[OSM-talk] Fwd: Fix maritime borders of Ceuta and Melilla (Spain)

Andy Townsend ajt1047 at gmail.com
Sat Mar 13 11:31:20 UTC 2021

Hi Bert,

 > Is it OK with you I move on with this ?

Yes, but it probably won't be a simple import from 
"https://marineregions.org/" - it looks (from their map at 
https://www.marineregions.org/eezmapper.php ) like they claim not to 
have settled data around Ceuta and Melilla, and I'm not convinced there 
is enough accuracy there.  Obviously we're not directly interested in 
the EEZ data but the boundary data (although looking at EEZ overlaps on 
that site is useful),   It may be, that following processing and 
refining of their data it is what we would end up with what we would get 
by following a "median line" approach, following what Andrew Davidson 
said on 03/02/2021 10:44 (see history below), and using that data may be 
easier than drawing it manually (but their data would need a waiver to use).

What I think we definitely will need is to have a look at any data 
immediately before it is added to OSM - perhaps make it avaiilable as 
overlay tiles, or upload it to the dev server, or something else?

More widely, from looking at their data - I think that what OSM has at 
https://www.openstreetmap.org/way/8692572 is "more correct" then them.  
The outlines for the other islands immediately off the coast of Morocco 
(see https://www.openstreetmap.org/relation/1311341#map=8/35.140/-4.029 
to see them in OSM) may be helpful, but I'm not yet convinced there's 
enough detail.

Best Regards,

Andy (from the DWG)

PS: Apologies to digest readers for top-posting here, but I've done so 
to keep the history.  The end of this message is "but it might take a 
bit longer.", followed by my sig, followed by 6 list footers.

