Multidecadal change in the dissolved inorganic carbon in a long‐term ocean state estimation. (23rd November 2015)

• Record Type: Journal Article
• Authors: Doi, Toshimasa; Osafune, Satoshi; Sugiura, Nozomi; Kouketsu, Shinya; Murata, Akihiko; Masuda, Shuhei; Toyoda, Takahiro
• Is Part Of: Journal of advances in modeling earth systems. Volume 7:Number 4(2015:Dec.)
• Page Start: 1885
• Page End: 1900
• Publication Date: 2015-11-23
• Subjects: global ocean -- carbon cycle -- decadal change -- ecosystem model -- 4‐D‐VAR; Geological modeling -- Periodicals; Climatology -- Periodicals; Geochemical modeling -- Periodicals; 551.5011
• DOI: 10.1002/2015MS000462
• Languages: English
• ISSNs: 1942-2466

Abstract

Abstract: By using a four‐dimensional variational data assimilation system capable of estimating physical and biogeochemical variables for the global ocean, we investigated multidecadal changes in the dissolved inorganic carbon (DIC) in the ocean. The system was newly constructed with a pelagic ecosystem model and an oceanic general circulation model to integrate available ocean observations obtained with a wide range of observation tools. The distribution of estimated DIC was by and large consistent with previous reports. We validated the changes in DIC along the World Ocean Circulation Experiment (WOCE) Hydrographic Program sections. The correlation coefficients of the modeled versus observed decadal difference patterns of DIC ranged from 0.25 to 0.51 in the Pacific Ocean, from 0.36 to 0.62 in the Atlantic Ocean, and from 0.23 to 0.57 in the Indian Ocean, and were significant at the 95% confidence level. Thus, at basin scale, the reproducibility of long‐term climate change was similar. Estimation of vertical DIC fluxes in each basin showed that the fluxes changed on a multidecadal time scale in our system. These changes were possibly due to changes in the dynamical state of CO2 absorption and to changes in ocean circulation. Our integrated data set on the basis of a dynamically self‐consistent ocean state is a promising tool for examining long‐term changes in the ocean carbon cycle. Key Points: The multidecadal ocean state, including DIC distribution, was estimated Parameters for biological variables were optimized by Green's function approach Multidecadal change in the dissolved inorganic carbon was identified