pith. sign in

arxiv: 2606.28462 · v1 · pith:KQWD4O7Gnew · submitted 2026-06-26 · ⚛️ physics.ao-ph · physics.data-an

Investigation of regional variations in CO₂ growth rates : Integrating Emission Inventories and Atmospheric Observations

Yogesh Bali , Darja Cvetkovi\'c , Juan Gancio , Adri\'an Guti\'errez-Arroyo , Sofia Vazquez Alferez , Xuan Tung Vu , Jin Yan , Pietro Zgaga
show 2 more authors
Fakhteh Ghanbarnejad Nasrin Mostafavi Pak
This is my paper

Pith reviewed 2026-06-30 01:15 UTC · model grok-4.3

classification ⚛️ physics.ao-ph physics.data-an
keywords CO2 growth ratesregional variationsanthropogenic emissionsnatural variabilitycarbon cycle regimestop-down analysisatmospheric observationsemission inventories
0
0 comments X

The pith

Regional atmospheric CO2 growth rates are dominated by natural carbon-cycle processes that mask anthropogenic emission signals.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper integrates atmospheric reanalysis with emission inventories, biospheric activity measures, and a climate variability index to examine how CO2 growth rates vary across regions. It establishes that these rates are controlled mainly by natural carbon-cycle processes and global background trends rather than local human emissions. This dominance makes it difficult to detect changes in anthropogenic emissions from atmospheric data alone. The absence of consistent regional signals from 2020 emission reductions, even in a neutral climate-variability year, illustrates the masking effect. Clustering analysis further identifies five regimes in which spatial averaging leaves large-scale climate as the main driver except in zones of strong biogenic activity.

Core claim

Atmospheric CO2 growth rate varies substantially across space and time but is dominated by natural carbon-cycle processes and global background trends. Anthropogenic emission signals are frequently masked by natural variability, making regional top-down detection of human emission changes difficult. The COVID-19 emission reductions in 2020, despite occurring during a neutral ENSO year, were not consistently reflected in regional atmospheric CO2 growth rates. Using unsupervised clustering and persistence analysis, five characteristic carbon-cycle regimes are identified, with spatial averaging removing much of the regional variability and leaving large-scale climate as the dominant control in

What carries the argument

Unsupervised clustering and persistence analysis applied to combined atmospheric, emission, biospheric, and climate-variability data to identify five characteristic carbon-cycle regimes.

If this is right

  • Regional top-down detection of changes in human emissions is limited by the masking effect of natural variability.
  • Spatial averaging of data shifts emphasis to large-scale climate controls in most identified regimes.
  • Strong biogenic signals persist and dominate in active biosphere zones such as tropical forests.
  • Global background trends exert primary control once local variability is averaged out.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Strategies for emission verification may need to prioritize regions where natural signals are weaker or more predictable.
  • Higher-resolution transport modeling could help isolate anthropogenic contributions in future analyses.
  • The identified regimes suggest that monitoring networks focused on tropical areas might capture more persistent biogenic influences on CO2.

Load-bearing premise

The atmospheric reanalysis, anthropogenic emission inventory, biospheric activity dataset, and climate variability index each accurately capture their target components without systematic biases large enough to change the clustering outcomes or the conclusion that natural signals mask anthropogenic ones.

What would settle it

Observation that regional atmospheric CO2 growth rates in 2020 showed reductions matching the scale and pattern of COVID-related emission decreases across multiple clusters.

Figures

Figures reproduced from arXiv: 2606.28462 by Adri\'an Guti\'errez-Arroyo, Darja Cvetkovi\'c, Fakhteh Ghanbarnejad, Jin Yan, Juan Gancio, Nasrin Mostafavi Pak, Pietro Zgaga, Sofia Vazquez Alferez, Xuan Tung Vu, Yogesh Bali.

