FAO estimates of net forest emissions and removals, 1990–2025

Tubiello, Francesco N.; Branthomme, Anne; Obli-Laryea, Griffiths; Piccoli, Michelle; Ramadan, Nidal; Condor, Rocio Danica; Wanner, Nathan; Brivio, Rachele; Jonsson, Örjan; Conchedda, Giulia; Pekkarinen, Anssi

doi:10.5194/essd-18-6799-2026

Articles | Volume 18, issue 9

https://doi.org/10.5194/essd-18-6799-2026

© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/essd-18-6799-2026

© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 18, issue 9

Data description article

|

15 Sep 2026

Data description article |

| 15 Sep 2026

FAO estimates of net forest emissions and removals, 1990–2025

Francesco N. Tubiello, Anne Branthomme, Griffiths Obli-Laryea, Michelle Piccoli, Nidal Ramadan, Rocio Danica Condor, Nathan Wanner, Rachele Brivio, Örjan Jonsson, Giulia Conchedda, and Anssi Pekkarinen

Download

Final revised paper (published on 15 Sep 2026)
Supplement to the final revised paper
Preprint (discussion started on 22 Oct 2025)

Interactive discussion

Status: closed

RC1:
'Comment on essd-2025-635', Giacomo Grassi, 11 Dec 2025

The comment was uploaded in the form of a supplement: https://essd.copernicus.org/preprints/essd-2025-635/essd-2025-635-RC1-supplement.pdf

Citation: https://doi.org/10.5194/essd-2025-635-RC1
- AC1: 'Reply on RC1', Francesco N. Tubiello, 09 May 2026
  
  We thank this reviewer for very constructive comments, which helped us further develop the original manuscript into a much more complete set of data and analyses.
  We have greatly expanded all sections and included more detailed global, regional and country comparisons to UNFCCC and independent model data. A greatly exanded meethods section also focuses on improvements in FRA 2025 compared to the 2020 version.
  
  Citation: https://doi.org/10.5194/essd-2025-635-AC1
RC2:
'Comment on essd-2025-635', Anonymous Referee #2, 12 Dec 2025
Review of “New FAOSTAT forest emissions and removals estimates: 1990-2025”, Tubiello et al.
General comments:
This manuscript describes updated forest emissions and removals in FAOSTAT using the latest data in the FRA.
The manuscript is very brief. It left me with many questions, particularly about methods, study limitations, and how FAOSTAT fits with other data sources, particularly NGHGIs. In general, I think each section was terse to the point of lacking key information that would make this a more valuable contribution. I understand that being in ESSD, this is a data description paper (as opposed to an analysis and interpretation of data) but it still does not thoroughly describe the dataset and its production.
Below are some questions I would like to see answered:
This is an update of previous FAOSTAT emissions and removals data. What were the previous values? How do these compare with the current values? What changed in FAOSTAT’s methods and underlying country data from the previous version to this one?

How reliable is the time series of fluxes? Do all countries really have consistent timeseries back to 1990? I believe not, but this isn’t discussed anywhere. I know that the authors can’t control country data, but known inconsistencies by time or for specific countries should be mentioned, with potential impacts discussed. Not exploring this makes it hard to assess the validity of the (non-significant) trends in fluxes that the authors claim (e.g., line 166, 172-174). At the least, this needs to be mentioned in Section 2.2, as well as preferably in the results and discussion.

How do the FAOSTAT estimates relate to NGHGI submissions to the UNFCCC? A table describing their similarities and differences and what countries report in each (at least in principle) would help. Figure 5 shows some substantial differences in fluxes by country. I know you can’t get into all country differences but the main conceptual and methodological similarities and differences between FAOSTAT and NGHGIs are important to explain, as well as one would use one vs. the other. Finally, are NGHGI and FAOSTAT different enough from each other that they represent distinct lines of evidence for forest fluxes (and the forest sink), as implied in lines 179-180? All of this could be covered in the introduction and discussion.

Where does FAOSTAT fall in terms of covering fluxes from managed vs. unmanaged lands, direct anthropogenic effects, indirect effects, and natural fluxes? And does forest land include any emissions in it (e.g., fires, limited harvests)? This is important for understanding how FAOSTAT compares with other flux estimates and its scope relative to the Global Carbon Budget and NGHGIs. FAOSTAT’s position on this should be explained somewhere.

