Mapping of boundaries

Mapping of boundaries (admin_level=4,5,6,8,10 ; state to ward):
General problem: usage of other ways:

  • the coastline should be used as boundary
  • I would not use other mapped ways as boundaries, because roads, streams, rivers… are too often changed a bit and in consequence the boundaries are broken by Newbies. So seperate ways for boundaries are more secure.
  1. boundaries of admin_level=4,5,6,8 (state to panchayat) :
    There should not be any „nowhere land“. That means the districts fill the state area without gaps and intersection. At the borders the district boundries use the state boundary way. Like this also from district to subdistrict and subdistrict to panchayats…(level=8)
    These boundaries use the coastline and the middle of the rivers.

  2. Ward boundaries (admin_level=10):
    Specially in Kerala the mapping of wards has started. So there came up the following questions.
    a) Do the river and large water areas belong to the wards?
    The official documents in Kerala are ambiguous. Often they use the water edge. Then they let the boundaries go to the middle of the rivers. The maps are outdated; that means the river edges have been changed, new islands, changed islands…
    So how do we map these ward-boundaries near rivers and large water areas?
    Solution 1 : extend the ward boundaries to the level=8,6,5,4 in the middle of the river, lake…
    Solution 2 : ward boundaries finish at the water edge of the rivers, lakes…
    I prefer solution 2.

b) ward boundaries in land regions (no rivers, streams…):
Here the ward boundaries should not create gaps to higher boundaries but go to the level=8-boundaries. In no case do ward-boundaries cross other boundaries.

What is your opinion to these problems?

I have a basic question first. For villages, are we supposed to use admin level = 9 or 10?
https://wiki.openstreetmap.org/wiki/India/Administrative_Boundaries says level = 9
https://wiki.openstreetmap.org/wiki/Tag:boundary%3Dadministrative#admin_level.3D.2A_Country_specific_values says level = 10

Sorry but I am confused.

That is a difficult question. There has been a discussion in the Telegram group India but there is not a definite conclusion.
My opinion : village boundaries with admin_level=9 and boundary=administrative.
Do you have a good data source for the boundaries of villages?

Agree

Here, i feel we can distinguish between waterways which form natural land boundary vs other features like roads, building, farms which are not natural boundaries.

I support using waterways/riverbanks in boundary relations since they are the most common way to legally define a boundary and we should try to model those definitions wherever we can. These features also not commonly modified by new editors once mapped, so presents a lower risk of breakage.

Reusing road ways in the boundary relation definitely seems prone to breakage as they are continuously edited. It also becomes problematic when roads need to be detailed further with separate dual carriageways or have a detailed junction or roundabout.

I’m also for this solution to model ward as a basic unit of human representation and not land control. We can assume areas with ward gaps within a higher admin level are uninhabited natural areas.

In the telegram group we went through the various levels and seem to agree on admin_level=9 being the best fit for a revenue village boundary. Notes: https://etherpad.wikimedia.org/p/osm-in

Unless there is any opposition, i think its a good idea to update the wiki to admin_level=9 : revenue village, so there is no confusion for that.

At admin_level 5 (District) administrative boundary and local authority boundaries start to take different paths. On the administrative side it can go without gap and intersection; District split to Subdistrict (6); Subdistrict Split into Revenue Circles (7); Revenue Circles then Splits to Revenue Village(9)

Local Authority boundaries including Panchayath(8)/Corporation(8) need not follow that condition except at the level of district boundary.
While we assign admin_level values to Panchayath and Ward; they do not belong to administrative hierarchy or boundary in any manner. That need to be clearly stated.

While in Administrative boundary we take top down approach. For local authority and political boundary, we may go bottom up. To best of my understanding - as in discussions we had in telegram group - it follows like this;
Most basic unit is a ward.
In Rural area - A set of wards can form Panchayath; A set of wards also form Blocks which can overlap Panchayath; All wards (rural) together shall become District Panchayath (which is district boundary substracting all urban local authorities)

In Urban Areas - A set of Wards can form a corporation or municipality; In case the corporation is large; set of wards can become zones; zones together forms corporation.
If we add District Panchayath with with urban localbodies we get district boundary.

Wards are also the basic units combining which political boundaries of Assembly and Loksabha constituencies are formed.

Boundaries at local authority level, including ward, is primarily about defining who live in and who live out of that geography. It is not primarily about ownership
of a piece of land or authority over a piece of land. This is reason why they are drawn as if on river edge. Also often the boundary is defined with mountains; rivers or roads as edges. Then when it comes to who manages an controls the particular river etc, it is determined in quite a fuzzy manner. However this negotiation is quite important at panchayath level as to what is responsibility of local authority with regards to controlling people polluting water bodies, monitoring it etc. I am sure we may be mapping arbitrarily on our own initially. Slowly it can become part of the system.

Wards together must form District (5), considering that, I recommend Solution 1. There shouldnt be gaps; Every one should have a ward, even in forest. It does not really alter definition and purpose of ward boundaries, i feel;

What makes ward boundary more challenging is that it is politically negotiated and for the political purposes; few houses across a road or river can also be part of a ward which is largely on the otherside. (Which is not expected or acceptable). Addressing such anomalies in the future require ward maps. We dont really have one now.

I compared in the Kasaragod district of Kerala the already mapped boundaries and the ward/panchayat/block boundaries in the delimitation maps.

  1. the block boundaries cross the subdistrict boundaries
    example : the Kumbala panchayat belongs to the Kasaragod block but is not part of the Kasaragod subdistrict but it is in the Manjeswaram subdistrict. But our mapped Kumbala panchayat boundary is the same as in the delimitation map. That means the mapping of ward boundaries within this panchayat does not make problems.

  2. I found one case, where our mapped panchayat boundary does not fit to the boundary in the delimitation map : Bedadka panchayat
    Normally a panchayat is split in several wards. Since in this case the panchayat boundary is not clear, the wards con not be mapped without more exact informations.

A general problem is the low precision of the delimitation maps. POIs, rivers, streams, streets and railways are very unprecise. So the georefencing of the maps is difficult. It would be possible to map ward boundaries with a precision of about 200m. But there are same very small wards (only 200m extension).

If we have the chance to get better maps, then we should wait with the mapping of wards.