Figure 1
Figure 1. Figure 1: Global CO2 growth rate in 2019 estimated from deseasonalized DLM-fitted XCO2. Colors show annual growth rate in ppm yr−1 ; the global mean was 2.58 ppm yr−1 . In addition, we classify the different regions of the globe according to the two main drivers on CO2 growth rates, which can be linked to specific Earth systems and are related to specific stages of the carbon cycle. The remainder of this paper is or… view at source ↗
Figure 2
Figure 2. Figure 2: Summary of the persistence analysis of the clusters. a) Snapshots of the input data , DSe and DSb, from [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Observed and predicted CO2 growth rates for selected years. First column: observed growth rates from the [DSg] dataset. Second column: predicted growth rates from an ordinary least squares (OLS) linear regression on local emission [DSe] and the lagged, global annual SOI index [DSs], which agrees closely with the observed rates despite the model’s simplicity. Third column: residuals (observed − predicted). … view at source ↗
Figure 4
Figure 4. Figure 4: Year-by-year evolution of the observed (with error bars) and predicted CO [PITH_FULL_IMAGE:figures/full_fig_p010_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Five largest persistent clusters over the ten-year period (C1–C5). They are plotted using a normalized color [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Mean value of human-driven emissions (DSe) and bio-driven emissions (DSb) in the different clusters [PITH_FULL_IMAGE:figures/full_fig_p012_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Annual predictor contributions to modeled CO [PITH_FULL_IMAGE:figures/full_fig_p013_7.png] view at source ↗
Figure 1
Figure 1. Figure 1: Gridded regression coefficients and goodness of fit for the local growth-rate model. Upper left: β0(r), the local baseline growth rate (ppm yr−1 ). Upper right: βS(r), the ENSO sensitivity — the change in local growth rate (ppm yr−1 ) associated with a one-standard-deviation increase in the lagged SOI index. Lower left: βE(r), the emission sensitivity — the change in local growth rate (ppm yr−1 ) associate… view at source ↗
Figure 2
Figure 2. Figure 2: Gridded regression coefficients and goodness of fit for the local growth-rate model. Upper left: β0(r), the local baseline growth rate (ppm yr−1 ). Upper right: βS(r), the ENSO sensitivity — the change in local growth rate (ppm yr−1 ) associated with a one-standard-deviation increase in the lagged SOI index. Lower left: βE(r), the emission sensitivity — the change in local growth rate (ppm yr−1 ) associate… view at source ↗
Figure 3
Figure 3. Figure 3: Gridded regression coefficients and goodness of fit for the local growth-rate model. Upper left: β0(r), the local baseline growth rate (ppm yr−1 ). Upper right: βS(r), the GOSIF sensitivity — the change in local growth rate (ppm yr−1 ) associated with a one-standard-deviation increase in the GOSIF index. Lower left: βE(r), the emission sensitivity — the change in local growth rate (ppm yr−1 ) associated wi… view at source ↗
Figure 4
Figure 4. Figure 4: Gridded regression coefficients and goodness of fit for the local growth-rate model. Upper left: β0(r), the local baseline growth rate (ppm yr−1 ). Upper right: βS(r), the GOSIF sensitivity — the change in local growth rate (ppm yr−1 ) associated with a one-standard-deviation increase in the GOSIF index. Lower left: βE(r), the emission sensitivity — the change in local growth rate (ppm yr−1 ) associated wi… view at source ↗
Figure 5
Figure 5. Figure 5: Gridded regression coefficients and goodness of fit for the local growth-rate model using ∆emission and ∆SIF. Upper left: β0(r), the local baseline growth rate (ppm yr−1 ). Upper right: β∆E(r), the emission-change sensitivity — the change in local growth rate (ppm yr−1 ) associated with a one-standard-deviation increase in ∆emission. Lower left: β∆B(r), the biospheric-change sensitivity — the change in loc… view at source ↗
Figure 6
Figure 6. Figure 6: Year-by-year evolution of the observed (with error bars) and predicted CO [PITH_FULL_IMAGE:figures/full_fig_p026_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Elbow method: Intra-cluster variance as a function of [PITH_FULL_IMAGE:figures/full_fig_p027_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Silhouette method: Silhouette score as a function of [PITH_FULL_IMAGE:figures/full_fig_p028_8.png] view at source ↗
read the original abstract

Atmospheric carbon dioxide (CO2) growth rates reflects the combined influence of anthropogenic emissions, biospheric carbon exchange, and climate variability. While climate mitigation is primarily evaluated using bottom-up emission inventories within political boundaries, there is a need to validate these emission reductions using atmospheric measurements. Here, we present a global top-down analysis of atmospheric CO2 growth rates using CAMS atmospheric CO2 reanalysis, EDGAR anthropogenic emissions, GOSIF dataset and the Southern Oscillation Index (SOI) as a measures of biospheric activity, to quantify the relative influence of human and natural drivers. We find that atmospheric CO2 growth rate varies substantially across space and time but is dominated by natural carbon-cycle processes and global background trends. Anthropogenic emission signals are frequently masked by natural variability, making regional top-down detection of human emission changes difficult. The COVID-19 emission reductions in 2020, despite occurring during a neutral ENSO year, were not consistently reflected in regional atmospheric CO2 growth rates, highlighting the dominant roles of biospheric dynamics and atmospheric transport. Using unsupervised clustering and persistence analysis, we identify five characteristic carbon-cycle regimes. Spatial averaging removes much of the regional variability, leaving large-scale climate as the dominant control in most regimes. The active biosphere is the main exception, where strong biogenic signals persist, underscoring the critical role of tropical forests in shaping atmospheric CO2 variability.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript presents a top-down analysis of global atmospheric CO2 growth rates using CAMS reanalysis fields, EDGAR emission inventories, GOSIF GPP data, and the SOI index. It concludes that regional and temporal variations in growth rates are dominated by natural carbon-cycle processes and background trends, with anthropogenic emission signals (including 2020 COVID-related reductions during a neutral ENSO year) frequently masked by biospheric and transport variability. Unsupervised clustering identifies five characteristic regimes, with spatial averaging shown to suppress regional signals except in the active-biosphere regime.