You mention an important point with “Thirdly, the computation of net forest conversion emission, based on national-level statistics of area and biomass density as per equation (2), may lead to underestimates of carbon fluxes, so that actual deforestation emissions may be larger than estimated in FAOSTAT.” but don’t follow that thought. I think you’re saying that estimating fluxes over 5-year intervals based on carbon stock changes at a national level may underestimate fluxes. You should explore this point further. What is FAOSTAT missing in terms of emissions and removals by using stock-changes over 5 years instead of annually (e.g., interannual variability)? When are 5-year stock-changes a reasonable or unreasonable proxy for annual data (stock-change or gain-loss)?

Specific comments:
Gross vs. net: It would be helpful if net emissions and removals were identified as such, to contrast them with the net fluxes.

Where does non-forest converted to forest get assigned? Is it part of forest land? Either way, that should be explicitly stated. More generally, what is covered by forest land and net forest conversion? Does this vary by country (or over time)?

How do the FRA and FAOSTAT relate to each other? This is described very briefly in the intro. Is FAOSTAT derived from the FRA itself, or just from the same country data?

You say that “FAOSTAT statistics of CO2 emissions and removals from forests include estimates of CO2 fluxes on both forest land and on land lost due to net forest conversion. They are computed using carbon stock changes following the IPCC (2006) Tier 3, approach 1 guidelines (Tubiello et al., 2020), based on inputs of forest land area and biomass data from the FRA.” Has every country reported carbon stocks in 1990, 1995, 2000… (5-year intervals)? How does the FRA infer carbon stocks in years that countries don’t report? More generally, how is missing country data filled, if it is? This would be another place to clarify the relationship between FRA and FAOSTAT.

Please include the uncertainty propagation calculations (and results) mentioned in Section 2.2.

In “The FAOSTAT data show that forests acted globally as sinks of atmospheric CO2 during the period 1991-2025 (-4.0 Gt CO2 yr-1) and specifically during 2021-2025 (Fig. 2).”, I think this is actually forestland, not forests, which I take to be forestland and NFC combined. Although, I am confused about the difference between the -4.0 and the -3.6 in the next sentence (which is also apparently forestland). Is -3.6 for 2021-2025?

Where is the NGHGI data coming from, e.g., correlation with FAOSTAT (Section 3.2)? Why is the correlation between NGHGI and FAOSTAT for forestland included in the net forest conversion paragraph (line 142/143)? Aren’t FAOSTAT and NGHGI using similar data at their root (nationally reported data), which would mean that correlations between them are to be expected? A scatterplot of the forestland and NFC NGHGI-FAOSTAT correlations should be included if statistics are provided. Also, might not large forested countries skew this analysis?

I think that lines like “in particularly removing on average more than -1 Gt CO2 yr-1 during 2001-2025 (-0.4 Gt CO2 yr-1 since 1990)” are confusing. I think this is referring to forestland + NFC, but Table 1 doesn’t show -1 Gt/yr for 2001-2025. Where is -1 Gt/yr coming from?

What does “but never previously detected with this magnitude” mean (line 178)? Are you saying that this is a larger forestland sink than other approaches estimate? If so, you should explore this.

Figure 2: I suggest including another set of bars at the right side of the graph to show the 1991-2025 averages (like in Table 1).

Why does Figure 4 show 10 countries but Figure 5 only show 5 countries?

To what do you attribute the apparently high correlation between FAOSTAT and NGHGIs in Figure 4 and the lower correlation in Figure 5?

Given that this dataset ends in 2025, how are 2021-2025 values calculated? I imagine this varies by country but it needs to be explained. Are values from the most recent year extrapolated to 2025? Is the most recent year of data simply extended to 2025?

Forestland, net forest conversion, and their sum (sometimes you call it forests?) are not consistently used in the paper, table, and figures. Please ensure consistency between them.

Technical corrections:
Line 84: Missing “for” after “available” in “The updated FAOSTAT dataset makes values available forest emissions/removals estimates by country”.

Line 89: Zenodo link doesn’t look complete.

Use of homogenously could be replaced with “equally” (e.g., like 103).

