Hi folks.
I noticed that the boundaries of Northern Ireland’s 11 ‘super-councils’ were not mapped on OSM.
This seemed a pretty big omission, as these are the local government divisions used in NI since 2015 (see Local councils in Northern Ireland | nidirect ). These are the people who empty your bins
It was a fairly painless process that took me 3 or 4 days.
As I expect many of you know, nearly all boundaries on the island of Ireland use the historical townlands (Townland - Wikipedia) as the ultimate building block, and the Super-Councils are no expection. OSM already has pretty much all the townland boundaries already mapped, so this makes it fairly straightforward to create a new releation for a ‘super-council’, and add it to the various constituent ways that are already tagged with townland boundaries.
I wonder if any other mappers would be interested in mapping the remaining 10 super-councils?
However this legislation says the boundaries are those defined in “MAPS REFERRED TO IN THE LOCAL GOVERNMENT (BOUNDARIES) ORDER (NORTHERN IRELAND) 2012” and signed and sealed with reference to this Order and deposited at the office of the Department of the Environment at Goodwood House, 44-58 May Street, Belfast, BT1 4NN"
Obviously that not a very convenient reference source
As a more practical reference source, the supercouncils themselves usually have maps on their own websites, however these are of raster maps and can be a bit ambiguous on detail.
As a more hi-res source I took a peek at OSNI’s https://maps.spatialni.gov.uk/ which has layers showing townland boundaries and ‘Local Government Districts 2012’ (the latter being the 11 ‘super councils’). (I trust this is permissable, as I was not copying the shape files or anything like that, just using it to confirm which pre-existing ways and townland boundaries (that are already drawn on OSM) corresponded to which supercouncil? - in other words, just as an aid to labelling)
I opted to tag the supercouncil boundary relation as ‘boundary=administrative’ and ‘admin_level=6’.
This puts it on a par with local government authorities (country or city councils) in the rest of the UK, which is more or less the correct equivalent (albeit the demarcation of powers between NI local councils & Stormont isn’t quite the same as in other parts of the UK)
We happy with that?
(Note: in most renderers, this has the effect of it being labelled as a ‘County Boundary’. As we know, in Northern Ireland, the long-suffering ‘traditional counties’ have not served any official administrative function for many decades, despite being by far the most commonly used geographical devision for the man-in-the-street… These traditional counties are currently tagged as ‘admin_level=6’ but have ‘boundary=historic’ . I guess this is fair enough?)
other small quirks (which will be applicable to several of the other super-councils too):
a fair proportion of Newry, Mourne & Down District Council’s boundary is the international border between RoI & UK. I just used this as it stood. (I noticed it didn’t always 100% match where OSNI thinks the border is, but I didn’t fiddle with it )
sea boundaries: generally the supercouncil boundaries seem to stop at the shoreline (i.e. the don’t extend out into UK territorial waters. I therefore usually just tagged the ‘coastline’ ways with the super-council boundary.
The only exception to this was inside Strangford Lough, where I had to draw a new way to demarcate between Newry, Mourne & Down supercouncil and the Ards one.
i think we have always assumed that this is NOT a permissible source for OpenStreetMap which is why the boundaries have not been created, but I have not kept up-to-date with this.
In the past various people have had meetings with OSNI in the past. Various issues with persuading Land and Property Services as a whole to release open data were discussed at the OpenDataCamp held in Belfast back in 2017. Looking at various sites the actual licence under which the data are available is unclear: the UK Open Government Licence or the LPS version of the the OGL. Earlier versions of the LPS OGL were unsuitable.
One approximation can be made using the Food Hygiene Open Data as these have the local authority coded for each entry. I did this such a long time ago, I can’t recall where I put it.
Members of the OSM Ireland community seem to use Telegram these days, so you may get more current information via that channel.
Hmm … well, as I said, I wasn’t copying data from the OSNI: just using it to double check a few of the squigglier bits from the councils’ own maps. In 99% of cases, the Super Council boundaries are just following existing townland boundaries, which are all in OSM already and pre-date the creation of the OSNI by centuries, generally following field boundaries (hedges, rivers, tracks…) that are hundreds of years old and can clearly be seen on satellite imagery.
My recollection is that that’s an anti-spam measure, but my memory may be incorrect there (I’m an occasional visitor only, both to the Telegram channel and the island).
