A Streamlined, Mechanistic Particle Flux Model Designed For Earth System Models Of Intermediate Complexity: A Global Assessment Of Particle Transfer Efficiency To The Deep Ocean

The "transfer efficiency" of sinking particles through the mesopelagic zone and into the deep ocean is a key factor in controlling the atmosphere-ocean partitioning of CO₂, and has important implications for the ocean carbon cycle under glacial climate conditions or in a future warming climate. However, sinking particle fluxes are poorly understood and not well sampled, and earth system models often parameterize these fluxes as an empirical function of depth. Here we present preliminary results from a streamlined, mechanistic particle flux model that explicitly represents the effects of particle size, mineral ballasting, seawater temperature, and dissolved oxygen on particle sinking speeds and POC remineralization rates. Running at the same vertical resolution as the Bern3D earth system model of intermediate complexity, the particle flux model carries at least two tracers including two particle size classes (small and large particles). The POC remineralization rate depends on temperature, dissolved oxygen, and physical protection by mineral ballast content, whereas the particle sinking speed depends on particle size, particle density, and seawater viscosity following a modified and normalized version of Stokes’ law. We show that the predicted global distribution of the transfer efficiency of POC (T_eff) to the base of the mesopelagic zone is characterized by high values (~25%) in high-latitude, opal-dominated regions and low values (~5%) in low-latitude, CaCO₃-dominated regions. This latitudinal pattern is consistent with conclusions reached by upper-ocean, neutrally buoyant sediment trap studies as well as modeling studies, whereas an opposite pattern has been diagnosed from combined analyses of deep-sea sediment trap and thorium- and satellite-derived export flux data. Nonetheless, the robust pattern of predicted spatially varying T_eff suggests that one should use a process-based particle flux attenuation model in earth system models. In contrast to the commonly used, empirically derived “Martin curve” relationship, such a model allows the investigation of past and future changes in the environmental conditions influencing T_eff. For these reasons, the particle flux model presented here will be implemented in the Bern3D model and its application will be presented in future studies.