Talk:Proposed features/Maritime borders

From OpenStreetMap Wiki
Jump to: navigation, search

There are several forms of international maritime borders, the "territorial border", "base line", "fishing limit", and "economical zone" is the most commonly used. Each of these have a different definition and a different purpose. The 12nm "territorial border" is the limit of where a nation can claim sovereignty and keep control of foreign flanged vessels. All of these should in some way or another be available as map data. --Skippern 20:27, 30 November 2008 (UTC)

AFAIK There was no tag for it yet, but I didn't want to go through all this proposal stuff. So I simply tagged the Belgian EEZ with boundary=exclusive_economic_zone. --Cartinus 00:45, 1 December 2008 (UTC)

The approach taken in this proposal will give wrong results in a lot of places. Most coastline data in OSM is based on PGS data or aerial images. The first is an approximation of the mean high water line. With the second you don't know what tide it is. The maritime border however lies 12 nautical miles from the low-tide mark. There are lots of places with a significant distance between the low-tide mark and high-tide mark. I think creating a tool that works as proposed here is OK, because it gives people something to use in places where an accurate maritime border is not known. But putting that data into the database should not be done IMHO, because it is derived data. --Cartinus 00:45, 1 December 2008 (UTC)

To assume the international border is 12nm off the tagged coastline, and have generated a maritime border at that distance will not inflict on any copyrights, though there might be a lot of places where such border needs to be adjusted. A tool can easily generate this coast line, and later users can correct it where it is known to be different. Besides, in areas where you have a big difference between high and low tide, such should be marked with natural=wetland and wetland=tidal_flat or something similar. If such is done before this tool goes to work, than the error in the maritime border will be less. --Skippern 10:26, 1 December 2008 (UTC)


A maritime border should be tagged differently than a border on land, most maps will only mark the border from land and out to sea to mark that the border goes in the sea, and between features where meridian lines exists, other maps will render different types of maritime borders differently, i.e. fishery limits might be dashed lines with fish symbols in between, baseline might be alternating + and - symbols, and EEZ (Exclusive Economic Zones) might be just a dashed line. Maritime border does not include borders in inland lakes! --Skippern 14:24, 4 December 2008 (UTC)

I agree that tagging making rendering easy should be a priority. Today most renderers show a halo around all islands, not a very good solution IMHO. -- Gustavf 15:38, 31 December 2008 (UTC)
By agreeing on a tagging scheme than setting up the rules you want for rendering is no problem. The question we set about rendering is rather what do we want Mapnik to render. I would like to have Mapnik render maritime borders, but not in the same way as borders on land. --Skippern 21:23, 8 February 2009 (UTC)
Of course, what to render should be up to each renderer. What we can do with tagging, is just making making these choices as easy as possible. -- 21:43, 8 February 2009 (UTC)

Boundary relation

Also, how do we add this all to the country boundary relations? Would we need special roles "eez" or "territorial"? And would those new roles conflict with the "enclave" or "exclave" tags? It's also obvious that if we add both EEZ and territorial waters to the boundary relation, the ways made up by the relation won't form nicely enclosed areas anymore, would that be a problem? --Eimai 11:46, 31 December 2008 (UTC)

The national boundry should be the "base line", usually 12nm off the coast, the other boundries are of various administrative purposes, but EEZ is usually recognized as "international" waters even though resuorce exploitation belongs to a specific country. This means a ship can pass through a country's EEZ without restrictions while it might have to report to various authorities to pass through the territorial waters closer than 12nm. --Skippern 12:00, 31 December 2008 (UTC)
I agree with Skippern. The territorial waters boundary should go into the boundary relation. -- Gustavf 14:32, 31 December 2008 (UTC)
OK, that's how I've done it for Belgium, but we still have that EEZ as well. Should that be a part of the relation as well? Or should the EEZ be its own (enclosed?) boundary relation? --Eimai 15:44, 31 December 2008 (UTC)
I don't feel EEZ need to be in a relation, but if it does, than it is a completely different relation than the national border. EEZ should not inherit admin_level=* as it have no direct administrative purpose, except for what laws applies for resource exploitation and how it is taxed. Maybe boundary=economic is better, this type of boundary can also be used to show taxation zones within a country (such as tax exempted areas) --Skippern 13:03, 1 January 2009 (UTC)

