BRd 45(1)

CHAPTER 3

GEODESY

Para
0301. Introduction and Scope of Chapter
0302. The Need for Geodesy

SECTION 1 - SPHEROIDS

0303. The True Shape of the Earth

0304. The Surfaces of the Earth
0305. Geodetic and Geocentric Latitude
0306. Parametric Latitude
0307. Length of 1' in Latitude and Longitude
0308. Geodesics

SECTION 2 - DATUMS

0309. Datums
0310. Geodetic Datums
0311. Geodetic Datums Worldwide
0312. Datum Shifts
0313. Datum Transformation

SECTION 3- EARTH MODELS FOR NAVIGATION SYSTEMS

0314. Types of Earth Model

0315. Spheroidal Earth Models
0316. Spherical Earth Models
0317. Flat Earth Models

3-1
A .. ... , ~n-t Q. \/o. rc i nn 1
BRd 45(1)

CHAPTER3

GEODESY

0301. Introduction and Scope of Chapter

Chapter 2 introduced the concept of reference frames, and in particular the primary
terrestrial reference frame used for navigation, namely latitude and longitude. The shape of
the Earth was also discussed. This chapter takes the discussion further, by introducing the
concept of geodesy, and how the science of geodesy materially affects the way that
mariners navigate at sea.

0302. The Need for Geodesy

It has been explained that the Earth is not indeed a flat surface, and moreover it is
not even a true sphere; rather it is an oblate spheroid. Geodesy, or geodetics, is essentially
the science of measuring and representing the surface of the Earth (or indeed any other
planet or moon) mathematically, in such a way that useful calculations can be performed
with the necessary degree of accuracy without requiring unreasonable amounts of
computational power. It is essential for the modern navigator to understand geodesy as
now, with WECDIS, navigational calculations are not carried out on the flat surface of the
chart, but rather directly on the surface of the reference ellipsoid (which comes out of the
study of geodesy).
 BRd 45(1)

SECTION 1 - SPHEROIDS

0303. The True Shape of the Earth

As described in Chapter 2, the Earth is not a true sphere but rather it is an oblate
spheroid, with a distinct 'Equatorial bulge' and attendant flattening at the poles. Readers are
referred to Para 0204 for a full discussion of this subject.

0304. The Surfaces of the Earth

The oblate spheroid, the Earth's true shape, is not in fact a regular shape. The
oblate spheroid is overlain by the bathymetry and topography of the Earth's surface, the
various bodies of water, and various other factors such as local variations in density and
gravitational anomalies. A number of distinct surfaces can be identified, either in reality or
conceptually. These are illustrated in Fig 3-1 and discussed in more detail below.

Fig 3-1. The Surfaces of the Earth:

The Physical Surface, the Geoid, and the Spheroid

GEOID
UNDULATION

'liEROfD (f:((fi

(This Is the 'local vertical". u

Indicated by to
observer on the Earth's surface)

a. The Surfaces of the Earth. At least three distinct surfaces of the Earth can be
defined.

(1) The Physical Surface. The physical surface of the Earth, self-evidently,
is the land and water which physically makes up the Earth. This is not a perfet
sphere as it sits upon the oblate spheroid, and is then further deflected from the
mean sphere by the topography (mountains, bathymetry) and the fluid surfaces
of the oceans and ice caps. The physical surface of the Earth is thus both in
constant motion, and highly irregular in shape with local departures of several
kilometres from a pure sphere or even spheroid. The physical surface of the
Earth is thus not a suitable surface on which to base calculations, even with
present-day computing technology.
BRd 45(1)

(2) The Geoid. The geoid is a more regular shape, which is often considered
the 'true figure of the Earth'. It is the shape which would be defined by the
surface of the oceans under the influence of Earth's gravity and rotation alone, if
they extended in an unbroken line around the Earth (it assumes hypothetical
narrow canals through the continents rather than a 'water world'). Whilst it is
conventional to think of the geoid at sea level, it can be defined at any arbitrary
level of gravitational potential because what specifically defines the location of
the geoid is that it is an 'equipotential surface' meaning that all points on the
surface have the same effective potential (the sum of gravitational potential
energy and centrifu§al potential energy). Gravity is always perpendicular to the
geoid, which means that, due to the many density and gravitational anomalies
in the Earth, the geoid is still a highly irregular shape. The geoid is therefore still
not suitable to use as a computational surface. In fact, whilst it has long been
identified as a concept, it is only in recent decades with the advent of satellite
technology and sophisticated gravity sensing systems that the geoid has been
identified with any sort of confidence.

