[Tagging] [Talk-us] Large fire perimeter tagging?

[stevea]

On Sep 30, 2020, at 5:27 AM, Paul Allen <pla16021 at gmail.com> wrote:
> BTW, ordinary polygons won't do for this.  You'll need a multipolygon
> to exclude the Mount Wilson observatory and some campgrounds that
> were saved from the fires burning all around them. :)

Perhaps I have not been clear or remain misunderstood:  the polygon is not an exact delineation of "this (and exactly this) has all burned."  It is the perimeter of the fire, inside of which the fire was "fought" or allowed to burn, outside of which, "not."  If there are areas (like Mt. Wilson Observatory and campgrounds) inside of a perimeter that were "saved," the polygon should not explicitly exclude these areas with role "inner" as part of a multipolygon relation — that isn't the semantic of the perimeter.  Many areas inside the perimeter DID burn and will need their existing land cover (natural=wood, natural=scrub...) tags removed, some areas (trees, houses which were saved...) did NOT burn.  It would not be correct to re-map the fire=perimeter as a multipolygon with not-burned areas with role "inner."  As newer imagery becomes available, it is correct to leave not-burned areas alone, adjusting them up to the edge of where they DID burn, and in areas which did burn, removing / adjusting-to-smaller-areas land cover tags as they exist now.  This area was almost exclusively heavily wooded in the real world and OSM well maps fairly precisely where these "woods" were.  Only now, much of them burned.  Where, exactly?  "Somewhere inside of" the polygon denoted fire=perimeter.  OSM contributors await newer imagery, we will better (re-) map landcover and other data that are inside of the polygon when they become available.

At the completion of this process, the usefulness of the polygon diminishes to zero (perhaps there remain closed roads and dangerous areas, these can be mapped "differently," although "no-go" area tagging remains unclear) and the polygon can be removed, having exhausted its usefulness.

Some might complain that such an "improve existing map data HERE" polygon overlaps with small projects like a localized import or a Mapping Party to improve a particular city or county, saying a Tasking Manager or similar tool can and should be used to manage this.  But while "a particular city or county" have well-defined, largely in-the-map boundaries, an area devastated by major fire has no such boundary.  Unless and until a polygon tagged fire=perimeter is entered, to describe an "area of interest for improvement of existing map data" (rather than "this is all burned").  A Tasking Manager could be used for this, but it needs such a polygon to identify the area of interest.

SteveA