Talk:List of highest mountains on Earth

Original research?

While I appreciate the amount of effort that clearly has gone into this list, I'm concerned that a large portion constitutes WP:OR. I say this because one of the notes states "" and another states "". I understand that working with primary sources is necessary here, but an analysis and interpretation of raw data seems to suggest that this list was generated using original research. Are there are any other sources that could be used to help support this article? --NoGhost (talk) 17:55, 17 April 2017 (UTC)

References

Yep, it was a lot of (fun) work. At the time there was no proper "list of highest mountains" out there; newer ones may be derivatives of wikipedia's list. We do give a link to the "List of highest mountains down to 6750 metres" by Eberhard Jurgalski, which he compiled separately a little later, though also in collaboration with user:viewfinder I believe. Eberhard's is pretty close but not identical to this one. Lists are particularly prone to original research assaults and there may be a separate guideline on this somewhere. If so, I can't find it right now. In WP:OR's own talk pages I found only this relevant discussion. Someone argues, advocating for the devil perhaps, that because a list's scope is necessarily arbitrary, "the only “Lists of” that Wikipedia should contain are those where the list is compiled and maintained by a reliable source outside of Wikipedia". This, combined with WP:PLAG would result in a nicely uncontaminated wikipedia. The replies ("Every piece of writing within and without Wikipedia explicitly or implicitly defines the scope of the writing" and "None of that process [of defining the scope] involves WP:OR - it's just planning what you are going to present and how you are going to do it. The actual data you present has to be verifiable, but the criteria and format are yours to decide.") seem adequate to me and we do spell out our criteria here. With respect to sourcing, maps probably are considered primary (though they're really an interpretation of primary data), but "only the interpretation of primary source material requires a reliable secondary source for that interpretation". I hope that reading the heights and coordinates of peaks and key cols from the best available maps is not considered an interpretation. My experience with WP:OR is that it is overused and rarely constructive (e.g. I've previously had to defend addition and subtraction as not being original research), so I hope I didn't come across too defensive;-) Afasmit (talk) 03:12, 18 April 2017 (UTC)

There is a guideline that requires that the selection criterion for a list be supported by reliable sources. But NoGhost's objection does not seem to be based on selection criteria, but on the data in the list itself. The two notes highlighted by NoGhost do not seem (to me) to be evidence of original research nor synthesis of published sources, because they are not presenting an unsupported conclusion. Rather, the list combines the most reliable data available for each list element. This strikes me as permissible under WP guidelines. —hike395 (talk) 03:50, 18 April 2017 (UTC)

@Afasmit: Thanks for commenting. I agree that lists of verifiable and sourced information are a worthy undertaking on WP and I even believe most synthesis of existing information does not constitute OR. Where I have concerns, however, is primarily with the 'prominence' information. While sources may differ on the elevation of mountain summits, there are at least sources and a clear definition of where/what the summit is (for the most part). The prominence is tricky though and for many of the peaks I can't seem to verify the info. Not only is there no way to verify the prominence due to a lack of sources, there is no way to verify the data using maps with no information of key col coordinates. As you know - the identification of key cols is not as easy as just citing a height from a map, it can become difficult and vague with overlapping massifs and endoheric basins. This kind of data is ripe for differing analyses and misinterpretations. --NoGhost (talk) 04:33, 18 April 2017 (UTC)

@NoGhost and Afasmit: I agree that the prominences were unverifiable. I found a table of prominences for peaks above 7200 meters at Peakbagger and used those. Whether Peakbagger is a reliable source has been a matter of debate for years now, but we use their data in many mountain infoboxes, so I decided go with their table. I dropped two peaks that had prominences below 500 meters and whose prominences I could not verify at Peakbagger nor from references at their corresponding WP articles. —hike395 (talk) 10:47, 18 April 2017 (UTC)

