[Proj] Egyptian Datum Parameters

[Irwin Scollar]

>Cliff Mugnier wrote:

>There are numerous academics and government agencies in Egypt that 
>can supply what you want, but nobody is likely going to do so.  I 
>have tried to get such information for years, and have never even 
>received the courtesy of a reply.  Note that such information is 
>considered a military secret in some countries, and I suspect that 
>Egypt is one of those countries.

I thought that this might be the case which is why I asked on this 
list.  My dear departed friend Yigal Shiloh told me that before and 
during the 6 days war, he and his reconnaissance squadron flew the 
length of the Nile Valley at low altitude and archived nearly 
complete vertical aerial cover.  But military minds are usually two 
wars or more behind in their thinking, even if Keyhole and Quickbird 
imagery, both public and non-public, makes it all absurd.

>The NIMA/NGA transformation parameters are from ED50 - European 
>Datum 1950 to WGS84 as well as Old Egyptian Datum 1930 to WGS84.
>
>If you need better accuracy, I recommend you perform the 
>observations yourself and publish your derived transformation 
>parameters on this list for everyone else.

The reason behind my request was simply that in my program AirPhoto, 
I recently implemented the use of Google Earth at enhanced high 
resolution, and for less belligerent parts of the world have 
incorporated the ability to search GE using a local grid system 
converted to the lat/lons needed by Google on the assumption that 
over small areas of interest, the error is not significant.  After 
linear correction of the result, I get apparent precision to better 
than a meter, and in some places like the UK, Switzerland and Germany 
even in the low decimeter range as tested by using some of the known 
coordinates of the more than 6000 UK Ordnance Survey primary trig 
point pillars, many of which are visible in Google Earth.  For rough 
GE search in Egypt, the error in Molodensky/Bursa Wolfe won't be 
troublesome.  What will be a bit annoying will be the inability to do 
a local correction down to the error range of a dual frequency receiver.

This project using GE data raised the interest of the Egyptologists 
in Berlin who did not have access to accurate imagery or maps.  Hence 
my request to this list.

>I would like to point out that 7-parameter transformations are 
>really applicable only to small provincial areas that are 
>significantly smaller in areal extend than a single Egyptian 
>Transverse Mercator Belt.  Working with a crew of a dozen men all 
>with dual-frequency receivers, you should get the job done within a 
>decade or so ...    :-)

I haven't the slightest intention of ever travelling south of the 
Balearics >:-} . Seriously though, I didn't expect to get a single 7 
parameter Helmert transform value for a country as large as Egypt.  I 
had hoped that like Spain, where four such sets are openly available 
on the national mapping service web site, that a set of such 
parameters might exist and be available.

>The point to this is that what you desire is likely a government 
>secret, and you are asking for geodetic-quality transformations for 
>an entire country.  That sort of thing is implemented with 
>algorithms such as NTv2 or NADCON - not with a Molodensky or 
>Bursa-Wolfe model.  I have used such models in the past, but for 
>tiny regions such as a single international boundary line or for a city.

Molodensky/Bursa Wolfe used to be adequate in the days of GPS 
"selective availability". Today, the way I have followed in temperate 
highly occupied parts of the world  is to create a dense Delaunay 
trianglulation on visible points and do affine interpolation within 
that grid to a precision digital map or to primary and secondary trig 
points if visible rather than the use of high order polynomials. I 
obtained fairly good results matching old first edition Ordnance 
Survey maps from about 1793-1805 to to modern digital maps of the 
Isle of Wight that way.  Lantmateriet in Sweden has published a 
similar technique to connect up the large number of idiosyncratic 
communal grids to the national grid.

>You likely will have to "make do" with what you have unless you 
>receive Egyptian Government authorization to do otherwise.

Never fear! I shan't ask. :-)

Irwin Scollar

>From: proj-bounces at lists.maptools.org on behalf of Irwin Scollar
>Sent: Tue 26-Aug-08 09:03
>To: proj at lists.maptools.org
>Subject: [Proj] Egypt Datum Transformation Parameters
>
>Can anyone supply Helmert 7 parameter datum parameters between WGS84
>Ellipsoid and the Helmert 1906 Ellipsoid used in the "Red Belt, Blue
>Belt, Purple Belt, Purple Belt Extended"  Old Egyptian 1907 areas or
>a link to where they may possibly be found?
>
>I have only the 3 parameter Molodensky transform data in the NIMA
>1987 Madtran and the later Geotran2 programs which may not have
>errors as low as those of a modern dual frequency GPS receiver.
>
>Irwin Scollar