13

Before January 1999 the geode c datum used in South Africa was the Cape Datum based on the Clarke
1880 spheroid, a er January 1999 the Datum changed to the Hartebeesthoek94 Datum based on the
WGS84 spheroid. There is a mispercep on that the SA coordinate system changed in 1999, but this is
not true, only the Datum changed as the Lo Gauss projected coordinate System is s ll in use. The
reason for this mispercep on might be rooted in the fact that SACRS users use terminology
interchangeably such as using “Hartebeesthoek” as the name of the coordinate system and not Lo
Gauss.
In prac ce, many GIS users choose to use the Universe Transverse Mercator projected coordinate
system and not the Lo Gauss projected coordinate system. This means that even though they use the
Hartebeesthoek94 Datum they cannot claim that they are using the SACRS.

The South African Lo Gauss Coordinate System

The official coordinate system for South Africa that forms part of the SACRS is the Lo Gauss projected
coordinate system. Note that the South African coordinate system did not change in 1999, but only
the Datum.
The Lo Gauss Coordinate system consists of belts running north and south, 2 degrees of longitude
wide, the central meridians being every odd meridian example 21°, 23°, and so on. Each of these 2-
degree belts is projected using the Gauss Conformal projec on. This projec on is a transverse
cylindrical projec on, and the cylinder touches the globe at each odd meridian.
The belt is referred to as a Lo, in other words, if the central meridian is 21° the belt will be referred to
as Lo21 (some users refer to it as WG21, this is probably because of them using WGS84)

Figure 8: The Lo Gauss Coordinate System

In Figure 8 you can visualize the bands running across South Africa. Note that the equator has a value
of 0 meters and that the central meridian, for example, 21° also has a value of 0 meters. Different from
normal mathema cs, your Y values will be nega ve to the right and posi ve to the le of the central
meridian.

Therefore, if we want to calculate the Lo Gauss coordinate of a feature you will do it as follows: (See
Figure 9)
 14

 Distance from the Equator to A is

3 720 013m
 Distance from the Equator to B is
3 741 101m
 Distance from the Equator to C is
3 750 503m
 Distance from the central meridian to
A is 21 967m
 Distance from the central meridian to
B is 60 279m
 Distance from the central meridian to
C is 61 001m

This Diagram is not drawn to scale

Figure 9: Calculate Lo Gauss coordinates

Below are the Lo Gauss coordinates for point A. La tude is calculated by adding the distance to point
A from the equator to 0 meters (the value of the Equator). Longitude is calculated by adding or
subtrac ng the distance to point A from the central meridian to 0 meters (the value of the meridian).
Lo Gauss Coordinates for A:
La tude: Equator + Distance to A
0m + 3 720 013m
+3 720 013mX
Longitude: Central Meridian +/- Distance to A
0m - 21 967m (minus because A is to the right of 0)
-21 967mY
-21 967mY; +3 720 013mX;

Lo Gauss Coordinates for B:

La tude: Equator + Distance to B
0m + 3 741 101m
+3 741 101mX
Longitude: Central Meridian +/- Distance to B
0m - 60 279m (minus because A is to the right of 0)
-60 279mY
-60 279mY; +3 741 101mX

Lo Gauss Coordinates for C: (Note C is to the le of the central Meridian)

La tude: Equator + Distance to C
0m + 3 750 503m
+3 750 503mX
Longitude: Central Meridian +/- Distance to C
0m + 61 001m
+61 001mY
+61 001mY; +3 750 503mX
 15

All Lo Gauss coordinates are in meters.

The Universal Transverse Mercator Coordinate System (UTM)

The Universal Transverse Mercator (UTM) is a projected CS that divides the world into 6°x 8° zones,
each zone being projected using the Transverse Mercator Projec on (TS). The UTM system divides the
earth into 60 longitudinal bands, each 6 degrees of longitude wide, see Figure 7. These bands extend
from a la tude of 80° South to 84° North and from 180° West to 174° West and are centered at 177°
West. UTM longitudinal bands are numbered from 1 to 60, star ng at the interna onal date line,
longitude 180°, and proceeding east. Each band is divided into horizontal bands spanning 8 degrees of
la tude. These bands are le ered south to north, beginning at 80° South with the le er C and ending
with the le er X at 84°North. The le ers I and O are skipped. The band le ered X spans 12° of la tude.
As the longitudinal bands of 6° and the la tudinal bands of 8° crisscross the Earth, it produces a square
grid with zones of 6°x 8°. A square grid is superimposed on each zone, which has a central meridian
called a false Eas ng. UTM grid coordinates are expressed as a distance in meters to the east, referred
to as the Eas ngs, and a distance in meters to the north, referred to as Northings. Note that the
Eas ngs are longitudes (running parallel with the central meridian) and the Northings are la tudes
(running parallel with the equator).

