On Sep 27, 2020, at 12:51 AM, Mateusz Konieczny via Tagging 
<tagging@openstreetmap.org> wrote:
> I am a bit dubious about value of updating fire=perimeter

It isn't anticipated to be "updated."  It is a static boundary where "inside of 
this polygon, there might be burned / destroyed landcover (and perhaps some 
buildings, if they are / were there), outside of this polygon, there is / was 
no fire."

> It is something that changes extremely quickly, we should
> not encourage people to survey perimeter of ACTIVE fire,
> OSM is doomed to be strictly worse source of fire perimeter
> than alternative sources....

I'm not surveying this (though I would like to, many areas are inaccessible or 
dangerous).  I'm saying the polygon tagged fire=permiter is a useful data 
structure in the map to delineate where the bounds (perimeter) of a fire was 
(now it is "was," for a while it was "is").  So, use ground truth, 
personally-gathered sources, satellite data... to better characterize that what 
was once there (landcover in the form of trees and scrub, largely) is quite 
likely no longer there.  So, "map this area better."

> Obviously, we should (try to) update map where situation changed.

Maybe it wasn't obvious from my posts about this, yes.  This is the whole 
reason for entering these data.

> Delete building that will not be rebuild (mark them as destroyed:building=*
> until aerial imagery will update)
> [deleting buildings and remapping them as they get reconstructed may
> be viable in cases of heavy mapper presence]
> Delete other permanently destroyed objects and so on.

It is still too early to make final determinations of buildings:  some people 
who lost homes to fire will rebuild, some will not.

> > Do we have landcover tags which could replace landuse=forest
> or natural=wood with something like natural=fire_scarred?
> 
> AFAIK nothing established, see
> https://lists.openstreetmap.org/pipermail/tagging/2018-March/035435.html
> for related discussion about wind damage.

Thank you, that is interesting and relevant!  So, preliminary results are that 
such tagging is rare, but it does happen.

SteveA
_______________________________________________
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Reply via email to