Ocean Biogeochemistry in GFDL’s Earth System Model 4 and its Response to Increasing Atmospheric CO2

Charles A Stock1, John P Dunne2, Songmiao Fan3, Paul A Ginoux4, Jasmin G John5, John P Krasting6, Charlotte Laufkötter7, Fabien Paulot1 and Niki Zadeh8, (1)NOAA Geophysical Fluid Dynamics Laboratory, Princeton, United States, (2)NOAA Geophys Fluid Dynamic, Princeton, United States, (3)Princeton Univ-NOAA GFDL, Princeton, NJ, United States, (4)NOAA/Geophysical Fluid Dynamics Laboratory, Princeton, NJ, United States, (5)NOAA/GFDL, Princeton, United States, (6)Geophysical Fluid Dynamics Laboratory, Princeton, NJ, United States, (7)University of Bern, Zürich, Switzerland, (8)NOAA/Geophysical Fluid Dynamics Laboratory, Princeton, United States
Abstract:
The Geophysical Fluid Dynamics Laboratory’s Earth System Model 4 (GFDL-ESM4) was developed to holistically study past and future earth system changes under different scenarios for natural and anthropogenic drivers (e.g., greenhouse gases). The response of the ocean’s vast carbon and heat reservoirs to such forcing greatly reduces atmospheric and terrestrial impacts. Provision of this service, however, profoundly effects ocean environments and the living marine resources they support. We describe the formulation of ocean biogeochemical dynamics within GFDL-ESM4 and show that simulations capture many aspects of observed seasonal, inter-annual and century-scale trends in large-scale carbon and nutrient cycles. We then characterize the response of the biological pump and ecosystem properties (oxygen, acidity, productivity) to CO2 accumulation, highlighting positive and negative feedbacks. Finally, we present two prominent limitations of GFDL-ESM4: over-estimation of the vast hypoxic region in the Eastern Equatorial Pacific, and over-expression of phosphate limitation in the Atlantic and Indian ocean. The causes of these biases, their consequences for the ocean’s response to global change, and potential remedies linked to advances in ocean physical and ocean ecosystem modeling will be discussed.