nvkelso / natural-earth-vector

A global, public domain map dataset available at three scales and featuring tightly integrated vector and raster data.
https://www.naturalearthdata.com/
Other
1.78k stars 369 forks source link

Feature request: Mountain Passes #465

Open Maxszik opened 3 years ago

Maxszik commented 3 years ago

I wish passes were a part of natural Earth. The Khyber Pass, The Brenner Pass, Torugart and so many others carry historical, cultural and contemporary importance. If we have peaks and capes in the dataset, for example, I think mountain passes deserve a place as well.

Andrettin commented 3 years ago

I definitely agree. And I think the project would also benefit from having forest geopolygons for the physical label geographical areas.

nvkelso commented 3 years ago

Great idea! Do you have any data or research time to contribute for mountain passes?

Andrettin commented 3 years ago

@nvkelso Would geocoordinates set out in 19th century books be a permissible source of information for the project (considering they are in the public domain)? I have some experience with researching them, so I could see if I manage to find something useful. I imagine this could be helpful to get the geocoordinates of points such as mountain passes and settlements.

nvkelso commented 3 years ago

Yes, that should be fair game. But we'd want to adjust to match the SRTM based relief, admin boundaries, and roads that are in Natural Earth.

Is there a source similar to https://peakbagger.com/ but for passes?

ImreSamu commented 3 years ago

Wikidata info:

Map view: image

Andrettin commented 3 years ago

I found this book with some geocoordinate information on mountains passes:

https://books.google.at/books?id=JHbXAAAAMAAJ&pg=PA342&dq=mountain+pass+latitude&hl=de&sa=X&ved=2ahUKEwiRy-by24vuAhWGy4UKHUjSDFcQ6AEwAXoECAQQAg#v=onepage&q=mountain%20pass%20latitude&f=false

I was thinking of entering its data manually in a geodata file, but I suppose the wikidata query by @ImreSamu could collect that information much more efficiently?

ImreSamu commented 3 years ago

@Andrettin

I don't know what is the best method .. but in the end we should match the results with the Wikidata ( so need a WikidataID )

imho: the original "20631" record is too much ... for manual checking Maybe we can limit, and keeping only the wikidata records - with English Wikipedia page. so " mountainPass AND has an english wikipedia page AND has a Coordinate! "

nvkelso commented 3 years ago

This should probably number in low hundreds of features, some small fraction of mountain peak count?

On Jan 11, 2021, at 13:53, ImreSamu notifications@github.com wrote:

 @Andrettin

I don't know what is the best method .. but in the end we should match the results with the Wikidata ( so need a WikidataID )

imho: the original "20631" record is too much ... for manual checking Maybe we can limit, and keeping only the wikidata records - with English Wikipedia page. so " mountainPass AND has an english wikipedia page AND has a Coordinate! "

https://w.wiki/tzZ this is: ~1558 record ... probably need more filtering .. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

Maxszik commented 3 years ago

Hello,

Great idea! Do you have any data or research time to contribute for mountain passes?

I definitely have the time and desire to contribute. Experience and skills, unfortunately, no. I have no idea about the methodology and how to process data. If someone could point me to some instructions that would be great.

I agree with @nvkelso that the above example with all those wikidata passes is too much. This is just my personal preference, of course. I imagine something like the mountain peaks example, which would be not all passes but those one would consider "epic" or "classic". Totally subjective, obviously.

What about this: there a Wikipedia entry called List of mountain passes. What if I should begin there, assembling a csv with coordinates and a number of features? Is that a good idea?

nvkelso commented 3 years ago

That list looks legit, if too exhaustive (almost 1000 passes). BTW, the red links don't have a full page, the blue links do. That's sometimes a good indicator for how "significant" the feature actually is.

I'd cross reference this list with lists of "top 10 passes" (or 20) by continent and see what's common.

ImreSamu commented 3 years ago

The good news: There are already 2 "mountain pass" in the NE database. { Khyber Pass,Chilkoot Pass}

ne_id min_zoom name featurecla elevation region wikidataid wdid_score name_en name_fr name_de
1159105727 5 Khyber Pass pass 1070 Asia Q191333 4 Khyber Pass Passe de Khyber Chaiber-Pass
1159106213 6.7 Chilkoot Pass pass 1067 North America Q1072958 2 Chilkoot Pass col du Chilkoot Chilkoot Pass
Maxszik commented 3 years ago

I raised this issue once. Since I've been searching and collecting mountain passes far and wide. Collecting a massive set of passes is no problem, but to do it in a way that makes sense seems just impossible. There's no way for me to have an idea what passes are vital, important, culturally, economically and historically significant, and so on. So it's a fail so far.