Jump to content

Wikipedia talk:WikiProject Weather

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

This is an old revision of this page, as edited by Rlandmann (talk | contribs) at 12:17, 13 October 2024 (→Copyright of content from National Weather Service: Reply). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.


2024–25 WikiProject Weather Good Article Reassessment

I would like to announce that a new task force has been created to re-examine the status of every GA in the project. Many good articles have not been reviewed in quite a while (15+ years for some) and notability requirements have changed quite a bit over the years. The goal of this task force is to save as many articles as possible. Anyone not reviewing an article may jump in to help get it up to par if it does not meet the GA requirements. The process will start officially on February 1 and will continue until every article has been checked and either kept or delisted. The task force may be found at Wikipedia:WikiProject Weather/2024–25 Good Article Reassessment. Noah, AATalk 15:22, 26 January 2024 (UTC)[reply]

Articles under review

Merging vs. breaking out storm articles into/from parent articles

I reviewed 2022 Pembroke–Black Creek tornado at FAC, and commented there that I thought it should be merged into the parent season article, Tornado outbreak sequence of April 4–7, 2022. I know I've seen some storm articles merged into their season articles -- Tropical Storm Erick (2007) used to be an FA, for example, but was merged into 2007 Pacific hurricane season in 2013. Does this project have any guidelines on when splits or merges should happen? I'm considering nominating 2022 Pembroke–Black Creek tornado for merging but wanted to check with this project first since I suspect there must have been discussions on this sort of thing in the past. Pinging WeatherWriter, the nominator of that article. Mike Christie (talk - contribs - library) 21:09, 14 September 2024 (UTC)[reply]

It doesn’t need merged. It was broken out due to coverage vs the rest of the outbreak. That single tornado’s article is about 22,000 bytes in size while the outbreak article is 115,000 bytes. The outbreak article is for 89 tornadoes. The “small” (quotes for a reason) section in the outbreak article for that tornado is 5,700 bytes long. Roughly, if you subtrack what is duplicate/near duplicate from the two articles, you are left with about 16,000 bytes worth of content. So that is about 16,000 bytes of content that would be merged…meaning the outbreak article would be 131,000 bytes. Out of that 132,000 bytes, over 16% of the article would be about this single tornado. Note, a merge would mean over 16% of an article related to four days worth of tornadoes…with 89 total tornadoes…would be about a single tornado. That is why it was split out. I have been creating GA-worthy articles for tornadoes when they start reaching 10+% of a large outbreak’s worth of content. One tornado should not be over 16% of an article which is about 89 tornadoes. Right now, due to it being split out, it is actually 5% of the total article. The strongest tornado of the year should have additional coverage (obviously based on RS), but 5% for an article about 89 tornadoes sounds a lot better than over 16% coverage. The Weather Event Writer (Talk Page) 21:23, 14 September 2024 (UTC)[reply]
I am just mentioning this as how it is. If one tornado is indeed the “main” tornado that defines an outbreak, it should not be split out. But instead, the outbreak article is renamed. A good example of this is the 2020 Nashville tornado outbreak. The 2020 Nashville tornado, despite not being the strongest tornado of the outbreak, is 20% of the entire article’s length. Due to it being the tornado of the Tornado outbreak of March 2–3, 2020, consensus formed to rename the article. A similar thing exists for the 1953 Waco tornado outbreak. So, if a merge was ever done, given that single tornado would be +16% of the entire article length, the entire article should be renamed, which consensus almost certainly would not fall behind, as other notable (and well RS-covered) tornadoes occurred that day as well. Basically, there isn’t any clear-cut guidelines on when to split vs not split, but my personal “go-to” is that 10% mark in terms of byte-length. The Weather Event Writer (Talk Page) 21:32, 14 September 2024 (UTC)[reply]

