A translator in a conference hall tries to convey the speaker's words, but the hall is noisy: someone coughs, chairs creak, and the street outside is loud. If the speaker lowers their voice, the translator still keeps translating, but their voice drowns in the noise. The atmosphere is like that: it "translates" our carbon dioxide emissions into rising concentrations, but that signal is drowned out by the breathing of forests, the ocean, and global climate swings.
Researchers took data from 2015 to 2024 and aligned it to a common grid of 1°×1°. Atmospheric CO₂ growth was calculated from the CAMS reanalysis, which provides three-hourly fields of XCO₂ (the column-averaged dry-air mole fraction of carbon dioxide). After weekly averaging and fitting a dynamic linear model, they obtained annual growth rates. Anthropogenic emissions were taken from the EDGAR database, biosphere activity was estimated from solar-induced fluorescence GOSIF (the glow of plants that reflects photosynthesis), and climate variability was represented by the Southern Oscillation Index with a seven-month lag. For each land cell, they built linear regressions where CO₂ growth was explained by local emissions and either SOI or GOSIF. Then they applied K-means clustering to the pairs of "emissions–fluorescence" and, using stability analysis, identified five main carbon regimes that persisted over the decade.
The clustering identified five regimes, ranging from nearly inactive polar regions to anthropogenic hotspots and active biospheric sinks. When averaging within clusters, unique regional features are smoothed out, and the global climate signal remains the main driver almost everywhere. The exception is the active biosphere cluster, where the biogenic signal is so strong that it survives averaging. The model with SOI explained 43% to 89% of the variance within clusters, whereas the model with GOSIF performed significantly worse, especially in cross-validation.
For climate policy, this means that verifying emission reduction commitments cannot rely solely on atmospheric observations without considering the regional context. If the translator cannot hear the speaker's quiet remark, it does not mean the speaker is silent: it is just that the noise in the hall is too strong. To understand who is saying what and how much, one must account for the acoustics of each hall.
The Southern Oscillation, which the authors use as a climate driver, is measured by the difference in atmospheric pressure between Tahiti and Darwin. This index was proposed back in 1987 and remains one of the simplest and most reliable indicators of El Niño.