France and Channel Islands

There seems to be a problem with the 12NM-maritime border of France in the Channel: the Channel Islands (Jersey, Guernsey, Alderney, Sark...) are included as French dependancies rather than self-governing "bailiwicks" (cf. Wikipedia), or at least British Crown dependencies... Damouns 10:00, 6 February 2009 (UTC)

In such cases as the Channel Islands, a meridian line should be used rather than an actual 12NM zone. A meridian line should be used whenever one of the maritime borders cannot be used because of too short distance between two entities. The rule about meridian line also applies for the other maritime borders. --Skippern 15:25, 6 February 2009 (UTC)

The boundries in the sea and air above the channel between Britain and France and around the Channel Islands have been formally agreed between allthe goverments concerned so it would be beter to check offical sea and air charts for where the boundries actually run. This fits with the law of sea through the allowance for mutal agreements to resolve overlaping situtaions. The EEZ's and Shipping lanes follow similar but indepentant agreements the British Empire already has most borders worked out like this such as between Burmuda and the USA etc.--Govanus (talk) 16:37, 19 April 2015 (UTC)

Original proposal 1


If the coast twists and turns sharply, around bays and inlets, a simple offsettting by 22224m to the right may well end up with lines that cross over each other, or go over land, a method is required to simplify the maritime boundary where it loops back on itself.

the same process that is used to trim the circles/lines where they meet can be used here. i do this sort of thing every day as part of being a draughtsman, it's fairly trivial to do manually, and shouldn't be difficult to do programatically.


Coastlines are made up of two components, lines and ways. This method proposes to handle them separately and then 'join' the results to form a continuous line.


For a given processed area, each way tagged 'natural=coastline' will have a corresponding way created


Each point that is part of a way tagged 'natural=coastline' will have a circle created around it, with the point it's centre, with a radius of 22.224km.


A set of three components (a point, and it's adjoining ways) are then processed, to form a smooth, continuous, tangential line. An existing library may be used for this processing, or if the maths is trivial enough, it will be created from scratch.

  • The maths near a node is easy enough: You just go from the node at 90 degree angles to the segments. I think what's more difficult is to determine which segments and node actually contribute to the maritime border. Robx 09:25, 13 January 2008 (UTC)
  • i'm not sure what you mean, "determine which segments and..."? Myfanwy 18:44, 13 January 2008 (UTC)
shouldn't be any more difficult than actually trace the GSP coastlines, as was done to form most of the coastlines in the system. Only thing you need the bot to clean up the segments afterwards to remove multiple borders due to islands and such. Also when the border overleaps a border from another country, such will be the case in English channel, a meridian border should be calculated. In such cases, the border should be marked as dirty and retraced later as the border doesn't follow an actual meridian, but a set of waypoints agreed between the two nations. --Skippern 14:20, 4 December 2008 (UTC)

Proposal 3 discussion

I don't like switching from maritime to administrative for territorial waters, but apart from that this I like the approach. -- Gustavf 16:04, 4 January 2009 (UTC)
Maybe we should change it to maritime, but still keep admin_level. (updated proposal) --Skippern 14:06, 10 January 2009 (UTC)
Looks very good now. Easy to tag, easy to render, easy to extend. -- Gustavf 16:51, 26 January 2009 (UTC)
Added ref tag on baseline. --Skippern 16:05, 10 February 2009 (UTC)

Rendering rules for Kosmos, using this proposal, is available at Maritime borders/Kosmos 3. The border between Norway and Russia is tagged according to this proposal, but please not that parts of this border is disputed and somewhat difficult to follow. This is espceially true for the EEZ. -- Gustavf 18:36, 12 February 2009 (UTC)

