Adding Custom Datum to ArcGIS

[KevinDanaher]

I would like to plot several coordinates that are on a datum not included in the ArcGIS coordinate system library (10.4.1 for Desktop). The datum is the Israel-Jordan Boundary Datum of 1994, which was created between the two countries in order to delimit their boundaries with accuracy and precision.

The technical documents submitted to the UN in regards to their boundaries define the datum very well, however there are no transformation parameters provided to get to WGS84 or any other standard. The following is an excerpt from the documents, where the datum is defined:

IJBD '94 DEFINITION

THE GEODETIC DATUM IJBD'94 WAS DETERMINED BY FIXING THE COORDINATES OF POINT IJBDO9 (ONE OF THE 12 DATUM POINTS), ADOPTING THE WGS'84 ELLIPSOID, AND FIXING THE REFERENCE ELLIPSOID IN THE GEOCENTER ACCORDING TO THE PRECISE GPS VECTORS WHICH WERE MEASURED BETWEEN THE 12 DATUM POINTS.

CONCERNING THE VERTICAL DATUM, THE JTE AGREED TO ADOPT THE ELLIPSOIDAL HEIGHTS (FOR ALL THE BOUNDARY COORDINATES) WITH REFERENCE TO THE IJBD'94 DATUM AND TO THE WGS'84 REFERENCE ELLIPSOID. THE DECISION SIMPLIFIED AND FACILITATED THE COMPUTATION SINCE ONLY ONE 3 DIMENSIONAL DATUM HAD TO BE DETERMINED. SO, NO ATTEMPT WAS MADE TO DETERMINE THE GEOID, OR, THE SEA LEVEL SURFACE AS THE DATUM OF THE VERTICAL COMPONENT OF THE COORDINATES.

THE COORDINATES OF POINT IJBD09 WERE COMPUTED BY AN AVERAGE BETWEEN THE RESULTS OF THE ABSOLUTE POSITIONING WHICH WAS CALCULATED BY EACH SIDE USING BROADCAST EPHEMERIS.

THE FOLLOWING ARE THE AGREED COORDINATES:

POINT: IJBDO9

LATITUDE: 31 45 04.37499

LONGITUDE: 35 36 13.70799

HEIGHT: -272.150 M (ELLIPSOIDAL HEIGHT)

THE REFERENCE ELLIPSOID PARAMETERS ARE:

ELLIPSOID: WGS84

SEMI MAJOR AXIS: 6,378,137.000 M

I/FLATTENING: 298.257223563

GM: 3986005 * 10^8 M^3 S^-2

In the document itself, there are even more details given as to the exact series of points (12) used to define the datum and their locations in terms of local settlements/cities.

My question is, how would I go about accurately plotting these points? I have not been able to find a way to add a custom datum to ArcMap using a custom point of origin. And even if I did, the accurate plotting of these points would only be correct relative to one another without a proper transformation to WGS84 datum or something else.

I'm attaching the treaty PDF, which is also available at the UN site.

[NeilAyres]

As far as I am concerned, you don't need to define a new datum for this.

It is equivalent to GCS_WGS_1984.

But perhaps someone with more experience in this area like MKennedy-esristaff might like to comment

[KevinDanaher]

Any comment is very welcome.

It may be that differences between this and WGS84 datum are negligible.

However I'd imagine this datum was created for high level of precision in

surveys of that local area, while the WGS84 datum is suitable for use

globally.

[NeilAyres]

But, all these details are doing is pinning down the principal point in terms of the WGS84 ellipsoid. So therefore it is based on that. That is all. It is WGS84 based, that's all you need.

You cannot define a "Custom Datum". You can however define a custom datum transformation between known and defined datums. That is how transformations are calculated between old "classical" datums and more modern ecef type datums (chiefly WGS84) are defined.

So, you still don't need a custom datum.

[NeilAyres]

If you have a series of measured points with datum 1 X, Y, Z and datum 2 X, Y, Z, you can calculate a geocentric transform between the 2.

Most survey software will do this and we can do it here as well.

[KevinDanaher]

I guess I assumed since the datum was so well-defined in the technical report that it could be classified as a "known datum," or at least have a documented transformation. Perhaps this information exists but is proprietary and held closely between these two states? These points, plotted in GCS_WGS_1984 coordinate system, certainly appear to plot close to where expected, anyhow.

In terms of a geocentric translation between the two. I could probably use high resolution satellite imagery to attempt to create one, but at that point I could be introducing errors because of flaws in imagery orthorectification, etc. 

Neil, I appreciate your insightful replies. Thank you.

[MelitaKennedy]

Hmmm, I'm not enough of a surveyor/geodesist to fully interpret and understand what they're saying. I found similar information in a book by Ron K. Adler, https://books.google.com/books?id=dJECrHJRkUsC&pg=PA73&lpg=PA73&dq=ijbd+datum&source=bl&ots=aN2qDW0Z...

To me, it looks like it was WGS84 more or less. The fact that they say "broadcast ephemeris" implies there was no post-processing. They didn't recalculate the results using the precise orbits. Checking around, I'm finding accuracy (error?) estimates of 1-3 meters for broadcast ephemeris data. See this thesis for example: http://www.colorado.edu/ASEN/asen6090/broadcast_vs_precise.pdf

If you can get decent imagery and identify the pillars, that's probably all you need.

Melita