PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
May 2017
339
escarpments and peaks including the Inyanga Mountains
and the Chimanimani Mountains. The parameters published
by DMA/NIMA for the Zimbabwe Arc 1950 Datum to WGS84
are: ΔX = –142m ± 5m, ΔY = –96 m ±8m, ΔZ = –293m ±11m.
The Moçambique–South Africa boundary, which is about
488 km (305 miles) in length, consists of two discontinuous
parts. The longer part extends southward from the confluence
of the Limpopo River and the Luvuvhurivier River for 408
km (255 miles) along straight line segments to the northern
tripoint with Swaziland and the M’Pundweni Beacon. The
remainder of the boundary from the southern tripoint with
Swaziland to the Indian Ocean follows the Maputo River
(Great Usutu) downstream for about 17 miles and then
continues by straight-line segments for another 53 km (33
miles). The parameters published by DMA/NIMA for the
South Africa Cape Datum to WGS84 datum shift are not as
accurate as those are according to Rens & Merry. Professor
Charles Merry of the University of Cape Town found a
substantial improvement with a 4-Parameter transformation
such that: X = –190.0m ± 7.7m, DY = –137.9m ±3.7m, Z =
–257.4m±4.7m, scale = +11.0x10-6, and the aggregate
positional rms = ±2.9m. Note that the values quoted are the
most reliable for points within Moçambique.
The Moçambique–Swaziland boundary has two tripoints
with South Africa. Northward from the Maputo River it
extends along the summit of the Lebombo Mountains for
approximately 106 km (66 miles) to M’Pudweni Beacon.
The boundary consists of various straight-line segments
demarcated by trigonometrical and/or boundary beacons. The
original treaty in this area dates back to 1869, and portions
remained in dispute until 1927. The M’Ponduine Geodetic
Station (MGM 677) was used as the origin of a local bound-
ary grid that, based on an ellipsoidal polyhedric projection,
has the origin as:
f
o
= 25° 56´ 47.19˝ S,
l
o
= 31° 58´ 40.46˝ E.
No false origin was used, and the M’Uguene Geodetic Station
(MGM 658) given as a computational check is:
f
= 26° 07´
25.98˝ S,
l
= 32° 16´ 31.58˝ E and X = –29,757.40 meters, Y
= +19,691.48 meters. There is no doubt that this boundary
Grid is on the Cape Datum, Clarke 1880 ellipsoid. Profes-
sor Merry’s solution for Swaziland is a 3-parameter shift to
WGS84 Datum where: ΔX = –136.8m ± 0.5m, ΔY = –106.6m
±0.5m, ΔZ = –293.0m ±0.5m, and the aggregate positional
rms = ±0.7m.
There are three classical geodetic Datums in Moçambique,
and all are referenced to the Clarke 1866 ellipsoid where a
= 6,378,206.4 meters and
1
/
f
= 294.9786982. This ellipsoid
was a favorite of the Portuguese in Africa — they used it in
Angola, too! The oldest is the Madzansua Datum of 1904 with
its origin at point MGM 2 (near the village of Zumbo), where:
Φ
o
= 15° 35´ 20.7˝ S, Λ
o
= 30° 28´ 09.3˝ E and H
o
= 1010.9
meters.
The most often quoted classical datum for Moçambique
in Western literature is the Observatorio Campo Rodrigues
Datum of 1907. Its origin is near Lourenço Marques (Maputo)
at station MGM 650 where: Φ
o
= 25° 58´ 06.99˝ S, Λ
o
= 32° 35´
37.75˝ E. The defining azimuth is unknown. I was surprised to
learn that before 1971 it wasn’t used for much other than the
1:50,000 topographic mapping from the southern border to no
further north than the coastal village of Moebase. The Cape
Datum was connected to the Observatorio Datum of 1907
by the M’Ponduine and Ypoy (MGM 675) stations through
the Transvaal triangulation. Both of these old datums were
established by Captain Coutinho.
The largest classical horizontal geodetic datum is the
Tete Datum of 1960. It’s origin is at the station at the NW
Tete Baseline (MGM 799) where: Φ
o
= 16° 09´ 03.058˝ S,
Λ
o
= 33° 33´ 51.300˝ E. The reference azimuth to station
Caroeira (MGM 40), α
o
= 355° 50´ 21.07˝ from south, and
H
o
= 132.63 meters. A French memo from Maputo observed
that “the astronomic coordinates = geodetic coordinates, but
it is suspected (according to the results of the computations)
which is such that the deviation of the vertical are at the point
chosen for the ‘datum.’ In effect for each zone north and to
the NW, there are concentrated deflection forces due to iron
mineral deposits, whereas to the south and SE coal and less
dense minerals are found. If that hypothesis is likely, then
the coordinates of the ‘datum’ of all the points in Moçambique
will need correction.” The Tete Datum of 1960 coordinates
of MGM 2 are:
f
= 15° 35´ 15.349˝ S,
l
= 30° 28´ 15.057˝ E.
The Tete Datum of 1960 coordinates of MGM 650 are:
f
= 25°
58´10.359˝ S,
l
= 32° 35´ 40.056˝ E.
In 1995, a comprehensive readjustment of the entire
geodetic network of Moçambique was initiated by Norway
Mapping in collaboration with the government. The project
was concluded in January of 1998, and the result was a
32-point constrained adjustment of 759 two-dimensional
triangulation points throughout the country and the
designation of a new datum called MOZNET/ITRF94, com-
patible with the WGS84 Datum. A 7-parameter Bursa-Wolf
transformation was developed, but the national model yields
residual errors as high as 30 meters. Four “regional” models
were developed, but these accuracies vary between 1 to
10 meters, depending on the “region.” This seems to be an
ideal country for the development of a multiple regression
equation model for a single national datum shift model.
Probably a “NADCON” sort of solution would be even better!
The MOZNET 98 adjusted coordinates of MGM 2 are:
f
= 15°
35´ 18.7529˝ S,
l
= 30° 28´ 12.3667˝ E, of MGM 650 are:
f
=
25° 58´ 12.7520˝ S,
l
= 32° 35´ 38.4687˝ E, of Base Tete NW
are
f
= 16° 09´ 07.0480˝ S,
l
= 33° 33´ 49.7778˝ E, and of
M’Ponduine are:
f
= 25° 56´ 52.6154˝ S,
l
= 31° 58´ 39.6248˝
E. The government of Moçambique adopted the use of the
UTM Grid in 1954, and, with the exception of the M’Ponduine
Polyhedric Grid, no other Grids are used.
Thanks for much of the historical and geodetic details
contained herein go to Mr. J. Carvalho of Maputo.