While tagging, I noticed a couple of things that probably should be clarified: How to add names to boundaries where you have the same country on both sides (most typically baseline)? I would suggest adding right:country=* or left:country=* for the shore side of the boundary. It might also be necessary to add something like border_type=internal for cases where some subnational entity (state, county) extends into the sea, but not to one of the other boundaries. -- Gustavf 18:39, 12 February 2009 (UTC)
There is no need to tag left:country=* and right:country=* when it is not a boundary between two countries, as such there are no need for left/right tagging on border_type=baseline, if a state division extends into the sea, than it is a national definition, and can probably be solved with tagging boundary=maritime together with admin_level=* of corresponding level. If it is needed with a separate border_type=internal or border_type=administrative for these cases, than this could easily be amended later, and should be solved on a later stage. --Skippern 19:32, 12 February 2009 (UTC)
This should at least be specified in the proposal. What about country name tagging on the territorial line? This is where the admin_level tag is supposed to be used. -- Gustavf 20:20, 12 February 2009 (UTC)
The territorial line should be tagged with left:country or right:country, even if it isn't a border between two countries, besides the border relation of the country should also follow this. --Skippern 10:19, 13 February 2009 (UTC)

Proposal 4

All the here discussed lines references on "Baselinie". The Baselinie is a governmental definition and serves as reference for to define the international boundary. It follows:

International Boundary = territoial_waters

(or 12-nm-line, 3-nm, 6-nm or something else)

+ admin_level=2

The Baselinie we could mark as:

there ReferencePoints as ref=Nr/LAT/LON

The Coastal Zones are also referenced to the Baseline. So we can describe it by an additional key:

boundary:type="200-nm-zone" etc.

I prefer continuously same terms (boundary). For type-declarations, it is useful to mark the type by the upper term, divided by a ":".

Gruss, --Markus 12:51, 4 February 2009 (UTC)

This proposal is a problem for renderers, taht do not want to make halos around all coastlines, as there is no easy way to separate a land border from the territorial waters. Gustavf 12:35, 8 February 2009 (UTC)
Hi Gustav, please can you explain exactly what you mean: what is a "halo"? what is the difference between a land border and a sea border? what is the rendererproblem? Thanks, --Markus 22:20, 8 February 2009 (UTC)
With a "halo" I mean a coloured ring around all islands and coastlines. You can see this (at least now) around Ireland, Isle of Man and France.
There are many differences between sea borders and land borders (and many similarities). What I think are the three main differences are
  1. A sea border normally is the end of some kind of jurisdiction, but without a similar entity on the other side. That is, not a border between two countries, but between a country and "international".
  2. A sea border does not indicate exactly the same jurisdiction as a land border. The best example is probably "innocent passage", which is allowed in territorial waters, but not over land. On land it is easy to say exactly where a country ends, but at sea this is mostly at the territorial waters, but also to a limited extent at the end of internal waters (baseline), the end of the EEZ and the end of the continental shelf. (The Wikipedia article is pretty good at explaining this).
  3. For most people, I think, the maritime border is a much more vague thing than a land border. If you ask a German to "draw a map of Germany", I think he or she would attempt to draw the coastline and the land borders, not the land borders and the end of German territorial waters.
For rendering: If you look at almost any non-OSM map, that be an Atlas of the World from you bookshelf, a tourist map of Europe or most (if not all) online maps, you will not see halos around islands and coastlines. This is not because the data to make them have been unavailable for the mapmakers, but because the mapmakers have made a choice not to show these borders or show them differently (perhaps as a thin blue line). If we tag territorial waters the same way as land borders, it will be very difficult for someone using OSM data to avoid drawing halos, with todays renderers I would even call it impossible. I think we should make it easy to follow long established cartographic conventions for general purpose maps using OSM data, and at the same time making it fairly easy to make a special purpose map.
I think that all of this is easiest using Proposal 3, which also follows a naming convention that is closer to the United Nations Convention on the Law of the Sea and not specific about distances in sea miles. -- Gustavf 23:18, 8 February 2009 (UTC)
It's only a problem for renderers if you implement your renderer in a certain way, i.e. if you render all boundary=administrative borders without looking at admin_level=*, or if you don't check whether two distinct entities share the same border before rendering it, i.e. where you only want to render maritime borders where two countries meet. I don't think tagging schemas should be designed around renderers that want to blindly render things only based on the primary key-value pair without looking at anything else, especially when it reduces the consistency of tagging for editors. --Ævar Arnfjörð Bjarmason 10:20, 9 February 2009 (UTC)
As far as I read this proposal, the territorial waters line should be tagged exactly as a land border? The only exception being one country name missing if two territorial waters do not meet. -- Gustavf 10:56, 9 February 2009 (UTC)
There should be a clear difference between land and sea borders, independent if it is between two entities or on the edge of one. My proposal 3 contains the information needed to link the two together as well as differ between the two, so it can be up to the renderer or search software, or whatever the user needs to link or differ the two. Is it a problem for you that we don't use boundary=administrative on maritime borders? Look for admin_level=* instead! --Skippern 15:07, 9 February 2009 (UTC)

