[Tagging] Planning route in the shade during hikes either in urban areas or forests

[Bert -Araali- Van Opstal]

This fits in the discussion we recently had about other highway related 
hazards 
(https://lists.openstreetmap.org/pipermail/tagging/2021-April/061315.html 
for gravel smoothness etc.. but applicable for other hazards).

The presence of shade or not is a possible mitigation for several 
hazards.  I think about 2 very distinct hazards: heat and sunlight. Heat 
is an issue for most people, sunlight not necessary, as people with 
darker skin are not necessarily bothered by excessive sunlight. 
Excessive sunlight, thus the absence or presence of shade can also be a 
hazard in cold climates during winter, in regard to snow blindness.
Heat can extend to the periods when a surface is shaded, think about an 
asphalt road which has overheated and keeps heating the direct air above 
long time after the period it was exposed to direct sunlight. A serious 
hazard, even for pedestrians or hikers with darker skin or drivers 
without airco.

So instead of inventing some kind of a shade tag I would go for 
hazard=heat and hazard=sunlight. Add sufficient data to allow external 
programs to do an appropriate risk analysis so a user can make a 
decision based on the residual risk and his use case.
The cases where shade matters are so variable and of importance for 
different users, not just for hiking, it is unfeasible to find an 
appropriate tagging scheme that fits all or let all OSM users in all 
places all over the world come to the same conclusions.

Further defining or quantifying the hazards with severity and frequency 
(like seasonal tag, periods of a day) and eventually a threshold (like 
average air temperature) can provide just sufficient data to allow 
decision making by a user and make the tag viable for a long period, 
having climate change in mind.
The level of detailing, like splitting a highway in ridiculous small 
detailed sections is also not feasible as already identified in this 
discussion.  It comes to the same principle as f.i. we use for 
potholes.  We don't create a highway section for every pothole but 
incorporate it in a severity classification or one could add a certain 
percentage.

Hazard can be applied on ways but also more detailed on areas, thus you 
can use it for larger areas with varying shade or heat exposure due to 
absence of shade. Specifying hazards like heat and sunlight will also 
make this data interesting for use with other applications like sun 
blinding in evening periods, planning of buildings, buffer zones, sun 
screens, camping spots, etc... .

