A marine carbon monoxide (CO) model with a new parameterization of microbial oxidation
Introduction
Carbon monoxide (CO) plays two key roles in the atmosphere: 1) it impacts on climate forcing by competing with the methane in the reaction with the hydroxyl radical (Daniel and Solomon, 1998), the main atmospheric oxidant, and 2) it is involved in the production of ozone (Logan et al., 1981), which in turn leads to the photochemical smog, reducing atmospheric visibility. Therefore, considering its impact on the chemical properties of the atmosphere, CO is regarded as one of the most important trace gasses (Stocker et al., 2013).
Although most of the atmospheric CO is emitted from the continent, outgassing from the sea surface could be a significant source in the remote marine environments and in the southern hemisphere where the CO in the surface ocean is supersaturated with respect to the overlying air (Bates et al., 1995; Conrad et al., 1982; Khalil and Rasmussen, 1990; Logan et al., 1981; Rhee, 2000; Stubbins et al., 2006; Zafiriou et al., 2003). Understanding CO dynamics in the marine upper layer is thereby crucial to assess the role played by this gas in the global climate regulation.
The concentration of dissolved CO in the surface ocean results from the balance between photochemical production (Conrad et al., 1982; Redden, 1982; Zuo and Jones, 1995), microbial oxidation (Conrad et al., 1982; Jones and Amador, 1993; Jones and Morita, 1984), air-sea gas exchange (Bates et al., 1995; Conrad et al., 1982; Park and Rhee, 2016; Zuo and Jones, 1995), and vertical mixing (Doney et al., 1995; Gnanadesikan, 1996; Johnson and Bates, 1996; Kettle, 1994) (Fig. 1). Among these processes, photochemical production and microbial oxidation are the major processes contributing to the CO budget in the ocean (Zafiriou et al., 2003). Indeed, photolysis of chromophoric dissolved organic matter (CDOM) in the euphotic zone is the only known source of CO, while microbial oxidation is by far the dominant sink, destroying more than 80% of the CO pool. Therefore, an accurate parameterization of microbial oxidation is essential to estimate CO flux from the upper ocean (Moran and Miller, 2007). Microbial CO oxidation is described conventionally as a first-order kinetics (Conrad et al., 1982; Johnson and Bates, 1996; Jones, 1991; Jones and Amador, 1993; Xie et al., 2005; Zafiriou et al., 2003), assuming that CO oxidation rate increases linearly with the increase of dissolved CO concentrations, regardless of microbial abundance and community composition.
Several modeling studies of marine CO dynamics were carried out to understand the CO budget in the ocean. Kettle (1994) developed a model to understand the surface diurnal pattern of dissolved CO concentrations at the Bermuda Atlantic Time-series Study (BATS) station in the Sargasso Sea. The model adopted the Price-Weller-Pinkel (PWP) vertical mixing scheme (Price et al., 1986) to simulate the physical mixing and transport in the surface mixed layer and subsurface layers. CO production was simulated by using a modified version of the photochemical production module previously developed for dissolved hydrogen peroxide (H2O2) in the marine environments by Sikorski and Zika (1993). In subsequent studies (Kettle, 2000, 2005b), an optimization technique was used to reduce the discrepancy between observed and simulated dissolved CO concentrations. Based on the CO observation by Kettle (1994), Gnanadesikan (1996) developed a simple model coupling CO dynamics with a bulk mixed layer model to identify nine marine regimes corresponding to different interactions between physical mixing and (photochemical) productions. The regimes were defined depending on the ratios between physical length scales, i.e. the depths to which vertical mixing, ventilation, and photochemically active radiation penetration occur. The critical difference between the models of Gnanadesikan (1996) and Kettle (1994) is the way they describe CO photochemical production: while Kettle (1994) considered the variation of CO photoproduction rate depending on the spectral irradiance, Gnanadesikan (1996) assumed that the photoproduction is proportional to a fraction of the total irradiance.
Simple box models were also used to determine the ratio between photochemical production and microbial oxidation (Johnson and Bates, 1996; Kitidis et al., 2011). Johnson and Bates (1996) determined the photoproduction and oxidation rates using the exponential fits of their observed diurnal variations of CO. Kitidis et al. (2011) took the production and oxidation rates from their experimental measurements and considered the vertical gradient of CO concentration due mainly to the light attenuation with depth.
All the above mentioned modeling studies used a first-order kinetics to describe CO microbial oxidation implicitly assuming that the CO microbial oxidation rate is constant in the ocean. However, this assumption is not consistent with experimental studies reporting that the CO oxidation rates (kCO) varies dramatically (0.003 - 1.11 h−1) both temporally and spatially (Johnson and Bates, 1996; Jones, 1991; Jones and Amador, 1993; Kwon, 2015; Xie et al., 2005). Moreover, Xie et al. (2005, 2009) reported complex influences of various biotic and abiotic variables on this process showing that kCO is dependent on temperature, primary production and salinity. These studies suggest that previously used formulations can only be reliable in specific conditions (i.e. specific location and time of the year), but cannot be used in modeling work dealing with large spatial (e.g. global models) and temporal scales (from seasonal upward).
