Talk:Instrumental temperature record/Archive 3

Expert review (Feb 2022)

I've received comments about this article in a Word document by Blair Trewin, one of the authors of the IPCC Sixth Assessment Report. This is done as part of this project for 3 SDGS. I will try to enter his comments into the Wikipedia article over the coming days/weeks and also raise discussion points here on the talk page. In case someone is watching this page and has worked on this Wikipedia article a lot and would be interested to collaborate, please reach out to me. In that case, I could e-mail you the marked up Word document for your careful consideration. I am a complete non-expert on this topic, so will struggle. EMsmile (talk) 22:49, 23 February 2022 (UTC)[reply]

Here are some of the main comments from the Word document that Blair provided to me. Do people broadly agree with his suggestions? (I am happy to forward the Word document to anyone who's interested, just send me an e-mail through the system):

Change citation to IPCC AR6 WG1 (these numbers are also in SPM). Think it best to present consolidated assessment here rather than individual data sets for consistency.
AR6 findings. Since AR6 uses changes from one period to another (rather than linear trends) as the primary temperature change metric, there is not an assessment finding re: linear trends but results are reported in a table. Suggest citing Chapter 2, Table 2.4 - the final chapters are still being typeset so page numbers are not necessarily stable, but table numbers should be.
Regarding the figure that is currently under "warmest years": This figure is about records at the regional level, not in global means - I would move it to the regional temperatures section and introduce a discussion on extremes there.
I think we want to structure this in such a way that it can be updated with each new year's data with fairly minimal disturbance. I've suggested a replacement paragraph below. I also suggest we focus here on WMO and IPCC as consolidated sources of information, while noting that they draw on a number of data sets (which I then suggest we discuss in detail in the 'Global temperature since 1850' section).
Regarding the paragraph that starts with "While record-breaking years can attract considerable public interest": I'm not sure this paragraph really adds very much. The key points we want to capture here are that there is a long-term warming trend, that there is variability about this trend because of natural sources of variability (e.g. ENSO, volcanic eruption), that not every year will set a record but record highs are occurring regularly.
Regarding the table "Top 10 warmest years (NOAA)": We will need to take a decision on which dataset to use here. IPCC uses the mean of four different datasets - there are advantages in doing this (including that it is expressed relative to 1850-1900), but at the moment there is a loss of transparency since the consolidated dataset itself will be accessible through the IPCC data portal which is not yet operating. Is archiving the time series on Wikipedia itself an option?
Regarding the image under "warmest decade": I like the format of this image, but it will need to be updated for the latest dataset versions. It's also unclear which underlying dataset is used (presumably not NOAA given the 1850 start date). (NB the source data is explain when clicking on the image but perhaps a reference in the caption itself would be good?)
Chapter 2 executive summary (currently p6), also in the main body of the chapter but expressed slightly differently. This is the only AR6 finding which compares the last decade specifically to paleoclimate results, although there are other relevant findings on paleoclimate timescales (e.g. rate of warming in last 50 years).
Regarding the table with temperature anomalies: Whichever dataset we end up using here, I suggest we use years ending in 0 to align with the IPCC reporting to 2011-2020 (and so we don't have a "decade" with only 1-2 years of available data.
Regarding the graph where the caption starts with "Colored bars show how El Niño years": This looks like something that's been produced specifically? (WMO has from time to time published something conceptually similar in its annual State of the Climate reports but hasn't distinguished strong from weak events). If it doesn't have a defined source then we probably need to have a link somewhere to how the strength of El Nino/La Nina events is defined.
Regarding the paragraph about El Nino: There will be papers which can be cited here. (As it happens, I've got one myself in review, but there are others which have been published). The new text can largely be inferred from the graph (assuming it stays in its current form). Note that the 1982-83 El Nino has a weak temperature response because it was offset by the El Chichon eruption.
Presumably there will be links here to pages on the Pinatubo and Agung eruptions? Larger volcanic signals are sometimes quoted but these are typically either not for a full year, or are model simulations rather than observations. The effect of Pinatubo was also partially offset by El Nino conditions which prevailed for much of the 1991-94 period.
Either here, or in the 'Temperature anomaly' article, there needs to be a discussion of which baseline periods are in use, the purposes they serve, and why the choice of baseline has no impact on long-term trends.
A picture of a Stevenson screen might be useful here. The MMTS name covers both the instruments and the shelter so shouldn't be used for the shelter alone.
Buckets are almost out of use now but were heavily used historically. There are papers which can be cited here.
I think either as an addition to this section, or a new section (I prefer the latter), there should be a discussion of how global temperature datasets are developed from the base observations, and a listing of the main datasets. Reanalyses, and how they are used/not used, could also be introduced here - I see there is a Wikipedia article for 'Atmospheric Reanalysis', although it could do with some more content. (I'm not a reanalysis expert but can point you to some if you want that article strengthened).EMsmile (talk) 17:36, 24 February 2022 (UTC)[reply]