Hi Gustav, I see your point. The "halo" you see is the official boundary=administrative, like it is announced to UNO. The sea-border is a part of boundary=administrative + admin_level=2, same as land-border. The differences are marked by the attribute "is on water" or "is on land". If you want you coin a additional tag for.

ad 1) I don't think that we need a tag for "opposite is another country" ore "opposite is international". You can see it directly in the map.

ad 2) Yes we have some other rules for sea-border and land-border. But both are boundary=administrative, and I don't think that we need a tag for jurisdiction, you can see the border in the map, and you have to now the meaning when you are sailing. But if you want you coin a additional tag for. (boundary:type="12-nm-zone"?).

ad 3) May be people learn about borders by OSM. Most seaside borders in common maps are wrong...

Hi Ävar, I agree: we dont tag for the renderer.

Hi Skippern, There is a clear difference between sea- and land-borders: one is on sea and the other on land. I don't think, that we need to tag it additionally. But if you want, you can coin a additional tag? --Markus 23:01, 9 February 2009 (UTC)

I already did, it is Proposal 3 --Skippern 00:41, 10 February 2009 (UTC)
Nope, Proposal 3 replaces the "top level" tag boundary=administrative with boundary=maritime. Using an additional tag would be something like adding the "lower level" tag maritime=yes along with boundary=administrative to the same object. --Cartinus 01:47, 11 February 2009 (UTC)
I wrote Proposal 3, I like it, if you have a better proposal, write it. Personally I like boundary=maritime better than maritime=yes because of the nature of these, and for rendering purposes. The 12nm territorial border have more in common with the other maritime borders than it have with boundary=administrative + admin_level=2 --Skippern 10:21, 11 February 2009 (UTC)
We need to tag maritiem borders differently because they are different (part of a hiearchy of border types) and because most maps want to show them differently. This is not "tagging for renderers", this is tagging in such a way that it is possible to make a usable stylesheet for renderers. Tagging for renderers would be to tag maritime borders as highway=footway so they would show up as a dashed line. What you can "see from the map", a rendering engine must be able to see from the data. -- Gustavf 14:58, 10 February 2009 (UTC)

Why does this proposal use boundary:type=* insted of the established border_type=*? Why does this proposal try to invent new names for the EEZ and contingency zone, when these already have established names in UNCLOS? -- Gustavf 15:06, 10 February 2009 (UTC)

Hi Gustav, we use boundary as master-key, so we should not switch to "border". Sub-keys are build with master-key and ":" and sub-key. Of course: use the right UNCLOS-terms. --Markus 15:20, 3 March 2009 (UTC)
border_type=* was established from long before there was a discussion about maritime boundaries. --Skippern 19:14, 3 March 2009 (UTC)


