[OSM-newbies] rivers and coastlines

[David Groom]

>
> ----- Original Message ----- 
> From: "Peter Childs" <pchilds at bcs.org>
> To: <newbies at openstreetmap.org>
> Sent: Monday, October 05, 2009 8:05 AM
> Subject: Re: [OSM-newbies] rivers and coastlines
>
>
>
> 2009/10/4 Ævar Arnfjörð Bjarmason <avarab at gmail.com>:
>> On Sun, Oct 4, 2009 at 8:07 AM, Robert Helvie <alimamo at gmail.com> wrote:
>>> I have been correcting some of the coastline of South Korea (NK border
>>> south
>>> through to Busan provisionally finished) using the Yahoo imagery (yes, I
>>> know it is old and not always perfect, but it is a darnsite better than
>>> what
>>> was there).
>>>
>>> My question is ... What is the common practice for when rivers meet the
>>> ocean? Here I am talking about the actual way of the river. Not the
>>> "riverbank" ways. Should the river way run out and butt against the
>>> coastline with a node?

That's how I do it.
http://www.openstreetmap.org/?lat=29.9324&lon=48.6163&zoom=14&layers=B000FTF


>>>Cross the coastline and run out into the ocean?
>>> ...
>>> Something else??
>>>
>>> Additionally, coastline segments often (and correctly in my view) run 
>>> "up
>>> river" of wider rivers. How far up river is a judgment call for which I
>>> am
>>> not looking for an answer right now. But I am wondering about the river
>>> way
>>> running out. If the "coastline" runs up river, should the river way end
>>> at
>>> or just past that coast line or should it run out between the banks into
>>> deeper water?
>>>

I'd say it should run out to the point where people would no longer cal it a 
river, see my point below.  In practice there are many rivers where the 
riverbank is still marked as coastline, but probably should be changed to 
waterway = riverbank.  The main reasons for this are:

i) the initial import for that data would have been from PGS which tagged 
everything as natural=coastline, and the data is waiting for someone to come 
along and tag it correctly

(ii) in some places where there is a particularly complex estuary system it 
used to be a lot easier to leave things as natural=coastline, but ever since 
multipolygon relations have been allowed it would be better to go back and 
correctly tag complex estuaries as waterway = riverbank.


>>> I couldn't seem to find anything on the wiki.
>>>
>>> Just looking for a consenses or perhaps some examples in practice.
>>> Thanks. If you do have an example, send a link, please.
>>
>> I draw a line between the two points where the riverbanks meet the
>> ocean on their far sides. That's my coastline:
>>
>> Riverbank: http://www.openstreetmap.org/browse/way/28101869
>> Coastline: http://www.openstreetmap.org/browse/way/28101860
>>
>> I think that's a sensible way to do it.
>>
>
> Strictly speaking, I think its to do with Tide. If the Bank is
> effected by Tide its a Coast, If its not its a River Bank. But I don't
> suspect this really makes that much sence in every case and since the
> coast and river banks are usually broken in to short segments anyway.
> Its a judgment call.

I agree, the principle I've tended to work to is would someone looking at 
the land / water boundary think of it as coast or river.  Take the River 
Thames as it passes through London, it is tidal, but would most Londoners 
think they lived beside the coast or beside the river?  I suspect they would 
think a river.  From a tagging perspective I'll admit this is a subjective 
judgement call, but then so would any that doesn't stick to a pure is this 
tidal or not basis.

David

>Hence if its a Tidal River bank in needs to be
> marked as such, But Coasts always have tides, Lakes are different
> again....
>
> Peter.