A GOSAT-Based Assessment of Interannual Variability in Terrestrial CO₂ Fluxes and Its Hydroclimatic Drivers

Abstract. Understanding interannual variability (IAV) in terrestrial CO₂ fluxes is essential for constraining ecosystem responses to climatic anomalies and future carbon-climate feedbacks. A 14-year GOSAT record spanning 2009 to end of 2022 now enables a robust global assessment of regional IAV. Here, we analyse annual net land CO₂ fluxes from nine in situ-constrained inversions and three GOSAT-based inversions, and compare them with the TRENDY v13 ensemble mean of dynamic global vegetation models (DGVMs). We find that DGVMs show weaker IAV than the inversion-based flux estimates in almost all regions. In Australia, boreal and temperate North America, and Eurasian Temperate, differences in NEE anomalies between GOSAT-based inversions and TRENDY are strongly related to hydroclimatic anomalies, with soil moisture and temperature emerging as dominant drivers in different regions. These results suggest that DGVMs underestimate the sensitivity of NEE to climate variability in key regions. The extended satellite record makes it possible to use the hydroclimatic anomalies of the last 14 years to identify deficiencies in ecosystem parameterizations and thereby reduce uncertainty in the global carbon budget.