Regional and seasonal partitioning of water and temperature controls on global land carbon uptake variability.

Global fluctuations in annual land carbon uptake (NEE) depend on water and temperature variability, yet debate remains about local and seasonal controls of the global dependences. Here, we quantify regional and seasonal contributions to the correlations of globally-averaged NEE against terrestrial water storage (TWS) and temperature, and respective uncertainties, using three approaches: atmospheric inversions, process-based vegetation models, and data-driven models. The three approaches agree that the tropics contribute over 63% of the global correlations, but differ on the dominant driver of the global NEE, because they disagree on seasonal temperature effects in the Northern Hemisphere (NH, >25°N). In the NH, inversions and process-based models show inter-seasonal compensation of temperature effects, inducing a global TWS dominance supported by observations. Data-driven models show weaker seasonal compensation, thereby estimating a global temperature dominance. We provide a roadmap to fully understand drivers of global NEE and discuss their implications for future carbon-climate feedbacks.