(3) The Spheroid or Reference Ellipsoid. As neither the physical surface

nor the geoid are sufficiently regular for practical purposes, geodesists further
define the spheroid or reference ellipsoid . This is an ellipse which has been
rotated about its semi-minor axis. Calculations based upon the semi-major and
semi-minor axes provide values for the spheroid's flattening and eccentricity.
The spheroid (or reference ellipsoid) forms a mathematically convenient surface
on which to base positions and positional calculations. Whilst there is only one
physical surface of the Earth, and only one geoid, there are an infinite number
of spheroids which can be specified to support local, regional or worldwide
geodetic datums; by changing the mathematical parameters of the spheroid it
can be made to (for example) more closely fit the geoid in a region at the
expense of less accuracy elsewhere on the globe. WGS 84 is the spheroid
used by GPS, ECDIS, and the vast majority of navigational systems in use in
the Royal Navy - see Para 0311.

a Note. The term 'spheroid' is used in this book, as it clearly aligns with the
concept of the spheroid Earth rather than the spherical Earth. However, in
much of the literature relating to WECDIS and ECDIS systems it is
increasingly common to use the term 'reference ellipsoid' and it must be
empha§i_sed that the two terms have exactly the same meanina.

b. Flattening. The degree of 'flattening' (sometimes referred to as 'ellipticity'), f, of

the spheroid is an important factor in geodetic datum conversion calculations. The
reciprocal of flattening (i.e. 1/f) and the Earth's semi-major Equatorial axis a are
defined values, and all other values are derived from them. The relationship
between f, a, and b (the semi-major polar axis) is:

f = a-b
b

3-4
 BRd 45(1)

Fig 3-2. Flattening ('f) of the Spheroid

a = Semi-major (Equatorial) Axis

b = Semi-minor (Polar) Axis
f = Flattening (ellipticity)

<------ a

I= a -b
a

8 Note. The ellipse above is exaggerated for clarity.

c. Eccentricity. The 'eccentricity' of the spheroid is another key parameter in

geodetic datum conversion calculations. If point M moves so that its distance from a
fixed point S (the focus of the ellipse) is always in a constant ration e (less than
unity) to its perpendicular distance from a fixed straight line AB (known in geometry
as the 'directrix'), then the locus of Mis an ellipse of eccentricity e. The eccentricity
(e) of the spheroid may be calculated from the semi-major (Equatorial) axis a and
semi-minor (polar) axis b as follows:

e =("':.Irr thus ez = az + bz
a2

Using flattening/: e = (2f - f 2 ) 112

Fig 3-i The Eccentricity ('e') of the Spheroid

A y

*
I a = Semi-major (Equatorial) Axis
I b = Semi-minor (Polar) Axis
b e = Eccentricity

I
X
s
,,,- j
c
e = ( a2 a~ b2 1/2

a2 -b2
e2
7
B
3-5
August 2018 Version 1
BRd 45(1)

0305. Geodetic and Geocentric Latitude

Fig 3-4 below shows a 'meridional' section of the spheroid. A meridional section is a
section through a sphere or spheroid oriented along a meridian (i.e. North/South); its
converse is a zonal section, which is orientated along a parallel of latitude in the case of the
Earth. It is used to demonstrate the distinct differences between geodetic and geocentric
latitudes. M is a point on the meridian PAP', and MK is the tangent to the meridian at M.
Whilst the navigator will not practically deal with these two sorts of latitude themselves, they
are important with regards to how WECDIS conducts navigational calculations.

