Wikipedia talk:WikiProject Climate change

MainParticipantsPopular articlesRecommended sourcesStyle guideGet started
with easy editsArticle alertsTalk

Climate non-profit looking for review of History updates

Hi, I'm an employee of ClimateWorks Foundation, acting on the organization's behalf. I'm looking for help from editors to review updates to the History section.

On the Talk page, I've explained more about what I'm proposing in a request but in short: the current History is outdated, overly focused on the founders, and missing information about key milestones. For example, the most recent detail is from 2018, and most of the current section is sourced to a news piece from 2009. I've written a succinct draft for editors to look at. Would someone have time to review? Thanks! TDM for ClimateWorks (talk) 17:44, 10 June 2026 (UTC)

Hi all, thanks to the help from a member of this WikiProject, the article is in much better shape - thank you! I'm returning here to ask if anyone has a moment to look at a request? This would fix a detail in the infobox and streamline the intro for the ClimateWorks page. TDM for ClimateWorks (talk) 23:32, 13 July 2026 (UTC)

Edit request at Talk:Balcony solar power

I have posted a COI edit request at Talk:Balcony solar power proposing a short, sourced addition on advocacy organizations in the United States section. Review from uninvolved editors would be appreciated. There is also an older unanswered request there (from June) about updating legislation figures, from a different editor. (I have a declared conflict of interest; details in the request.) Zhoo22 (talk) 18:38, 15 July 2026 (UTC)

Draft:Bogslide

Hello! I've been trying to create this article, however, I'm having trouble with it as it is quite technical, and I usually make articles on more specific things, like shopping centres or public sculpture, and am having trouble with something more broad. Would anyone be interested in helping create it? Someone recommended I reach out to specific wikiprojects so im doing that here now Froglord (Finnfrog99) (talk) 23:33, 31 July 2026 (UTC)

Rebuilt Holocene Temperature Variations figure, feedback wanted

I have rebuilt the 2005 Global Warming Art figure from current published reconstructions, keeping its layout and replacing its contents. I would like comment before proposing it anywhere.

• 
  Current figure, prepared by Robert A. Rohde in 2005
• 
  Proposed replacement (updated version shown below)

Bringing it here because the 2005 file is used on around a dozen articles and on Template:Holocene sidebar, so it seemed better to have the discussion once in a venue that spans them than to open the same thread repeatedly.

The case for replacing it. The original plots eight individual local proxy records with their unweighted average, on a mid-twentieth-century baseline. It has held up remarkably well as science communication, but it predates the Temperature 12k database, the Last Glacial Maximum Reanalysis and the PAGES 2k Common Era reconstruction, and its own Commons file page describes its average as "only a rough, quasi-global approximation" while noting that better data "could significantly alter the apparent long-term behavior".

In this version, every thin line is a published reconstruction plotted as its authors published it: Kaufman et al. 2020, Osman et al. 2021, PAGES 2k 2019 and Leclercq and Oerlemans 2012 as global means, with Marsicek et al. 2018 and Viau et al. 2006 drawn dashed as continental records and excluded from everything else. All are on 1850-1900, the IPCC pre-industrial period, using the conversions published in Kaufman and McKay 2022 for exactly these archived files, so values can be read directly against the way modern warming is normally reported.

There is no editor-computed average. The shaded band is the union of the published 5th-95th ranges of the two global full-Holocene reconstructions. Those ranges do not overlap across roughly half the record, including the whole mid-Holocene, so the band shows genuine disagreement between reconstructions rather than a single confident line through the middle of it.

On the splicing question, the instrumental marker is a decadal mean drawn as a separate point and never joined to any curve, following the convention in IPCC AR6 Figure SPM.1 where both reconstructed and observed series are decadally averaged. The resolution limit is stated on the figure itself: these reconstructions resolve roughly 100 to 200 years, so a decadal excursion anywhere in the Holocene would leave no trace on them.

The full source list, every numeric offset applied with its citation, and the known limitations are on the file's talk page on Commons. The build code and processed data are not hosted yet; I will post a link once they are, and I am happy to send them to anyone who wants to check a number in the meantime.