Significance. If the central attribution holds after addressing dataset-bias and robustness concerns, the work would usefully quantify the practical limits of regional top-down detection of emission changes against natural variability, with direct relevance to verification of mitigation policies. The regime classification could provide a framework for interpreting carbon-cycle controls, though its value depends on demonstrated stability under alternative processing choices.

major comments (3)
  1. [Results and Discussion (clustering and persistence analysis)] The claim that anthropogenic signals are masked (including the 2020 COVID reductions not appearing in regional growth rates) rests on the assumption that CAMS, EDGAR, GOSIF, and SOI contain no systematic biases large enough to alter clustering outcomes or attribution; no sensitivity tests to alternative inventories, reanalysis products, or transport-error scenarios are reported, leaving the masking conclusion vulnerable to dataset-specific artifacts.
  2. [Methods (clustering and persistence analysis)] The unsupervised clustering procedure that yields the five carbon-cycle regimes is presented without reported hyperparameters (e.g., number of clusters chosen, distance metric, initialization), data-exclusion criteria, or cross-validation against random seeds or subsamples; this directly affects the robustness of the regime map and the subsequent claim that spatial averaging leaves large-scale climate as the dominant control.
  3. [Results (growth-rate maps and time series)] No uncertainty quantification (error bars, bootstrap intervals, or ensemble spread) is attached to the reported growth-rate time series, regime assignments, or the statement that COVID reductions were "not consistently reflected"; without these, it is impossible to determine whether the absence of an anthropogenic signal exceeds the expected variability from natural terms.
minor comments (2)
  1. [Figures] Figure captions should explicitly state the temporal averaging window and any spatial smoothing applied to the growth-rate fields.
  2. [Abstract and Methods] The abstract states that the analysis uses "measures of biospheric activity" but the text does not clarify how GOSIF GPP is converted into a growth-rate contribution; a short methods paragraph would improve clarity.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their constructive comments, which highlight important aspects of robustness and presentation. We address each major comment below and outline revisions to strengthen the manuscript.

read point-by-point responses

  1. Referee: [Results and Discussion (clustering and persistence analysis)] The claim that anthropogenic signals are masked (including the 2020 COVID reductions not appearing in regional growth rates) rests on the assumption that CAMS, EDGAR, GOSIF, and SOI contain no systematic biases large enough to alter clustering outcomes or attribution; no sensitivity tests to alternative inventories, reanalysis products, or transport-error scenarios are reported, leaving the masking conclusion vulnerable to dataset-specific artifacts.

    Authors: We agree that the lack of reported sensitivity tests represents a limitation in demonstrating robustness. While the datasets employed are standard and widely validated in the carbon-cycle literature, we will add sensitivity analyses in the revised manuscript using alternative emission inventories and reanalysis products to evaluate the stability of the clustering outcomes and the attribution of masked anthropogenic signals. revision: yes

  2. Referee: [Methods (clustering and persistence analysis)] The unsupervised clustering procedure that yields the five carbon-cycle regimes is presented without reported hyperparameters (e.g., number of clusters chosen, distance metric, initialization), data-exclusion criteria, or cross-validation against random seeds or subsamples; this directly affects the robustness of the regime map and the subsequent claim that spatial averaging leaves large-scale climate as the dominant control.

    Authors: The clustering procedure details were not fully specified in the original submission. In the revised Methods section we will report the number of clusters (selected via silhouette analysis), the distance metric, initialization approach, data-exclusion criteria, and results across multiple random seeds and subsamples to confirm the stability of the five-regime classification and the spatial-averaging conclusions. revision: yes

  3. Referee: [Results (growth-rate maps and time series)] No uncertainty quantification (error bars, bootstrap intervals, or ensemble spread) is attached to the reported growth-rate time series, regime assignments, or the statement that COVID reductions were "not consistently reflected"; without these, it is impossible to determine whether the absence of an anthropogenic signal exceeds the expected variability from natural terms.

    Authors: We concur that uncertainty quantification is necessary to support statements about signal masking. The revised manuscript will include bootstrap-derived intervals or ensemble spread estimates for the growth-rate time series and regime assignments, enabling a quantitative evaluation of whether the absence of the COVID-related signal is distinguishable from natural variability. revision: yes

Circularity Check

0 steps flagged

No significant circularity; analysis relies on independent external datasets and unsupervised methods

full rationale

The paper performs a top-down analysis by ingesting four independent external datasets (CAMS CO2 reanalysis, EDGAR emissions, GOSIF GPP, SOI index) and applying unsupervised clustering plus persistence analysis. No equations, fitted parameters, or self-citations are used to derive the reported spatial regimes or the conclusion that natural variability masks anthropogenic signals. All load-bearing inputs are externally sourced and not redefined from the paper's own outputs, satisfying the self-contained criterion.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the fidelity of four standard public datasets and the validity of unsupervised clustering to identify meaningful regimes; no new free parameters, axioms beyond domain-standard assumptions, or invented entities are introduced.

axioms (1)
  • domain assumption CAMS reanalysis, EDGAR, GOSIF, and SOI each provide sufficiently unbiased representations of their target quantities for the purpose of regime identification.
    This premise is required to interpret the clustering results as evidence that natural processes dominate and mask anthropogenic signals.

