Coordinate system transformation (local (1970) to UTM - 34N; 35N)

The steps required to edit a Geographic Coordinate System are show in the Slides at: 
 

http://communities.bentley.com/products/geospatial/desktop/m/geospatial_desktop_gallery/218386

Thank you!

I managed to create a editable coordinate system.

Now it looks it comes hard part... The Bulgarian 1970 coordinate system is very old and "stinky"... Very little people know how to convert from it. And I don't know the correct parameters.  If someone know how to find correct parameters it will be great.

This is the information that I know about the Bulgarian 1970  coordinate system. If someone knows other information that he may need to help me I will try to find it.

"A type of projected coordinate system from the conical projections class. Four types of Lambert zones are defined under the coordinate system. Each of the zones is determined by a preset origin, by the reference latitude scale and by the range of F. The maps’ grid reference system is the 1970 Coordinate System. The four existing zones are: К - 3 (North-west), К - 7 (North-east), К - 9 (South-west) and К - 5 (South-east). Each of the zones is part of the Lambert coordinate system, i. e. its planar part is rotated to match map’s north, at a relatively small, but zone specific angle, selected in such a way as to preserve map’s conformance.Constants specific to each zone are added to the to the planar coordinates, so that the final coordinates’ format is Gaussian. The real coordinate values are Y=8 500 000 m и Х=4 500 000 m."

www.google.bg/.../@42.7104605,25.5902966,8z

It uses Ellipsoid [KRASSOVSKY 1940]

I can provide points transformed from 1970 from all zones converted to WGS84 N34 and WGS84 N35 if it will be in help.

Did someone know which are the correct parameters ?

You may have to build your Custom Coordinate System from Placemarks using the MicroStation tools for that.  This PDF document provided by the ASPRS indicates that...

In the 1960s it was decided to use

the “System 50” only for military applications

and to introduce a set of plane

coordinates for civilian use. The new system

consists of four Lambert Conformal

Conic projections on the Krassovsky

1940 ellipsoid; the four overlapping

zones are K3, K7, K5, and K9. Each zone

is defined with a latitude of origin and a

scale factor at origin, and each zone has

a different initial azimuth of the central

meridian in order to rotate (and obfuscate)

the grid. The parameters remain

secret to this day, and this practice (including

zone rotation) is typical of many

of the former Soviet Bloc countries.

Here are some online Bentley Communities remarks and references for Placemarks:

How to define a custom projection?

Geocoordinate Systems, back to Placemarks?

 

otherwise use the MicroStation HELP instructions under Setting Up Projects | Geo-Coordination.

Hi again!

I menage to create GCS by place marks and re project to WGS.

In the text files you can see the Placemarks-70.txt file are my 1970 CS points. In Placemarks_trWGS.txt are points converted from a converting program.

My placemark set looks like this:

And the placemark 2 is in far left. 

As you can see in the GCS from Placemarks dialog box, there is some error:

That means the program did not create the correct GCS (1970) but some close to it... (i think the program did not recognize the placemarks as Lambert Conformal Conic projection)

Is there a chance to make exactly the 1970 GCS by placemarks ?

Aleksandar,

The functionality to create a geographic coordinate system from placemarks always creates one based on the projection method "Azimuthal Equal Area", and never on any other type of projection method. The computation of a coordinate system from a set of point pairs can be very complex or impossible for any other method than this Azimuthal Equal Area method. Looking at your result I would be happy such accuracy could be obtained. Errors of the order of a meter is rather good concerning the conditions of your data (unknown coordinate system parameters). Of course it is an approximation but I would be rather surprised you could obtain better with any other method. If you want you can also try an Helmert transformation that can be set using the method detailed in the following article: communities.bentley.com/.../setting-up-in-the-real-world-georeferencing
But I would be really surprised you would obtain better results. At least the method using Placemarks may compensate for some non-linear distortion and Earth curvature.