This tool is very helpful for seeing what boundary relations exist all over the world (it has a tree structure that lets you drill down level by level in each country).
I discovered that I am not the first to try and add NI Supercouncils to OSM: last year someone already set the ball rolling with Fermanagh & Omagh District Council (FODC). https://www.openstreetmap.org/relation/4183318
So, I was able to enrich the tagging for that one a bit, and harmonise the tagging with the one I did for Newry, Mourne & Down (NMDDC). Now we have just 9 more supercouncils to go
One interesting decision the contributor for FODC made: they selected ‘admin_level=7’.
On reflection, and with all due deference to Tag:boundary=administrative - OpenStreetMap Wiki, I would agree with this choice, and have revised the admin_level I originally chose for NMDDC (6).
The main reason I think this works best is that (as https://osm-boundaries.com/ shows) is that admin_level 6 has already been bagged for the ‘traditional’ 6 counties.
It could be argued that these 6 counties are now only ‘historical’ and don’t belong admin_level hierarchy but (a) the grubbings of civil servants notwithstanding, pretty every human on the island of Ireland is well aquainted with them and (b) it’s slightly ‘niche’, but they do still exist as administrative areas in their 1898 boundary forms (that mean 6 counties plus ‘county boroughs’ for Belfast and Derry), as ‘Lieutenancy Areas’ for the British Crown (see Lieutenancy area - Wikipedia). These Lieutenancy Areas don’t do very much (they’re involved in things like the honours system, and providing dignitaries to preside at citizenship cermonies) but they are nonetheless still a thing.
OSM Wiki suggested admin_level 6, but noted that it hadn’t been decided yet. Using level 6 would place NI Supercouncils on a par with other local government councils throughout the UK, but the fact is they are not really on a par in terms of their administrative functions anyway, as lot of things done by councils in GB (roads, libraries, schools, social housing) are done by central government (i.e. Stormont) in NI.
Neither do I, but this subcategory doesn’t see much use.
I agree that this is because of the role of the traditional counties as admin_level=6. These keeps admin level tagging across Ireland (e.g., Fingal), but not across the UK (although admin_level=6 is used for both top-tier councils and combined authorities).
Thanks SK53.
Re: your blog post - FWIW, I’ve been doing hundreds edits in my local (fairly rural) area in the past year, and from peeking at OSNI maps in comparison, my completely-unscientific-and-anecdotal impression is that the correspondance between OSM and OSNI townland boundaries in these parts in excellent.
The vast majority involve field boundaries and roads, and appear to be correct. (The exact geospatial positioning of those features may not always be super tight if satellite imagery is to be believed, but in terms of logical positioning - e.g. along a road or field boundary - they’re generally correct). Where streams are involved, the accuracy is more wooly - I would imagine this is because streams and ditches on agricultural land often change their courses (or have them changed) over the decades.
I could count on the fingers of one hand the number of times were OSNI and OSM disagreed about particular fields being in a different townland, and I felt justified in making a correction.
I would concur that the OSM coastlines are a bit hit and miss at times. But it’s something of a moot point as regards townland boundaries, as the boundary is still ‘the coast’, however exactly the map might chose to demarcate that.
Where you note that:
Blockquote Islands. Some offshore islands appear to be missing from the OSNI data (see The Skerries N of Portrush)
I would speculate that this is because in centuries gone by, may small off-shore (or intra-lough) islands on Ireland were farmed, and thus as relevant to the townlands system as any other part of the land -
However in modern times they are nearly all uninhabited, and so have no residential addresses on them containing persons who can vote or pay tax - so this may have contributed to them being omitted from some modern legally-defined local government and electoral areas, and hence not shown as part of those areas in ONSI data?
Very much so. At one stage I contemplated a project to improve the coastline county by county (and incidentally try and add missing offshore islands, skerries & drying rocks). Unfortunately we lack a decent open data reference, and therefore must fall back on the GSGS 3906. You can see the issue with my example of Umbra.
Particularly in Northern Ireland streams often run in wooded glens and it’s really hard to pick them out. Where my cousin lives on Inishowen I only realised that one boundary of his local townland is a stream by hearing it.
Other things to look or are the mistaken assumption because something has an acreage figure on the old 6 inch maps it must be a townland: islands have areas reported too. There may be one or two spurious townlands created this way.
)