A comment to a vote said that this proposal does not even metion relations. That is not entirely true, as it is said that the territorial waters: "This line should also be member of the border relation surrounding the country." It is of course up to the definition of any relation, what should be part of that relation. This proposal merely suggests how a way being part of a border relation should be tagged when it is a maritime border. -- Gustavf 19:12, 17 February 2009 (UTC)

"Contingency" should be "Contiguous"

The 24nm zone, which we are currently calling the "contingency" zone, is actually called the "contiguous" zone, per UNCLOS. (See also Wikipedia.) I suggest that border_type=contingency should be changed to border_type=contiguous. - Htonl 21:24, 2 March 2009 (UTC)

Rewrite / New Vote

I have started to look into rewriting this proposal in order to get a new vote going. I am not satisfied with the outcome of the vote, specially since many have voted against and argued with things I saw as part of the idea behind this proposal. Last count I had was 9/8 in favor, so the popular vote was for, but I will still rewrite and open for a new vote, in hope that we get a better result. --Skippern 03:55, 24 March 2009 (UTC)

1st Vote Results

  • I approve this proposal I approve this proposal. -- Gustavf 18:36, 17 February 2009 (UTC)

  • I approve this proposal I approve this proposal. Sletuffe 18:49, 17 February 2009 (UTC)
Given the comments down there, I think I haven't enough knowledge to decide if it is good or not, therefore I prefer to abstain Sletuffe 00:54, 4 March 2009 (UTC)
  • I approve this proposal I approve this proposal. --Skippern 18:51, 17 February 2009 (UTC)
  • I oppose this proposal I oppose this proposal. - I firmly believe that using simple ways with "country:left" and "country:right" and the like is outdated for any type of border. Relations should be used to describe the various zones, and tags should be applied to relations, not participating ways. However this proposal does not even mention relations. --Frederik Ramm 18:59, 17 February 2009 (UTC)
Than you havn't read the discussions where relations have been mentioned several times, IMO the territorial border should be a part of the national border relation. --Skippern 19:07, 17 February 2009 (UTC)
Left and right is continuing the currently used scheme, its use doesn't hurt the proposal, but shouldn't be required by it. Relations are the best approach to this. --Thomas Wood 23:23, 17 February 2009 (UTC)
By reading this proposal I have assumed that relations are usable as well as left: and right:. Though, mentionning it would be better. Sletuffe 10:17, 18 February 2009 (UTC)
Ok, not clearly stating that relations should be used on territorial waters is maybe a weakness of the proposal, though it is not a reason to reject it. As long as the base tags are approved, additional features can be amended later. --Skippern 11:13, 18 February 2009 (UTC)
  • I oppose this proposal I oppose this proposal. - I agree completely with Frederic. I tried to implement parsing the old concept (left/right) and it does not work (downloading the border is difficult at best and the borders cannot be stiched together because they where broken all the time). Grouping the parts of a border via a relation works much better. --MarcusWolschon 11:18, 3 March 2009 (UTC)
  • I approve this proposal I approve this proposal.--Walley 19:50, 17 February 2009 (UTC)
  • I approve this proposal I approve this proposal. Pleased to see a sane approach to these types of border. --Thomas Wood 23:22, 17 February 2009 (UTC)
  • I approve this proposal I approve this proposal., but wonder how you'll draw the line between Greece and Turkey... --katpatuka 06:11, 18 February 2009 (UTC)
  • I approve this proposal I approve this proposal.--Patou 06:51, 23 February 2009 (UTC)
  • I approve this proposal I approve this proposal.--Nickvet419 Flag of United States 17:27, 25 February 2009 (UTC)
  • I approve this proposal I approve this proposal.--Bathizte 17:46, 01 March 2009 (UTC)
  • I approve this proposal I approve this proposal. As long as border_type=contingency is changed to border_type=contiguous as per UNCLOS and as I have explained on the talk page. -- Htonl 21:25, 2 March 2009 (UTC)
I will read over UNCLOS again, and if it is only a spelling mistake, than of course it will be corrected before implemented. --Skippern 11:10, 3 March 2009 (UTC)
It's Part 2, Section 4, Article 33: "In a zone contiguous to its territorial sea, described as the contiguous zone, the coastal State may exercise the control necessary..." - Htonl 13:02, 3 March 2009 (UTC)
To be corrected in approved feature. --Skippern 13:53, 3 March 2009 (UTC)
  • I oppose this proposal I oppose this proposal. - this should be solved by using relations --xylome 11:00, 3 March 2009 (UTC)
This is needed to be solved by your relations, at the moment there are nothing showing the maritime boundaries, and using the coastline is completely wrong, at least for national boundaries. --Skippern 11:10, 3 March 2009 (UTC)
  • I oppose this proposal I oppose this proposal. - National boundaries - and the 12-nm-boundary is part of it - are attributed as relation.
    So the sea-side national boundary should be attributed with a sub key. May be boundary=territorial_waters or 12-nm-zone, or 3-nm-zone.
    For all the other national zones we need other relations:
boundary=baseline, boundary=custom, boundary=contiguous_zone, boundary=exclusive_economic_zone, boundary=continental_shelf