A common discussion in merge discussion is counting the size in bytes (which includes references and coding) versus counting the number of words (readable prose). The outbreak sequence article has around 7,500 words, including 672 words (8.96% percent of the article as is). The 2022 Pembroke–Black Creek tornado article is only 1,273 words, which includes the lead, the synopsis for the event, aftermath, and image captions. If you add the ~600 new words by merging, then the parent article would be 8,100 words, with about 1,200 words dedicated to the EF4 tornado, or 15% of the total word count. I don't think that's unreasonable for the only EF4 of the event, so I agree that a merger is appropriate. The benchmark I think about is the 1999 Oklahoma tornado outbreak and the 1999 Bridge Creek–Moore tornado, which has lengthy analysis and coverage of the single event. That doesn't seem to be the case for the Pembroke tornado. In general, I don't think individual tornadoes should get their own articles if they're part of a bigger outbreak, unless they're extraordinary on their own. ♫ Hurricanehink (talk) 22:05, 14 September 2024 (UTC)[reply]

Instead of pouting, you could work on something that's important enough that won't be merged, such as the outbreak article that the article is the merger target. This isn't targeted toward you WeatherWriter (talk · contribs). You don't own the article, but at the same time, your contributions are generally appreciated. Try putting that editing energy toward something more important, which will likely mean that even more people will appreciate your edits. ♫ Hurricanehink (talk) 22:19, 14 September 2024 (UTC)[reply]
I’m not pouting. In the last 2 months, non-Weather editors have done a 180 to weather articles. Nothing personally against Mike Christie, but at the same time, my dumb choice to do an FAC brought non-weather editors to the article. A lot of out tornado photos are going to be gone from the Commons due to non-weather editors, several long-time weather editors (myself included) are too tired/busy to make articles. Just really annoyed that we have no clear-cut guidelines and everything has changed dramatically in the last 2-3 months. Now, one of things I was proud of is being more or less deleted. Yes, I know it isn’t deleted, but I was proud of getting that article to GA status…now y’all are about to take that pride 100% away. The Weather Event Writer (Talk Page) 22:26, 14 September 2024 (UTC)[reply]
That's part of the process, figuring out better guidelines so they're in line with longstanding Wikipedia policies. It wasn't a dumb choice to do an FAC either, since you felt it was a good article. I'm sorry you went through that effort and now you're finding out that your work might not be kept on Wikipedia. That does suck, and there are a lot of users who can probably relate to that feeling. At the same time, there is a right way of doing things, and sometimes that takes figuring things out, even if that includes input from meddling outside editors :P If you want to salvage as much as you can, then you can help participate in the process of making the outbreak a featured article. You're already invested in that topic. I don't think it will be that difficult, considering you've already tackled the most important part of the outbreak. Try not to be discouraged though. ♫ Hurricanehink (talk) 22:30, 14 September 2024 (UTC)[reply]

I've gone ahead and proposed the merge, here; please add !votes and comments there. WeatherWriter, Hink says above that plenty of other users can sympathize with your situation, and I agree. Plenty of the articles I've created have been deleted; it's annoying, but you might think about it as a graduation process -- it's certainly part of how I learned the rules here, and it was no fun for me either to realize I'd wasted my own time on some things. I've been here a long time and I think I understand the rules but I still had a new article of mine taken to AfD not that long ago. It happens. And don't feel like your content is at risk of being destroyed; if the merge happens much of what you wrote will survive in the target article. Mike Christie (talk - contribs - library) 01:44, 15 September 2024 (UTC)[reply]

Regenerating Storms

I'm proposing that {{Infobox tropical cyclone}} be modified to have formed1, dissipated1, formed2, dissipated2, and formed3, dissipated3. It'd be rendered as "Formed", "First dissipation", "Reformed", or something like that. The idea is to better show the durations of storms that regenerated, like John. What do we think of this idea? Jasper Deng (talk) 07:07, 26 September 2024 (UTC)[reply]