a. Geodetic Latitude. If the normal LM to the tangent MK cuts OA at L, the angle

ML.A describes the geodetic latitude of M, and is conventionally denoted by <t>. ML
does not pass through the centre of the spheroid 0, except in the special cases
where M is on the Equator or at the poles. Unless specified otherwise, when the
term 'latitude' is used without qualification then it refers to geodetic latitude.

b. Geocentric Latitude. The angle M0A which, by definition, must pass through
the centre of the spheroid 0 , describes the geocentric latitude of M and is
conventionally denoted by 9.

c. Connection between Geodetic Latitude and Geocentric Latitude. The

geodetic latitude <l> and geocentric latitude 9 are connected by the following formula:

b2
tan 0 = ---:a-2--·tan <p
(1- f}2 tan.fJ

= (1- e) 2 tan0
d. Difference between Geodetic and Geocentric Latitude. There is no
difference between geodetic and geocentric latitudes at the Equator and the poles,
but elsewhere there is a non-zero difference between the two. This difference is at
its greatest when cp = 45°. When using WGS 84, the maximum difference is 11.54
minutes of arc.

3-6
August 2018 Version 1
 BRd 45(1)

Fig 3-4. Geodetic and Geocentric Latitude

Geocentric Lat 9 K

A1I ~rY~ I

P1

0306. Parametric Latitude

Parametric latitude is another definition of latitude, which is relatively obscure in that
it is not used practically for navigation (it is not part of the science of chart projections for
example), but it is used mathematically for calculations on the reference ellipsoid. It is
therefore relevant to WECDIS and command systems which carry out navigational
calculations on the spheroid rather than on the displayed representation of the chart. Fig 3-5
below shows a meridional section of spheroid WPE and sphere WBE. The polar axis of
spheroid WPE is OP and its shape and size are defined by the radii OE a and OP b.= =
WBE is the meridional section of a sphere with centre 0, polar axis 08 and radii OE= OB =
a (hence a sphere). Mis a point on the spheroid with geodetic latitude <P. HM is parallel to
OP and produced to cut the circle WBE at U. The radius OU makes an angle /3 with the X
axis.

a. Definition of Parametric Latitude. Angle /3 is the parametric latitude

(sometimes referred to as reduced latitude) of point M. The parametric latitude is
often used for geodesic calculations on the spheroid.

b. Connection between Geodetic Latitude and Parametric Latitude. The

geodetic latitude cp and parametric latitude /3 are connected by the following formula:

b
tan/3' = a~tan({l

The difference between the geodetic latitude and parametric latitude is zero at the
Equator and at the poles, and has greatest value when <P= 45°. Using the WGS 84
spheroid, where f = 1/298.257223563, the greatest value is approximately 5.85

minutes of arc.

3-7
August 2018 Version 1
BRd 45(1)

Fig 3-5. Parametric Latitude.

W ( ±~-a 1H JE
Geodetic Lat 0,

Pl

0307. Length of 1' in Latitude and Longitude

As explained in Para 0207, the length of the sea mile varies between the Equator
and the poles because of the Earth's changing radius of curvature.

a. Latitude Formula. The length of one minute of latitude may be found from the
formula po<P where p is the circular radius of curvature in the meridian and o(f) is a
small increase (measure in radians) in the geodetic latitude (f)_ It may be shown
that:

a(l - e) 2
p = (1 - e 2 sin2 q> ) 3 12

When o<P = 1 minute of arc:

a(l - e)2
1; of latitude
(1 - e.r2 ~ini <p ) 312 sin 1'
When <P = zero:

1'' of la:titude at the Equatar =cz(l - e2}sin '1 1

3-8
August 2018 Version 1
 BRd 45(1)

Fig 3-6. Length of One Minute of Latitude

p
p'de

WI nl /~\ IE
Geodetic Lau,

P,

8
Note. Position C is the centre of a circle and arc pd<P is part of its
circumference. This on the spheroid (reference ellipsoid), the position of C
will move relative to O as arc E_d<P moves over the surface of the spheroid.

b. Longitude. 1' of longitude at the Equator= a sin 11• (minutes of arc). Thus, at
latitude <P:

a cos <p . ,
1' of longitude =-(1--e--;:;.2;-s-:-in---:;2;-<P:-:;);-;-1n.12 sin 1

0308. Geodesic
Just as a great circle gives the shortest distance between two points on the surface
of a sphere, a geodesic gives the shortest distance between two points on the spheroidal
Earth.