On 09/03/2021 16:08, Bert -Araali- Van Opstal wrote:
> Hello Andy, all
> Is it OK with you I move on with this ?
> On OSM Africa we recently had a request for a reliable source to map 
> boundaries of Natural Parks.
> It was considered as a good source, since their data is also used by 
> the UN for their public maps, for marine but also national boundaries 
> and national parks are included. We are asking the UN Mappers for 
> feedback on experience with the data.
> I did some random checks on differences and accuracy regarding EEZ. In 
> general there were some offsets limited to 1 mile on average with up 
> to like 2 to 3 miles max occasionally. OSM EEZ mostly had less nodes 
> so was less accurate. Also compared with public available data from 
> the EU and NOAA and those seem mostly identical, but less nodes, so 
> marineregions.org more favourable.
> I want to ask them for the waiver, and then use the data for a manual 
> edit for Ceuta / Melilla. For the African Natural Parks we can make a 
> proposal then for Organised or Automated edit, based on that experience.
> Greetings,
> Bert Araali
> On 20/02/2021 19:26, Colin Smale wrote:
>> One other thing to watch out for: Baselines are based on 
>> interpretations of the coastline, which in this case is defined as 
>> low-water. In OSM the concept of coastline is linked to high-water. 
>> So therefore we can't extract or calculate maritime baselines from 
>> OSM data, except in areas where low-water marks are present, such as 
>> the UK, where the administrative jurisdiction normally extends to 
>> low-water so we get this data for free when the admin boundaries are 
>> imported from Ordnance Survey data. There are however anomalies in 
>> certain specific areas (like the Bristol Channel and Brighton Marina 
>> to name but two).
>>> On 02/20/2021 4:49 PM Colin Smale <colin.smale at xs4all.nl> wrote:
>>> Looks like their data is licensed as CC-BY 4.0 which is very 
>>> permissive but I can't say if it is enough for inclusion in OSM.
>>> https://marineregions.org/disclaimer.php
>>> It actually looks like it may be possible, with some additional 
>>> documentation:
>>> https://wiki.openstreetmap.org/wiki/Import/ODbL_Compatibility
>>> https://blog.openstreetmap.org/2017/03/17/use-of-cc-by-data/
>>> They also have ready-made median lines available for download, 
>>> licensed as above.
>>>> On 02/20/2021 2:54 PM Bert -Araali- Van Opstal 
>>>> <bert.araali.afritastic at gmail.com> wrote:
>>>> Can this be a possible source (although also simplified vector lines):
>>>> https://www.marineregions.org/about.php
>>>> They provide downloadable shape, KML and low resolution data files. 
>>>> Someone could ask them to provide us with high resolution data as 
>>>> they seemed to have done all the calculations according to the UN 
>>>> regulations and International Standards. They used PostGIS to 
>>>> produce it.
>>>> https://www.marineregions.org/downloads.php
>>>> Didn't look at any License restrictions or conflicts though.
>>>> Greetings,
>>>> Bert Araali
>>>> On 20/02/2021 16:32, Colin Smale wrote:
>>>>> The Median Lines are defined in terms of the Baselines - a 
>>>>> simplified representation of the coastline doing specific things 
>>>>> around islands, bays and estuaries. Perhaps we can get hold of the 
>>>>> Baseline data and create our own median lines mathematically from 
>>>>> that?
>>>>>> On 02/20/2021 2:13 PM Bert -Araali- Van Opstal 
>>>>>> <bert.araali.afritastic at gmail.com> wrote:
>>>>>> Dear Andy, list members,
>>>>>> Do you have or are there resources which we can use to define 
>>>>>> these Median Lines and which cause no licensing conflicts ?
>>>>>> Do you have the intention to create a wiki page to describe this 
>>>>>> issue and how you plan to implement it ?
>>>>>> If you need some help, I am very willing to allocate some time to 
>>>>>> participate.  As I have mapped 325 (mostly administrative) 
>>>>>> boundaries already in JOSM.
>>>>>> Greetings,
>>>>>> Bert Araali
>>>>>> On 20/02/2021 14:50, Andy Townsend wrote:
>>>>>>> On 03/02/2021 10:44, Andrew Davidson wrote:
>>>>>>>> The Spanish law says that in the absence of an agreement with 
>>>>>>>> another country, their territorial sea shall not extend beyond 
>>>>>>>> the median line:
>>>>>>>> https://www.state.gov/wp-content/uploads/2020/11/LIS149-Spain.pdf
>>>>>>>> Moroccan law also has the same arrangement:
>>>>>>>> https://www.un.org/Depts/los/LEGISLATIONANDTREATIES/PDFFILES/MAR_1973_Act.pdf 
>>>>>>>> As there is no agreement between the two countries it would 
>>>>>>>> appear that they both think the boundary is the median line. 
>>>>>>>> However Morocco amended their law last year:
>>>>>>>> https://www.moroccoembassy.co.za/morocco-in-the-news/260-morocco-updates-its-maritime-legislation-to-un-standards 
>>>>>>>> Details in English are sparse (they haven't sent the required 
>>>>>>>> paper work to the UN yet), but the changes appear to be about 
>>>>>>>> their Atlantic maritime boundaries.
>>>>>>> Thanks Andrew.
>>>>>>> I was waiting here just in case anyone posted another point of 
>>>>>>> view, and no-one has.  Unfortunately the US document (which 
>>>>>>> describes the Spanish point of view) says "The Kingdom of Spain 
>>>>>>> also includes its exclaves located on the northern coast of 
>>>>>>> Africa, which are beyond the scope of this study".  It does, 
>>>>>>> however, describe what Spain does elsewhere, which is presumably 
>>>>>>> a "general intent".  As you say, the Moroccan article 2 ("In the 
>>>>>>> absence of a specific agreement on the subject, the breadth of 
>>>>>>> the territorial waters shall not extend beyond a median line 
>>>>>>> every point of which is equidistant from the nearest points on 
>>>>>>> the baselines of the Moroccan or adjacent coasts") is clearer.
>>>>>>> Does anyone see a problem with the "median line" approach?
>>>>>>> If not, does anyone fancy volunteering to actually make the 
>>>>>>> change?  It'd need to be someone familiar with boundary editing, 
>>>>>>> which probably means familiarity with JOSM's "validator" to 
>>>>>>> check that the resulting boundaries are both valid multipolygons 
>>>>>>> and match the documentation above.  If not, I can do it, but it 
>>>>>>> might take a bit longer.
>>>>>>> Best Regards,
>>>>>>> Andy (from the DWG)
>>>>>>> _______________________________________________
>>>>>>> talk mailing list
>>>>>>> talk at openstreetmap.org
>>>>>>> https://lists.openstreetmap.org/listinfo/talk
>>>>>> _______________________________________________
>>>>>> talk mailing list
>>>>>> talk at openstreetmap.org
>>>>>> https://lists.openstreetmap.org/listinfo/talk
>>>>> _______________________________________________
>>>>> talk mailing list
>>>>> talk at openstreetmap.org
>>>>> https://lists.openstreetmap.org/listinfo/talk
>>>> _______________________________________________
>>>> talk mailing list
>>>> talk at openstreetmap.org
>>>> https://lists.openstreetmap.org/listinfo/talk
>>> _______________________________________________
>>> talk mailing list
>>> talk at openstreetmap.org
>>> https://lists.openstreetmap.org/listinfo/talk
>> _______________________________________________
>> talk mailing list
>> talk at openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/talk
> _______________________________________________
> talk mailing list
> talk at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20210313/b5fc4c44/attachment-0001.htm>

More information about the talk mailing list