Line 131: Should this heading also be “Emissions/removals on forest land”, like in line 98?
Citation: https://doi.org/10.5194/essd-2025-635-RC2
- AC2: 'Reply on RC2', Francesco N. Tubiello, 09 May 2026
  
  We thank this reviewer for similarly very constructive comments, which helped us further develop the original manuscript into a much more complete set of data and analyses.
  We have greatly expanded all sections and included more detailed global, regional and country comparisons to UNFCCC and independent model data. A greatly exanded meethods section also focuses on improvements in FRA 2025 compared to the 2020 version.
  
  Citation: https://doi.org/10.5194/essd-2025-635-AC2

Peer review completion

AR – Author's response | RR – Referee report | ED – Editor decision | EF – Editorial file upload

AR by Francesco N. Tubiello on behalf of the Authors (09 May 2026) Author's response Author's tracked changes

EF by Anna Mirena Feist-Polner (11 May 2026) Manuscript

ED: Referee Nomination & Report Request started (11 May 2026) by Hanqin Tian

RR by Giacomo Grassi (25 Jun 2026)

RR by Anonymous Referee #2 (06 Jul 2026)

Suggestions for revision or reasons for rejection

General comments:
1. This version has much more detail but content is often in unexpected sections. For example, some Results are in the Methods, some Discussion (limitations) is in the Methods. Overall, this manuscript needs heavy reorganization. The amount of material in unexpected places makes it hard to follow the manuscript. Numerous examples are below.
2. The use of the three FAO estimates is not introduced well and is applied inconsistently. Sometimes you use one, sometimes all three. I understand wanting to use a specific set of carbon pools for different purposes, but which version is used where and why is generally not explained. In some ways, it might be easier if you say in the methods that you calculated these three versions but stick with showing the results for just one of them, except for specific cases where you think another one is more appropriate. Furthermore, I don’t think you ever explain how these three versions relate to what is on FAOSTAT or in the FRA.
3. I appreciate the comparison with the Global Carbon Budget, but it still leaves much to be desired. At the very least, the specific components of the GCB that are being compared to in lines 494-498 need to be explicitly stated (Eluc, Sland, a component of Eluc, or something else). Even a sentence or two about differences in scope between the GCB and FAO would help. For example, do the GCB estimates you’re comparing to include peat emissions and other non-forest CO2 fluxes?

Specific comments:
1. Overall, this paper assumes that readers are familiar with FAO data and the FRA that may confuse or deter readers. I encourage the authors to try to lay out the relationship between country submissions to the FAO (and UNFCCC) and what the FAO team does with country data more thoroughly in the abstract, introduction, and methods.
a. The abstract should more clearly articulate the relationship between the FRA and the values presented in this paper. In general, the abstract would be clearer if it started with one sentence of background about the FRA or FAOSTAT.
b. For example, is it correct that countries submit forest area and carbon stock to the FAO (per line 71), but the FAO is the one that calculates emissions and removals (per line 70)? Please clarify which party is doing what.
c. What are the main differences in scope and objective between the FRA data and the NGHGIs?
2. “the FAO results indicated strong removals on forest land, ranging from -3.6 to -6.0 Gt CO2 yr-1—with lower and upper boundaries reflecting inclusion of stock change for living biomass only or for all pools. The forest removals were partially counterbalanced by emissions from net forest conversion of 2.8 to 3.8 Gt CO2 yr-1”: These are net removals and gross emissions, right? Can the difference between them be presented as -0.8 and -2.2 Gt CO2 yr-1? If so, why not present that range of net values in the abstract? Also, see above comment about how the three FAO estimates are not conveyed clearly. The range of values here (e.g., 2.8 to 3.8) appears to be purely due to the inclusion of different carbon pools. You could potentially present the results as X emissions from biomass, Y from litter and dead wood, and Z from soil, because that’s essentially what the range of values is providing.
3. Explain why you created your own dataset of country submissions to the UNFCCC rather than using the JRC LULUCF Data Hub (lines 93-100)? What is the advantage of creating a new dataset instead of using the existing, more comprehensive one from the Data Hub?
4. Much of the material in section 2.1.2 (Forest land area) should go in the results, e.g., similarity between FRA 2025 and FRA 2020 (with associated graph) and changes in Argentina’s data. Everything between Tables 2 and 3 should be in the results or discussion. (Also, more generally, is it correct that the correlation between forest area submissions to the UNFCCC and forest area submissions to FAO aren’t a concern of FAO carbon estimates specifically? This comparison is relevant to FAO in general, not just this paper.)
5. Similarly, much of the material in section 2.1.3 (Forest biomass and carbon stock change) should be in the results in a section that could be called something like “Comparison of FRA 2020 to FRA 2025”. What I would expect to see in section 2.1.3 is where the non-biomass data (litter, dead wood, and soil) came from this time, since you said that those other carbon pools are new to FRA 2025. This is covered in a later section but could at least be introduced here.
6. NFC and FL are introduced in Section 2.2 (Net forest emissions/removals estimation) but it seems like these categories should be covered in 2.1.2.
7. “We did not report estimates together with their uncertainty values. While each estimate carried in principle 50% uncertainty, we used the range of the three FAO models to provide a heuristic, practical if imperfect way to communicate uncertainty.” I understand you can’t do a formal uncertainty analysis, but the range between the 3 models isn’t any kind of uncertainty. It’s a range of estimates with different carbon pool scopes. It’s not really a sensitivity analysis, either, because you’re not changing some input or parameter. Please rethink how you’re communicating your three estimates.
8. To what do you attribute the difference between FAO and UNFCCC values for China? I know you can’t explore every country’s difference, but this is the largest of all differences in Table 8.