Any feedback, questions, comments appreciated. Dawei20 (talk) 02:57, 4 August 2026 (UTC)

Thanks for your obviously substantial efforts. On substantive notes: your chart conveys a greater sense of confidence in the science, compared to the ~ramdomness that the 2005 chart projects to lay readers. Kudos for choosing a pre-industrial value as a baseline. On formal notes: a general graphical guideline is that important text (like that you now have in a separate legend) should be comparable in size to the adjacent narrative text in Wikipedia articles. Caveats, sourcing, micro-techy-explanations, etc. can be either omitted, placed in the separate textual caption or on the Commons file description page, or in tiny text in non-distracting places such as the very bottom border: we have a lay audience that we don't want to distract or overwhelm. Related: if possible, data visualizers tend to favor labels being closely associated with the data traces they represent (simple example) rather than being in a separate legend (avoiding whiplash of viewers' heads). Increasing the width of the data traces (lines) would improve readability when readers first see it (in thumbnail view). Thanks for tackling this very important topic. —RCraig09 (talk) 03:29, 4 August 2026 (UTC)

Good points. Something like this a little closer? I removed extra text, thickened the lines, put the dataset names right on the graph, moved more things to the main file description.

Dawei20 (talk) 17:02, 4 August 2026 (UTC)

Thanks for your work. The remaining issue I see is font size. By convention, significant text should be at least the same size as the narrative text surrounding the graphic. That should be a quick change; you can simply upload a new version of the same chart, so there's no need to form a completely new file for minor changes. —RCraig09 (talk) 19:50, 4 August 2026 (UTC)

Sure, I just updated a version with uniform font size .. gets a bit crowded on the right, I may try to tweak it a bit more, but anything else you see? Also, if I wanted someone knowledgeable to really review the methodology and assumptions, like someone who has experience building these kinds of paleoclimate graphs, is there any regular contributor here that you think might be willing? Dawei20 (talk) 00:22, 5 August 2026 (UTC)

I ping @Femke: a (busy) subject matter expert and experienced Wikipedia editor. Aside: the text is still not large enough to compare to surrounding text. —RCraig09 (talk) 04:39, 5 August 2026 (UTC)

Great thank you. Sorry just to make sure I'm clear, you mean like all text should be exactly the same size? Or which part specifically should I update? I'm not familiar with the formatting rules. Dawei20 (talk) 19:57, 5 August 2026 (UTC)

There is no specific "rule". However, I try to make all significant text as big as adjacent narrative text. If text is not significant, I make it very small and put it at the bottom (like a footnote), so that it is not distracting to casual readers. —RCraig09 (talk) 22:35, 5 August 2026 (UTC)

I just pushed another update, hopefully the font is coming through better now. It looks like Femke is rather busy, anything else you think I should do? Such a well-known chart, I hesitate to just directly edit and replace it myself. Dawei20 (talk) 02:11, 7 August 2026 (UTC)

Thanks; the font size is adequate. On Talk:Climate change or Talk Pages of other pertinent articles, you might try posting a link to this discussion. (Don't start separate, parallel discussions.) You could also post a link on User talk:Dragons flight as a courtesy, though he is undoubtedly busy—with only two Wikipedia posts so far in 2026. I personally prefer your image as being clearer (less ~frenetic-appearing) for our lay readers, but I'm not expert enough to judge the relative substances of each. —RCraig09 (talk) 16:43, 7 August 2026 (UTC)

First of all, super excited to see people delve into graph making of such an important topic! Trying the image out at an increased size, to see if the fonts become legible. Feels still a bit too small, the text.

It's been a while since I delved into Holocene temperature reconstructions. I'm somewhat surprised by the large uncertainty band. I looked into the extent to what the 2021 superseded previous research, but I don't believe that is true, after reading this lovely review paper. In particular, the 2021 paper explains why the differences between the methodologies (e.g. proxies-only studies cannot deal well with sparse and low-quality data, while combining physics + proxies can). But the review paper argues that there may be similar biases in the chosen ESM.

That leaves us with how to combine the two uncertainty ranges. Should we have a lighter colour where they do not overlap? I think that might convey more clearly that the outlier of 1.8C is only supported by a paper that is a bit of outlier. Perhaps the colouring in of the last 2k/instrumental temperature record can be removed to keep the graph cleaner? In solidarity, —Femke (talk) 🐦 17:30, 7 August 2026 (UTC)

Thank you for checking it! Yes the uncertainty band was a combination of Kaufman and Osman's own uncertainties, the union of the two published 5th - 95th ranges, the lower of the two 5th percentiles up to the higher of the two 95th percentiles.

