[Talk-us] [Talk-us-newyork] Highway classification guidelines for New York State

[Brian M. Sperlongano]

On Sun, Sep 12, 2021 at 7:52 PM Eric Patrick <txemt1 at gmail.com> wrote:

> I have a thought regarding trunks in regards to functional classification,
>> and this is going to come from a different viewpoint. Why not make Primary
>> Arterials trunk roads, from a coding standpoint. They're coded lower than a
>> motorway, but higher than everything else. Any GPS system will use the
>> highest coded route, with the fewest penalties, they can get to between
>> points A and B. I understand that in doing this, a lot of downtown areas
>> will look like a sea of red, due to the density of the Primary Arterials
>> within those areas (cue Brian's groan about this). Functional
>> classification isn't going for looks, though, it's going for function. The
>> states have spent a lot of time and effort since FC was first introduced
>> about a decade ago.
>
>
I want, as I think we all do, to be able to defer decisions about what
classification value should be applied to a road based on a standard rule,
external authoritative source, or clearly defined methodology.  However, it
is important to point out that in order to achieve the goal of
machine-generated mapping, it requires that tags mean roughly the same
thing all over the world.  That does not mean that the definition of
trunk/primary/etc should be defined by the density of those roads in a
particular area.

It is instructive to compare the density of trunk roads (or any other
class) between two cities of roughly the same size in different countries
of similar infrastructure quality.  If these comparisons are significantly
different, that's a good indicator that we haven't chosen definitions for
highway classes that are good approximations for international equivalence.

I am currently in the process of updating highway classifications in
Providence, RI.  My point of comparison is the similar-sized city of
Ipswich, England.  Ipswich has four trunk roads leading into the city, 5
primary routes leading in and out of the city center, and both an inner
primary and outer trunk loop.

Providence meanwhile, has five motorways (one of which becomes trunk after
leaving the city), one motorway loop, seven primary routes leading in and
out of the city center, and 2-3 other primary roads that effectively
function as inner primary loops.

When I look across these comparisons, I find that this density is
comparable, and tells me that I've come up with a ruleset that is roughly
comparable with other cities of a similar size internationally.

If I were to upgrade all of Providence's FHWA principal arterials to trunk,
this would now leave me with 12 trunk or motorways routes in and out of the
city, compared to Ipswich's four.  This tells me that principal arterials
are not significant enough to qualify as trunk roads based on
international equivalence.

Additionally, it should be noted that in British cities (my point of
comparison), it is common that primary roads converge in the city center,
for example as seen at:
https://www.openstreetmap.org/#map=16/52.0554/1.1655

This convergence is similar to the behavior of the FHWA's principal
arterials in US cities, including Providence.  However, you never see trunk
roads laid out in a "ball of yarn" configuration in city centers.

All of this leads me to believe that the correct mapping, at least in Rhode
Island, is for FHWA's principal arterials to be tagged at least primary.
The only question remaining is which of those "at least primary" ought to
be trunk.

For my state, there were only a few that made sense.  The first was US-6
between Providence and Hartford.  This was the originally planned alignment
for I-84 before it was cancelled due to neighborhood opposition.  As this
key long distance route between the two cities, this is a clear case for
trunk.  The second is a route starting from the end of the RI-24 expressway
in Portsmouth, running down RI-114 through Aquidneck Island, and meeting
back up with the RI-138 expressway on the other side.  This was also the
route of a planned I-895 interstate that was cancelled before being built,
but resulted in upgraded roads for most of the alignment, and represented a
clear loop through Providence, Fall River, and Newport.  The last trunk
road is US-1, which is the principal link between Newport and New London
through South County and is built to expressway quality for most of its
length, and thus anything lower than trunk wouldn't really be appropriate.

This was my thinking for RI; it made sense here, was consistent with what
my other northeast neighbors have been doing, and almost perfectly lines up
with cross-border roads into Massachusetts (the CT border is mostly woods).
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-us/attachments/20210913/0010852b/attachment.htm>