📄 earth_science

An unusually delayed and persistent atmospheric CO₂ response to the 2023/24 El Niño

This study reveals that the 2023/24 El Niño triggered an unusually delayed and persistent atmospheric CO₂ growth response driven by a prolonged land–atmosphere carbon imbalance, which resulted in a cumulative carbon anomaly twice as large as that of the 2015/16 event despite weaker initial forcing.

Original authors: Xiaoyu Sun, Denghui Ji, Sharon Patris, Darko Dubravica, Mathias Palm, Frank Hase, Minqiang Zhou, Marcus Rex, Justus Notholt

Published 2026-08-07

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Xiaoyu Sun, Denghui Ji, Sharon Patris, Darko Dubravica, Mathias Palm, Frank Hase, Minqiang Zhou, Marcus Rex, Justus Notholt

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the Earth's atmosphere as a giant, invisible bathtub. Normally, water flows in from the tap (carbon dioxide released by people burning fuel) and drains out through the pipes (plants and oceans soaking it up). The water level rises slowly and steadily. But sometimes, the weather gets weird. Every few years, the Pacific Ocean throws a tantrum called an El Niño. Think of this as a giant, warm hug from the ocean that changes the weather everywhere. When this happens, the "drain pipes" of the planet—especially the forests and the ocean surface—start to clog up or even reverse, spitting carbon back into the air instead of swallowing it. Scientists have long known that these weather tantrums cause the water level in our atmospheric bathtub to rise faster for a while. The big question is: how long does that extra water stay there, and why does it sometimes linger much longer than the tantrum itself?

This paper investigates a very recent, strange tantrum that happened in 2023 and 2024. The researchers, who act like cosmic detectives tracking carbon dioxide, found something surprising. Usually, when the El Niño "peak" (the hottest, most intense part of the event) passes, the extra carbon in the air starts to calm down pretty quickly. But for the 2023/24 event, the atmosphere acted like a stubborn teenager who refuses to go to bed even after the party is over. The carbon levels didn't just spike; they stayed high for a shockingly long time, long after the ocean had cooled down and the weather had returned to normal.

The team compared this recent event to the two biggest "super El Niños" of the past few decades (1997/98 and 2015/16). They expected the 2023/24 event to be weaker because the ocean warming wasn't as extreme as those previous giants. And they were right: the ocean didn't release as much carbon, and there were fewer massive forest fires than in 2015/16. So, why was the carbon in the air so high and so persistent?

The answer, the paper suggests, lies in the land. While the ocean and fires were quiet, the land was acting up. The researchers found that the forests and soils didn't just stop absorbing carbon; they kept releasing it for months after the El Niño peak. It's as if the land was holding its breath during the storm and then let out a giant, slow sigh of carbon dioxide long after the storm clouds had cleared.

Here is what the data actually shows:

The Delay: In previous big events, the carbon growth peaked just 2 to 4 months after the ocean got hottest. In 2023/24, the peak didn't happen until seven months after the ocean peak.
The Duration: The carbon levels stayed above half of their maximum height for 10 months in 2023/24. That's much longer than the 6 months seen in 2015/16.
The Amount: Even though the ocean warming was weaker, the total extra carbon that piled up in the atmosphere (the "cumulative anomaly") was about twice as big as the 2015/16 event.
The Cause: The paper explicitly rules out the ocean and fires as the main culprits for this delay. The ocean signals were actually weaker than before, and the fire signals were much smaller. Instead, the evidence points to a "persistent late-phase imbalance" in the land. The land kept spitting out carbon long after the event should have ended.

The researchers used a clever method of lining up all the El Niño events from 1980 to the present to see the patterns clearly. They checked their findings against multiple different computer models and satellite data to make sure they weren't just seeing a glitch. While they can't say exactly which specific tree or patch of soil was responsible (the data mixes everything together), they are very confident that the land was the reason the carbon stayed in the air so long.

In short, this paper tells us that the Earth's carbon cycle is more complicated than we thought. A weather event might end, but the land's reaction can drag on for months, keeping our atmospheric bathtub fuller for longer. This suggests that as the planet gets warmer, these "hangover" effects might become more common, making it harder for us to slow down the rise of carbon dioxide in the air.

Technical Summary: An Unusually Delayed and Persistent Atmospheric CO₂ Response to the 2023/24 El Niño

Problem Statement
The El Niño–Southern Oscillation (ENSO) is a primary driver of interannual variability in the global carbon cycle, modulating atmospheric CO₂ accumulation through changes in tropical temperature, precipitation, and ecosystem productivity. While the general "ocean-first, land-later" sequence of the ENSO–CO₂ response is established, the timing, persistence, and cumulative magnitude of the post-peak atmospheric response have not been systematically quantified across multiple events. The 2023/24 El Niño presented a unique case: it occurred under exceptionally warm global background conditions and was followed by record atmospheric CO₂ growth, yet its peak Oceanic Niño Index (ONI) forcing was significantly weaker than the super El Niño events of 1997/98 and 2015/16. The central problem addressed is whether the peak amplitude of ENSO forcing determines the timing and integrated magnitude of the atmospheric CO₂ response, and specifically, what mechanisms drove the unusually delayed and persistent CO₂ accumulation observed in 2023/24.