pith-pipeline@v0.9.1-grok · 5835 in / 1173 out tokens · 52726 ms · 2026-06-30T01:15:20.661895+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

82 extracted references · 46 canonical work pages

  1. [1]

    Thoning, K. W. and Tans, P. P. , number =. 1989 , journal =. doi:10.1029/JD094iD06p08549 , issn =

    work page doi:10.1029/jd094id06p08549 1989

  2. [2]

    2020 , journal =

    Sussmann, Ralf and Rettinger, Markus , number =. 2020 , journal =. doi:10.3390/RS12152387 , issn =

    work page doi:10.3390/rs12152387 2020

  3. [4]

    and O'Dell, C

    Lindqvist, H. and O'Dell, C. W. and Basu, S. and Boesch, H. and Chevallier, F. and Deutscher, N. and Feng, L. and Fisher, B. and Hase, F. and Inoue, M. and Kivi, R. and Morino, I. and Palmer, P. I. and Parker, R. and Schneider, M. and Sussmann, R. and Yoshida, Y. , number =. 2015 , journal =. doi:10.5194/acp-15-13023-2015 , issn =

    work page doi:10.5194/acp-15-13023-2015 2015

  4. [5]

    and Kenea, Samuel Takele and Lindqvist, Hannakaisa and Kim, Jinwon and Li, Shanlan and Byun, Young Hwa and Goo, Tae Young , number =

    Labzovskii, Lev D. and Kenea, Samuel Takele and Lindqvist, Hannakaisa and Kim, Jinwon and Li, Shanlan and Byun, Young Hwa and Goo, Tae Young , number =. 2021 , journal =. doi:10.3390/RS13193868/S1 , issn =

    work page doi:10.3390/rs13193868/s1 2021

  5. [6]

    2024 , journal =

    Hachmeister, Jonas and Schneising, Oliver and Buchwitz, Michael and Burrows, John P and Notholt, Justus and Buschmann, Matthias , pages =. 2024 , journal =

    2024

  6. [7]

    2016 , journal =

    Tollefson, Jeff , number =. 2016 , journal =. doi:10.1038/nature.2016.19216 , issn =

    work page doi:10.1038/nature.2016.19216 2016

  7. [8]

    2019 , journal =

    Li, Xing and Xiao, Jingfeng , number =. 2019 , journal =. doi:10.3390/rs11050517 , issn =

    work page doi:10.3390/rs11050517 2019

  8. [9]

    doi:10.1038/nclimate2911 , issn =

    2016 , journal =. doi:10.1038/nclimate2911 , issn =

    work page doi:10.1038/nclimate2911 2016

  9. [10]

    2019 , journal =

    Rogelj, Joeri and Huppmann, Daniel and Krey, Volker and Riahi, Keywan and Clarke, Leon and Gidden, Matthew and Nicholls, Zebedee and Meinshausen, Malte , number =. 2019 , journal =. doi:10.1038/s41586-019-1541-4 , issn =

    work page doi:10.1038/s41586-019-1541-4 2019

  10. [11]

    2023 , journal =

    Comber, Alexis and Brunsdon, Christopher and Charlton, Martin and Dong, Guanpeng and Harris, Richard and Lu, Binbin and L. 2023 , journal =. doi:https://doi.org/10.1111/gean.12316 , issn =

    work page doi:10.1111/gean.12316 2023

  11. [12]

    1979 , booktitle =

    Hartigan, J A and Wong, M A , number =. 1979 , booktitle =

    1979

  12. [13]

    and Gasser, Thomas and Grassi, Giacomo and Forsell, Nicklas and Janssens, Iris and Lamb, William F

    Gidden, Matthew J. and Gasser, Thomas and Grassi, Giacomo and Forsell, Nicklas and Janssens, Iris and Lamb, William F. and Minx, Jan and Nicholls, Zebedee and Steinhauser, Jan and Riahi, Keywan , number =. 2023 , journal =. doi:10.1038/s41586-023-06724-y , issn =

    work page doi:10.1038/s41586-023-06724-y 2023

  13. [14]

    2018 , journal =

    Xiao, Qingyang and Chang, Howard H and Geng, Guannan and Liu, Yang , number =. 2018 , journal =. doi:10.1021/acs.est.8b02917 , issn =

    work page doi:10.1021/acs.est.8b02917 2018

  14. [15]

    1987 , journal =

    Ropelewski, Chester F and Jones, Phil D , number =. 1987 , journal =

    1987

  15. [16]

    and Griffith, David W

    Mostafavi Pak, Nasrin and Hachmeister, Jonas and Rettinger, Markus and Buschmann, Matthias and Deutscher, Nicholas M. and Griffith, David W. T. and Iraci, Laura T. and Lan, Xin and McGee, Erin and Morino, Isamu and Pollard, Dave and Roehl, Coleen M. and Strong, Kimberly and Kivi, Rigel and Wennberg, Paul , month =. 2025 , journal =

