Pre-Industrial Spin-Up of Dissolved Oxygen in the NASA GISS Model E2.1, with Emphasis on the Major ODZs.

Paul Lerner1, Anastasia Romanou2, David P Nicholson3, Maxwell Kelley4 and Reto Ruedy4, (1)NASA Goddard Institute for Space Studies, New York, NY, United States, (2)NASA Goddard Institute for Space Studies, and Dept of Applied Phys. and Applied Math., Columbia University, New York, NY, United States, (3)Woods Hole Oceanographic Inst., Woods Hole, MA, United States, (4)NASA Goddard Institute for Space Studies, New York City, NY, United States
The inventory of marine dissolved oxygen has declined by ~2% over the past several decades, due a combination of the decreasing solubility of O2 and changes in ocean dynamics associated with increasing ocean temperature. This deoxygenation has resulted in an expansion of tropical oxygen deficient zones (ODZs). Aside from the restricting the habitable zones of many marine organisms, expanding ODZs can have important impacts on the ocean carbon cycle, limiting the rate at which carbon is transferred from organic to inorganic forms. The cycling of marine nitrogen, iron, and other elements whose speciation depends on dissolved oxygen concentration, may also be affected by changing O2 concentrations.

In this poster, we present the pre-industrial (750 yr) spin-up of dissolved oxygen in simulations of the NASA GISS modelE2.1, a coupled atmosphere-ocean model with 1o x1.25o resolution for the ocean. In this model, the O2 dynamics are driven by air-sea gas exchange, photosynthetic production, detrital degradation, autotrophic and heterotrophic respiration, and bacterial degradation of dissolved organic carbon. Due to the spin-up of the model being relatively short (<1000 yrs), we restrict our comparison of the model O2 field to GLODAPv2 and WOA2013 climatologies in the upper ocean (top 500 m). The model skill for the upper ocean is generally high (RMSE_GLODAP = 0.021, RMSE_WOA2013 = 0.022; compared to the average upper ocean O2 concentration of 0.232 mmol/kg). There is regional variability in the ability of the model to fit these climatologies, e.g. the model performs better in the north Atlantic (RMSE_GLODAP = 0.018, RMSE_WOA2013 = 0.019) than in the north Pacific (RMSE_GLODAP = 0.021, RMSE_WOA2013 = 0.026). Focusing on the ODZs, the model is able to reproduce the general structure of the equatorial pacific ODZ, contained between 100-500 m in the tropical South Pacific and 100-1000 m in the tropical North Pacific in both the model and in observations, although their lateral extent is smaller than those observed. On the other hand, the model Arabian OMZ is constrained to a very narrow depth range between 350-700 m, compared to that observed in GLODAPv2 of 100-1000 m. Work is ongoing to determine the leading cause of the of the ODZ volume biases in both regions.