A towering rain cloud lit by the late-afternoon sun rises above the rainforest canopy in the Mamirauá reserve, Amazonas, Brazil
Global Carbon Project
Laboratoire des Sciences du Climat et de l'Environnement
THRAC3E
CALIPSO
Future Earth
Carbon Monitor

Key Results from Budget 2025

Photo: Adriana Brandao (Adrisub), CC BY-SA 4.0

These are preliminary results. The paper describing this budget is in review.

The main findings for 2025.

A strong rebound of the land sink slowed the rise of atmospheric CO2 in 2025

The CO2 growth rate fell by 45% from the 2024 record, even though fossil emissions rose to a new high. Bottom-up models and four OCO-2 inversions agree that the land sink rebounded strongly after the 2023–2024 weakening, while the ocean sink changed little. The recovery was strongest in the tropics and incomplete across the northern extratropics.

  1. The CO2 growth rate from the NOAA marine boundary layer network fell by 45%, from a record 3.76 ± 0.09 ppm yr−1 in 2024 to 2.06 ± 0.09 ppm yr−1 in 2025, below the 2015–2022 mean of 2.47 ppm yr−1, even as fossil CO2 emissions rose by 0.7% to 10.38 GtC yr−1.
  2. The global net land sink rebounded to 2.36 ± 0.16 GtC yr−1, a strengthening of 2.81 ± 0.31 GtC yr−1 from 2024, and 0.71 ± 0.13 GtC yr−1 above the 2015–2022 mean.
  3. The ocean sink (3.11 ± 0.36 GtC yr−1) was similar to 2024: the rebound of the land sink was the dominant change behind the drop in CO2 growth.
  4. Where the land sink had weakened substantially in 2023–2024, about 80% of the land area showed some recovery in 2025; carbon-weighted recovery reached 87.3% (DGVMs) and 99.5% (inversions) of the earlier loss.
  5. Recovery exceeded 100% in the tropics (106.4% DGVMs, 116.4% inversions) but remained incomplete in the northern extratropics (59.6% and 90.7%).

The budget at a glance

QuantityJanuary–December 2025Notes
Atmospheric CO2 growth rate2.06 ± 0.09 ppm yr−1NOAA marine boundary layer; 3.76 ± 0.09 in 2024 (−45%)
Fossil CO2 emissions10.38 GtC yr−1+0.7% compared with 2024
Net land sink2.36 ± 0.16 GtC yr−1Mean of DGVMs (2.04 ± 0.24) and inversions (2.68 ± 0.20); +2.81 ± 0.31 compared with 2024
Ocean sink3.11 ± 0.36 GtC yr−1Emulators 3.15 ± 0.65; inversions 3.06 ± 0.32
Budget imbalance0.81 ± 0.69 GtC yr−195% interval −0.66 to 1.91, not significantly different from zero

Uncertainties are ±1σ; changes and anomalies use member-wise paired differences; the imbalance uncertainty comes from Monte Carlo propagation.

Models and data used

  • Fossil CO2 emissions. Carbon Monitor and Global Carbon Budget 2025 projections (range 10.34–10.42 GtC yr−1)
  • Atmospheric growth rate. NOAA MBL; GRESO (OCO-2 and GOSAT) and inversion-implied growth as cross-checks
  • Land, bottom-up. 3 DGVMs driven by ERA5: JULES, ORCHIDEE-MICT, LPJ-EOSIM; fire emissions from GFED4.1s and GFAS
  • Ocean, bottom-up. Machine-learning emulators of 10 ocean models and 9 data products (GCB 2025)
  • Top-down. 4 OCO-2 inversions: CAMS, CMS-Flux, CTE, GONGGA, with a riverine transport correction
  • Anomaly reference. 2015–2022

Publication

Ciais, P., Ke, P., Tian, X., Sitch, S., Li, W., Du, X., et al. (2026). Low latency global carbon budget reveals strong land sink recovery in 2025. Preprint, arXiv:2609.34226. In review at National Science Review. https://arxiv.org/abs/2609.34226

Back to the 2025 budget