ALBERT

Description: Marine aggregates are the vector for biogenically bound carbon and nutrients from the euphotic zone to the interior of the oceans. To improve the representation of this biological carbon pump in the global biogeochemical HAMburg Ocean Carbon Cycle (HAMOCC) model, we implemented a novel Microstructure, Multiscale, Mechanistic, Marine Aggregates in the Global Ocean (M4AGO) sinking scheme. M4AGO explicitly represents the size, microstructure, heterogeneous composition, density and porosity of aggregates and ties ballasting mineral and particulate organic carbon (POC) fluxes together. Additionally, we incorporated temperature-dependent remineralization of POC. We compare M4AGO with the standard HAMOCC version, where POC fluxes follow a Martin curve approach with (i) linearly increasing sinking velocity with depth and (ii) temperature-independent remineralization. Minerals descend separately with a constant speed. In contrast to the standard HAMOCC, M4AGO reproduces the latitudinal pattern of POC transfer efficiency, as recently constrained by Weber et al. (2016). High latitudes show transfer efficiencies of ≈0.25±0.04, and the subtropical gyres show lower values of about 0.10±0.03. In addition to temperature as a driving factor for remineralization, diatom frustule size co-determines POC fluxes in silicifier-dominated ocean regions, while calcium carbonate enhances the aggregate excess density and thus sinking velocity in subtropical gyres. Prescribing rising carbon dioxide (CO2) concentrations in stand-alone runs (without climate feedback), M4AGO alters the regional ocean atmosphere CO2 fluxes compared to the standard model. M4AGO exhibits higher CO2 uptake in the Southern Ocean compared to the standard run, while in subtropical gyres, less CO2 is taken up. Overall, the global oceanic CO2 uptake remains the same. With the explicit representation of measurable aggregate properties, M4AGO can serve as a test bed for evaluating the impact of aggregate-associated processes on global biogeochemical cycles and, in particular, on the biological carbon pump.

Description: Oceans play a major role on the exchange of carbon with the atmosphere and thereby on past climates with glacial/interglacial variations of the CO2 concentration. The melting of ice sheets during deglaciations lets the sea level rise which leads to the flooding of coastal land areas resulting in the transfer of terrestrial organic matter to the ocean. However, the consequences of such fluxes on the ocean biogeochemical cycle and uptake/release of CO2 are poorly constrained. Moreover, this potentially important exchange of carbon at the land-sea interface is not represented in most Earth System Models. We present here the implementation of terrestrial organic matter fluxes into the ocean at the transiently changing land-sea interface in the Max Planck Institute for Meteorology Earth System Model (MPI-ESM) and investigate their effect on the biogeochemistry during the last deglaciation. Our results show that during the deglaciation, most of the terrestrial organic matter inputs to the ocean occurs during Meltwater Pulse 1a (between 15–14 ka) which leads to additional 21.2 GtC of terrestrial origin (mostly originating from wood and humus). Although this additional organic matter input is relatively small in comparison to the global ocean inventory (0.06 %) and thus doesn’t have an impact on the global CO2 flux, the terrestrial organic matter fluxes initiate oceanic outgassing at regional hotspots like in Indonesia for a few hundred years. Finally, sensitivity experiments highlight that terrestrial organic matter fluxes are the drivers of oceanic outgassing in flooded coastal regions during Meltwater Pulse 1a. Furthermore, the magnitude of outgassing is rather insensitive to higher carbon to nutrients ratios of the terrestrial organic matter. Our results provide a first estimate of the importance of terrestrial organic matter fluxes in a transient deglaciation simulation. Moreover, our model development is an important step towards a fully coupled carbon cycle in an Earth System Model applicable for simulations of glacial/interglacial cycles.

Description: Direct comparison between paleo oceanic δ13C records and model results facilitates assessing simulated distributions and properties of water masses in the past. To accomplish this, we include a new representation of the stable carbon isotope 13C into the HAMburg Ocean Carbon Cycle model (HAMOCC), the ocean biogeochemical component of the Max Planck Institute Earth System Model (MPI-ESM). 13C is explicitly resolved for all existing oceanic carbon pools. We account for fractionation during air-sea gas exchange and for biological fractionation εp associated with photosynthetic carbon fixation during phytoplankton growth. We examine two εp parameterisations of different complexity: εpPopp varies with surface dissolved CO2 concentration (Popp et al., 1989), while εpLaws additionally depends on local phytoplankton growth rates (Laws et al., 1995). When compared to observations of δ13C in dissolved inorganic carbon (DIC), both parameterisations yield similar performance. However, with regard to δ13C in particulate organic carbon εpPopp shows a considerably improved performance than εpLaws, because the latter results in a too strong preference for 12C. The model also well reproduces the oceanic 13C Suess effect, i.e. the intrusion of the isotopically light anthropogenic CO2 into the ocean, based on comparison to other existing 13C models and to observation-based oceanic carbon uptake estimates over the industrial period. We further apply the approach of Eide et al. (2017a), who derived the first global oceanic 13C Suess effect estimate based on observations, to our model data that has ample spatial and temporal coverage. With this we are able to analyse in detail the underestimation of 13C Suess effect by this approach as it has been noted by Eide et al. (2017a). Based on our model we find underestimations of 13C Suess effect at 200 m by 0.24 ‰ in the Indian Ocean, 0.21 ‰ in the North Pacific, 0.26 ‰ in the South Pacific, 0.1 ‰ in the North Atlantic and 0.14 ‰ in the South Atlantic. We attribute the major sources of the underestimation to two assumptions in Eide et al. (2017a)'s approach: a spatially-constant preformed component of δ13CDIC in year 1940 and neglecting 13C Suess effect in CFC-12 free water.