Greetings,

Bert Araali

On 07/07/2021 02:17, bkil wrote:
>> - in case of placing lamps the goal is of
>> making entire stretch of easy lit,
>> something else happens in case of some
>> weird plans or deep incompetence
>>
> Imagine that endangered trees can be protected in many places, so
> trimming them is not allowed and they can thus grow unhindered. On
> installation, the lamps provided enough light everywhere, but in a few
> years, even the growth of a few branches that took a wrong turn could
> block a significant portion of light.
>
>> - lamps put light in one direction, not
>> subject of rotation like most shades are
>>
> You are again not focusing on the problem at hand and what the
> customer (map user) wants and what an MVP means.
>
>> Mapping shade=* would require unusually
>> complex conditional syntax and ridiculous way splitting.
>>
> It would not make sense to add it to such a level of detail. This was
> only considered as a time saver approximation. If you want something
> more precise, you should definitely go for full 3D.
>
> An approximation and heuristics are only that: they help a hiking
> router make you the decision where you should go if two roughly
> similar lengths of paths both lead to the same destination, but one
> path has shade=no, while the other one has shade=limited.
>
> Also, if you are unsure, just don't map it. If it is clear that a
> given path is never shade in the summer months around noon, why not
> put something like shade=no on it? If trees completely cover a given
> stretch of footway and provide shade during summer months for many
> hours around noon, why not put something like shade=yes on it?
>
> What if several possible paths/sidewalks/cycleways exist in the same
> area and long stretches of them can be easily compared in some cases
> and in case of one being clearly inferior, mark it with shade=limited?
> In many company cultures (and many open and volunteer communities)
> many code of conduct contain clauses to the effect of "trust your
> peers that they can make educated decisions for themselves as the
> mature adults they are". We are intelligent humans. Please trust
> mappers that they can make usable decisions in fuzzy situations - this
> is what differentiates us from robots. The map is made by humans for
> humans.
>
>> For what? In case of shade from buildings
>> where shade depends on time of year
>> and day I am curious how you would even
>> tag using it with shade.
>>
> Not sure where you live, but in this region, heat is only problematic
> in very specific months and very specific hours of a given day.
>
>> In my area deciduous trees, changing
>> sun location etc result in shade changing
>> massively during year.
>>
> In my area, in the few months of the warmest part of the summer where
> shading matters at all in real life, the sun shines in a well defined
> angle and deciduous trees have their foliage on. I would be surprised
> if you lived on a planet where this goes differently.
>
>> - width of the sidewalk/footway
>> Doable with measuring tape, though not
>> sure why it would be needed
>>
> In order to see where people can walk exactly, you either need to map
> the sidewalk as an area, or it needs to have its width=* added. In
> case a sidewalk is added as a linear feature, it makes a big
> difference whether the sidewalk is 1m wide or if it is 5m wide to see
> exactly how far from a given fence or tree a router can make one walk.
> Surely, I wasn't talking about marking whether a sidewalk is 1.10m or
> 1.15m.
>
>> Tree size should be estimatable by measuring
>> it's trunk diameter and species.
>>
> I've surveyed thousands of trees around the area, so I know exactly
> how much work it is. It is unfeasible even for me, let alone laymen
> mappers to do this. The godfather of the iD editor called me names
> some years ago when I suggested that it should be possible to add the
> taxon or at least the genus of a tree, reasoning along the way that
> mappers aren't smart enough to tell apart an apple tree from figs (and
> if some of them can, they should be denied the possibility to mark it
> anyway just because).
>
> Also, I'm not sure if you ever had a garden or been to one, but in
> towns or gardens where they trim trees (basically in many urban
> areas), the trunk diameter keeps growing annually, but the height and
> the number of branches and the volume of foliage stays approximately
> the same every year (along with its cast shadow). I'd say at least
> half of urban trees are trimmed as a method of precaution. My
> botanical observation tells me that weather, the genetic makeup of the
> given individual, its health and life record can cause a lot of
> variance in crown diameter and foliage density (along with accidents
> in the past and trimming patterns), so I think the approximation based
> on the trunk would only be useful in the wild and in a statistical
> (population) context, not for such micromapping.
>
> It would surely be better than nothing, but I think compared to how
> much it would cost, just estimating the height and crown diameter via
> various methods would be much more useful for this question.
>
>> Mapping approximate height of building
>> can be done by surveying.
>> See StreetComplete
>>
> How many towns have you yourself drawn on an empty map, encircling
> every structure and then walking through every street and adding the
> height and shape of every building and fence? We regularly organize
> parties over here where many do this, I have drawn towns from the
> ground up and many 3D stuff, so I know exactly how much work it is (a
> lot, hence I prefer not to do that any more). I see you have lots of
> edits (mostly single changes, though, so I can't feasible review
> them), but I don't know how much 3D you have done.
>
> What you must understand is that unless you come over here and fill in
> the map for us, it will never contain the needed information to a high
> enough coverage that would make a service usable based on that. I
> would guess that the situation is the same for most parts of the world
> (especially where people walk and finding shade matters at all).
>
> Hence you are basically saying that "if you can't do it 100%
> precisely, don't do it!" or "you're either doing overtime and do it my
> way or it's the highway!". This is toxic to the community.
> Volunteering does not work like that. The world does not work this
> way. People have needs, there are many viable and useful solutions to
> their problems, and people will definitely solve their problems one
> way or another if they need to be solved. It simply does not cut it
> that naysayers say that they should not solve their problems.
>
>> Still easier than mapping whether it shades
>> road depending on time of day and year.
>>> But how common is it that one would have all required data?
>> More likely than mapping shade= in way
>> that would actually work given that
>> shade changes during day
>> Single 10 storey building would require
>> footways split every 10cm around it
>> (towards East, West, North from it)
>> or such shade=* tags would be useless.
>> That sounds like horrific idea that should
>> be considered as ridiculous, unusuable,
>> not worth encouraging and bad idea.
>>
> Yes, I agree that your recommendation is ridiculous and unusable. Lots
> of straw man arguments here - I'd prefer if we wouldn't do that.
>
> We have none of this data available in the whole country as I have
> already stated, and I have shared a possible way of approximate
> surveying that takes a similar amount of time as adding interpolated
> house number ranges - stopping for a minute at each street corner.
> There could exist other approximations and many others ways to tag
> such approximation as well.
>
> It was suggested on our local list that even an average percentage or
> expected probability of shade could be marked per longer stretch of
> path. I think that would be more difficult to verify exactly, but this
> also shows that we could come up with lots of useful methods for
> approximation if we really put some effort and a positive attitude in
> it together.
>
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20210707/61c4fa34/attachment.htm>