It just seemed like a clear example of "they can't both be right" which I feared would confuse people, like of course the answer could be somewhere in the middle, but here is a version where the respective ranges are split out again. I also removed the shaded areas on the right.

• 
• 

Dawei20 (talk) 19:26, 8 August 2026 (UTC)

FYI, made it live on Holocene Climatic Optimum article (only). Dawei20 (talk) 00:49, 17 August 2026 (UTC)

Since there *was* no Holocene Climatic Optimum, according to the Osman et al (2021) analysis, putting this figure on that page demands a substantial re-write of the text. Further, if the consensus of experts is that the Osman et al result is misleading, it probably should not be included in the plot, but just mentioned in the text as a result disagreeing with the consensus view. PaddyLeahy (talk) 10:07, 18 August 2026 (UTC)

To my understanding, there isn't really a consensus that Osman et al. 2021 is misleading. It does make the idea of a well-defined warm period that "ended" around 5,500 BP (as the article claims), seem less certain. But this is new evidence, and we should go where the evidence leads us.

From what I can gather, it is an open question. Kaufman & Broadman, 2023 published in Nature called it a "conundrum". From the abstract:

"From a large variety of available evidence, we find support for a relatively mild millennial-scale global thermal maximum during the mid-Holocene, but more research is needed to firmly resolve the conundrum and to advance our understanding of slow-moving climate variability."

But yes, definitely a lot of the article may need to be updated. I can help with that.

Dawei20 (talk) 02:04, 21 August 2026 (UTC)

Relevant deletion discussion on Wikipedia:Articles_for_deletion/Frontier_Climate

It would be useful to have a few more thoughts on the notability criteria there. Know its in the monitoring list, but folks aren't always checking Sadads (talk) 22:36, 4 August 2026 (UTC)

Representing estimates

I'm doing a bit of editing on 2026 United Kingdom heatwaves. When it comes to reporting on heat-related deaths there are a couple of slightly different estimates available. Is there consensus on how to present them?

Estimate 1 comes from a group of researchers from Imperial College London, the Met Office, and the London School of Hygiene and Tropical Medicine and estimate 2 comes from the UK Health Security Agency so they're both reliable sources.

Estimate 1's 'main findings' presented rounded numbers and when it comes to news reporting on the results the figures are given to two significant figures. Presumably this is to help people understand that it's an estimate. However, estimate 2 does not round the figures at all. In the main points the totals are presented on their own, while in the main body of the report they are presented with confidence intervals.

• Is it acceptable to use the same level of rounding so that one does not appear to be more precise than the other even though this may not be how the report presents them?
• Should we include confidence intervals and if so where (infobox, lead, main body of the article)?

🏰 Richard Nevell (talk) 10:38, 9 August 2026 (UTC)

@Richard Nevell estimating heat deaths is notoriously variable, I don't have a strong recommendation one way or another. One place to look and ask questions would be somewhere like WP:WikiProject Hurricanes which has more experience compiling estimates of damage across larger disaster events, Sadads (talk) 23:03, 16 August 2026 (UTC)

I personally, would round with heat deaths because most hospital systems don't have an accurate way to count -- and unless there is an accepted standard for the geography, I would put both estimates in the 100s or thousands range, -- these numbers are going to be reevaluated 2 or three times probably, Sadads (talk) 23:05, 16 August 2026 (UTC)

Thank you, that's helpful. Rounding to the nearest hundred is an improvement but is still more precise than the range. The figure of 753 deaths in the May heatwave given by the Environment Agency is in the range of 529 to 977, so 800 works.

I think where there are slightly different estimates they'll have to be presented separately. I considered using the rounded midpoints as the upper and lower bounds of a range but that wouldn't represent either of the ranges presented by the estimates. 🏰 Richard Nevell (talk) 20:18, 21 August 2026 (UTC)

And you can explain how you got there in the footnote (i.e. include a quote from the original source), Sadads (talk) 20:31, 21 August 2026 (UTC)

Sadads's suggestions reminded me that Wikipedia talk:WikiProject Medicine is another good place to ask this kind of question. Clayoquot (talk | contribs) 02:36, 17 August 2026 (UTC)