    2025

  16. [17]

    2017 , booktitle =

    Lesnikowski, Alexandra and Ford, James D and Berrang-Ford, Lea and Biesbroek, G R , pages =. 2017 , booktitle =

    2017

  17. [18]

    2021 , booktitle =

    Lourens, Lucas J , editor =. 2021 , booktitle =. doi:https://doi.org/10.1016/B978-0-12-821575-3.00028-1 , keywords =

    work page doi:10.1016/b978-0-12-821575-3.00028-1 2021

  18. [19]

    2019 , journal =

    Lenton, Timothy M and Rockstr. 2019 , journal =

    2019

  19. [20]

    2022 , journal =

    Zhu, Qindan and Laughner, Joshua L and Cohen, Ronald C , number =. 2022 , journal =. doi:10.1021/acs.est.1c05636 , issn =

    work page doi:10.1021/acs.est.1c05636 2022

  20. [21]

    2022 , journal =

    Lamb, William F and Grubb, Michael and Diluiso, Francesca and Minx, Jan C , number =. 2022 , journal =

    2022

  21. [22]

    2021 , journal =

    Grassi, Giacomo and Stehfest, Elke and Rogelj, Joeri and van Vuuren, Detlef and Cescatti, Alessandro and House, Jo and Nabuurs, Gert Jan and Rossi, Simone and Alkama, Ramdane and Vi. 2021 , journal =. doi:10.1038/s41558-021-01033-6 , issn =

    work page doi:10.1038/s41558-021-01033-6 2021

  22. [23]

    2026 , journal =

    Lamb, W F and Andrew, R M and Jones, M and Nicholls, Z and Peters, G P and Smith, C and Saunois, M and Grassi, G and Pongratz, J and Smith, S J and Tubiello, F N and Crippa, M and Gidden, M and Friedlingstein, P and Minx, J and Forster, P M , number =. 2026 , journal =

    2026

  23. [24]

    2023 , journal =

    Vernooij, R and Eames, T and Russell-Smith, J and Yates, C and Beatty, R and Evans, J and Edwards, A and Ribeiro, N and Wooster, M and Strydom, T and Giongo, M V and Borges, M A and Menezes Costa, M and Barradas, A C S and van Wees, D and der Werf, G R , number =. 2023 , journal =

    2023

  24. [25]

    2009 , journal =

    Yeh, Sang-Wook and Kug, Jong-Seong and Dewitte, Boris and Kwon, Min-Ho and Kirtman, Ben P and Jin, Fei-Fei , number =. 2009 , journal =

    2009

  25. [26]

    2020 , journal =

    Isson, T T and Planavsky, N J and Coogan, L A and Stewart, E M and Ague, J J and Bolton, E W and Zhang, S and McKenzie, N R and Kump, L R , number =. 2020 , journal =. doi:https://doi.org/10.1029/2018GB006061 , issn =

    work page doi:10.1029/2018gb006061 2020

  26. [27]

    1996 , journal =

    Brunsdon, Chris and Fotheringham, A Stewart and Charlton, Martin E , number =. 1996 , journal =. doi:https://doi.org/10.1111/j.1538-4632.1996.tb00936.x , issn =

    work page doi:10.1111/j.1538-4632.1996.tb00936.x 1996

  27. [28]

    2004 , journal =

    McMillen, Daniel P , number =. 2004 , journal =. doi:10.1111/j.0002-9092.2004.600

    work page doi:10.1111/j.0002-9092.2004.600 2004

  28. [29]

    and Frame, David J

    Allen, Myles R. and Frame, David J. and Friedlingstein, Pierre and Gillett, Nathan P. and Grassi, Giacomo and Gregory, Jonathan M. and Hare, William and House, Jo and Huntingford, Chris and Jenkins, Stuart and Jones, Chris D. and Knutti, Reto and Lowe, Jason A. and Matthews, H. Damon and Meinshausen, Malte and Meinshausen, Nicolai and Peters, Glen P. and ...

    work page doi:10.1038/s41586-024-08326-8 2025

  29. [30]

    2017 , journal =

    Kennedy, John and Dunn, Robert and McCarthy, Mark and Titchner, Holly and Morice, Colin , number =. 2017 , journal =. doi:10.1002/wea.3042 , issn =

    work page doi:10.1002/wea.3042 2017

  30. [31]

    2025 , journal =

    Friedlingstein, Pierre and O'Sullivan, Michael and Jones, Matthew W and Andrew, Robbie M and Bakker, Dorothee C E and Hauck, Judith and Landsch. 2025 , journal =

    2025

  31. [32]

    2025 , journal =

    Grassi, Giacomo and Peters, Glen P and Canadell, Josep G and Cescatti, Alessandro and Federici, Sandro and Gidden, Matthew J and Harris, Nancy and Herold, Martin and Krug, Thelma and O’Sullivan, Michael and Pongratz, Julia and Sanz, María J and Schwingshackl, Clemens and van Vuuren, Detlef , number =. 2025 , journal =. doi:10.1038/s41893-025-01565-1 , issn =

    work page doi:10.1038/s41893-025-01565-1 2025

  32. [33]

    2024 , journal =

    Forster, P M and Smith, C and Walsh, T and Lamb, W F and Lamboll, R and Hall, B and Hauser, M and Ribes, A and Rosen, D and Gillett, N P and Palmer, M D and Rogelj, J and von Schuckmann, K and Trewin, B and Allen, M and Andrew, R and Betts, R A and Borger, A and Boyer, T and Broersma, J A and Buontempo, C and Burgess, S and Cagnazzo, C and Cheng, L and Fr...