Figure 7: The UTM projected coordinate system h ps://www.onestopgis.com/GIS-Theory-and-

Techniques/Coordinate-Systems/Projected-Coordinate-Systems/posts/UTM-Grid-System/Figure-
shows-Universal-Transverse-Mercator-System.png

The center longitudinal line of the zone is known as the central meridian. This central meridian is
assigned an eas ng value of 500 000 meters East. Since this 500 000m value is an arbitrary value, this
central meridian is referred to in this case as a False Eas ng. An Eas ng of zero will never occur, since
a 6° wide zone is never more than 674 000 meters wide. See Figure 8.
 16

Figure 8: UTM 6°x 8° Zone

UTM northings are measured rela ve to the equator. For loca ons north of the equator (for example,
Europe, Asia, and North America) the equator is assigned the northing value of 0 meters North. To
avoid nega ve coordinates, loca ons south of the equator are calculated with the equator assigned
an arbitrary value of 10 000 000 meters North and known as a false Northing. Always remember to
avoid confusion, specify if the loca on is north or south of the equator.

Therefore, if we want to calculate the UTM coordinate of a feature you will do it as follows: (See Figure
9)

Equator
A (False Northing)

Central Meridian (False Eas ng)

Zone UTM 35J

Figure 9: Calculate UTM coordinates

 Distance from the Equator to A is 3 720 013m

 Distance from the Equator to B is 3 750 503m
 Distance from the central meridian to A is 28 290m
 Distance from the central meridian to B is 28 370m

UTM Coordinates for A:

Northing (La tude): Equator – Distance to A
10 000 000m – 3 720 013m
6 279 987m
Eas ng (Longitude): Central Meridian + Distance to A
500 000m + 28 290m
528 290m
 17

UTM zone 35J, 528 290mE; 6 279 987mN

Note that in GIS we usually do not use the zone designator, we just indicate that the coordinate is
North or South of the Equator. Therefore, the coordinate will be wri en as follows:
UTM zone 35S, 528 290mE; 6 279 987mN

UTM Coordinates for B:

Northing (La tude): Equator – Distance to B
10 000 000m – 3 750 503m
6 249 497m
Eas ng (Longitude): Central Meridian – Distance to B
500 000m – 28 370m
471 630m
UTM zone 35J, 471 630mE; 6 249 497mN
Note that in GIS we usually do not use the zone designator, we just indicate that the coordinate is
North or South of the Equator. Therefore, the coordinate will be wri en as follows:
UTM zone 35S, 471 630mE; 6 249 497mN.

Georeferencing and Geometric Transforma ons

Spa al data that is used in applica ons as layers, is used as analy cal tools for spa al analysis or
modelling and will need a coordinate reference system (CRS) and its proper es. Without a CRS, layers
of spa al data will not overlay properly, and analysis will be difficult or impossible. If you have spa al
data without a spa al reference (in other words without a CRS) you will need to Georeference your
spa al data. If your Spa al data layers have different spa al references, for example, use different
coordinate systems or datums you will need a Geometric Transforma on.

There might exist some confusion between the terms Georeferencing and Geometric Transforma ons.
Chang (2019) differen ates between two types of Geometric Transforma ons, Map-to-Map, and
Image-to-Map Transforma ons. He explains that Map-to-Map Transforma ons are newly digi zed
spa al data that are based on computer units and the geometric transforma on converts the newly
digi zed map into projected coordinates. Furthermore, Image-to-Map Transforma ons apply more to
remotely sensed data (in other words raster data). This transforma on changes rows and columns into
projected coordinates. Chang (2019) goes on to say that Image-to-map transforma ons are also known
as georeferencing.

A geometric transforma on requires a set of control points and transforma on algorithms to register
a digi zed map, a satellite image, or an aerial photograph onto a projected coordinate system. It is also
commonly used in a GIS to transform maps between coordinate reference systems. Control points play
a key role in determining the accuracy of a transforma on, the root mean square error (RMSE) is
calculated to find the “goodness” of the control points. The selec on of control points for a map-to-
map transforma on is rela vely straigh orward. What we need are points with known real-world
coordinates. Control points for an image-to-map transforma on, also called ground control points
(GCP), are points where both image coordinates (in rows and columns) and real-world coordinates can
be iden fied. GCPs are selected directly from a satellite image.

Chang (2019) proposed different geometric transforma on methods for transforming one coordinate
system to another. Each method is dis nguished by the geometric proper es it can preserve and by
the changes it allows, see Figure 10.