Back in March, changeset 179,365,728 (OSMCha) introduced a sawtooth pattern in the California–Oregon state line where Interstate 5 crosses it. I just spotted and reverted the change. At first I thought it was another novice user accidentally snapping things together in iD, but it turns out to have been intentional.
@Flap_Slimy_Outward explains that they settled on this sawtooth shape in order to make the boundary match what’s on the ground. Specifically, aerial imagery shows a change in pavement quality where repaving projects on either side of the border actually stopped.
This change is counterfactual. The actual state line is defined in the California Constitution:
The boundaries of the State are those stated in the Constitution of 1849 as modified pursuant to statute.
The relevant part of the boundary description in the 1849 Constitution reads as follows:
to the forty-second degree of north latitude; thence on the line of said forty-second degree of north latitude to the place of beginning
There are boundary markers along the state line and they don’t follow the legal description exactly, but there are no markers in this area because it goes in a straight line and was never disputed. There was no provision in 1849 for following pavement projects from this decade along a stretch of road that didn’t exist before 1966.
I have often trotted out pavement quality transitions as a counterpoint when mappers claim that administrative boundaries are inherently unverifiable and unmappable. Even so, they are at best a crude heuristic. Highway contractors can stop short or color outside the lines depending on what needs fixing.
For that matter, even the state border signs aren’t guaranteed to be precisely on the state line, because traffic engineers have to avoid unsafe practices like sign bunching. For example, there’s a California entrance sign along Nevada State Route 373, followed by Nevada exit sign along California State Route 127. This doesn’t represent a condominium or border dispute.
I look for pavement quality transitions when I’m cleaning up a TIGER-imported boundary that I don’t trust and don’t have a more reliable source on hand. If we do have more precise information about part of an administrative boundary, we should use that information. As much as we value verifiability, I don’t think this project guideline should cause us to spread misinformation.
Does the community agree that we should take these on-the-ground heuristics with a grain of salt? Do we think there’s enough of a risk of misunderstanding that we need to document these considerations on the wiki? Should we map the location where the pavement quality usually drops off, to dissuade mappers from fudging the boundaries? If so, how should it be tagged?