    2024

  33. [34]

    Forster, P M and Smith, C and Walsh, T and Lamb, W F and Lamboll, R and Cassou, C and Hauser, M and Hausfather, Z and Lee, J.-Y. and Palmer, M D and von Schuckmann, K and Slangen, A B A and Szopa, S and Trewin, B and Yun, J and Gillett, N P and Jenkins, S and Matthews, H D and Raghavan, K and Ribes, A and Rogelj, J and Rosen, D and Zhang, X and Allen, M a...

    2025

  34. [35]

    2017 , journal =

    Chatterjee, A and Gierach, M M and Sutton, A J and Feely, R A and Crisp, D and Eldering, A and Gunson, M R and O’Dell, C W and Stephens, B B and Schimel, D S , number =. 2017 , journal =

    2017

  35. [36]

    2024 , journal =

    Crippa, Monica and Guizzardi, Diego and Pagani, Federico and Schiavina, Marcello and Melchiorri, Michele and Pisoni, Enrico and Graziosi, Francesco and Muntean, Marilena and Maes, Joachim and Dijkstra, Lewis , number =. 2024 , journal =

    2024

  36. [37]

    Crippa, D

    Crippa, Monica and Guizzardi, Diego and Pagani, Federico and Schiavina, Marcello and Melchiorri, Michele and Pisoni, Enrico and Graziosi, Francesco and Muntean, Marilena and Maes, Joachim and Dijkstra, Lewis and Van Damme, Martin and Clarisse, Lieven and Coheur, Pierre , number =. 2024 , journal =. doi:10.5194/essd-16-2811-2024 , issn =

    work page doi:10.5194/essd-16-2811-2024 2024

  37. [38]

    2024 , journal =

    BONGAARTS, JOHN , number =. 2024 , journal =. doi:https://doi.org/10.1111/padr.12632 , issn =

    work page doi:10.1111/padr.12632 2024

  38. [39]

    2024 , journal =

    Marongiu, Alessandro and Distefano, Gabriele Giuseppe and Moretti, Marco and Petrosino, Federico and Fossati, Giuseppe and Collalto, Anna Gilia and Angelino, Elisabetta , number =. 2024 , journal =. doi:10.3390/air2040022 , issn =

    work page doi:10.3390/air2040022 2024

  39. [40]

    2020 , journal =

    Stirnberg, Roland and Cermak, Jan and Fuchs, Julia and Andersen, Hendrik , number =. 2020 , journal =. doi:https://doi.org/10.1029/2019JD031380 , issn =

    work page doi:10.1029/2019jd031380 2020

  40. [41]

    2018 , journal =

    Craft, Brianna and Fisher, Susannah , number =. 2018 , journal =. doi:10.1080/14693062.2018.1485546 , issn =

    work page doi:10.1080/14693062.2018.1485546 2018

  41. [42]

    1979 , journal =

    Cleveland, William S , number =. 1979 , journal =. doi:10.1080/01621459.1979.10481038 , issn =

    work page doi:10.1080/01621459.1979.10481038 1979

  42. [43]

    1987 , booktitle =

    Rousseeuw, Peter J , pages =. 1987 , booktitle =

    1987

  43. [44]

    2013 , journal =

    Solanki, Sami K and Krivova, Natalie A and Haigh, Joanna D , pages =. 2013 , journal =

    2013

  44. [45]

    2019 , journal =

    Pauw, Pieter and Mbeva, Kennedy and van Asselt, Harro , number =. 2019 , journal =. doi:10.1057/s41599-019-0298-6 , issn =

    work page doi:10.1057/s41599-019-0298-6 2019

  45. [46]

    2016 , journal =

    Mann, Michael E and Rahmstorf, Stefan and Steinman, Byron A and Tingley, Martin and Miller, Sonya K , number =. 2016 , journal =. doi:10.1038/srep19831 , issn =

    work page doi:10.1038/srep19831 2016

  46. [47]

    2019 , booktitle =

    Delbeke, Jos and Runge-Metzger, Artur and Slingenberg, Yvon and Werksman, Jake , pages =. 2019 , booktitle =

    2019

  47. [48]

    2010 , journal =

    Cole-Dai, Jihong , number =. 2010 , journal =. doi:https://doi.org/10.1002/wcc.76 , issn =

    work page doi:10.1002/wcc.76 2010

  48. [49]

