Re: UCR ISSUE-70: add a requirement for avoiding coordinate transformations? [SEC=UNCLASSIFIED]

Dear Frans,

> But if there are cases where publishing data with multiple CRSs is
> a hard rule that must be complied with it would be good to have them in
> the UCR document. And that would probably mean we have to formulate
> another requirement in the UCR doc, e.g. 'it must be possible to publish
> geometric data with multiple CRSs'.

I have a good example, that I used already many years ago during my 
presentation at the workshop that precedes the creation of this joint 
OGC-W3C working group.

Have a look at this dataset published on the French Government Open Data 
Portal: 
https://www.data.gouv.fr/fr/datasets/adresse-et-geolocalisation-des-etablissements-denseignement-du-premier-et-second-degres/. 
This dataset is provided in the CSV format. It lists all primary and 
secondary schools in France, with their geo-coordinates, name, address, 
state, etc. While there is a single file, the various values of the 
columns 'coordonnee_x' and 'coordonnee_y' MUST be interpreted 
differently. The explanatory note adds the following information:

"The coordinates X and Y are provided by IGN according to various CRS / 
projections being used:"
  - France métropolitaine:  RGF93 - Lambert 93
  - Guadeloupe:            WGS84 - UTM Nord fuseau 20
  - Martinique :                 WGS84 - UTM Nord fuseau 20
  - Guyane :                  RGFG95 - UTM Nord fuseau 22
  - Réunion :                  RGR92 - UTM Sud fuseau 40

This means that depending on which places you are on earth, you have to 
interpret geo-coordinates differently, otherwise, you will plot schools 
in the middle of the sea ...
Best regards.

   Raphaël

-- 
Raphaël Troncy
EURECOM, Campus SophiaTech
Data Science Department
450 route des Chappes, 06410 Biot, France.
e-mail: raphael.troncy@eurecom.fr & raphael.troncy@gmail.com
Tel: +33 (0)4 - 9300 8242
Fax: +33 (0)4 - 9000 8200
Web: http://www.eurecom.fr/~troncy/

Received on Monday, 5 September 2016 13:42:44 UTC