I can see it working in the NHC areas where regeneration happens all the time, but not in the non NHC areas where systems are more likely to be carried as a tropical depression etc.Jason Rees (talk) 17:25, 26 September 2024 (UTC)[reply]
Don't we treat dissipation as being the dissipation of the extratropical cyclone after the fact (as in Hurricane Wilma) and not when it loses tropical characteristics? John as a disturbance never died. I think the existing infobox is fine, we don't need to address every nuance in the infobox. It is meant to be an at a glance summary. DJ Cane (he/him) (Talk) 18:04, 26 September 2024 (UTC)[reply]
Not quite. A system retains a low-level cyclonic circulation if it undergoes ETT. John didn't do that–its LLCC completely dissipated and it opened into a trough. That's why its (first) dissipation was declared with a "Remnants of John" and not "Post-tropical Cyclone John" heading.--Jasper Deng (talk) 19:27, 26 September 2024 (UTC)[reply]
I still think this is overly detailed for an infobox. That information belongs in the prose. Note that even as a trough, John was still a weather disturbance and the vast majority of people associated the system's impacts with that name. DJ Cane (he/him) (Talk) 13:32, 27 September 2024 (UTC)[reply]
Saying it "belongs in the prose" on its own is a non-argument. It is more important to be correct than to be brief, and I would not say the "vast majority" of people associated the intermediate rains with "John". Also, John is just one example. Another is Harvey of 2017. Making data structured this way will also facilitate its import into Wikidata; prose by definition is not machine readable, unlike infoboxes.--Jasper Deng (talk) 00:09, 28 September 2024 (UTC)[reply]
@Jasper Deng: Infobox tropical cyclone is no longer used so I assume you meant this for {{Infobox weather event}}? Noah, BSBATalk 19:20, 26 September 2024 (UTC)[reply]
Yes, I mean infobox weather event.--Jasper Deng (talk) 19:27, 26 September 2024 (UTC)[reply]
I'd be fine with this being an option. Noah, BSBATalk 00:51, 30 September 2024 (UTC)[reply]

RSN discussion about AccuWeather's damage estimates

I started an RSN thread at Wikipedia:Reliable sources/Noticeboard#AccuWeather for damage estimates about the efficacy of AccuWeather's damage estimates in the context of Wikipedia. Feel free to participate. GeorgeMemulous (talk) 01:19, 29 September 2024 (UTC)[reply]

Excessive Rainfall Outlooks

Per @ChessEric’s reply on the List of SPC high risk days talk page. I would like to propose creating a list of WPC high risk excessive rainfall days. We don’t get that many of them and they weren’t introduced but maybe a decade ago. Hurricane Clyde 🌀my talk page! 13:48, 2 October 2024 (UTC)[reply]

New tornado articles and the news

I was wondering if it's a good idea to change the project's ideas for notability for new tornado articles that are considered newsworthy. The example I'm immediately going to point to is the 2024 Greenfield tornado, which took until 8 July 2024 to exist in mainspace. SirMemeGod nominated a previous mainspace article for deletion on 30 May: This article was created way too prematurely. See Wikipedia:Articles for deletion/2024 Sulphur tornado for a good example of this “too soon” policy.

WP:TOOSOON is an argument I see all too often when it comes to articles on recent individual tornadoes and I'm getting a little sick of it, especially here. The tornado had already killed 5 and was in the news cycle for its DOW-measured windspeeds of 263–271 mph (423–436 km/h), not the devastation to Greenfield. If BEST (the FARM research project that put the DOW fleet in Iowa that day) wasn't happening in 2024, it wouldn't have an article, but since it was, we had factual evidence that it was much more powerful than its assigned rating. Don't rush to delete articles is part of WP:DEADLINE, which also applies, as everyone knew that it would get news coverage for its DOW measurements eventually, and sure enough, by 24 June 2024, the DOW team calculated the figure of 309–318 mph (497–512 km/h) inside the tornado.

Although Greenfield wasn't as sensationalized as I would have expected, it also did have significant news coverage after it happened. This leads me to my point. If, say, an EF5 tornado moves through a city, and DOW catches 320mph+ wind measurements / calculations / whatever, under the current guidelines, the article is likely to be deleted until well after it occurred. My proposal is to institute a specific set of in-project guidelines for notability so that way a repeat of this situation doesn't occur.

From my understanding, the articles for 2013 El Reno tornado and 2021 Western Kentucky tornado are the best examples of tornadoes that were obviously notable enough for an article. They had a concrete death toll, record-breaking qualities, and were significant in the weather community. 2023 Rolling Fork-Silver City tornado and 2021 Tri-State tornado were both created well after they occurred. If a tornado is in the news cycle, just the one tornado, and it has record-breaking qualities or a high death toll or something else about it, like Greenfield did, then it shouldn't wait to get an article, after the significance is established.

