Extended Data Fig. 1: Two alternative approaches to investigate the unexpected nature of the ocean carbon sink in 2023.

(a) The slope of a linear regression between the air-sea CO₂ flux anomalies and SST anomalies in previous decades (1990-2022), which was multiplied with the actual spatial distribution in the SST anomalies in 2023 (Fig. 1a) to obtain (b) a predicted CO₂ flux anomaly map for 2023 that considers only the SST anomaly. (b) The unexpected part of the global non-polar ocean carbon sink, obtained from comparing the actual observed sink strength in each year to an expected value obtained from (left panel) the linear regression baseline from Fig. 1b and the expected CO₂ flux anomaly based on the global mean SST anomaly and (right panel) a multiple linear regression model that was fitted with the atmospheric CO₂, atmospheric CO₂ growth rate and SST in the equatorial Pacific as predictor variables and for the annual mean ocean carbon sink from 1990 through 2022 as target variable. The unexpected CO₂ flux anomalies in (c) represent the mean (black line) and standard deviation (grey ribbon) across our ensemble of four fCO₂ products.