Hi User:RCraig09 what are your thoughts regarding the comments by Blair Trewin which I posted above? Would you have time to involve yourself to address them? (you appear to be a content expert on this topic as well as a Wikipedia expert which is an ideal combination). Feel free to send me an e-mail through the system if you'd like to see the mentioned Word document. EMsmile (talk) 18:43, 4 March 2022 (UTC)[reply]

@EMsmile: I'm not a subject matter expert (just an engineer who is interested in climate change). You may want to post a note at Wikipedia talk:WikiProject Climate change to get a broader, more expert response. I can respond to issues on the graphics I created:

3. "Warmest years" graphic: I think that the graphic is appropriate where it is, since it tallies temperature records worldwide. The already-crowded Regional temperature section is actually less appropriate since the graphic isn't concerned with regions per se.

7. "Warmest decade" graphic: Source data is HadCRUT from Met Office (UK), with specific links in the Wikimedia file description. I didn't put a footnote in the graphic's caption here because the source provided the data but not the caption's content that can be deduced by looking at the graphed data. I think it's superfluous to cite source data in this instance; so I didn't. Separately, it shouldn't be a big worry updating data, since a new bar won't be added until 2030, and dataset variations are so small that they probably won't visually change the shape of the existing bars in the meantime.

10. The "El Nino etc" graphic: Yes, the graphic was specially made from WP:Routine calculations (adding within timewise groupings of monthly data in a Microsoft Excel spreadsheet). The sources for the El Nino/La Nina year data, and for the temperature data, are provided on the Wikimedia file description page. Most Wikipedia readers won't be interested in the source data.

Again, Wikipedia talk:WikiProject Climate change may inspire more informed responses on other issues. —RCraig09 (talk) 19:54, 4 March 2022 (UTC)[reply]

Thanks, have written at Wikipedia talk:WikiProject Climate change now. I feel bad towards the reviewer that I haven't been able to address his comments yet. Maybe I'll start by sending him your explanations about the graphics as a starting point. EMsmile (talk) 13:09, 16 March 2022 (UTC)[reply]

Update: I have now addressed all the changes that Blair Trewin had made (as listed above). I need a few more references in some cases; have asked him about that now. I have also culled some of the content that was in the last quarter of the article as it was outdated and is better available at effects of climate change and climate change.EMsmile (talk) 15:34, 17 March 2022 (UTC)[reply]

How global temperature datasets are developed from the base observations

Comment by Blair Trewin: "I think either as an addition to this section, or a new section (I prefer the latter), there should be a discussion of how global temperature datasets are developed from the base observations, and a listing of the main datasets. Reanalyses, and how they are used/not used, could also be introduced here - I see there is a Wikipedia article for 'Atmospheric Reanalysis', although it could do with some more content. (I'm not a reanalysis expert but can point you to some if you want that article strengthened)." EMsmile (talk) 14:29, 17 March 2022 (UTC)[reply]

Continued review with content expert

I am continuing the article review with content expert Blair Trewin. He sent me the following: "I think there's two significant points to make a decision on:

Now that the IPCC temperature data set is available in a publicly accessible archive, I suggest that we should use that where appropriate, e.g. in the two tables that currently use NOAA data. If you agree I can provide the numbers.
I think the section that's currently about satellite temperature data should instead be reframed as upper atmospheric temperature, referring to the main sources of information – satellite (with link to the article on that), reanalyses and radiosondes. This would largely draw on the material in IPCC AR6 section 2.3.1.2. If you agree I can draft some text on this.