I propose the following guidelines for establishing the notability for articles on recent (i.e. less than 1 month since they occurred) tornadoes to avoid the old WP:TOOSOON deletion arguments and all that. Note that not all of these criteria must be met.

  • 1. The tornado causes over 20 deaths. (i.e. 2021 Western Kentucky tornado)
  • 2. The tornado impacts a smaller-scale location, where significant devastation occurs. (none have happened recently, but think the Indiana State Fair stage collapse of 2011 if the tornado was confirmed, or the 1967 Belvidere tornado which caused 300 injuries at a school. If the only notable impact is at the location, then the article should be about the impact there, not the tornado itself.)
  • 3. The tornado causes a significant amount of damage or devastation in a city of over 25,000. (i.e. 2011 Joplin tornado. The population figure is arbitrary, as is the amount of damage, but Joplin would still have been worthy of its own article before surveys were complete and before the death toll was calculated.)
  • 4. The tornado sets a record for path length, path width, or another objective indicator. (i.e. 2013 El Reno tornado)
  • 5. The tornado has a measured or calculated wind speed of over 300 mph (480 km/h). (i.e. 2024 Greenfield tornado)
  • 6. The tornado is rated EF5, IF5, F5, or equivalent. (i.e. 2013 Moore tornado)
  • 7. The tornado receives significant news coverage from non-weather and international sources. (i.e. 2021 South Moravia tornado. This is a catch-all clause that can save any tornado article as it is the clearest example of setting notability. However, the articles must be on the individual tornado, not on the parent outbreak.)

Some other notes:

  • Any records must be confirmed beyond a reasonable doubt to have been surpassed. Wind speed measurements or calculations should be confirmed by the agency that produced them, and any calculated figures must be calculated by an expert and not synthesized by Wikipedia editors. (WP:SYNTH)
  • The tornado does not need to have a final NCEI event database entry to have its own article.
  • Tornadoes rated EF5 or equivalent must be confirmed by the relevant agency that rates them.

Tornadoes that meet any of these criteria will be spared from WP:TOOSOON deletions, as they will have clearly shown their notability already. This is assuming, of course, that everything is confirmed by independent coverage; if it is only assumed that the criteria are met, then a TOOSOON AfD can go through.

Some examples of tornadoes that don't quite meet these criteria:

  • 1. 2023 Rolling Fork-Silver City tornado. It only caused 17 deaths as a high-end EF4 tornado.
  • 2. Tornado outbreak of March 31 – April 1, 2023#Belvidere, Illinois. I'm referring to the EF1 tornado that caused the collapse at the Apollo Theatre in Belvidere, Illinois that killed one. It falls short of the objective devastation I propose, and doesn't even have its own section.
  • 3. 2023 Little Rock tornado. Yes, over 600 injuries were reported in a major city of 200,000, however that number fell steeply to 54, with only one indirect fatality occurring.
  • 4. 2021 Quad–State tornado. If we listened to the initial reports of what would be the Tri-state and Western Kentucky tornadoes, this would have had a path length of 230 miles. While perhaps the Western Kentucky part did set a record, it wasn't the figure of 230 that would have easily beaten the 1925 Tri–State tornado.
  • 5. 2024 Harlan, Iowa tornado. Before Greenfield, this was a big news story in the weather community with wind speeds of 224 mph (360 km/h) calculated at ground level. That's IF4 on the International Fujita Scale and not quite newsworthy in this context.
  • 6. 2023 Keota, Iowa tornado. This had an EF5 DI on the preliminary Damage Assessment Toolkit (DAT) that Wikipedia editors use so often that was confirmed to have been an error. It was, however, confirmed to have done exactly one point of low-end EF4 damage, which does technically make it violent per the EF scale.
  • 7. 2024 Hollister tornado. It was the subject of much local coverage that stated it was "the most powerful tornado of all time" but of course it only did EF1 damage and all we had to go off of was radar. Its parent outbreak doesn't have its own article, and when Trey Greenwood did an analysis he found that it was not the most powerful tornado of all time and indeed probably wouldn't even have been a violent one.