Please discus this under boundary=administrative. We don't need a new master-key for national boundaries and corresponding borders. --Markus 14:51, 3 March 2009 (UTC)

  • I oppose this proposal I oppose this proposal. I agree with Frederik and Markus. Maritim borders are sorts of national related boundary with different legal consequences. Actually 12 nm is properly shown if tagged with boundary=administrative adminlevel=2 - please have a look at the southern Northsee. Additional tagging rules for other maritim borders are OK but do not need boundary=maritime as a main key. -- nkbre 00:39, 4 March 2009 (UTC)
  • I oppose this proposal I oppose this proposal. A maritime border is an administrative boundary and should be tagged as such (boundary=administrative). If the main purpose of the proposal is to have maritime borders rendered differently to land borders then use something like maritime=yes as a hint to renderers. Polderrunner 08:58, 4 March 2009 (UTC)
  • I oppose this proposal I oppose this proposal. I agree with Polderrunner. --Cartinus 19:16, 4 March 2009 (UTC)
  • I oppose this proposal I oppose this proposal. -- for the same reasons described above by Frederik Ramm and MarcusWolschon Dieterdreist 01:47, 9 March 2009 (UTC)
  • I oppose this proposal I oppose this proposal. relations should be used and not "country:left" and "country:right" --Ckruetze 00:49, 11 March 2009 (UTC)

Proposal in August

Ich denke, das ist eine Verschlechterung von "boundary".
Die seewärtige Grenze der Länder ist nicht international gleich. Zwar liegen einige auf 12 sm ab Basislinie (Basislinie /= Küstenlinie), aber es gibt viiiele Sonderregeln.

Die aktuelle Regelung:

+ adminlevel=2

ist gut, sie erfasst alle Variationsmöglichkeiten. Sie sollte nicht geändert werden.

Andere seewärtigen Grenzen (Basislinie, 24-Meilenzone etc.) könnten im bestehenden Schema abgebildet werden:

+ boundary:type=*         ( Baseline | 24_nm_zone | 200_nm_zone | 350_nm_zone )

Es macht m.E. wenig Sinn, einen zusätzlichen top-level-Schlüssel für Grenzen zu erfinden, im Gegenteil: das würde eher Verwirrung stiften.

Denkbar wäre, seewärtige Grenzen in den Namensraum "seamark" zu verschieben und sie nur auf nautischen oder anderen Spezialkarten darzustellen, und nur die seewärtige Staatsgrenze in den allgemeinen Karten zu rendern.

Siehe auch: de:Key:boundary#Seewärtige_Grenze : please translate! (vielleicht kann das jemand übersetzen?)

