IPCC AR7: What Will Change, and Why Nobody Can Tell You When
The Seventh Assessment Report will eventually replace the global warming potential values that sit under every corporate carbon inventory in the world. The obvious question is when. The honest answer, as of today, is that the IPCC’s own member governments have not agreed a delivery date for the Working Group reports, and have now deferred that decision at four consecutive plenary sessions. This page sets out what is actually fixed, what is not, where the report has genuinely got to, and what a reporting team should do in the meantime.
Two AR7 dates are agreed and neither is the one you want: the 2027 methodology reports, and the Synthesis Report “by late 2029”. The three Working Group reports have no agreed delivery date. Working Group I carries the GWP values, and its first draft is in expert review right now. Governments said at the March 2026 plenary that they want the timeline settled at the next session. Until then, treat any published AR7 date as somebody’s estimate, pin your data version, and record which assessment basis each factor uses so the eventual switch is a lookup rather than a rebuild.
What is actually decided
The seventh assessment cycle formally began in July 2023. At its first plenary of the cycle, in Istanbul in January 2024, the Panel settled the programme of work: three sequential Working Group reports, and a set of dated side products. Those decisions have held.
| Product | Delivery | Agreed? |
|---|---|---|
| Methodology Report on Inventories for Short-Lived Climate Forcers | 2027 | Yes |
| Methodology Report on Carbon Dioxide Removal Technologies, Carbon Capture, Utilization and Storage | 2027 | Yes — scientific content agreed at the October 2025 plenary |
| Special Report on Climate Change and Cities | By 2027 | Yes |
| Working Group I — Physical Science Basis (this is where GWP values live) | — | No date agreed |
| Working Group II — Impacts, Adaptation and Vulnerability | — | No date agreed |
| Working Group III — Mitigation of Climate Change | — | No date agreed |
| Synthesis Report | Finalised by late 2029, after all three Working Groups | Yes |
Read the last two rows together and the shape of the problem appears. The Synthesis Report is anchored to late 2029 and must follow all three Working Group reports, so that is a ceiling. There is no floor.
The timeline, deferred four times
This is the part that is almost never reported, and it is a matter of public record in the IPCC’s own session documents. The Working Group delivery schedule has been on the agenda at every plenary of this cycle and has not been agreed at any of them.
IPCC-60 — Istanbul, January 2024
The products are decided and the 2027 and 2029 anchors are set. The Panel asks its Bureau to come back with the month and year of delivery for the Working Group reports. That request is the origin of everything below.
IPCC-61 — Sofia, July–August 2024
Divergent views on the overall timeline and on alignment with the UN climate process. Discussion moves to a contact group. In the Panel’s own words, “time limitations and insufficient convergence prevented agreement”. Deferred to the next session.
IPCC-62 — Hangzhou, February 2025
The outlines of all three Working Group reports are agreed, which is real progress. On timing, two proposals emerge, the second of which would shift Working Group approvals into later months in 2028 and early 2029. The Panel is again unable to reach consensus, and defers the workplan to the next session.
IPCC-63 — Lima, October 2025
Governments remain divided. Some consider the revised schedule a workable compromise; others hold that it is too compressed, citing overlapping back-to-back review processes and limited time for national coordination. Deferred again.
IPCC-64 — Bangkok, March 2026
The timeline is not given its own agenda item, which delegations object to. A contact group is convened. The outcome, verbatim from the IPCC: “While the Panel made no formal decision, the member governments clearly indicated that they want the timeline agreed at its next session.”
A delivery date is not a scheduling detail for a reporting team. It decides whether a base year gets restated inside the current target period or the next one, whether an assurance provider will accept a figure computed on the older basis, and how long a hard-coded GWP constant has left. Anyone publishing a confident AR7 release date is estimating, and should say so. The disagreement is not about science; it is about review capacity and national coordination time, which is why it has proved hard to close.
Where AR7 has actually got to
Absent a schedule, the honest way to answer “how far along is it?” is to look at what has already happened rather than what is promised.
| When | What |
|---|---|
| December 2024 | Scoping meeting, Kuala Lumpur, to draft the outlines. |
| February 2025 | Outlines of all three Working Group contributions agreed. |
| December 2025 | First Lead Author Meeting, held jointly for all three Working Groups, Paris. |
| April 2026 | Working Group I Second Lead Author Meeting, Santiago. |
| May 2026 | Working Group III meets in Riyadh; Working Group II in Nassau. |
| 10 August – 2 October 2026 | Expert review of the Working Group I First Order Draft — the first of several review stages. |
| Later in 2026 | Working Group I Third Lead Author Meeting, being planned. |
As this page is published, the Working Group I First Order Draft is in expert review, and that window closes on 2 October 2026. If you have relevant expertise, registration is open through the IPCC. This is the draft that will eventually carry the AR7 global warming potentials, and expert review is the stage at which the numbers can still be argued with. It is also the clearest available signal of progress: a report with a first draft in review is further along than the absence of a schedule suggests.
