How can I setup a new datum for a new coordinate system

Hi

I need to define a new coordinate system to add to the Geographic Coordinate systems.

There is no problem in defining the projection, which just another Transvers Mercator, the problem is that I need to define a new Datum

How do I do that ?. As far as I can see I can only chose among the allready defined datums. Are there a possibility for adding a new datum ??

Why do I need a new Datum! Because the Fixed Link across the Fehmarn Belt between Denmark and Germany has decided to use ITRF2005 as Datum for it project coordinate system. And since ETRS89 differs around 20 centimeters from ITRF2005, we need to have a simple datum transformation between these two datums allthough they both are 'WGS84' datums

Thanks

Erik Wirring

Parents
  • Erik,

    The functionality for adding new datums will be added in the next scheduled release of MicroStation (MicroStation v8i SS3) which is scheduled for sometime in 2011.

    I can also add this datum as part of the pre-defined datums for the release. Can you provide the necessary documentation (hopefully an official government link) for this datum definition please?

    Alain Robert

  • Erik wrote:

    =============================================================================

    Hi Alain

    I don't have the final values at this time. But it is going to be tricky if you wants to add the ITRF2005 to the official Datum list.

    Let me try to explain. ITRF2005 is timeframe for the ITRS system -that is the geodectic use of 'WGS84' so it is the mean values for WGS84 in the year 2000 thru 2005.

    Basically ETRS89 is a fixed version of WGS84 in Europe at the year 1989. The problem is that ETRS89 in Denmark and Germany is not excatly the same. So to build the fixed Link over Fehmarn Belt someone who thought they were wise, chose to use ITRF2005 instead. I believe it create more errors  than chosing one of the two ETRS89 definitions.

    Since the 'basic datum' of the transformation system, and here I am guessing, is Wgs84 in all it different implementation around the world (NAD83, NZ2000, ETRS89 etc.) It will be highly confusing to implement a datum transformation from ETRS89-danish version to ITRF2005 and one from ETRS89-german version too !! In the end both ETRS89 and ITRF2005 is 'Wgs84'

    I hoped we were able to define our 'own' datum 'ITRF2005' with its datumtransformation (just a xyz move in the ECEF coordinates) and keeping it locally just for this project.

    Best Regards

    =============================================================================

    Hi Erik,

    I have given this a little thought and it seems to me that what you need is some sort of transformation to be applied once. This does not fall under the responsibility of the Geographic Coordinate System engine. This engine can only support transformation that follow clearly identifiable paths with consistency to WGS84. Since two different transformations relate to the same datum then it cannot be clearly be defined in the GCS engine.

    One possibility to apply the required transformation is to “import” the German data into Demark using a transformation to be applied once. This transformation could be a simple rotation/translation/rotation that approximate the relation from German ITRF2005 to Denmark ETRS89 if you can derive the transformation parameters with sufficient accuracy for your application.

    The other possibility would be as you appear to suggest, to define a “transformation datum” “GermanITRF2005ToDenmarkETRS89DatumTRF” or some other shorter name that would serve to define GCS meant for transformation purposes only. You would define all GCS needed upon this datum then perform a “Select Geographic Coordinate System…” with option “Correcting the Geographic Coordinate System – do not reproject the data” using these transformation GCS definition immediately followed by a “Select Geographic Coordinate System …” with option “Reproject the data to the new Geographic Coordinate System” upon the Danish GCS.

    Defining a new [transformation] datum will require MS v8i SS3 and defining new GCS will require Bentley Map (user-defined GCS can be used on any MS v8i site but definition requires a Bentley Map).

    I will be posting a stripped version of the present mail and your mail below for the benefit of the Be Community.

    Alain

Reply
  • Erik wrote:

    =============================================================================

    Hi Alain

    I don't have the final values at this time. But it is going to be tricky if you wants to add the ITRF2005 to the official Datum list.

    Let me try to explain. ITRF2005 is timeframe for the ITRS system -that is the geodectic use of 'WGS84' so it is the mean values for WGS84 in the year 2000 thru 2005.

    Basically ETRS89 is a fixed version of WGS84 in Europe at the year 1989. The problem is that ETRS89 in Denmark and Germany is not excatly the same. So to build the fixed Link over Fehmarn Belt someone who thought they were wise, chose to use ITRF2005 instead. I believe it create more errors  than chosing one of the two ETRS89 definitions.

    Since the 'basic datum' of the transformation system, and here I am guessing, is Wgs84 in all it different implementation around the world (NAD83, NZ2000, ETRS89 etc.) It will be highly confusing to implement a datum transformation from ETRS89-danish version to ITRF2005 and one from ETRS89-german version too !! In the end both ETRS89 and ITRF2005 is 'Wgs84'

    I hoped we were able to define our 'own' datum 'ITRF2005' with its datumtransformation (just a xyz move in the ECEF coordinates) and keeping it locally just for this project.

    Best Regards

    =============================================================================

    Hi Erik,

    I have given this a little thought and it seems to me that what you need is some sort of transformation to be applied once. This does not fall under the responsibility of the Geographic Coordinate System engine. This engine can only support transformation that follow clearly identifiable paths with consistency to WGS84. Since two different transformations relate to the same datum then it cannot be clearly be defined in the GCS engine.

    One possibility to apply the required transformation is to “import” the German data into Demark using a transformation to be applied once. This transformation could be a simple rotation/translation/rotation that approximate the relation from German ITRF2005 to Denmark ETRS89 if you can derive the transformation parameters with sufficient accuracy for your application.

    The other possibility would be as you appear to suggest, to define a “transformation datum” “GermanITRF2005ToDenmarkETRS89DatumTRF” or some other shorter name that would serve to define GCS meant for transformation purposes only. You would define all GCS needed upon this datum then perform a “Select Geographic Coordinate System…” with option “Correcting the Geographic Coordinate System – do not reproject the data” using these transformation GCS definition immediately followed by a “Select Geographic Coordinate System …” with option “Reproject the data to the new Geographic Coordinate System” upon the Danish GCS.

    Defining a new [transformation] datum will require MS v8i SS3 and defining new GCS will require Bentley Map (user-defined GCS can be used on any MS v8i site but definition requires a Bentley Map).

    I will be posting a stripped version of the present mail and your mail below for the benefit of the Be Community.

    Alain

Children
No Data