Gruss, --Markus 07:51, 13 August 2009 (UTC)

I am able to read German, but unfortunately my knowledge of the language is not good enough to write it. It seems your knowledge og English is similar, so I will answer in English.
First of all, you will notice that the proposal talks about maritime borders as defined by the United Nations Convention on the Law of the Sea. That is why a term like "territorial waters" is used, and not something like "12 nm line".
The proposal does not change the boundary=administrative + adminlevel=2 usage, but amends it with maritime=yes. It clarifies that the boundary tagged this way should be the territorial waters line, and not the coastline or the baseline. I think this is a very useful clarification in itself.
In your vote comment you suggest that a rendering rule is used instead of maritime=yes. I suppose you mean something like "select the part of the border over the sea and render it as a blue line". This does not work, since the border is considered a land border out to the baseline. An example could be the border between Sweden and Norway ( Here the land border is over the sea from the bottom of the fjord and out the the outermost islands. Using only a rendering rule would not convey this important detail. I think adding maritime=yes to the sea border is a very unintrusive way to tag this information.
If a new "top level" key is needed can certainly be discussed. Could boundary:maritime=yes or boundary:type=territorial_waters or a number of other names been used? Certainly, but so far it has been common to use "top level" keys for most values (electrified and frequency for railways, dispensing for pharmacies and admin_level for borders are just a few examples). I would certainly not mind moving to more tags using prefixes to avoid confusion, but this proposal uses what is currently the most common scheme.
I do not agree that using boundary=administrative for more than one level of maritime borders is a good idea. That would, in my opinion, be more likely to cause confusion than introducing a new tag. My original idea was to tag the baseline this way, but I was convinced that the terriorial waters was a better idea (see previous comments).
Should the other boundaries (baseline, continous zone, eez) be tagged using boundary or seamark? I think it is quiet clear that they are boundaries, and not navigational aids. Using something like seamark=maritime_border + seamark:border=eez is hardly any better that this proposal.
Based on all of this, I think that this proposal is a good way to tag maritime borders.
-- Gustavf 18:04, 13 August 2009 (UTC)
Hallo Gustav! This is a nice proposal, helping in national boundaries. I like your following to the UN-Conventions.
Yes, "territorial waters" sounds more understandable then "12 nm line". But sometimes the 12nm-line has her own local meaning and therefore we need also a tag. "territorial waters" (what for) and "12nm-line" (where) are two different classes. We should separate it.
Creating new tags, I prefer a hierachical or relational style, who refers to the tags we have already (or create a better model instead of the used one). So I regret, that here is created a "new thing" for boundaries. I prefer consistent models. Here we discuss administrative boundaries - and the boundary is one class seaward and landward, made by two nightbeour countries. The proposal does not well show this context. All "maritime borders" here are boundaries with admin_level=2 (and I don't like to learn a new term "border").
But altogether, the proposal is a win right now. --Markus 11:14, 15 August 2009 (UTC)
"seamark:" is a namespace for OSM-tags in the "blue" part of the map. We use it to separate geodatas in land and sea. With this we are able to show different views of land- and sea maps without disturbing interferences. --Markus 11:14, 15 August 2009 (UTC)


  • Note that the treaty says that 12nm territorial waters can go beyond 12nm from shore within some bays and bights where a straight line can be drawn across it. As with countries closer together than 24nm, often the real boundary is a political compromise not an exact 12nm buffer from the land. Bad assumptions based on a strict interpretation of 12nm buffer from shore has led to many a lawsuit ... --Hamish 10:21, 29 April 2010 (UTC)
I am fully aware of these caveats, but this proposal is written in the light of the UN treaty of maritime boundaries, which are followed by almost all coastal states. There are some gray areas where the treaty doesn't quite cover the real picture, such as the "Smutt Hole" (Iceland/Norway) and "Smutt Ocean" (Norway/Russia), and areas where straights and canals dividing countries are less than 24nm. All of these exemptions are resolved with political agreements rather than UN treaties and resolutions. --Skippern 18:56, 29 April 2010 (UTC)