Later: Unfortunately, the prominences from Peakbagger demote Changtse and Nangpai Gosum below the 500 m cutoff, which would remove them from the ranked list. This invalidates the annotated satellite images. I would re-rank the table and simply drop those images, but want to get feedback from other editors first. Re-ranking also provides an opportunity to use heights derived from secondary sources (e.g., the 7200-m table at Peakbagger), rather than the primary sources that the table currently uses. Thoughts? —hike395 (talk) 10:55, 18 April 2017 (UTC)

@Hike395: Thanks for that. I realize it is unrealistic, but ideally all prominence data could be individually sourced. It's definitely better than nothing, but is Peakbagger (and Bivouac in the PNW) the only source that measures prominence? I would love to be able to include more info about key cols on WP and verify some of these prominence measurements. --NoGhost (talk) 17:12, 18 April 2017 (UTC)

@NoGhost: It's not unrealistic, see references in List of mountain peaks of Colorado. It's just a lot of work! We could do it gradually. —hike395 (talk) 04:12, 19 April 2017 (UTC)

The Peakbagger site is not the best source for these data. For the US they're (still error prone but) an excellent source as the peaks and cols are superimposed on very good topographic maps. For these Asian mountains, they only have google's relief data to show, which fails to reflect the true topology in steep places. Outside Nepal they seem to use SRTM data alone for these estimates, and still give a prominence down to the meter, whereas they should be given a ±100 m error bar or more. An example of a simple but all too common mistake is Sherpi Kangri, where they used the col between it and its lower East peak as key col (giving a 900m prominence), rather than the one between it and the slightly higher Ghent Peak (which gives it a 1320m prominence)

In contrast to Peakbagger, the Peaklist list was very well vetted in 2007. These people take their prominences very seriously;-) I'm not sure how much work has been done to update the numbers based on newly-published better maps, but I recommend sticking to it for prominences over 1450m. Viewfinder and Eberhard's list of "all" mountains over 6650m (original cutoff 450m prominence, but now much lower) has been updated as late as 2015. This list primarily diverges from our current table for mountains in the Hispar Karakoram and surroundings, where Eberhard used 1950s American military maps, while we used 1970s Soviet military maps (which we found to correspond much better with the SRTM data). It's unfortunately unlikely that real topographic maps will be published of that region any time soon. Another difference is that I rounded prominences, as realistically they have errors of dozens of meters, while Eberhard gives precise numbers. For this "heightist" table, only the ballpark prominence is of importance, so we might as well copy Peaklists's and Eberhard's prominences exactly to avoid any more OR charges (I did that just now) even if the implied precision is illusory.

The heights are critical to the order of course, so we could for each give individual references to the maps we used. Since recently, the Nepal maps can be linked to, so that's great. Others would just have to be a reference.

Changtse's height (7543 m) and prominence in our list comes, as stated, from the excellent Chinese Qomolangma map. The spot height for "Bei' ao", the North Col connecting it to Mount Everest, is 7042 m. This spot appears to be on a bump on the col, which would explain the discrepancy with the widely reported North Col height of 7020 m (as used in the wikipedia article). Eberhard gives a 514 m prominence. Perhaps from the National Grographic map? At any rate, the prominence is over 500 m. Peakbagger claims a height of 7553m for Changtse and 7056 for the North Col. No idea where they get that from; as data source they give "SRTM data", which of course doesn't give exact heights.

I agree that according to our rules, Nangpai Gosum falls below the cut-off. The Nepal map shows 7350m for the peak, and 6924m for presumably the lowest point in the col connecting it to Cho Oyu. My data came from the Chinese map, which has 7351m and a col height of just over 6840m. I should have updated it once the Nepal maps came out, and we decided to use those data for all peaks in or on the border of Nepal.

While both Peakbagger and my own observations give Teram Kangri a promince between 500 and 520 m, Eberhard's list has it at 488 m, so this one would drop off, if we follow my suggestion above, though, as I wrote above, I think his estimates in this region are less conclusive.