Technical corrections:
1. “The FAO net forest emissions/removals estimates were further developed” is an abrupt way to start an abstract. Something like “The FAO net forest emissions and removals estimates were refined and updated…”
2. Briefly explain what AI and NAI countries are in the abstract when they are first mentioned.
3. “The forest net emissions/removals dataset” (line 41): Presumably CO2 emissions and removals, but this should be explicit.
4. “regularly updated” (line 44): meaning every 5 years, when a new FRA is released? Or at a different cadence?
5. “FRA data sources may include national forest inventories, remote sensing, administrative data, academic and specialized studies and, in the lack of other data, expert opinion.” Were all of these kinds of data used for the carbon stock-change estimates in this paper?
6. What are “direct reports” (line 85)? This is an example of this paper assuming reader familiarity with the FRA process.
7. Table 1 needs all acronyms explained.
8. The 1:1 line in Figure 1 doesn’t look like it’s going through (0,0) because it’s intercepting the x-axis and not the bottom-left corner. Maybe this is a trick of log-log axes. If that point on the x-axis is indeed (0,0), the figure should have a note confirming that. (I’m likely not going to be the only one confused by it.)
9. In Table 2, I don’t understand what the Global-FAO and Report to both FAO and UNFCCC-FAO columns represent and why they’re different. Add a note explaining this.
10. Not all terms in equations 1-6 are defined. For example, I don’t see a definition for ΔB. αp is defined much further after the equations (line 226), complicating interpretation of the equations.
11. Why is section 2.3 (Limitations and uncertainty) in the Methods? This would traditionally be in the Discussion.
12. The fact that the 2025 data in the FRA is projected by countries needs to be clarified up front in the methods.
13. “The FAO data showed increasing carbon stocks, hence negative carbon stock change fluxes…”, i.e. “The FAO data showed increasing carbon stocks in forests (i.e. net carbon increases)…”?
14. What FAO scenario number (1, 2 or 3) is associated with -1.8 Gt CO2 yr-1 on line 331? Table 4 indicates that value is from FAO 1, but the scenario associated with every statistic should be stated in the text where the number is used.
15. “the CSC fluxes were consistently negative”, meaning they were consistently sinks? Rather than referring to the flux by sign (positive or negative) in writing, say whether it’s a source or sink.
16. As with pretty much all other figures and tables, Table 4 needs explanation of the abbreviations used, including what is included in each FAO model and what CSC, NFC, FL, AI, and NAI are.
17. There are too many colors for too many lines in Figure 3. All panels of this figure are very hard to understand. Some possibilities include: separating into more panels; dropping some lines altogether from the figure (maybe the less relevant FAO models); categorizing the legend entries more clearly (with category headers); otherwise distinguishing the legend entries better by color, line type, or thickness. As it is, I can’t tell what this figure is saying.
18. You sometimes cite Friedlingstein et al. 2025 and sometimes Friedlingstein et al. 2026. Which version of the GCB are you using?

Hide

ED: Reconsider after major revisions (06 Jul 2026) by Hanqin Tian

AR by Francesco N. Tubiello on behalf of the Authors (22 Jul 2026) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (23 Jul 2026) by Hanqin Tian

RR by Giacomo Grassi (19 Aug 2026)

Suggestions for revision or reasons for rejection

Review of Tubiello et al "FAO estimates of net forest emissions and removals, 1990–2025" (essd-2025-635) – v3

General assessment
I thank the authors for a thorough and constructive revision. The dataset is valuable and I would like to see it published. However, a few points still need fixing. It should not be difficult.

