PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
September 2016
673
transferred from Eskisehir). Major Niyazi finally determined
the coordinates of Aya Sofia as: (Φ = 41º 00´ 30.0709˝ N, Λ =
28º 58´ 52.7238˝ West of Greenwich), but there was a problem.
The major published his findings in
Old Turkish!
A certain
couple of German Photogrammetric Technicians knew their
photogrammetry and Zeiss C-5 Stereoplanigraphs very well,
but they were apparently oblivious to geodetic publications
published in
Old Turkish
. During WWII, there was a
catastrophic military defeat of German and Axis forces in
the European Turkish region of Edirne. The secret
Planheft
of the Wehrmach mentioned that there was an unexplained
shift in the longitudes (and Eastings) of native maps in the
region of Edirne! I do not think this was just a “coincidence”
that happened to affect indirect artillery fire control.
The following is another fascinating excerpt from Major
Niyazi’s description of the enormous field work associated
with the Aya Sofia Dome Survey: “The geometric equipment
of the post consisted of one large Hildebrant theodolite, six
heliotropes, five geodetic projectors, twelve storage batteries
and one small Wild theodolite, while the participating staff
consisted of assistants Lieutenants Hilki and Kemal, one
permanent sergeant and twenty-one soldiers for general duty
and heliotrope communications, besides myself. We also had
tents and animals. Assigning three soldiers to each heliotrope
we placed them at designated points, and to insure the long
stay of these troops on the mountaintops beside the first-
order points, we gave them whatever supplies they needed.
We also constructed caves as shelter in their long vigil. On
some of these points we found caves already constructed in
1933, so we utilized them. In the matter of transportation,
motorized transports first brought our equipment to the
slope of the mountains, then horses, donkeys, ox and horse
carts took it to the mountaintop. During the transportation
of equipment the officers of the post paid most attention to
the transportation of the large theodolite and they saw to it
that there was always an officer in charge of this instrument.
After the instrument was taken out of the spring carriage
one or more soldiers always carried it on their shoulders to
the top of the mountain.” The entire unit spent about a year
in the field while observing the triangulation for the dome!
For Anatolia proper, the national system was the Mesedag
Datum (near Ankara) where: Φ
o
= 39º 52´ 10.451˝ N and Λ
o
= 32º 34´ 38.430˝ East of Greenwich and H
o
= 3255 m. Other
local datums included the Balikesir Datum of 1933 for the
region south of the Sea of Marmara (Mysia) where: Φ
o
~ 39º
37´ N and Λ
o
~ 27º 57´ E, and the Söke Datum for the region
of southwest Turkey (Caria) where: Φ
o
~ 37º 45´ N and Λ
o
~
27º 22´ 30˝ E. I believe it is likely that all of these old datums
were referenced to the Clarke 1880 ellipsoid as modified by
General Şevki.
According to a 1981 personal communication I had with
Professor I. Kasim Yasar of the Middle East Technical
University in Ankara, the parameters of the (1932-1946)
Gauss-Krüger Transverse Mercator were: “Latitude of
Origin,
j
o
= Equator, Longitude of Origin,
l
o
= 27º E - 45º
E in 3
g
intervals with Zones for all of Turkey and was used
with a scale factor at central meridian m
o
= 0.99992 only for
cadastral purposes and then abandoned. False Easting at
origin = Zone number × 10
6
+ 500,000 meters, False Northing
at origin = zero, Measurement unit = meter, and Ellipsoid
name = International, Madrid 1924. Since 1946 this system
was replaced with the identical system in all respects except
that the zone width was changed to 6° and m
o
= 0.9996 is
still in use (
as of 1981 – Ed.
) and the system of projection is
Gauss-Krüger for all of Turkey to a map scale 1/25,000. In
addition to … this and only for cadastral maps of scale 1.5000,
we employ again the Gauss-Krüger projection system in 3°
interval width with the scale factor at the central meridian
m
o
= 1.0000.”
The U.S. Army Map Service began computing the “Central
European Adjustment” in the late 1940s after WWII. The
adjustment was carried to Turkey through Greece and
Bulgaria, in part using the Czarist Russian Zapiski journals
of the 19
th
century in which the local origin was at the minaret
of the main mosque in Kyustendja (now Constanta, Rômania)
where: Φ
o
= 44º 10´ 31˝ North, Λ
o
= 28º 39´ 30.55˝ East of
Greenwich. Note that this longitude is a
correction
from
that published for Rômania (
PE&RS
, May 2001), thanks
to Dr. Momchil Minchev of the Bulgarian Geoinformation
Company. “The Central European Adjustment” was later
renamed the
European Datum of 1950
. For the remainder
of the 20
th
century, information on the geodetic foundation
and the ED50 network in Turkey remained a military secret
and its use was denied to all,
including to official civilian
government surveyors!
The cadastral agencies were forced to
establish their own networks, independent of the National
Triangulation and National Leveling networks that were
military secrets. In this respect, Turkey was definitely
following old European custom.
The times are changing for geodetic Turkey.
Official
transformation parameters for Turkey are now published
by the International Association of Geodesy (IAG), the
Bundesamt für Kartographie und Geodäsie
(German Federal
Office for Cartography and Geodesy), and Eurographics. Note
that the European sign convention for rotations is opposite
from the United States (and Australian) standard. Therefore,
the U.S. standard sign convention for rotations is listed in
the following parameters. In Turkey, for ED50 to ETRS89:
Δ
X = –84.1 m,
Δ
Y = –101.8 m,
Δ
Z = –129.7 m, R
x
= +0.0˝,
R
y
= +0.0˝, R
z
= +0.468˝,
Δ
s = +1.05 ppm. This 7-parameter
transformation cannot be truncated to just a 3-parameter
translation only, without complete recalculation of the least
squares solutions for only 3 parameters. Do not truncate the
abovepublished rotation and scale change parameters! The
transformation with the above 7 parameters for Turkey is
expected to yield positions of about 2 meters accuracy. An
example test point published for Turkey on the European
Datum 1950:
j
= 37º 08´ 35.8˝ N,
l
= 28º 28´ 25.32˝ E, which
transforms to ETRS89:
j
= 37º 08´ 32.07˝ N,
l
= 28º 28´ 23.79˝
E. The mathematical equations for all the projections and
