[Gvsig_english] No problems reprojecting from GK to UTM usingBETA2007

[Juan Lucas Dominguez Rubio]

Hello, Heinz:
 
I have done the test you describe and indeed the points in the CSV file are not a subset of the GSB grid, but the error is almost zero. If we are honest, we'll have to look for an explanation, no? Here is my guess: the CSV file was created using the GSB grid. I misunderstood the goal of that verification file (perhaps the Australian file too?). They are not used to check if the grid works fine. They are used to check that your GIS software knows how to deal with the grid. This is my second hypothesis :-)
 
As for the transformation with code 1780 available in gvSIG, I have used it and get the expected errors (< 5m):
 
3.43 m
2.85 m
0.15 m
etc.
 
Perhaps you have not chosen the right options? I have used gvSIG 1.1.2 because I am on Windows Vista.
 
 
Regards,
 
Juan Lucas Domínguez Rubio
---
Prodevelop SL, Valencia (España)
Tlf.: 96.351.06.12 -- Fax: 96.351.09.68
http://www.prodevelop.es <http://www.prodevelop.es/> 
---

________________________________

> De: gvsig_internacional-bounces at listserv.gva.es en nombre de Heinz Mundhenk
> Enviado el: jue 10/12/2009 18:07
> Para: gvsig_internacional at listserv.gva.es
> Asunto: Re: [Gvsig_english] No problems reprojecting from GK to UTM usingBETA2007
> 


> Dear list,
> 
> here is another working grid-based reprojektion for Germany.
> 
> Reprojecting from GK to ETRS89/UTM using grid BETA2007
> 
> In the next years coordinates from different old SRS in
> Germany will change to the new projection ETRS89/UTM
> with EPSG:25832 (Westgermany) and 25833 (Eastgermany).
> 
> Most old SRS in former Westgermany are based on the Gauss-Krueger-
> Projektion (GK, EPSG:31466, 31467 and 31468).
> 
> The transition from GK to UTM can be done with the integrated
> "transformation on the fly" in gvSIG for errors lower than 5 meters.
> 
> If you use the grid "BETA2007.gsb", the error is lower than 1 meter.
> 
> Here you can get the grid and coordinates for a tests :
> 
>   www.crs-geo.eu/BeTA2007
> 
>   files BETA2007.gsb (85 kByte)
>         BETA2007testdaten.csv (12 kByte)
> 
> On this website you find lots of additional informations.
> 
> For a test make a dfb-file from the csv-file by selecting all records
>   with Quell-Bezugssystem (source) = DE_DHDN_3GK3
>   and  Ziel-Bezugssytem (target) = ETRS89_UTM32.
> 
> From this dbf-table create the shape (with testpoints)
>   GK3_31467.shp using Quell-X and Quell-Y for X- and Y-coodinates.
> 
> How to integrate the *.gsb - file in gvSIG 1.9 ?
> 
> Make a view with SRS-property EPSG:25832 (ETRS89/UTM32) and add the
> testpoint-shape with the SRS-Property EPSG:31467, using grid-transformation
> (with german-program-surface/skin : NTv2 Format Raster- Datei).
> In the next window click the button Import and select the
> previous downloaded grid BETA2007.gsb.
> 
> Then Export the reprojected shape under the name UTM32_25832.shp and add it to
> the view. Now you can add the fields X and Y and calculate x() and y().
> Then calculate the difference between Ziel-X and X / Ziel-Y and y.
> The Errors are lower 0.01 meters ! The testpoints are no grid-points !
> 
> Observations
> 
> 1. The reprojektion with the grid does not work in the reverse order
>    from UTM to GK.
> 
> 2. Reprojection from GK to UTM with gvSIG using the same testpoint-shape
>    and the integrated EPSG-transformations 1780 or 15948 result with
>    errors between 100 and 200 meters !  Why ?
> 
> Greetings
> 
> Heinz Mundhenk
> hmundhenk at gmx.de
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://listserv.gva.es/pipermail/gvsig_internacional/attachments/20091211/c70719cb/attachment.htm