I know this is a lot of text, but these guidelines will help us if an individual tornado gets nominated for ITN (most likely, I'd be the one to put it there), or if it gets enough news coverage that readers would go out of their way to find an article on it specifically, independent of its parent outbreak. Let me know what you think of this proposal. GeorgeMemulous (talk) 13:57, 4 October 2024 (UTC)[reply]

Pinging a few editors that I'd guess would be interested in this sort of thing: @ChessEric @Sir MemeGod @Hurricane Clyde @Tails Wx @TornadoLGS @WeatherWriter @Алексеев Н. @Hoguert @Hurricane Noah
Summary: I've written a proposal above concerning WP:TOOSOON nominations in WPW space. If agreed to, tornadoes meeting any of the 7 criteria outlined above will be exempt from TOOSOON nominations. However, they still must have reliable sources proving they meet the exemption and can be deleted for other reasons; this is merely a way to have articles about new tornadoes that have a high-impact or newsworthiness, such that they can appear on ITN or elsewhere, where traffic to the individual tornado's article warrants it to be independent of its parent outbreak. This does not affect any existing articles, but will affect any future tornado articles about tornadoes that are created under one month after they occur. GeorgeMemulous (talk) 18:57, 4 October 2024 (UTC)[reply]
I agree to all the suggestions except 4 and 5. Also, I think the Little Rock tornado could have its own article. The 2023 Wynne–Parkin tornado article is not needed though. ChessEric 03:26, 5 October 2024 (UTC)[reply]
Comment: I will formally have a vote later tonight, but I will note that a certain set of articles can't be "exempt" from a rule (with extremely rare cases obviously). TOOSOON is 100% a relevant policy, and a WikiProject can't overrule that. I'm even going to be bold and say that the Greenfield tornado doesn't meet LASTING, but that's just my opinion and I won't start a second AfD because I'm not dead-set on deleting a well-written article. Anyways, that's aside from the point. :) SirMemeGod  19:06, 4 October 2024 (UTC)[reply]
Also courtesy ping @United States Man: and @TornadoInformation12:, since I know they'd be interested. SirMemeGod  19:10, 4 October 2024 (UTC)[reply]
Comment This isn't meant to override TOOSOON, rather to better define it within the project space. As it is, one of the main reasons the Greenfield tornado was voted to be deleted was it didn't have an NCEI entry, even though it had plenty of news coverage establishing notability. This proposal is meant to counter that.
Also, LASTING isn't a formal reason to delete I'd think. Moreso, a reason to keep articles that cause events beyond them to happen. GeorgeMemulous (talk) 19:11, 4 October 2024 (UTC)[reply]
LASTING is under WP:NEVENT, a formally-designated policy that can either be used as creation or deletions. I had just said that to add, it wasn't really anything of note (also don't worry, I won't target the Greenfield or Elkhorn articles). :) SirMemeGod  19:13, 4 October 2024 (UTC)[reply]
I would actually support having an exemption to WP:TOOSOON for high impact tornadoes. Especially ones that would be candidates for ITN.
But I also think tornado outbreak articles (only for outbreaks that have already happened) should have some clarification as to when TOOSOON applies (although maybe not an outright exemption). Hurricane Clyde 🌀my talk page! 21:09, 4 October 2024 (UTC)[reply]
But I’d like to also add to the above and courtesy ping @Hurricanehink and @Ks0stm as well; since they participate in weather-related projects. Hurricane Clyde 🌀my talk page! 21:10, 4 October 2024 (UTC)[reply]

I'm going to attach a table of tornadoes and their notability by this proposal, as well as general notability as gauged by my opinions and those of other editors.

