PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
October 2020
601
muth to Jelgava Church was adopted for the orientation of the
net. Although the Riga Church and Courland values are the
same as Scharnhorst value and the azimuth is identical, the
coordinates of Jelgava Church vary slightly by 0.002˝ in each
axis. This was due to the scales of the Liepaja and Jelgava bas-
es, which were adopted for the Courland System in preference
to the less reliable Scharnhorst scale. This became known as
the “Provisional Courland System (datum).” The Provisional
Courland System was divided into two Cassini-Soldner Grids:
The “Riga System” with its origin at Riga St. Peter’s Church
where
j
o
= 56º 56´ 53.919˝ N and
l
o
= 24º 06´ 31.898˝ East
of Greenwich, and the “Vardupe System” with its origin at
the Provisional Courland station Vardupe where
j
o
= 56º 51´
32.961˝ N and
l
o
= 21º 52´ 03.462˝ East of Greenwich. No false
origin was used for either grid. The Provisional Courland
System was immediately adjusted and computed before the
triangulation of central and east Latvia was completed. This
Provisional Courland System was first adjusted within itself
and then adjusted to the Latvian part of the Baltic Ring. The
lower-order control as far east as 24º 20´ East of Greenwich
was adjusted and computed in terms of this system.
The General Latvian Triangulation Net of first-order sta-
tions covers practically all of Latvia, including Courland.
The lower-order control east of 24º 20´ East of Greenwich
was computed in terms of the General Latvian Triangulation
Net. Actually, there is a small overlap around 24º 20´ East
of Greenwich for which the coordinates of all stations, of all
orders, were computed in both the Provisional Courland Sys-
tem and the General Latvian Triangulation Net. The 1924 net
was adjusted in stages to fit the following eight bases: Pu-
ci-Sarmen Jelgava, Jekabpils-Daborkalns, Garmaniski-Viski,
Kangari-Jamilova, Kirbbisi-Akija, Duorno Sielo-Dziedzinka
(Polish Base), Arula-Urkaste (Estonian Base), and Liepa-
ja-Paplaka. The chain Puci-Sarmen to Jekabpils-Daborkalns
forms the backbone of the modern net from which the adjust-
ment started. The origin of the General Latvian Triangula-
tion Net is Riga St. Peter’s Church (top of the Riga Church
spire) where Φ
o
= 56º 56´ 53.919˝ N, Λ
o
= 24º 06´ 31.898˝ East
of Greenwich, and the reference azimuth to Mitau German
Church
α
o
= 215º 24´ 04.38˝. The value for Riga St. Peter’s
Church approximates the Dorpat II System (datum). The Lat-
vian control was computed in terms of four Cassini-Soldner
Grids. The grid names and the coordinates of the respective
origins are as follows: Vardupe Cassini-Soldner Grid where
j
o
= 56º 51´ 32.961˝ N and
l
o
= 21º 52´ 03.462˝ E; the Riga Cas-
sini-Soldner Grid where
j
o
= 56º 56´ 53.919˝ N and
l
o
= 24º
06´ 31.898˝ E; the Gaizinkalns Cassini-Soldner Grid where
j
o
= 56º 52´ 15.031˝ N and
l
o
= 25º 57´ 34.920˝ E; and the
Vitolnieki Cassini-Soldner Grid where
j
o
= 56º 40´ 08.447˝ N
and
l
o
= 27º 15´ 12.252˝ E. These grid systems cover zones of
about 1½º to 2º wide, and overlap slightly. The Vardupe Grid
is computed from the geographics of the Provisional Courland
System, while the Gaizinkalns and Vitolnieki Grid values cor-
respond to the General Latvian Triangulation Net. The Riga
Cassini-Soldner Grid coordinates are computed from both the
Provisional Courland System geographics and the General
Latvian Triangulation Net. However, care is taken in the Lat-
vian “Trig” Lists to show from which geodetic system the Riga
Cassini-Soldner coordinates are computed.
During these inter-war years, the Russians were also ac-
tively recomputing their survey information in the Baltic
states. Prior to 1932 the Russian horizontal control of the
Baltic States was always referenced to Dorpat Observatory
at Tartu in Estonia. In 1932 the Russians set up Pulkovo Ob-
servatory 1932 as their horizontal datum and origin reference
to the Bessel 1841 ellipsoid, and later revised this to Pulkovo
1942, now properly termed “System 42” (datum) referenced to
the Krassovsky 1940 ellipsoid.
Thanks to E.A. Early of the U.S. Army Map Service, “In
1942 the German Army undertook the conversion of the Lat-
vian Soldner coordinates to DHG Pulkovo.” (Deutches Heeres
Gitter – German Army Grid) “The first phase of the conver-
sion embodied the change of projection from Soldner to Gauss-
Krüger. The Latvian Geodetic Engineer Mensin set up formu-
las and tables to convert the four Latvian Soldner systems
to the German Gauss-Krüger system. However, upon check-
ing these formulas at the boundaries of the Soldner systems,
inadmissible gaps were discovered. Mensin’s formulas were
then abandoned and new ones were derived following the
method given in Jordan-Eggert’s
Handbuch der Vermessung-
skunde
. Since there were no reliable geodetic connections to
the Pulkovo system available at that time, the conversion of
the Latvian system to the Pulkovo 1932 system could only be
approximated. The value of the datum point of the general
Latvian triangulation net approximates the Dorpat II system
value. The necessary formulas converting Dorpat II system
values to the Pulkovo 1932 system were available in the offi-
cial Russian work of Brigade Engineer O. A. Sergjew,
Making
and Editing of Military Maps
, Moscow 1939. In the absence
of better data, these formulas were taken as a basis for the
conversion to Pulkovo 1932 datum. The German Preliminary
DHG Pulkovo 1932 coordinates resulting from this conversion
were published in the form of
Ausgabe Koordinatenkartei
by
the Kriegs-Karten und Vermessungsamt Riga, in 1943.
In 1943 extensive surveys were executed along the Lat-
vian-Russian border for the final connection of the Latvian
triangulation with the Pulkovo system. At the conclusion of
these surveys, the Latvian system (already in terms of the
preliminary DHG Pulkovo 1932 system) was converted to the
Pulkovo 1932 system by a rigidfield adjustment. The Rus-
sian Pulkovo 1932 system coordinates used in this adjust-
ment were taken from Russian Catalogs. As a consequence
of this adjustment, new conversion constants were comput-
ed to convert from the four Latvian Soldner systems to the
Final DHG Pulkovo 1932 system. As mentioned previously,
the triangulation of Latvia is not completely uniform, since
the triangulation in Courland is based on the Provisional
Courland System adjustment. Only the first-order stations
in Courland are available in terms of the General Latvian
Triangulation Net. The lower-order trig in Courland was con-