Main other things outstanding from my perspective:

Drafting some further text for the lead, and a few other bits where flagged
Adding a wider range of citations more generally"

Comments/objections on these suggestions changes by anyone? If not, we'll go ahead. EMsmile (talk) 10:27, 22 March 2022 (UTC)[reply]

Comment from Blair: " I see from the talk page that there were no objections to our proposed restructurings/additions, so can go ahead with those. (The upper atmosphere part also addresses what you mentioned in the talk page under the heading 'removed content about global temperatures')." EMsmile (talk) 08:45, 30 March 2022 (UTC)[reply]

Questions about some of the graphs

About the graph in the section on warmest years: "Global area reaching record temperatures"

Comment by Blair about this graph: "This figure is about records at the regional level, not in global means - I would move it to the regional temperatures section and introduce a discussion on extremes there." Answer by RCraig09 above was: "I think that the graphic is appropriate where it is, since it tallies temperature records worldwide. The already-crowded Regional temperature section is actually less appropriate since the graphic isn't concerned with regions per se." Answer by Blair: "It's a global aggregation of regional results (i.e. it is the percentage of gridpoints which set a record in each year). It doesn't relate directly to the global average temperature which is what is being discussed at this point." Do we follow Blair's argument? EMsmile (talk) 15:18, 24 March 2022 (UTC)[reply]

The thrust of this graphic is global area. Granted, that's not exactly the same as global average, but it's much more relevant to the global viewpoint than to the distinctions between regions that is the subject of the tiny "Regional temperature changes" section. It's OK to maintain the current mix of global area and the global average temperatures "that is being discussed at this point." —RCraig09 (talk) 16:39, 24 March 2022 (UTC)[reply]

Answer from Blair: "on the first graph - if it's going to stay where it is, it probably just needs a bit of text to better link the graph to the text and explain what it's telling us - I can do this. EMsmile (talk) 08:44, 30 March 2022 (UTC)[reply]

About the graph with the caption "Colored bars show how El Niño years (red, regional warming) and La Niña years etc.

Comment by Blair about this graph: "This looks like something that's been produced specifically? (WMO has from time to time published something conceptually similar in its annual State of the Climate reports but hasn't distinguished strong from weak events). If it doesn't have a defined source then we probably need to have a link somewhere to how the strength of El Nino/La Nina events is defined". My reply to him: "The full explanation of the graph is accessible when clicking on the graph: https://commons.wikimedia.org/wiki/File:20210827_Global_surface_temperature_bar_chart_-_bars_color-coded_by_El_Ni%C3%B1o_and_La_Ni%C3%B1a_intensity.svg But I don’t understand it fully: does it explain the definition of strength for these events?" - His next reply was: "It does, although the text has a "very strong" category, which looks as if it is grouped with "strong" on the graph." EMsmile (talk) 15:30, 24 March 2022 (UTC)[reply]

In making this graphic, I followed the data categories in the original source. I didn't group "strong" with "very strong"—they're distinct colors for distinct categories. —RCraig09 (talk) 16:45, 24 March 2022 (UTC)[reply]

Answer from Blair: "on the second one - I see what's happening now - there's no 'Very strong' label for La Nina since no year has yet reached the 'Very strong' criteria for La Nina (whereas some have for El Nino). I still think it might be better to add 'Very strong' to the La Nina legend for balance even though it isn't used (yet), but don't feel strongly about this." EMsmile (talk) 08:44, 30 March 2022 (UTC)[reply]

Hi RCraig09, I am just wondering where we stand with this now? See my comments here and just above from 30 March. EMsmile (talk) 10:19, 4 May 2022 (UTC)[reply]

Absent specific substantive reasons, I will not change the graphic away from what is disclosed in the reference. In particular, I do not plan to add a 'Very strong La Nina'—"for balance", as suggested—as that would wrongly suggest such years exist. —RCraig09 (talk) 15:37, 4 May 2022 (UTC)[reply]