[Tagging] Feature Proposal - RFC - boundary=forestry(_compartment) relations (Was "Feature Proposal - RFC - boundary=forest(_compartment) relations")

[David Marchal]

Bert,

Reading you, I'm starting to think that it will be near impossible to manage what you're talking about in the proposal, but I'll try. Don't think I make harsh comments, but I'm trying to unmess my mind, because what you describe is very, very different of what I know.

Sent with [ProtonMail](https://protonmail.com) Secure Email.

‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
Le lundi, 15. février 2021 15:31, Bert -Araali- Van Opstal <bert.araali.afritastic at gmail.com> a écrit :

> Thank you for your answer David. It is clear to me and your justifications underline it, that it is your intention to present a solution in mapping (boundary driven) and tagging to present a "westerner forestry" philosophy. But I remain with my previous vision that your proposal, which is in no way incorrect, has too much conditions and limitations to be considered as a top level boundary class. In this format, as you can already see from the many questions and objections, it is not suitable to be used on a top level as was done with hazard and others, which cover a more broader spectrum of management / use cases and were specific issues are addressed with attribution keys.

These conditions are there to mitigate possible contestations about the area purpose/limits/use as a forestry area. You probably saw that I'm a westerner, and our customs and habits are often way more strict and organized than what you describe; proposing something not structured enough would, in my mind, give a stillborn proposal.

> As I see it all your forestry proposals would fit under a protect_class 15, or one of the IUCN classes. Your proposal says that if a forestry area is also a protected_area, an additional protect_class should be added, meaning that you propose to tag the whole area as forestry and protect_area with IUCN classification is of minor importance, since your primary boundary identification remains Forestry. I think that is definitively wrong and should be the opposite. Protected_area with IUCN classification is always more significant.

If the local laws does not consider the area protected, I think it will be difficult to explain to mappers that it should still be mapped as a protected area. In my country, protected areas are generaly seen as something very important and very different of non protected areas, so telling people to boundary=protected_area on most wooded areas would most likely lead to something like "What's this weirdo? I'll stay with landuse=forest.". Again, I'm not being harsh, I try to explain why I think your reasoning would be inapplicable.

> This leaves us with areas which, if you allow me to refer back to protect_class 12 / 14 / 15 / 19 that apply to resource-protected areas. In stead of defining very specific "forestry" boundaries, what is against defining a more general therm, like "natural-resource-area" or any other more general term, under which the "western forestry" is defined with a new or existing atrribution key:value. Much the same way as has been done with hazard zones.
> This would allow to use the same attribution keys for other areas, as species management, wetland management, water_resource_management (introduced without proposal !) or even more natural resources. As you already mentioned forestry can mean that other natural resources are managed under "forestry" besides wood extraction, but still to much limitations to be considered as a top-level boundary tag.
>
> I see most confusion and discussions arising from the very strict boundary requirements in your proposal. Boundaries are in the non western world not always, even I dare to say, mostly not clearly defined.
> Again this is going to be a long mail for which I apologise but I see no other way to make it clear. So allow me to illustrate with some examples from my home Uganda:
>
> 1. Most wetlands and forests are under the jurisdiction of the Uganda National Forests Authority (NFA). They have a mandate to maintain and exploit these wetlands and forests defined by law. All those forests and wetlands (the majority here) are thus "government land" and called "Central Forest Reserves", within these areas the "Uganda National Forestry and Tree Planting Act" has to be implemented by the NFA, in some cases even if there was never any "woody vegetation". So far so good, we have an act and a clear definition of forestry in the law. However, the government never gazetted the official boundaries of most of these Forest Reserves. Even more, they were so many that smaller forest were assigned to be managed by local governments and called "Local Forest Reserves", but also for these never clear boundaries were gazetted, even worse, the Local Authorities lack the resources or personnel to implement the Forestry act. So in practice very few boundaries are clearly marked in the field. Then if they are marked, due to encroachment and the communities argue about the fact that they were never gazetted, if the NFA puts boundary stones, often in peoples gardens, these are removed by the community and huge commotion starts. So if I would follow your proposal, none or very few could be mapped with boundary=forestry.
>
> 2. Because of 1., someone connected and the lack of official maps or even a cadastral register that overs the whole country, someone connected to the NFA decided they will use OSM to map all the Central Forest reserves, very rough and luckily with a fixme tag that they "are not official boundaries". All were mapped as landuse=forest, even those were by far no more tree is present ! In many cases we, the community already mapped other landuses from field checks, and natural tagging. But again lots of commotion since the NFA action caused all of these areas now to be rendered as forests (under the forestry act). We started to change their forest with relations with boundary=protected_area and protect_class=15. Still many unhappy users and unhappy NFA because the borders didn't render. So we changed it to protect_class=5 or 6 because it is rendered. We couldn't find any official document where the IUCN classes are connected to the local legislation. But still, "not official boundaries" remains, and except from some historical maps hard to determine were they are. In the field, for the few that are there they are often on private land and not verifiable without entering encroached "private" property and danger for life and limb. Still, if I refer to your proposal, none of this would be a candidate for your proposal !
>
> Example on OSM: https://www.openstreetmap.org/way/677824323 and https://www.openstreetmap.org/way/711261339#map=16/0.0825/32.4835.

