pmontra a day ago | next |

The animation is cool but I'd like a way to disable it and just be able to see the path on the globe, then tap to another place and see that other path too. It would be OK to reload the page to clean all paths. No need to have a way to share them if there are more than one path.

tetris11 2 days ago | prev | next |

If you drop a pin anywhere in south london, it takes several "unknown river"s which on the map are stated clearly as the River Wandle.

The US drops seemed a lot cleaner

steve_adams_86 2 days ago | root | parent | next |

Same with British Columbia, Canada. There are a lot of unidentified waterways. In Victoria, BC, virtually everything ends up going down a small inlet called "the gorge". It's not quite correct to consider it a stream, river, creek, etc.

Still a really cool idea, and when it works well it's good fun. I can see there being great practical applications if it was more accurate too (though I'm sure the people who could put it to use already have good data)

brazzy 2 days ago | root | parent | prev |

The river Isar through Munich, Germany, which is one of the larger tributaries of the mighty Danube, is also "unknown". The data outside the USA is definitely lacking in that regard.

I guess the names are taken from an underlying hydrologic dataset, and Google (or whoever) maps used purely for visualization.

returningfory2 2 days ago | prev | next |

My favorite fact about raindrop paths/watersheds is that they determine most of the political boundary between Argentina and Chile. If the rain at a point ultimately flows to the Pacific Ocean, that point is in Chile; if it flows to the Atlantic Ocean, it is in Argentina. This is article 1 of the original boundary treaty [0] from the 19th century.

The treaty was made before the Andes were fully explored, and so it doesn't handle some interesting edges cases like the Southern Patagonian Ice Field. The field has the property that when rain falls there, it freezes and goes nowhere!

[0] https://en.wikipedia.org/wiki/Boundary_Treaty_of_1881_betwee... [1] https://en.wikipedia.org/wiki/Southern_Patagonian_Ice_Field

gus_massa a day ago | root | parent | next |

It was later amended to consider the top of the mountains that define the watersheds. Watersheds are difficult to determine and may change from time to time. Mountaintops are easier to determine and more permanent.

jmilloy 2 days ago | root | parent | prev | next |

I think borders based on watersheds make so much more sense than those on rivers, both of which are common.

emmelaich 2 days ago | root | parent | prev |

Harder to find though.

londons_explore 2 days ago | root | parent |

The big bit of a river is easy to find, sure.

But up near the source where it splits into thousands of small tributaries which move and change a lot? That's a recipe for boundary disputes.

Whereas a watershed rarely changes far, even on geological timescales.

dredmorbius 2 days ago | root | parent | prev | next |

Watersheds and rivers are often key determinants of pre-industrial borders, if only because they served (variously) as obstacles (wide, fast rivers) and facilitators (narrow / slow) to navigation. Until the development of steam-driven railroads and internal-combustion-driven vehicles, if you wanted to move any considerable quantity of goods or people, your best option was a ship over either a (reasonably) calm sea, or inland rivers, lakes, and canals. Ice ran a close second, and ice roads remain vital transport routes in Siberia and Alaska. Where run over rivers and lakes, those tend to be (near) level terrain, and frictional losses are minimised.

China built over 1,000 km of canals 1,000 to 2,000 years ago. Inland travel within England and Great Britain was revolutionised by the development of canals. The US opened up its interior with the Erie Canal, on which a single mule could haul an 80-foot barge laden with 40 tonnes of goods or people. Not only was the travel efficient (in terms of motive power and drovers) but it was comfortable. A four-foot-deep ditch (the Erie Canal) was immune to the rutting and mud of a road, and though the speed of travel was low, in a matter of a few days people could travel in great comfort across New York from the Hudson to Lake Erie and from there to points inland via the Great Lakes, Chicago, and Mississippi / Missouri / Ohio / Red, and other, rivers.

I'd realised the political boundaries aspect looking at a map of European watersheds created by Robert Szucs:

<https://www.seattlepi.com/local/science/article/Viral-maps-s...>

(Europe is in the carousel at the top, third image, isolated here: <https://s.hdnux.com/photos/54/21/22/11605137/3/960x0.webp>)

It's not that national borders are specifically identified as watersheds, but in large part you can patch together current state boundaries out of watersheds and rivers, or find internal divisions (e.g., UK counties, German states, French provinces) out of watersheds.