The aim of this paper is to provide a novel model formulation able to simulate the variability of CO oxidation rate described in literature. To this end, we tested the hypothesis that microbial CO oxidation is a function of not only dissolved CO concentration but also bacterial biomass. Our CO model was implemented in a widely used marine ecosystem model, the European Regional Seas Ecosystem Model (ERSEM; Butenschön et al., 2016). Since ERSEM only accounts for heterotrophic bacteria, we assumed that the activity of CO oxidizing bacteria is proportional to the biomass of the heterotrophic bacteria community. This assumption is supported by the several studies (Gonzalez and Moran, 1997; Gonzalez et al., 2000; Moran et al., 2004; Suzuki et al., 2001; Tolli et al., 2006) which show that CO oxidizing bacteria (Roseobacter-associated clade) are ubiquitous in the ocean and that account for a relatively constant fraction of the heterotrophic bacterial biomass.
Section snippets
CO model with a new formulation of microbial oxidation
Temporal and spatial variability of dissolved CO concentrations ([CO]) in the water column was formulated as a function of depth (z) and time (t) associated with photochemical production (J), air-sea gas exchange (F,z = 0), vertical mixing (V), and microbial oxidation (M):
The detailed formulation of the four processes on the right hand side of Eq. (1) are described in the following sections.
Nine-day simulation with optimum kbio
The microbial oxidation rate coefficient, kbio, in the new CO-GOTM-ERSEM model determined by the procedure described in Section 2.3 was 5.7 ± 0.2 (μg C m−3)−1 h−1. We present here the model performance focusing on the surface CO concentrations as well as their vertical distributions. Sensitivity results of the kbio to the perturbations of CO sources and sinks are also presented.
Meaning of the second-order loss kinetics
Microbial oxidation is the dominant sink of CO in seawater overwhelming air–sea gas exchange under normal turbulent conditions at the sea surface (Gnanadesikan, 1996; Zafiriou et al., 2003). It is therefore crucial to accurately parameterize this process if we want to simulate dissolved CO concentrations in a reliable way. In this work, we propose a new model including variable bacterial biomass in the formulation of CO oxidation. Our model is supported by literature findings showing that the
Conclusions
A CO model was developed with a new parameterization of microbial oxidation, the dominant sink of CO in the ocean. We suggested a new parameterization implying a second-order loss kinetics depending on bacterial biomass other than CO concentration. The new parameterization introduces a universal constant kbio which describes the bacterial biomass specific CO oxidation rate. By optimizing CO simulations against the 9-day observations of surface CO concentrations at BATS (Kettle, 1994), kbio was
Credit_Author_Statement
Young Shin Kwon: Conceptualization, Writing-Original draft preparation, Software, Data curation Visualization, Formal analysis, Funding acquisition
Hyoun-Woo Kang: Supervision, Methodology, Writing- Original draft preparation, Funding acquisition
Luca Polimene: Writing-Reviewing and Editing, Funding acquisition
Tae Siek Rhee: Writing-Reviewing and Editing, Supervision, Funding acquisition
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments, samples, and data
The original data set from BATS station can be accessed at the BIOS database http://bats.bios.edu/data/. This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. NRF-2018H1A2A1060886) and by Korea polar programs (PM20050 and PE20150). Hyoun-Woo Kang was supported by the KIOST in-house project (PE99811). Luca Polimene was funded through the UK NERC grants NE/N001974/1 and NE/R011087/1.
References (81)
- et al.
Linking CDOM spectral absorption to dissolved organic carbon concentrations and loadings in boreal estuaries
Estuar Coast Shelf Sci
(2012) - et al.
The microbial food web in the european regional seas ecosystem model
Netherlands journal of sea research
(1995) - et al.
Ecosystem dynamics at six contrasting sites: a generic modelling study
Journal of marine systems
(2004) - et al.
Models of turbulence in the marine environment - a comparative study of two-equation turbulence models
Journal of marine systems
(1999) - et al.
Bacterial abundance and production in epipelagic and mesopelagic waters in the Subantarctic and Polar Front zones south of Tasmania
Deep Sea Research Part II: Topical Studies in Oceanography
(2011) The relationship between chromophoric dissolved organic matter and dissolved organic carbon in the European Atlantic coastal area and in the West Mediterranean Sea (Gulf of Lions)
Mar Chem
(2000)- et al.
Relationship between the optical properties of chromophoric dissolved organic matter and total concentration of dissolved organic carbon in the southern Baltic Sea region
Mar Chem
(1996) - et al.
Biogeochemistry of total organic carbon and nitrogen in the Sargasso Sea: control by convective overturn
Deep Sea Research Part II: Topical Studies in Oceanography
(2001) Evaluation of a seawater equilibrator for shipboard analysis of dissolved oceanic trace gases
Anal. Chim. Acta
(1999)- et al.
Summary diagrams for coupled hydrodynamic-ecosystem model skill assessment
Journal of Marine Systems
(2009)