What AR7 is likely to move
Nobody outside the author teams knows what AR7 will say. What is knowable is how much the last two transitions moved, because those numbers are in this corpus and can be read live. That record is the only defensible basis for planning, and it says the changes are material rather than cosmetic.
| Gas | AR5 GWP-100 | AR6 GWP-100 | Change |
|---|---|---|---|
| Methane, fossil | 28 | 29.8 | +6.4% |
| Methane, biogenic | 28 | 27 | −3.6% |
| Nitrous oxide | 265 | 273 | +3.0% |
| Sulphur hexafluoride | 23500 | 25200 | +7.2% |
| Nitrogen trifluoride | 16100 | 17400 | +8.1% |
| HFC-134a | 1300 | 1530 | +17.7% |
| HFC-23 | 12400 | 14600 | +17.7% |
Three things in that table are worth carrying into an AR7 plan. The direction is not uniform: biogenic methane fell while every other gas here rose, so “GWPs go up” is not a safe planning assumption. The refrigerants moved most, by nearly a fifth, which matters disproportionately to anyone with significant fugitive emissions. And a split appeared that had not existed before: AR5 gave one methane value, AR6 gives different figures for fossil and biogenic methane. A future assessment can introduce a distinction as easily as it can change a number, and a schema that assumes one value per gas will not survive that.
The full gas-by-gas comparison across assessments is on the AR5 versus AR6 GWP page, and the reporting consequences are set out on the AR6 standards page.
What to do before it lands
The useful preparation is not guessing the date. It is making the eventual switch cheap, which is a data-engineering problem with a known answer.
| Do this | Because |
|---|---|
| Record the assessment basis on every stored factor, not just the value | The single most expensive AR7 outcome is an inventory where nobody can tell which rows are AR5 and which are AR6. Then the migration starts with an audit instead of an update. |
| Stop hard-coding GWP constants | A constant with no expiry is a constant nobody revisits. Resolve them, and if you must hard-code, put the basis and a re-check date in the comment beside it. |
| Pin the data version your report used | So a figure filed today still reproduces after the switch. This is the practice in emission factor version control, and AR7 is exactly the event it exists for. |
| Do not restate early | Reporting frameworks adopt a new assessment on their own schedules, and most still specify AR5 or AR6 explicitly. Moving before the framework you report under does is a divergence you will have to explain. |
| Model one gas per row, per basis | AR6 split methane into fossil and biogenic. Assume AR7 can split something else, and a schema keyed one-value-per-gas becomes a migration rather than an update. |
| Know your base-year policy in advance | A GWP change is a methodology change, which usually triggers base-year recalculation above a materiality threshold. Decide the threshold before the number moves, not after. |
Applying a new GWP to a factor that already carries an old one. Most published emission factors are pre-aggregated carbon dioxide equivalent figures with the assessment’s GWPs already inside them. Multiplying such a factor by an AR7 methane GWP does not update it; it double-counts the methane component. A GWP basis belongs to the factor, and the only correct update is to take the factor whose publisher has restated it — which is why the basis has to be recorded per row.
What we will do when it does
This corpus already holds 44 gases across five bases, including the AR6 20-year and 500-year horizons and the AR4 identities for the 31 gases that have one, all derived from canonical rows rather than transcribed by hand. When AR7 publishes, the work is ingestion and a new basis alongside the existing ones, not a replacement: the older bases stay, because a report filed on AR5 must still reproduce on AR5.
Two commitments follow from that. Every existing value keeps working at the version it was published under, through the reproducibility archive. And the transition itself becomes a machine-generated diff of the same kind published for the annual UK factor release — every gas, old value, new value, and the size of the move, rather than an announcement that the values have changed.
Until then this page states what the IPCC has decided and what it has not. If you want the underlying evidence rather than our summary, the session documents are public, and the wording quoted above is theirs.
Frequently asked questions
There is no agreed date for the three Working Group reports, which is where the GWP values sit. The Panel has deferred that decision at four consecutive plenary sessions, most recently in March 2026, when it made no formal decision but indicated it wants the timeline agreed at the next session. Two AR7 dates are fixed: the methodology reports in 2027, and the Synthesis Report finalised by late 2029. Because the Synthesis Report must follow all three Working Groups, late 2029 is a ceiling on the Working Group reports rather than a schedule for them.
Almost certainly, though nobody outside the author teams knows by how much. The last transition is the guide: AR5 to AR6 moved fossil methane about 6% up, nitrous oxide about 3% up and the common refrigerants by nearly 18%, while biogenic methane went down. It also introduced a fossil-versus-biogenic methane split that had not existed before. Plan for material changes in both directions, and for the possibility of a new distinction rather than only new numbers.
No. Report on the basis the framework you report under specifies, which today is generally AR5 or AR6 stated explicitly. Publication of an assessment and adoption by a reporting standard are separate events, historically years apart. Switching early creates a divergence from your own framework that you then have to explain to an assurer, and it breaks comparability with your own prior years.
The disagreement is procedural rather than scientific. Some governments consider the proposed schedule a workable compromise consistent with past practice. Others hold that it is too compressed, citing overlapping and back-to-back review processes, limited time for national coordination, and the need for a more inclusive approach. The IPCC works by consensus, so a genuine split of that kind produces deferral rather than a vote.
They are part of the AR7 cycle and their dates are agreed, unlike the Working Group reports. Two are due in 2027: one on inventories for short-lived climate forcers, and one on carbon dioxide removal technologies and carbon capture, utilization and storage, whose scientific content was agreed in October 2025. These matter to national inventory compilers and to anyone accounting for removals, and they arrive well before any Working Group report.
One row per gas per basis, with the assessment recorded as data rather than implied by which column it sits in, and the data version pinned on anything you have published. Then adopting AR7 means adding a basis, not rewriting a table, and last year’s report still reproduces on last year’s basis. The expensive alternative is an inventory of bare numbers whose provenance nobody recorded, where the migration begins with an audit.