[Talk-us] US Trunk road tagging

Minh Nguyen minh at nguyen.cincinnati.oh.us
Wed May 5 16:40:36 UTC 2021


Vào lúc 08:15 2021-05-05, Adam Franco đã viết:
> What the NHS maps highlight to me is that importance and connectivity do 
> *not* always match the highway sign. This has also been my experience as 
> a road user. Signed US routes may wind out of the way and have segments 
> that aren't so important and State routes may have portions that are 
> critical infrastructure, but never got included in the federal system 
> for some variety of historic reasons.
> 
> With respect to predominantly rural regions, I would be very happy if 
> highway=trunk was used for the segments of roads listed in NHS and 
> primary/secondary for US & State routes not in NHS. I'd also be happy 
> with mapper discretion to include/exclude a few edge cases not reflected 
> in NHS, but NHS seems like a very good baseline for regional 
> connectivity importance.

NHS is itself a functional road classification scheme, just not one that 
maps cleanly to other classification schemes, because it's very 
explicitly about funding. The law literally says which streets are part 
of the system.

Tagging the entire NHS, sans Interstates, as highway=trunk would result 
in basically every signposted local truck route becoming a trunk road. 
(As in MUTCD R14-1, not M4-4.) It wouldn't technically add much 
information to the map, because NE2 already went around tagging NHS=* on 
the whole system as it was back then.

It would be more reasonable to tag the Other Principal Arterials class 
as highway=trunk, but NHS is not the only scheme that includes an Other 
Principal Arterials class. That terminology comes from the FHWA's 
Highway Functional Classification system. It's unlikely that Congress 
and FHWA agree on every individual principal arterial classification, so 
there could still be persistent disagreements about highway=trunk.

> Lastly, using high connectivity-importance for trunk on down doesn't 
> preclude renderers from giving a visual indication of expressway-grade 
> physical condition. If we can get the highway=* hierarchy actually 
> meaning importance, then a later feature-request on current and future 
> renderers could be to give more prominence to ways tagged with 
> expressway=yes. For example, just giving the casing an extra pixel of 
> weight may be sufficient to highlight expressways, regardless of their 
> connectivity-importance.

Lately there's been a lot of discussion about an Americentric renderer 
on OSMUS Slack. [1] It's an idea OSMUS has been kicking around forever, 
but the enthusiasm around highway classification and shields may help 
make it a reality.

It would be feasible for an Americentric renderer to apply an orthogonal 
rendering style to expressway=yes, but I'm sober about the ability of 
such a renderer to resolve highway=trunk disputes on its own. After all, 
osm.org's openstreetmap-carto style is still the front door for mappers 
everywhere. Maybe better editor support could help tip the scales, but 
expressway=yes may be a hard sell to globally relevant editors and data 
consumers, being so peculiar to the U.S.

[1] https://wiki.openstreetmap.org/wiki/United_States/Map_style

-- 
minh at nguyen.cincinnati.oh.us




More information about the Talk-us mailing list