    2012 , journal =

    Levitus, Sydney and Antonov, John I and Boyer, Tim P and Baranova, Olga K and Garcia, Hernan Eduardo and Locarnini, Ricardo Alejandro and Mishonov, Alexey V and Reagan, James R and Seidov, Dan and Yarosh, Evgeney S , number =. 2012 , journal =

    2012

  49. [50]

    1861 , journal =

    Tyndall, John , number =. 1861 , journal =. doi:10.1080/14786446108643138 , issn =

    work page doi:10.1080/14786446108643138

  50. [51]

    1896 , journal =

    Arrhenius, Svante , number =. 1896 , journal =. doi:10.1080/14786449608620846 , issn =

    work page doi:10.1080/14786449608620846

  51. [52]

    and Calvin, K

    Lee, H. and Calvin, K. and Dasgupta, D. and Krinner, G. and Mukherji, A. and Thorne, P. and Trisos, C. and Romero, J. and Aldunce, P. and Barret, K. and Blanco, G. , doi =

  52. [53]

    and Schimel, David S

    Liu, Junjie and Bowman, Kevin W. and Schimel, David S. and Parazoo, Nicolas C. and Jiang, Zhe and Lee, Meemong and Bloom, A. Anthony and Wunch, Debra and Frankenberg, Christian and Sun, Ying and others , journal =. 2017 , publisher =. doi:10.1126/science.aam5690 , url =

    work page doi:10.1126/science.aam5690 2017

  53. [54]

    and Tadi

    Shiga, Yoichi P. and Tadi. Geophysical Research Letters , volume =. 2018 , publisher =. doi:10.1002/2017GL076630 , url =

    work page doi:10.1002/2017gl076630 2018

  54. [55]

    2022 , publisher =

    Zhang, Shaoqing and Lei, Liping and Sheng, Mengya and Song, Hao and Li, Luman and Guo, Kaiyuan and Ma, Caihong and Liu, Liangyun and Zeng, Zhaocheng , journal =. 2022 , publisher =. doi:10.3390/rs14195024 , url =

    work page doi:10.3390/rs14195024 2022

  55. [56]

    2016 , publisher =

    Hakkarainen, Janne and Ialongo, Iolanda and Tamminen, Johanna , journal =. 2016 , publisher =. doi:10.1002/2016GL070885 , url =

    work page doi:10.1002/2016gl070885 2016

  56. [57]

    and Griffith, David W

    Mostafavi Pak, Nasrin and Hachmeister, Jonas and Rettinger, Markus and Buschmann, Matthias and Deutscher, Nicholas M. and Griffith, David W. T. and Iraci, Laura T. and Lan, Xin and McGee, Erin and Morino, Isamu and others , journal =. 2026 , publisher =. doi:10.5194/bg-23-1477-2026 , url =

    work page doi:10.5194/bg-23-1477-2026 2026

  57. [58]

    Atmospheric Chemistry and Physics , volume =

    Buchwitz, Michael and Reuter, Maximilian and Schneising, Oliver and No. Atmospheric Chemistry and Physics , volume =. 2018 , publisher =. doi:10.5194/acp-18-17355-2018 , url =

    work page doi:10.5194/acp-18-17355-2018 2018

  58. [59]

    doi:10.1017/9781009157964 , url =

    2019 , pages =. doi:10.1017/9781009157964 , url =

    work page doi:10.1017/9781009157964 2019

  59. [60]

    Technical note: The CAMS greenhouse gas reanalysis from 2003 to 2020 , JOURNAL =

    Agust\'. Technical note: The CAMS greenhouse gas reanalysis from 2003 to 2020 , JOURNAL =. 2023 , NUMBER =

    2003

  60. [61]

    and Krivova, Natalie A

    Solanki, Sami K. and Krivova, Natalie A. and Haigh, Joanna D. , title =. Annual Review of Astronomy and Astrophysics , volume =. 2013 , publisher =. doi:10.1146/annurev-astro-082812-141007 , issn =

    work page doi:10.1146/annurev-astro-082812-141007 2013

  61. [62]

    and Jin, Fei-Fei , title =

    Yeh, Sang-Wook and Kug, Jong-Seong and Dewitte, Boris and Kwon, Min-Ho and Kirtman, Ben P. and Jin, Fei-Fei , title =. Nature , volume =. 2009 , doi =

    2009

  62. [63]

    and Andrew, Robbie M

    Friedlingstein, Pierre and O'Sullivan, Michael and Jones, Matthew W. and Andrew, Robbie M. and Bakker, Dorothee C. E. and Hauck, Judith and Landsch. Earth System Science Data , volume =. 2026 , publisher =. doi:10.5194/essd-18-3211-2026 , url =

    work page doi:10.5194/essd-18-3211-2026 2026

  63. [64]

    2026 , month = jan, url =

    2026

  64. [65]

    and Berrang-Ford, Lea and Biesbroek, Robbert , title =

    Lesnikowski, Alexandra and Ford, James D. and Berrang-Ford, Lea and Biesbroek, Robbert , title =. 2017 , isbn =

