[GRASS-dev] Datum-addition

Dear list,

I have found several datum definition which are currently not supported
by GRASS yet. However, I would like to implement them into GRASS.

for example:
DATUM|Ellipsoid Code|Region of Use|DX| DY |DZ |EX|EY|Ez
India Bangladesh|EA |Bangladesh |282|726|254|10|8 |12

This values are found here[1].

I found out that this needs to go into datum.table in lib/gis/.But the
datums there only contain dx,dy,dz, not the ex,ey,ez values.

What I am not sure about is the addition into datumtransform.table.
Is it nessessary to provide more information there? How to transform
into PROJ4-style (towgs84=)?

Thanks for your help.

  Stephan

[1]
http://earth-info.nga.mil/GandG/coordsys/onlinedatum/CountryAsiaTable.html#INDI

Hello Stephan

On Mon, 15 May 2006, Stephan Holl wrote:

Dear list,

I have found several datum definition which are currently not supported
by GRASS yet. However, I would like to implement them into GRASS.

for example:
DATUM|Ellipsoid Code|Region of Use|DX| DY |DZ |EX|EY|Ez
India Bangladesh|EA |Bangladesh |282|726|254|10|8 |12

This values are found here[1].

I found out that this needs to go into datum.table in lib/gis/.But the
datums there only contain dx,dy,dz, not the ex,ey,ez values.

What I am not sure about is the addition into datumtransform.table.

See the comment at the top of datum.table:
# If 99999 or above is given for all 3 parameters, then the parameters
# listed in this file will not be offered to the user and there
# must be at least one set of parameters for the corrseponding datum
# in the datumtransform.table file

If you only have a 7-parameter transform, then put 99999 for each parameter in datum.table and make sure and put the transformation (in the format towgs84=dx,dy,dz,rx,ry,m) in datumtransform.table

Is it nessessary to provide more information there? How to transform
into PROJ4-style (towgs84=)?

See above. Put in as much information as you can about the accuracy, area covered, etc. But the paramters you have shown don't seem to be accompanied by a scale factor m, so it looks supicious! Are you sure they work? Do you have a test point? It's important to be very sure the parameters are correct (preferably you have used them) before adding them to CVS as people will use them and assume they are correct because they are in GRASS!

Best regards

Paul