I'm sorry to say that, but I'm dubious about the fact that such a situation can be accurately mapped; it sounds that there is no agreement about who uses the land, what for, by which means, and along which limits. Please don't understand that I'm having some sort of racist point of view here, but I can't understand how to map something that isn't even clear to the locals. That is not simply related to my proposal, it is related to the simple possibility to map such a messy situation.

> 3. Regarding indigenous use. If there are people who clearly marked their territory it is them. They mark it by making marks on trees or other "natural means". However, again not official but they are allowed to harvest, hunt and extract wood from the forest. They do it in a sustainable way but in some cases, targetting specific valuable species like mahogany trees (by the way protected but they don't care, they have to make a living somehow) are extracted for commercial purposes, and replanted but the old monumental trees are all disappearing. Not all of them do this but some. There areas are sometimes located partially or in full in national parks but also on non-protected land, common, government land but in their eyes it is their land, not regulated as in reservations like in the US. So, again referring to your proposal: boundaries, officially not gazetted, check in the field yes they marked their boundaries where they are in their opinion (lots of conflicts with other surrounding communities) but they are officially mandated and allowed to harvest, hunt and extract wood in a sustainable way. So your proposal boundary=forestry, check, yes all conditions as described fulfilled although it is not, because the main activity there or main purpose is to allow them to continue their traditional lifestyle of harvesting and hunting, but hey, who doesn't like a nice smartphone or fancy bike to run through the forest and impress your neighbours... they are people like us and deserve to enjoy the same rights and luxuries.

Here, the limits are verifiable using the local marks, and one can see traces of the forestry works, so, if I told precendently that it is not subject to my proposal, it seems that I was wrong: what you describe here is clearly forestry and eligible to mapping under the current proposal terms.

> 4. The wetlands. All our wetlands are government land. In some of them the government allows extraction of papyrus reeds (for roofs), some have designated areas for rice fields, used a sewage filters, drinking water extraction and... the swamps, marshes and other wood covered areas as "forestry areas", used for wood extraction, charcoal burning etc... . Mandated mostly to the NFA who lacks financial and human resources to implement the act, and no clear boundaries. Some of them located in nature reserves, some of them ramsar some of them not. Your proposal: Clear boundaries no, one can question even if it is possible to define them because the wetlands are sometimes seasonal and some of them shift ! Wood extraction = yes, sustainable = sometimes, use of other natural resources yes, protected area = different interpretations, so I believe according to you no forestry ? Although some of these forest are like islands in the wetlands nd thus have a clear but ever changing boundary, not marked with other physical means in the field.

Well, again, if the limits are not clear, I'm dubious about the fact that it can be mapped, whatever the tagging. The simple fact of drawing entities here seems very difficult and subject to constant discord.

> 5. As last example, Ziwa Rhino Sanctuary. This sanctuary is located in a forest reserve, concession of NFA. They breed rhinos and also the rhinos maintain the scrub and forest in a natural way, but with support of human intervention. Because of the rhinos the whole sanctuary, and thus the forest reserve is fenced. Resources used: wood extraction / charcoal burning = yes, animal farm, breeding (species management), yes, nature reserve = no, compartments = yes (compartments, fenced are made to prevent overgrazing by the rhinos so they are forcefully moved through the forest). Clear boundaries = yes. SO where do we end up if we follow your advise ? Boundary=protected_area = no, boundary = forestry = yes but its an animal sanctuary / breeding farm, ok I can tag that additionally. But again practical problem: they also make a lot of money with tourism, this caught the eye of some opportunistic government people so they decided to stop the concession and take over the sanctuary, as justification: violations of the Forestry and Tree Planting act. SO are we again going to risk live and limb to define it as boundary = forestry ?
>
> That is why I think a less restrictive general new boundary value would be much more suitable, which allows us to carefully propose a general term for the management of natural reserves and where we use carefully and diverse chosen attribution to provide details, of which "western forestry" is just one variant.
> Please do comment and advise how we can handle these examples , I am sure many similar challenges exist throughout the world ! And yes be bold, I fully support deprecating landuse=forest because that is even worse !
>
> Greetings, Bert Araali

Well, for your last example, it is very tricky. It seems to vastly exceed the proposal scope, as it also have some aspects of landuse=meadow, and, again, locals do not agree about what the land is. I'm not sure I can give a sound advice in such a situation, but, then again, let's try. How is this currently mapped?

Regards.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20210215/dfa089ef/attachment-0001.htm>