Part A follows up on my previous comments; Part B collects observations on the revised text.

Part A — Follow-up on my previous comments
Previous comments 3 and 4: USA and Italy
I appreciate that the authors have chosen to document the case of the USA. The current text states that the carry-forward "implies for specific countries a FAO underestimate of 2020–2025 mean removals, had national carbon stock kept increasing past 2022—and an overestimate in the opposite case". If possible, please indicate how many countries carried a pre-2025 value forward in FRA 2025, and document this in the relevant FAOSTAT country notes.
I am glad the authors re-examined Italy, and I agree that a correction was warranted. The way it has been handled, however, leaves the reader with no way to reconstruct what happened.
According to the reply, the correction was made by carrying the 2020 forest-land fluxes forward to 2025. This is not a procedure described in Sect. 2.2, and it contrasts with the description of the dataset at lines 235–237 as relying on "straightforward, reproducible equations, grounded in official national statistics". I note also that the reply apparently describes the correction in two different ways — changing the country input carbon stock data on the basis of the biomass DM time series, and carrying the 2020 fluxes forward — which are not the same operation. Please state which was applied.
The authors write that they "preferred not to name the country affected directly in the manuscript—although we may publish a country note in FAOSTAT detailing the above changes". This is awkward for two reasons. The manuscript (lines 278–279) already states that the corrections "were documented in the relevant FAOSTAT country notes", in the past tense, while the reply puts that publication in the future and in the conditional. And by the authors' own account the adjustment is worth 100–110 Mt CO2 yr-1 under FAO 2 and FAO 3 and propagates to the southern Europe, Europe, EU27, OECD and Annex I aggregates — the last of which carries the paper's central UNFCCC comparison (Tables 4 and 5). Several Table 4 values have changed accordingly relative to the previous version, yet nothing in the manuscript records that input data were revised, and the Zenodo DOI is unchanged. The revised deposit should be given a distinct version, with the change noted in its description, and a sentence in Sect. 2.3 should record the correction. As things stand now, a reader cannot tell which country, which years, or what magnitude is affected. Please name Italy here, as the authors already do for the USA a few lines above.
Please also state the criterion that distinguishes cases where input data are corrected (Italy) from those where they are not (USA), since the reply notes a general preference not to correct FRA data.
Previous comment 5: Forest area in Russia
The authors note that the revision of the Russian data appeared in an NGHGI submission dated 25 June 2026, well after the cut-off date of 15 March 2026.
I think the wrong cells have been used to assess the area of managed land. The value of 739 Mha that I reported previously is already contained in Russia's 2025 submission https://unfccc.int/sites/default/files/resource/RUS-CRT-2025-V1.0-20250419-003747_awaiting%20submission.zip, well before the cut-off. The relevant entry is the cell "Forest land (managed)" in Table 4.1, which corresponds to the more detailed managed and unmanaged areas reported in 4.A.
The correct value affects Tables 2 and 3, and for Russia it brings the UNFCCC figure much closer to the FAO one.
Previous comment 6: Significant figures in Tables 2 and 3
My request for one further digit was declined on the grounds that 50% uncertainty permits at most two significant figures. The manuscript no longer states 50% (see B.1 below), so that justification no longer corresponds to the text. In any case these two tables report forest areas, which are considerably better constrained than fluxes — the text itself quotes country areas to the megahectare at lines 328–331. As currently rounded, Table 3 gives Namibia's FAO area as "0.0" Gha with a difference of "+0.1 (+800%)". Please report areas in Mha.
Previous comment 7: Managed-area boundaries
This is among the most important comments of this round.
I appreciate the added emphasis that the area difference matters for the affected countries, but one qualification is still needed. The conclusion at lines 562–564 rests on a comparison of aggregate totals. For countries such as Brazil (−150 Mha), Canada (−140 Mha), China (+85 Mha), the Russian Federation (+70 Mha) and Namibia (+60 Mha) the area differences between UNFCCC and FAO are large, and the good match at global level arises because these differences run in opposite directions and cancel. The country-specific explanations at lines 353–364 are useful, but they do not support the conclusion that "flux estimates made on the entire forest land area could be considered anthropogenic in the sense of the IPCC managed land use proxy, ensuing consistency and thus comparability of FAO and UNFCCC data". The Conclusions need a sentence acknowledging that for several important countries the difference in forest area limits the comparability of FAO and UNFCCC emissions.
Previous comment 10: Table 4 arithmetic
The arithmetic check I asked for is still outstanding. Several cells do not satisfy Eq. (6) by more than rounding:
Period Aggregate Model NFC + FL CSC reported
2010–2020 World FAO 2 3.2 − 4.7 = −1.5 −1.9
2010–2020 NAI FAO 3 3.3 − 3.1 = +0.2 +0.6
2010–2020 World FAO 3 4.0 − 6.6 = −2.6 −2.9
2010–2020 World FAO 1 2.9 − 4.1 = −1.2 −1.4
2010–2020 NAI FAO 2 3.0 − 2.4 = +0.6 +0.8
2000–2010 World FAO 3 4.3 − 7.1 = −2.8 −3.0
World = AI + NAI also fails for 2010–2020 CSC (FAO 1: −2.1 + 0.9 = −1.2 against −1.4; FAO 2: −2.4 + 0.8 = −1.6 against −1.9; FAO 3: −3.3 + 0.6 = −2.7 against −2.9). The 1990–2000 and 2020–2025 blocks seem consistent, so the problem looks localised. Please check again.
Previous comment 11: Extent of gap-filling in the input data
I had asked whether the interpolation and sub-regional gap-filling described in the FRA report — trends in growing stock used to fill below-ground, dead wood and litter; sub-regional average stock per hectare for non-reporting countries — carry through into the FAOSTAT estimates. The revision notes that FRA 2025 reduced the need for FAO imputations (line 90) and addresses gap-filling for the soil pool only (lines 252–254). For a data-description paper this is basic information for a user. A short statement of what share of values, by pool, derives from country reporting rather than from imputation would be a real addition, all the more so given that soil contributes a further 30–40% of removals on forest land (lines 388–390) and that FAO 3 is the model used in the GCB comparison (lines 189–190).