3-9
August 2018 Version 1
 SECTION 2- DATUMS

0309. Datums
Charted data, both vertical and horizontal, must be referenced to a datum. Charted
depths and heights are referenced to a vertical datum, and this subject is covered in more
detail in Chapter 11 . Horizontal data is also reference to a datum; the horizontal or geodetic
datum refers to the spheroidal shape of the Earth used to establish the positional framework
for that chart.

0310. Geodetic Datums

Strictly speaking, a position given solely in terms of latitude and longitude does not
define a position on the surface of the Earth unambiguously. To achieve this, it is also
necessary to quote the geodetic datum and the spheroid used to reference the position .
Differences between different datums can be significant; Fig 3-7 below illustrates the
physical difference on the Earth's surface between the same latitude and longitude reference
to two different geodetic datums.

Fig 3-7. A position Referred to Different Datums

-• "!!!l

l--'......,___.__.,..,,,,_..__.,,,-_,......, a,
.
S 11 9 \ \
----=..,, -,--o;-•.._...1
1°,
I.
When referenced to OSGB 36, position 'A' (S0° 21.S'N 004° 10.0'W) is 101 metres
from position 'B' (also so• 21.S'N 004° 10.0'W, when referenced to WGS 84 Datum)

a. Choosing a Geodetic Datum. Classically, a geodetic datum can be

considered as the surface to which positions are referred . A (horizontal) geodetic
datum by definition utilises a specific spheroid which best fits the geoid over the
area of interest. This is because the geodetic measurements (historically angles
obtained by theodolite and heights obtained by spirit levelling) are obtained with
respect to the local gravity surface which, as has been shown is the geoid (by
definition). Since calculations are performed on the spheroid, it is necessary that the
spheroid and geoid are closely aligned in order to minimise errors in calculation .

3-10
August 2018 Version 1
 BRd 45(1)

b. Local Geodetic Datums. In the simple case of a local geodetic datum, an

accurate position may be defined by precise astronomic observations (although will
likely by GPS today), with a spheroid chosen to fit the geoid exactly at this point.
Local geodetic datums only deal with discrete, bounded parts of the Earth's surface
and are only loosely related to the Earth's centre of mass.

c. Classical Regional Geodetic Datums. For 'classical' regional geodetic

networks (i.e. those devised prior to GNSS), the differences between observed
astronomic position and geodetic position computed through the triangulation
network at several points may be minimised, thus ensuring an overall minimal
separation between the geoid and the spheroid. As with a local geodetic datum,
classical regional geodetic datums only deal with discrete parts of the Earth's
surface, and are only loosely related to the Earth's centre of mass.

Note. A convenient analogy for local or regional geodetic datums is to

fi
consider the chosen spheroid as a rugby ball, and the local (or regional)
geodetic datum as a piece of orange peel placed over it where the position
and orientation fitted best. The precise observed position would then be a
pin stuck into the orange peel at the appropriate location. Elsewhere
however the oranqe oeel and ruqbv ball will be widelv seoarated.

d. The World Geodetic System (WGS). The World Geodetic System (WGS) is a
combined spheroid and datum. By design it is global in coverage, and is directly
related to the Earth's centre of mass; high accuracies in specific regions are
sacrificed in favour of a consistent accuracy across the entire planet.