    2017

  65. [66]

    and Jones, Phil D

    Ropelewski, Chester F. and Jones, Phil D. , title =. Monthly Weather Review , volume =. 1987 , doi =

    1987

  66. [67]

    , title =

    Thorndike, Robert L. , title =. Psychometrika , volume =. 1953 , doi =

    1953

  67. [68]

    , title =

    Rousseeuw, Peter J. , title =. Journal of Computational and Applied Mathematics , volume =. 1987 , doi =

    1987

  68. [69]

    and Planavsky, Noah J

    Isson, Terry T. and Planavsky, Noah J. and Coogan, Laurence A. and Stewart, Emily M. and Ague, Jay J. and Bolton, Edward W. and Zhang, Shuang and McKenzie, N. Ryan and Kump, Lee R. , title =. Global Biogeochemical Cycles , volume =. 2020 , doi =

    2020

  69. [70]

    and Smith, Chris and Walsh, Tristram and Lamb, William F

    Forster, Piers M. and Smith, Chris and Walsh, Tristram and Lamb, William F. and Lamboll, Robin and Hall, Bradley and Hauser, Mathias and Ribes, Aur. Earth System Science Data , volume =. 2024 , publisher =. doi:10.5194/essd-16-2625-2024 , issn =

    work page doi:10.5194/essd-16-2625-2024 2024

  70. [71]

    and Smith, Chris and Walsh, Tristram and Lamb, William F

    Forster, Piers M. and Smith, Chris and Walsh, Tristram and Lamb, William F. and Lamboll, Robin and Cassou, Christophe and Hauser, Mathias and Hausfather, Zeke and Lee, June-Yi and Palmer, Matthew D. and others , title =. Earth System Science Data , volume =. 2025 , publisher =. doi:10.5194/essd-17-2641-2025 , url =

    work page doi:10.5194/essd-17-2641-2025 2025

  71. [72]

    2025 , doi =

    Crippa, Monica and Guizzardi, Diego and Pagani, Federico and Banja, Manjola and Muntean, Marilena and Schaaf, Edwin and Quadrelli, Roberta and Risquez Martin, Ana and Taghavi-Moharamli, Parisa and K. 2025 , doi =

    2025

  72. [73]

    and Sutton, Adrienne J

    Chatterjee, Abhishek and Gierach, Michelle M. and Sutton, Adrienne J. and Feely, Richard A. and Crisp, David and Eldering, Annmarie and Gunson, Michael R. and O'Dell, Christopher W. and Stephens, Britton B. and Schimel, David S. , title =. Science , volume =. 2017 , doi =

    2017

  73. [74]

    and Andrew, Robbie M

    Lamb, William F. and Andrew, Robbie M. and Jones, Matthew W. and Nicholls, Zebedee and Peters, Glen P. and Smith, Christopher J. and Saunois, Marielle and Grassi, Giacomo and Pongratz, Julia and Smith, Steven J. and Tubiello, Francesco N. and Crippa, Monica and Gidden, Matthew and Friedlingstein, Pierre and Minx, Jan C. and Forster, Piers M. , title =. Ea...

    2026

  74. [75]

    and Wong, Manchek A

    Hartigan, John A. and Wong, Manchek A. , title =. Journal of the Royal Statistical Society: Series C (Applied Statistics) , volume =. 1979 , doi =

    1979

  75. [76]

    , title =

    Lan, Xin and Tans, Pieter and Thoning, Kirk W. , title =. 2026 , doi =

    2026

  76. [77]

    Remote Sensing , volume =

    Zhang, Yucong and Liu, Xinjie and Lei, Liping and Liu, Liangyun , title =. Remote Sensing , volume =. 2022 , publisher =. doi:10.3390/rs14163899 , issn =

    work page doi:10.3390/rs14163899 2022

  77. [78]

    Atmosphere , volume =

    Ji, Zhanghui and Song, Hao and Lei, Liping and Sheng, Mengya and Guo, Kaiyuan and Zhang, Shaoqing , title =. Atmosphere , volume =. 2024 , publisher =. doi:10.3390/atmos15030323 , issn =

    work page doi:10.3390/atmos15030323 2024

  78. [79]

    Atmospheric Measurement Techniques , volume =

    Mustafa, Farhan and Bu, Lingbing and Wang, Qin and Yao, Na and Shahzaman, Muhammad and Bilal, Muhammad and Aslam, Rana Waqar and Iqbal, Rashid , title =. Atmospheric Measurement Techniques , volume =. 2021 , publisher =. doi:10.5194/amt-14-7277-2021 , issn =

    work page doi:10.5194/amt-14-7277-2021 2021

  79. [80]

    Earth System Dynamics , volume=

    Dynamic savanna burning emission factors based on satellite data using a machine learning approach , author=. Earth System Dynamics , volume=. 2023 , publisher=

    2023

  80. [81]

    5 concentrations in China from satellite data , author=

    An ensemble machine-learning model to predict historical PM2. 5 concentrations in China from satellite data , author=. Environmental science & technology , volume=. 2018 , publisher=

    2018

Showing first 80 references.