In regions in which borders were more arbitrarily assigned, as with the inland US, Middle East (o hai Sykes-Picot), and Africa, the artificiality of imposed borders is highly evident, though regional affiliations become clearer.

South American countries are, as you note, more closely aligned with watersheds. It's also more evident why the north-eastern region of South America is so politically divided: it's largely composed of non-contiguous watersheds which have become their own polities.

<https://s.hdnux.com/photos/54/21/22/11605136/4/960x0.webp>

thaumasiotes 2 days ago | root | parent |

> obstacles (wide, fast rivers) and facilitators (narrow / slow) to navigation

Wide rivers are slow. Narrow rivers are fast. That is the nature of fluid dynamics.

Wide rivers are obstacles to movement on the land and facilitators of movement on the water; narrow rivers are worse at both of those things.

dredmorbius a day ago | root | parent |

I could have been clearer that both descriptions were meant to be or statements.

Fast rivers or wide rivers are challenging, the former as they're inherently treacherous, the latter as they're difficult to bridge.

Narrow rivers can be bridged. Slow rivers can be traversed by boats, ships, barges, etc.

Yes, wide rivers tend to be slow, and narrow rivers tend to be fast, all else constant, but all else is rarely constant such that there are narrow slow rivers and wide fast / treacherous ones.

jajko 2 days ago | root | parent | prev | next |

Even glaciers flow, otherwise they would over time become as thick as cloud layer that provides that rain drop or snow flake. Its just that speed of flow is... glacial.

1 year on glacier, once snow is compressed into ice which takes few years, is represented by maybe 1-2m layer in European Alps.

Thats why glaciers have crevasses - they bend as valleys flow (or over some hard bottom bumps that weren't grinded flat yet), and it cracks at those bends. Just like say chocolate on surface of cold Snickers bar if you try same bending

ch33zer 2 days ago | prev | next |

Tangentially related are lakes that flow out to two different oceans. Lake Isa in Yellowstone (https://en.m.wikipedia.org/wiki/Isa_Lake) is an example. Even more interesting the different sides of the lake flow to the 'wrong' oceans: the eastern side of the lake flows to the Pacific and the western to the Gulf of Mexico.

kaskakokos a day ago | prev | next |

I tested it on a peak we have in Spain called 'Pico Tres Mares,' which translates to 'Three Seas Peak,' and it was nice to see that depending on which side you drop the raindrop, the water flows to the Bay of Biscay, the Atlantic, or the Mediterranean.

Nice.

stewarts a day ago | root | parent |

Very similar named peak in Glacier National Park - Triple Peak Divide. Water can flow to the Pacific, Atlantic, or Arctic oceans.

PopAlongKid 2 days ago | prev | next |

Posted just a few months ago, the comments point to prior postings in last five years.

https://news.ycombinator.com/item?id=41752052

johann8384 2 days ago | root | parent |

So?

darth_avocado 2 days ago | root | parent |

Dupe?

dredmorbius 2 days ago | root | parent |

Not with only three comments and two votes, which is well below the "significant discussion" threshold.

<https://news.ycombinator.com/item?id=9416150>

darth_avocado 2 days ago | root | parent |

There has been significant discussion like the grandparent points out?

2022 (790 points, 115 comments) https://news.ycombinator.com/item?id=29841737 2021 (887 points, 128 comments) https://news.ycombinator.com/item?id=27297689

dredmorbius 2 days ago | root | parent |

HN's rules on dupes:

- No reposts within a year, unless ...

- Posts within the past year have not had "significant discussion", a value left provocatively vague by dang, but I figure as roughly >= 20 comments.

Reposts after 12 months or if there's been no significant discussion within the past 12 months are both perfectly OK.

You're welcome and encouraged to read the guidelines and FAQ (both linked on every HN thread and comment page) and/or see dang's many, many, many previous comments on this topic:

<https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...>

bagels 2 days ago | prev | next |

It doesn't work for me, but I'm going to assume the model does not include evaporation or ground water?

notcrazylol 2 days ago | prev | next |

Damn this is soo cool! I tried it at a random place and it worked like magic! But when I tried it near my house, the website broke :D I think it doesn't handle unidentified rivers.

psunavy03 2 days ago | prev | next |

"Unable to find a flowpath from that location, try something else." And that's most of greater Seattle.