Another change has an impact on the numbering: Muztagh Ata has dropped by about 40 meters in the most recent measurements, and it loses 5 spots in the ranking. I'll fix those NASA figures soon. Even after I added the GeoGroupTemplate I think they are a nice visual aid. Afasmit (talk) 11:14, 19 April 2017 (UTC)

As a major contributor to both the main sources, I must declare a personal interest in this discussion. Note that it is a considerable time since peaklist was last updated; in the future, peakbagger is likely to be more accurate. A few weeks ago, JAXA released more topographic data acquired by its Advanced Land Observation Satellite, from which new height and prominence data are likely to be researched and published. Viewfinder (talk) 15:43, 19 April 2017 (UTC)

Hi J, I'm glad though not surprised you've lent your expertise to Peakbagger. It's been more than a decade since we pored over maps for this table; do you recommend any new and improved maps for this region or should we now completely put our hopes on high-resolution DEMs? It's befitting that JAXA's home page uses a bit of poorly mapped Bhutan to demonstrate the increased (5m!) resolution obtained with ALOS. Impressive site. How do our Bhutan and Hispar Karakoram heights match up with these new data? Afasmit (talk) 21:33, 19 April 2017 (UTC)

I have not yet had time to research the above mentioned areas, I only acquired the new JAXA data very recently and I have not yet implemented it. Meanwhile I agree with your above comments about peakbagger. It's a great site, but I don't think its author has personally researched its prominence data, so the lack of any given sources raises questions about its reliability. Viewfinder (talk) 16:09, 20 April 2017 (UTC)

Do either Peakbagger or Peaklist accept user input? For example, if more accurate information about summit or key saddle altitude is found, can visitors suggest/make changes? --NoGhost (talk) 19:08, 20 April 2017 (UTC)

At the present time, Peaklist has not been edited for several years. Peakbagger, on the other hand, is being actively edited and updated in response to e-mails sent to its owner. User input may also be possible, although I am not sure to what extent. There have been reliability and sourcing issues with peakbagger in the past. These are being addressed, but the peakbagger heights for several 8000ers are credited to SRTM. This is nonsense because there are no SRTM data for these peaks. Also the 7050m peakbagger height for Chomo Lhari is wrong, the correct height is at least 7300m. Peakbagger's data source is probably EJ's list on my site, with which I collaborated, but we are not credited anywhere on peakbagger. It follows that I do not think we should be crediting peakbagger as a source until its author correctly credits his sources, but as EJ's list is on my site, I have to declare a conflict of interest. Viewfinder (talk) 14:18, 21 April 2017 (UTC)

Verifiability of key saddles

I can accept the argument that simply transcribing altitude information from a map or DEM may not constitute original research. As mentioned above, I am still concerned about the aspect of verifiability but I am sensitive to the amount of rigorous work that went into creating this list. In all likelihood, this list is the most accurate list available and references to other websites may decrease the quality of information. In this case, to help achieve verifiability standards, would it be possible to have another column listing the coordinates of each mountain's key saddle? Or perhaps just a note somewhere in each row containing coordinates? I think this would allow all information in the list to be verifiable using existing and available sources. --NoGhost (talk) 19:17, 20 April 2017 (UTC)

@NoGhost: I think your original objection is correct. The guideline isn't "no original research unless it's the most accurate list in the world". It's no original research, period. Identifying the key col for a mountain is a synthesis of information in a map. We have to cite to Peakbagger or Peaklist or some other source. We can use the wisdom of Afasmit and Viewfinder to help us choose such a source. But Wikipedia cannot make an independent decision about key cols. —hike395 (talk) 02:58, 21 April 2017 (UTC)

Jurgalski's table on 8000ers.com (columns 15 and 16) lists the location of most key cols and probably all the key cols for the mountains on our list. Many of these coordinates were originally determined by viewfinder. EJ has copyrighted the page, but I'm sure that is for the entire compilation and it's probably okay to copy the coordinates with a proper reference. Afasmit (talk) 07:14, 21 April 2017 (UTC)

@Hike395: Okay based on your comments, it would be best to include the sources as you mentioned. Hopefully an accurate and verifiable secondary source will emerge in the near future to help improve this list and topographic information across WP. --NoGhost (talk) 19:03, 21 April 2017 (UTC)

Blank Rank

Why is the "Rank" column blank for some of the rows in the table? Thanks. DH85868993 (talk) 09:01, 25 April 2019 (UTC)

Read the information at the top: "A drawback of a prominence-based list is that it may exclude well-known or spectacular mountains that are connected via a high ridge to a taller summit, such as Eiger or Nuptse. A few such peaks and mountains with nearly sufficient prominence are included but not numbered in this list." Bazza (talk) 09:10, 25 April 2019 (UTC)

Thanks for the explanation. DH85868993 (talk) 09:17, 25 April 2019 (UTC)

First recorded ascent?

