An intermediate‐complexity model for simulating marine biogeochemistry in deep time: Validation against the modern global ocean

Abstract

We present a new high‐resolution 1‐D intermediate‐complexity box model (ICBM) of ocean biogeochemical processes for paleoceanographic applications. The model contains 79 reservoirs in three regions that should be generally applicable throughout much of Earth history: (1) a stratified gyre region, (2) a high‐latitude convective region, and (3) an upwelling region analogous to those found associated with eastern boundary currents. Transport processes are modeled as exchange fluxes between boxes and by eddy diffusion terms. Significant improvement in the representation of middepth oxygen budgets was achieved by implementing nonlocal mixing between the high‐latitude surface and gyre thermocline reservoirs. The biogeochemical submodel simulates coupled C, N, P, O, and S systematics with explicit representation of microbial populations, using a process‐based approach. Primary production follows Redfield stoichiometry, while water column remineralization is depth‐ and redox couple–dependent. Settling particulate organic matter is incorporated into a benthic submodel that accounts for burial and remineralization. The C/P ratio of burial depends on bottom water oxygen. Denitrification takes place both by classical and anammox pathways. The ICBM was tested against modern oceanographic observations from the Global Ocean Data Analysis Project, Joint Global Ocean Flux Study, and other databases. Comparisons of model output with circulation tracers including θ, salinity, CFC‐12, and radiocarbon permit a test of the physical exchange scheme. Vertical profiles of biogeochemically reactive components in each of the three regions are in good agreement with observations. Under modern conditions the upwelling zone displays a pronounced oxygen minimum zone and water column denitrification, while these are not present in the high‐latitude or gyre regions. Model‐generated global fluxes also compare well to independent estimates of primary production, burial, and phosphorous and nitrogen cycling. The ICBM appears to adequately simulate the long‐term (kyr) evolution of several biogeochemical cycles and improves on previous box models in several important ways. In a companion paper, the model's performance under euxinic conditions is tested against modern Black Sea data. The simple and adaptable structure of the model should make it applicable to a wide range of paleoceanographic problems. The model source code is available in MATLABTM 7 m‐files provided as auxiliary material.