raw data variance (∼39 %) but strongly underestimates DMS in high-concentration regions. In contrast, the artificial neural network captures ∼66 % of the raw data variance in our database. Like prior climatologies our results show a strong seasonal cycle in surface ocean DMS with the highest concentrations and sea-to-air fluxes in the high-latitude summertime oceans. We estimate a lower global sea-to-air DMS flux (20.12±0.43 Tg S yr−1) than the prior estimate based on a map interpolation method when the same gas transfer velocity parameterization is used. Our sensitivity test results show that DMS concentration does not change unidirectionally with each of the environmental parameters, which emphasizes the interactions among these parameters. The ANN model suggests that the flux of DMS from the ocean to the atmosphere will increase with global warming. Given that larger DMS fluxes induce greater cloud albedo, this corresponds to a negative climate feedback.

show abstract

Global trends in marine nitrate N isotopes from observations and a neural network-based climatology

Cited by 34 publications

References 94 publications

Constraining the Global Ocean Cu Cycle With a Data‐Assimilated Diagnostic Model

Constraining the Global Ocean Cu Cycle With a Data‐Assimilated Diagnostic Model

Assessing Marine Nitrogen Cycle Rates and Process Sensitivities With a Global 3‐D Inverse Model

Global ocean dimethyl sulfide climatology estimated from observations and an artificial neural network

Contact Info

Product

Resources

About