Tornadoes since 2013 (updated 5 October 2024)
Tornado 20+ Fatalities Localized devastation Widespread devastation Objective record 300 mph recorded F5 / equivalent Significant coverage General notability Has an article
2011 Joplin tornado Yes No Yes Yes No Yes Yes Yes Yes
2013 Moore tornado Yes No Yes No No Yes Yes Yes Yes
2013 El Reno tornado No No No Yes Yes Maybe Yes Yes Yes
2013 Washington, Illinois tornado No No Maybe No No No Maybe Yes Yes
2020 Nashville tornado No No Maybe No No No Maybe Yes No
2021 Western Kentucky tornado Yes Maybe Maybe Maybe No No Yes Yes Yes
2021 Tri-State tornado No No No No No No No Yes Yes
2023 Little Rock tornado No No No No No No No Maybe No
2023 Keota, Iowa tornado No No No No No Maybe No No No
2024 Sulphur tornado No No Maybe No No No No No No
2024 Hollister, Oklahoma tornado No No No Maybe No No Maybe No No
2024 Elkhorn–Blair tornado No No Maybe No No No Maybe Maybe Yes
2024 Greenfield tornado No No No No Yes No Maybe Yes Yes

These are my view on these tornadoes, if you have a counterpoint to any of these entries feel free to reply. GeorgeMemulous (talk) 14:18, 5 October 2024 (UTC)[reply]

Draft:List of US tornado records by state

I promise this isn't WP:LISTCRUFT as specific as it might sound, but I've started work on a list of tornado records based on the states in which they occurred. Feel free to contribute. GeorgeMemulous (talk) 00:54, 12 October 2024 (UTC)[reply]

That really seems like listcruft. For example when I worked on List of California tornadoes, I had a lot of difficulty finding out the costliest tornado in the state. It seems like a random one in 2015, but I wasn't able to find proper sources for that. I mention that because I don't see any sourcing in the list, and that's probably going to be difficult going state by state. I do think there is a need for tornadoes by state, like List of Arkansas tornadoes, more than having a records list. ♫ Hurricanehink (talk) 15:21, 12 October 2024 (UTC)[reply]
This is mostly so tornadoes that have WP:NWEATHER's statewide record get a place, and plus, everything will be verifiable in time. I'm mostly waiting for the NCEI database to come back online. The only part of WP:LISTCRUFT this might break is the first, that being obscurity, but even then I'm sure there are way more lists that get a pass (cough cough Category:Lists of National Register of Historic Places by county for every single county in the USA, most averaging under one view per day, cough cough). GeorgeMemulous (talk) 15:45, 12 October 2024 (UTC)[reply]
I agree we should have those records, but it's going to be difficult doing every state without having lists for every state. I think you're kind of doing it backwards. If that list was just something like List of US tornadoes by state, I could get behind it. Such a page would have the individual lists for each state, and then could also have a table like this, which would be a useful comparison for, say, the deadliest in each state, or costliest, or strongest, or largest. But just having it as a page on its own seems a bit too specific IMO. ♫ Hurricanehink (talk) 16:05, 12 October 2024 (UTC)[reply]
I understand your position. Perhaps it isn't going to be its own standalone page in mainspace, but I'll keep it in draftspace until the aforementioned list gets created. GeorgeMemulous (talk) 18:52, 12 October 2024 (UTC)[reply]

Copyright of content from National Weather Service

Per the discussion at Commons:Requests for comment/Third-party images published by the National Weather Service, many files from the National Weather Service have been determined to have potential copyright issues. In particular, only some files submitted to the National Weather Service have actually been released into public domain. This means a lot of other files are at the risk of deletion from Commons; some may need to be transferred to Wikipedia for fair use.

@Rlandmann has put together a list of all known Commons files obtained from the National Weather Service. The author and provenance of some of those files have yet to be determined. If anyone here knows more about those files, then please share what you have. We appreciate any assistance! Ixfd64 (talk) 00:18, 13 October 2024 (UTC)[reply]

Thanks @Ixfd64 -- slight correction -- that list isn't all known Commons files obtained from the NWS (that would be a list of over 12,000 files!) Rather, it's a list of all known Commons files obtained from the NWS whose copyright status is currently under question.
Most of these are on the Commons via a purely innocent mistake of thinking that their publication on an NWS website meant that they had necessarily been released into the public domain. A very small number are blatant copyvios, and the copyright status of others is ambiguous for a range of reasons. (And Commons policy defaults to delete unless their free status can be demonstrated beyond significant doubt, via evidence) --Rlandmann (talk) 12:17, 13 October 2024 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:WikiProject_Weather&oldid=1250934628"