Decadal changes of terrestrial and ocean CO2 fluxes inferred by multiple top-down and bottom-up approaches

Understanding surface CO₂ flux variations is imperative for better projection of climate change and carbon-climate feedback and many top-down and bottom-up approaches have been employed to estimate surface fluxes. However, large uncertainties exist in each estimate due to limited knowledge of flux processes and sparse observation networks. In this study, we combined multiple top-down and bottom-up approaches to infer global terrestrial and ocean CO₂ fluxes. These approaches use a flux inversion system named NISMON-CO₂, the terrestrial biosphere model VISIT, atmospheric oxygen and stable carbon isotope measurements conducted by NIES, JMA and NIES ocean CO₂ flux mapping based on ship-board measurements, and the ocean biogeochemistry model based on the MRI Earth System Model ver.2. After adjustment of lateral fluxes, those multiple flux estimates agree with each other in the global land-ocean flux partitioning within ca. 0.3 Pg C yr^-1. Furthermore, their decadal changes from 2000 have similar trends. The ocean uptake is gradually increasing; the average of the global ocean uptake estimates is 2.37±0.16 Pg C yr^-1 for 2001–2010, 2.62±0.26 Pg C yr^-1 for 2006­­–2015, and 2.87±0.28 Pg C yr^-1 for 2011–2020. Meanwhile, the terrestrial uptake first increased from 1.49±0.09 Pg C yr^-1 for 2001­–2010 to 1.92±0.29 Pg C yr^-1 for 2006–2015, then decreased to 1.66±0.30 Pg C yr^-1 for 2011­­–2020. These features are commonly seen in flux estimates from the independent sources, including a sensitivity test of flux inversion by NISMON-CO₂ with a climatological (i.e., not year-by-year varying) prior flux. This agreement indicates that these approaches can form a good support for improving climate projection.