Hey folks,
I discovered that >1000 highway segments in Massachusetts are tagged as being 91.5 (implicitly meter) wide, that is approximately 300 feet; see this overpass query. I’m finding it hard to believe these are correct, but I don’t want to go and make changes of this magnitude outside my “home turf”. Does anyone (more) local want to check what’s up?
Thanks
You should be able to check the width with aerial imagery, especially in JOSM with the measurement plugin. I fixed something similar with bike paths marked with width=10, which is crazy wide.
I did a couple of spot checks and it seems like 91.5 isn’t even feet. The couple of roads I checked were about 50 and 35 ft across.
It looks great with the Lane style 
:
I don’t think the linked overpass query was what you meant to link, since that’s for “maxheight”=“13.91 feet”. But a lot of road widths in Massachusetts are from a 2007 import from MassGIS/MassDOT data. It wouldn’t surprise me if there were some wrong things in there.
I’d say that if the width data is obviously wrong, go ahead and fix it, or at least remove the tag. And I am “local”, for whatever that’s worth, though I haven’t personally done anything with editing or consuming width data.
Oops, fixed the link.
Here’s an example from Route 116. It appears all these width=91.5 tags came from the 2007 MassGIS import (the one that happened in Massachusetts instead of the TIGER import).
This attribute mapping says a field called RightOfWayWidth, representing the “ Right-of-way width in feet”, was translated to width=*. That’s unfortunate, since the same dataset came with other fields like SurfaceWidth, MedianWidth, and LeftSidewalkWidth that we could’ve used later on to reconstruct more detailed sidewalk and carriageway data. Instead, this right of way width would only be useful for mapping a landuse=highway area around the roadway, but it probably isn’t just a matter of buffering the roadway by a constant width.
I’m not sure there’s much we can do with these tags other than to drop them wherever we can tell that they came from the MassGIS import. We could also find the width information from the current MassGIS dataset and apply the width attributes more accurately (and with the proper unit symbol or conversion).