Is there any realistic prospect that any ascents were made — including by locals — in the past, but never recorded? If so, the column heading should change to "First known ascent" or "First recorded ascent". —DIV (1.129.109.51 (talk) 03:30, 21 June 2019 (UTC))

At 7,200 meters? No, there's not much realistic prospect that ascents were made.--Prosfilaes (talk) 04:33, 21 June 2019 (UTC)

See text at Highest unclimbed mountain: "Long before modern mountaineering commenced in the middle of the 19th century, evidence indicates that people did indeed travel up to the summits or near to the summits of major mountain peaks.". —DIV (1.129.109.51 (talk) 03:44, 21 June 2019 (UTC))

Lhotse Shar

An editor at Lhotse Shar has made the odd claim that it is "the 5.5th-highest mountain on Earth" based on where its elevation falls in this list. I have read the Considerations section to understand that (most) subsidiary peaks are excluded. So, what is the proper way do rank Lhotse Shar's height? The article originally claimed it to be the 11th tallest but I can't find any online source that ranks it separately from its parent peak, Lhotse. Would it be best to only list its height, and not mention its rank? Is it meaningful to describe it as a subsidiary of Lhotse, which the 4th tallest? Hoof Hearted (talk) 14:54, 21 May 2021 (UTC)

== Expansion of list to I recently wrote some code which found all the peaks over 7000m with a prominence greater than 500m. I was wondering if it would be a good idea to expand this page to include all mountains over 7000m? Just your average wikipedian (talk) 09:28, 2 August 2021 (UTC)

@Just your average wikipedian Late reply, apologies. But the list would be extremely long. PenangLion (talk) 09:50, 5 May 2022 (UTC)

Sort-ranking of sub-prominences

When sorting the § List of world's highest peaks table by the "Rank" column, one "S" (footnoted to mean "sub-prominence" comes before Rank #1 and the rest of the "S" are at the end (after Rank #108). The term sub-prominence is not defined anywhere, but I assume it means a separately-named peak that is a shoulder on another higher peak (like the central peak associated with the right-hand peak of File:Relative-height.png). Should these be keyed so they each sort immediately following to their adjacent main peak? DMacks (talk) 01:40, 20 June 2022 (UTC)

K2's parent mountain

Why is Everest listed as K2's parent mountain if they are not even in the same mountain range and over 1000 km apart? 77.65.140.40 (talk) 08:15, 7 July 2023 (UTC)

Read its article; in particular K2#Geographical_setting. Bazza (talk) 13:26, 7 July 2023 (UTC)

BULARUNG SAR

Bularung Sar, a 7200 m peak in the Karakoram. First ascent on 25 July, 1990. "Ch. 21, expeditions and notes:first ascent of Bularung Sar". The Himalayan Journal. 48. 1992. The height of the peak is equivalent to Lupghar Star, which is included in the list. Is this peak not fulfilling any criteria to not be included in the list. --Pheeca93 (talk) 08:24, 31 March 2024 (UTC)

@Pheeca93: The second paragraph of the article may explain why. Is Bularung Sar a peak in its own right, or subsidiary to another nearby but higher one (such as Distaghil Sar)? Bazza 7 (talk) 09:30, 31 March 2024 (UTC)

Documented sources[3] show it to have prominence 514m (7.2%). My interpretation of the data is that makes it a separate peak, even if nearby peaks are higher. Pheeca93 (talk) 10:01, 10 April 2024 (UTC)