Re: [OSM-talk-ie] Summits imported from mountainviews.ie

[Donal Hunt]

I'm a member of mountainviews.ie so familiar with their maps and
information. I do know they have members of the community that have access
to / borrowed high-quality surveying equipment which has resulted in the
comprehensive dataset they have.

I would not be surprised if there have been updates in the past 10 years
regarding elevation, location, etc.

Donal

On Sun, 1 Dec 2019, 13:38 moltonel 3x Combo,  wrote:

> On 29/11/2019, Donal Hunt  wrote:
> > The latest major revision of WGS 84 is also referred to as "Earth
> > Gravitational Model 1996" (EGM96), first published in 1996, with
> revisions
> > as recent as 2004. This model has the same reference ellipsoid as WGS 84,
> > but has a higher-fidelity geoid (roughly 100 km resolution versus 200 km
> > for the original WGS 84).
>
> Hopefully the 2004 revision is close enough to the 1996 one, as it
> seems that osm is defaulting to 1996.
>
> > The tl;dr is that you need to reference which model your peak height is
> > measured against. EGM96 is preferred (I'm assuming) because it has higher
> > definition and more refined model.
>
> AIUI from the wiki and taginfo, in an OSM context, the naked "ele" tag
> refers to EGM96, and there are subtags like "ele:wgs84" or even
> "ele:local" for "whatever's printed on the local signpost".
>
> I've figured out the overpass query I had in mind:
> https://overpass-turbo.eu/s/OCb but it finds basically every peak we
> have with an elevation.
>
> See also
> https://help.openstreetmap.org/questions/15397/elevation-of-irish-mountains-why-the-discrepancy-osmwikipedia
> which complains the the issue ireland-wide, and notes that the osm
> data in other countries is fine.
>
> All the mountainviews imports basically have
> ele:local/ele_local/ele:tm75 (3 keys with first value) and
> ele/ele:wgs84 (two tags with second value). I couldn't find a
> definition of the "tm75" coordinate system that we got from the
> mountainview export, I'm guessing this is an irish-specific
> coordinates system, and I have doubts that it is equivalent to the
> egm96 that we would like.
>
> At this point, I have a few questions:
> * Are the "tm75" values usable for the naked "ele" tag ?
> * If not, can we convert tm75 to egm96 ?
> * As a bonus, I think we should remove "ele_local" where we already
> have "ele:local"
> * Is it ok to edit 790 as a single changeset (once we agreed with the
> changes) or do we want to check each individually ?
>
> * Second bonus (probably for a second stage): is there's any point in
> keeping the iemv:* tags ? Might have been a requirement for the import
> ? It was done in 2009, maybe it's time to see if they have updated
> data.
>
>
>
> Cheers.
>
> ___
> Talk-ie mailing list
> Talk-ie@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-ie
>
___
Talk-ie mailing list
Talk-ie@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-ie

Re: [OSM-talk-ie] Summits imported from mountainviews.ie

[moltonel 3x Combo]

On 29/11/2019, Donal Hunt  wrote:
> The latest major revision of WGS 84 is also referred to as "Earth
> Gravitational Model 1996" (EGM96), first published in 1996, with revisions
> as recent as 2004. This model has the same reference ellipsoid as WGS 84,
> but has a higher-fidelity geoid (roughly 100 km resolution versus 200 km
> for the original WGS 84).

Hopefully the 2004 revision is close enough to the 1996 one, as it
seems that osm is defaulting to 1996.

> The tl;dr is that you need to reference which model your peak height is
> measured against. EGM96 is preferred (I'm assuming) because it has higher
> definition and more refined model.

AIUI from the wiki and taginfo, in an OSM context, the naked "ele" tag
refers to EGM96, and there are subtags like "ele:wgs84" or even
"ele:local" for "whatever's printed on the local signpost".

I've figured out the overpass query I had in mind:
https://overpass-turbo.eu/s/OCb but it finds basically every peak we
have with an elevation.

See also 
https://help.openstreetmap.org/questions/15397/elevation-of-irish-mountains-why-the-discrepancy-osmwikipedia
which complains the the issue ireland-wide, and notes that the osm
data in other countries is fine.

All the mountainviews imports basically have
ele:local/ele_local/ele:tm75 (3 keys with first value) and
ele/ele:wgs84 (two tags with second value). I couldn't find a
definition of the "tm75" coordinate system that we got from the
mountainview export, I'm guessing this is an irish-specific
coordinates system, and I have doubts that it is equivalent to the
egm96 that we would like.

At this point, I have a few questions:
* Are the "tm75" values usable for the naked "ele" tag ?
* If not, can we convert tm75 to egm96 ?
* As a bonus, I think we should remove "ele_local" where we already
have "ele:local"
* Is it ok to edit 790 as a single changeset (once we agreed with the
changes) or do we want to check each individually ?

* Second bonus (probably for a second stage): is there's any point in
keeping the iemv:* tags ? Might have been a requirement for the import
? It was done in 2009, maybe it's time to see if they have updated
data.



Cheers.

___
Talk-ie mailing list
Talk-ie@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-ie

Re: [OSM-talk-ie] Summits imported from mountainviews.ie

[Donal Hunt]

And for those (like me) who are not up to speed on how peaks are measured…


The World Geodetic System (WGS) is a standard for use in cartography
, geodesy
, and satellite navigation
 including GPS
.

WGS84 (elipsoid) and EGM96 (geoid) are two different models, with different
applications.

The latest major revision of WGS 84 is also referred to as "Earth
Gravitational Model 1996" (EGM96), first published in 1996, with revisions
as recent as 2004. This model has the same reference ellipsoid as WGS 84,
but has a higher-fidelity geoid (roughly 100 km resolution versus 200 km
for the original WGS 84).


When you see a height in a map, it is usually a height over the geoid
(orthometric height), but when you get one from a GPS device, it is usually
a height over the elipsoid. There are online tools to transform between
them as needed, using what it is called geoid undulation. E.g.
geoid-height-calculator


The tl;dr is that you need to reference which model your peak height is
measured against. EGM96 is preferred (I'm assuming) because it has higher
definition and more refined model.


Donal

On Fri, 29 Nov 2019, 19:00 moltonel 3x Combo,  wrote:

> Hi,
>
> I've just fixed Lugnaquilla's "ele" tag to use the EGM96 system, as
> opposed to the WGS84 one:
> https://www.openstreetmap.org/changeset/77737107
> https://wiki.openstreetmap.org/wiki/Key:ele?uselang=en
>
> That one was reported by an OSM note, but I wouldn't be surprised if
> most of the natural=peak have the same issue. I think it's possible to
> find those using an overpass query that would return all natural=peak
> where ele == ele::egs84, but I couldn't figure that query out straight
> away and there may be a lot of stuff to carefully fix, so I'm turning
> to the mailing list.
>
> Might get a solution tomorrow during the Kilkenny meetup, but I
> figured I might as well ask here so I don't forget about it.
>
>
> --
> Vincent de Phily
>
> ___
> Talk-ie mailing list
> Talk-ie@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-ie
>
___
Talk-ie mailing list
Talk-ie@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-ie