Skip to main content
U.S. flag

An official website of the United States government

Evaluating key uncertainties in ocean carbon cycle processes within CMIP6 models using the International Ocean Model Benchmarking package (IOMB)

Presentation Date
Friday, December 13, 2019 at 1:40pm
Location
Moscone South Poster Hall
Authors

Author

Abstract

Evaluating the representation of the ocean carbon cycle in Earth System Models is a fundamental task necessary for quantifying and reducing uncertainty in projections of future climate change. Here we used the International Ocean Model Benchmarking (IOMB) package to comprehensively assess the uncertainties in ocean carbon uptake and storage predicted by Earth System Models (ESMs) in the 6th Coupled Model Inter-comparison Project (CMIP6). We compared the air-sea CO2 flux, ocean dissolved inorganic carbon (DIC) concentration, transient tracers and physical variables in the historical runs with estimates from Ocean Carbon Inverse Modelling (OCIM) (DeVries, 2014) along with data products from Global Ocean Data Analysis Project version 2 (GLODAPv2) and the World Ocean Atlas (WOA). The CMIP6 models predicted a cumulative anthropogenic carbon uptake of 117-158 Pg C from 1850 to 2014. Most models had a low bias compared to the OCIM estimate after 1975. From 1994 to 2007, the CMIP6 models predicted a cumulative carbon uptake of 25.5-29.4 Pg C, which was considerably lower than the recent 34±4 Pg C estimate reported by Gruber et al. (2019) using GLODAPv2 data. Analysis of the DIC inventory revealed strong biases in the North Atlantic, Southern Ocean, North Pacific and Arctic Ocean. The evolving spatial variations of the bias suggested incomplete representation of processes regulating ocean circulation had a strong influence on ocean biogeochemistry. We used transient tracers such as chlorofluorocarbons (CFCs) to associate the uncertainties in ocean carbon cycle with ocean circulation. Comprehensive model-data benchmarking here could help to identify model deficiencies and subsequently lead to improved subgrid-scale parameterizations, thus accelerating ESM development.

Category
Biogeosciences
Funding Program Area(s)