Coordinates transformation hors zone RGF93

Les bases GNSS hors de la zone RGF93 ont des rapports de positionnement en ITRF de la part du service en ligne IGN ou NRCAN.
Il est important de mettre en place des outils afin de pouvoir convertir ces coordonnées dans un système de référence du pays en question, si il en a un.
C’est le cas pour la Hongrie et la Roumanie qui voient leurs nombres de bases installées grandir à vue d’oeil. En attendant d’avoir un outil développé propre à Centipede ( prj7 + interface web ?), il est importan,t de regarder si il n’existe pas des outils déjà existant et les tester.

GNSS bases outside the RGF93 zone have positioning reports in ITRF from the IGN or NRCAN online service.
It is important to set up tools to convert these coordinates into the reference system of the country in question, if it has one.
This is the case for Hungary and Romania, where the number of installed bases is growing by leaps and bounds. Until we have a tool developed specifically for Centipede (prj7 + web interface?), it’s important to see if there are any existing tools and test them.

GNSS bases outside the RGF93 zone have positioning reports in ITRF from the IGN or NRCAN online service.
It is important to set up tools to convert these coordinates into the reference system of the country in question, if it has one.

Peut-etre une piste ici:

Nous pouvons prendre en exemple le cas de la base ROMS1 en Roumanie avec son rapport de calcul:

One example is the ROMS1 gnss base in Romania and its calculation report:

“ROMPOS is a Romanian position determination system that ensures precise positioning in the European Terrestrial Reference System 1989 (ETRS89).”

La liste des CRS en usage est définie sur :

ITRF20 (2023.4)
Latitude (+n) 46° 31’ 39.12109"
Longitude (+e) 24° 32’ 32.59255"
Ell. Height 372.753 m

convert LLH degré to decimal : Latitude Longitude Data - Latitude Longitude Converter
lat 46.5275336361
lon 24.5423868194

convert LLH to ECEF :
3999029.62 1826036.87 4606057.2
SySense Inc. - ECEF-LLA Converter
3999029.62 1826036.87 4606057.20

il va falloir avoir les données de vélocité…

en renseignant 0 0 0 en vélocité

3999029.6693 1826036.9242 4606057.1589

46,5275329 24,5423872 372,77

dif en mètres 0.086

On est d’accord que ça ne fonctionne pas, n’est-ce pas ?

Pour les pays ayant défini leur CRS comme ETRS89-compatible (cas de la France mais aussi de pas mal d’autres pays européens), pas de problème.
La transformation a utiliser est en général ITRF2020@now vers ETRF2000@now, sans introduire de vitesse pour les stations. Le postulat est que la vitesse résiduelle en ETRF2000 est quasi nulle, et peut être négligée.
Pour le cas de la Hongrie : une transformation existe dans proj entre ETRF2000 et HD72. Elle est proche (mais pas tout à fait identique) à celle définie dans la page relative à la Hongrie dans sur le site
Dans le choix de la transformation, il faut être vigilant sur le fait que la transformation ne soit pas qualifiée de “ballpark” (voir le glossaire sur le site proj).

There’s a converter available that can transform between EOV and ETRS89: EHT 2.0 - API dokumentáció

The measures between IGN and the FÖMI VRS was 0.99XX meter.

POSITION ITRF2014 EPOQUE 2023.350 (08/05/23) :
DORI 99999S999
X: 4092387.2933 Y: 1417870.3962 Z: 4666662.5899

DORI 99999S999
LONGITUDE 19.109428581 ° LATITUDE 47.327968084 ° HELL 146.2775

With (This is the official one)

All these measurements were done on May 9th around 8pm.

So ~5cm error in the “ionoszféra” could be there.

A few weeks later we measured it back and it was still 1-2cm aligned.

Measured from KAR1 base:

Measured from TORS base: (This was the first I tried to align to the national grid I think it’s 1,5-2cm off as we only measured for 1 minute not 5… of course it’s the one we operate :smiley: )

(New user can’t put more than 1 picture to a post sorry about spamming…)

I wrote a proj pipeline in order to reproduce your calculation. The schema is basically ITRF2014@now->ETRF2000@now-> HD72.
I noticed that my results in ETRF2000@now are close to those in your screenshot.

1/ what is the CRS used in Hungary ? HD72 (as mentionned website) or ETRF2000 ?
2/ I do not understand why there are différents columns (“current”, “minimum”, “maximum”…)

Was your calculation ~2cm off by any chance?

We are using U-Center from Ublox. And I think the Hungarian VRS system is not super precise…
That min-max distance when I check it are within 2cm so it totally could be because of the GNSS receiver that’s only 1cm precise.
I basically start the process.
Wait for an RTK Fix to appear.

Then reset the database, wait 5+ minutes.
If min/max are within 1-2 centimeter I consider the measurement good.

This is KAPU where the VRS base station was within ~5km or so:

(Sometimes F9P was on 10Hz sometimes 1Hz so count might not reflect seconds.)

The NRCAN measurement for this one was:
ITRF20 (2023.3) 47° 35’ 30.52465" 17° 1’ 34.17059" 180.132 m

I’ve asked for help about those schemas. I believe ETRF2000 should be good.

Here are my différences :

For FOMI : dE = 0.023 m dN = -0.009 m dh = -0.027 m
for KAPU : dE = 0.044 m dN = 0.009 m dh = 0.026 m

I think that’s really close.

So the question is:
Can we execute the transformation ourselves?

If I have the steps I’ll ask in the facebook group to have a few folks to double check. That way we’d be able to check this on at least 5-10 bases.

Yes, if it is a difference between VRS+officiel transfomation vs static+PPP+Proj transformation, that looks good !

I would suggest to use NRCAN (with ITRF, not NAD83 for results) out of France. If you look at IGN reports, you will see that usually the new base is only calculated versus one or two know bases, hundreds of kilometres away.

Overall, I think this is a good suggestion to convert from ITRF to ETRF for European countries not suffering from internal deformation (Spain, Italy, Greece and Turkey).

So, NRCAN(ITRF) calculation. ITRF20@measurments_date to ETRF2000@measurments_date conversion (with zero base motion).

okay so I have a fresh rinex :slight_smile: I’d like to write down my steps and you tell me if it’s good or not.

I ran NRCAN: Precise Point Positioning
Then I click the summary, which starts like:
The estimated coordinates ITRF20 2023-06-06
This gives me the EPOCH that I can use here:
EUREF Permanent GNSS Network

I also need the X,Y,Z (cartesian coordinates) that’s in the same file as:

From this I do:
SBG1 4141140.548 1388894.707 4632493.710 0 0 0

And at the end I enter it and do the calculation:

which results in:
SBG1 ETRF2000 2023.43 4141141.1783 1388894.1532 4632493.3083 0.0169 -0.0176 -0.0105

Then I just need to do a cartesian to WGS84?
For which I tried to use an online converter: Free online coordinates converter

And ended up with:
46.87629736 18.54089816 161.938

Is this the right logic?