Part B — New observations on the revised text
B.1 Uncertainty
The revision replaces the previous 50% (95% CI) with "a range of about ±30% around FAO 2 values" (lines 284–285), derived from the spread of the three FAO models. As far as I can tell neither referee asked for this, and I would ask the authors to reconsider. The spread between FAO 1, 2 and 3 (defined at lines 180–183) reflects the completeness of the carbon pools included: a structural difference with a known direction, since FAO 3 is by construction the most complete. It is not a dispersion around a central value, and it cannot be read as an uncertainty range. Either keep a propagated uncertainty and present the model spread separately, as a sensitivity to pool coverage, or explain why the two can be equated.
B.2 Abstract
The abstract, rewritten in this round, reports good agreement at global, AI and NAI levels. The results state that the FAO range includes the UNFCCC value across all flux-category and aggregate combinations "except for FL fluxes for the NAI aggregate" (line 436). By my reading of Table 5 there are two exceptions rather than one: for World FL the FAO range is −3.9 to −6.7 against a UNFCCC value of −6.9, which falls outside it. Please check.
More generally, the non-Annex I sink gap of 1.2–2.4 Gt CO2 yr-1 (line 439) and its attribution to countries reporting constant carbon density (lines 440–442) is among the more valuable contributions of this paper. It deserves a place in the abstract.
Minor point: the abstract quotes 2015–2025 values (lines 24–27) that do not correspond to any period in Table 4.

Referee Report: PDF

Hide

ED: Publish subject to minor revisions (review by editor) (20 Aug 2026) by Hanqin Tian

AR by Francesco N. Tubiello on behalf of the Authors (20 Aug 2026) Author's response Author's tracked changes Manuscript

ED: Publish as is (29 Aug 2026) by Hanqin Tian

AR by Francesco N. Tubiello on behalf of the Authors (31 Aug 2026)

Download

Article (1285 KB)
Full-text XML

Short summary

Estimates of net forest emissions/removals were updated using the 2025 Forest Resources Assessment data, submitted by countries to the Food and Agriculture Organization. Results show that forests acted globally as sinks of atmospheric carbon, despite ongoing losses due to deforestation. Results were consistent with country data independently submitted under the Paris Agreement and with the most recent scientific evidence.