Methodology
The authors developed an event-aligned, observation-based framework to analyze 12 El Niño events since 1980, with a focus on six strong events (peak ONI ≥ 1.5°C). The study utilized a four-decade surface CO₂ record and several specific diagnostic metrics:

Event Alignment: All records were aligned to the central month of the maximum ONI season (lag 0).
Atmospheric Metrics: Three primary diagnostics were calculated for detrended CO₂ growth anomalies:
1. Peak Lag: The time (in months) from the ONI peak to the maximum CO₂ growth anomaly.
2. Persistence ( $D_{0.50}$ ): The duration the anomaly remained above 50% of its peak maximum.
3. Cumulative Anomaly ( $I_{18}$ ): The integrated detrended anomaly over the 18 months following the peak.
Data Sources:
- Atmospheric CO₂: NOAA global mean and Mauna Loa records, supplemented by American Samoa (SAM), Cape Kumukahi (KUM), and new EM27/SUN observations from Palau.
- Satellite Data: GOSAT XCO₂ retrievals for column-averaged CO₂ anomalies.
- Oceanic Data: Surface-ocean pCO₂ fields (Copernicus Marine Service) and TAO buoy observations to assess early oceanic contributions.
- Fire Data: IASI total-column CO anomalies to evaluate biomass burning contributions.
- Land Flux: Net Biosphere Exchange (NBE) anomalies derived from the CAMS v24r1 inversion system and cross-validated against 13 additional atmospheric inversion products from the Global Carbon Budget 2025.

Key Results

Delayed and Persistent Response: The 2023/24 event exhibited a CO₂ growth anomaly that peaked 7 months after the ENSO maximum, later than any other event studied (e.g., +4 months for 1997/98, +2 months for 2015/16). The anomaly remained above half its maximum for 10 months, the second-longest duration recorded.
Disproportionate Cumulative Impact: Despite a weaker peak ONI, the cumulative 18-month detrended anomaly for 2023/24 was 0.885 ppm, approximately twice that of the 2015/16 event (0.431 ppm). This indicates that peak ENSO amplitude alone does not dictate the integrated atmospheric response.
Exclusion of Ocean and Fire Drivers:
- Ocean: The early surface-ocean pCO₂ reduction in 2023/24 was weaker (minimum ~~−15 μatm) than in 2015/16 (~~−43 μatm), ruling out a stronger oceanic suppression of outgassing as the cause of the delayed peak.
- Fire: Fire-related CO anomalies in the western Pacific were substantially weaker in 2023/24 compared to the pronounced Indonesian burning episode of 2015/16.
Land-Flux Attribution: The delayed atmospheric response coincided with a sustained and spatially distributed positive land NBE anomaly (net carbon release).
- Unlike 2015/16, where land anomalies weakened in the late phase, the 2023/24 event showed a reversal where tropical land NBE increased from the early to the late post-peak phase.
- This shift was robust across 13 independent atmospheric inversion products, all of which showed a positive global late-minus-early land flux contrast for 2023/24. The anomaly extended from the tropics into extra-tropical regions.

Key Contributions

Quantification of Post-Peak Persistence: The study introduces a systematic framework to quantify the timing and duration of the atmospheric CO₂ response relative to ENSO peaks, moving beyond simple annual growth rate comparisons.
Decoupling of Forcing and Response: It demonstrates that a weaker peak ENSO forcing can result in a larger and more persistent atmospheric CO₂ accumulation if the underlying land-carbon imbalance is prolonged.
Identification of a "Late-Phase" Land Signal: The research identifies a specific "late-phase" land-carbon anomaly that persists well beyond the ENSO peak, contrasting with the more rapid recovery observed in previous events. This signal is spatially coherent across multiple land regions and inversion systems.

Significance and Claims
The paper claims that the persistence and cumulative magnitude of the atmospheric CO₂ anomaly serve as a system-level record of how long climate-driven carbon imbalances endure. The authors conclude that the 2023/24 event was characterized by a prolonged post-peak atmospheric response driven by a persistent late-phase imbalance in net land–atmosphere carbon exchange.

The study posits that this delayed recovery, potentially linked to asynchronous recovery of photosynthesis and respiration or water stress, can prolong CO₂ accumulation and may amplify positive carbon–climate feedbacks. The authors emphasize that while they do not uniquely attribute the delay to specific processes (e.g., specific respiration vs. photosynthesis mechanisms), the temporal consistency between the atmospheric tail and the land-flux inversions provides robust evidence that land-carbon recovery can be significantly delayed in a warming climate. The findings suggest that peak ENSO amplitude is an insufficient predictor of future carbon-cycle impacts, and that the duration of the post-peak carbon imbalance is a critical, previously under-quantified control on atmospheric CO₂ accumulation.

Technical Summary: An Unusually Delayed and Persistent Atmospheric CO₂ Response to the 2023/24 El Niño

More like this