(1) Concept. A true world geodetic datum/spheroid, fit for use with satellites
and satellite-derived positions, is defined by the system in which the satellite
orbital parameters are given. These in turn are dependent upon the precise co-
ordinates of the satellite tracking systems, the geopotential model of the Earth's
gravitational field, and a set of constants including GM (the Earth's mass M
times the gravitational constant G).

(2) WGS 84. From this satellite reference system data, a universal worldwide
spheroid and datum was established in 1984. The WGS 1984 (or almost
always just WGS 84) is the current system, and is NATO's preferred
spheroid/datum. It supersedes WGS 72, although some older model
navigational aids (such as legacy LORAN receivers in RN ships) may still
operate in WGS 72.

(3) Global Navigation Satellite Systems (GNSS). The adoption of WGS 84

has made it possible to define absolute position anywhere in the world by
common geodetic latitude and longitude, for use with GNSS. Further it enables
the use of WECDIS and electronic navigation charts, which are referenced to
WGS 84.

3-11
August 2018 Version 1
BRd 45(1)
e. Modern Regional Datums (Tied to GRS 80/WGS 84). In the modern era,
rather than rely upon networks of triangulation points ('trig' points, the familiar
concrete markers found on the tops of many hills in the UK), it is possible and
indeed common place to use satellite positioning systems with centimetric accuracy.
One of the consequences of the ubiquity and accuracy of this technology is that the
movement of the tectonic plates becomes both accurately detectable, and
geodetically significant. To account for this, regional geodetic datums tied to WGS
84 have been established at a specific time epoch, and then held fixed in relation to
the tectonic plate for the region . Whilst for many practical navigational purposes , the
modern regional datums may be considered to be equivalent to WGS 84, although
after three decades (or more) of divergence between the two then noticeable
differences are detectable even at mid-accuracy applications (precision farming, civil
engineering), so that the appropriate regional datum must be used (for example) for
terrestrial surveys. Provided the GNSS receiver and WECDIS are configured
correctly then this should be invisible to the operator, but may cause problems in
some circumstances.

a
Note. Strictly, many of the regional datums are based on the GRS (Geodetic
Reference System) 80 ellipsoid. WGS 84 was originally based on the GRS
80 ellipsoid when first defined, although since then the underlying parameters
of the two have drifted very slightly as WGS 84 has been refined. The
difference between the GRS 80 and WGS 84 spheroids amount to just over
0. 1mm on the semi-polar axis.

(1) ETRS 89 Regional Datum. The European tectonic plate is moving at

approximately 2.5cm per year in relation to WGS 84. The European Terrestrial
Reference System (ETRS) 89 was defined in 1989, at which point the
fundamental co-ordinates were fixed with respect to GRS 80 and held stable.
Due to the movement of the tectonic plate, WGS 84 and ETRS 89 have
diverged and in 2017 the differences between WGS 84 and ETRS 89 co-
ordinates of the same point were about 70 cm. ETRS 89 is the preferred
geodetic datum to be used for surveying and mapping purposes in Europe
including the UK, and European GNSS position data is theoretically tied to
ETRS. One advantage of using ETRS 89 is that it provides higher accuracies in
Europe, and ETRS 89 positions remain fixed whereas, for maximum accuracy,
WGS 84 positions would need adjustment for secular change.

(2) Other Regional Datums. A number of other regional geodetic datums

have been established or proposed, including NAD 83 (North America), AFREF
(Africa) and SIRGAS (South America). Whilst at the time of definition these
reference frames would (will) be broadly equivalent to WGS 84 (GRS 80) by
definition, increasingly accurate satellite-based survey data, and secular
movement of the tectonic plates, has meant that for the longer-established
datums such as NAD 83 significant differences can exist between the regional
datum and WGS 84. This can be more than a metre for NAD 83.

3-12
August 2018 Ver!':inn 1
 BRd 45(1)

0311. Geodetic Datums Worldwide

As discussed above, WGS 84 is a global reference framework (TRF), which is
designed for use anywhere on Earth with a certain degree of accuracy. However, in order to
improve accuracies in smaller regions of the world (countries or continents), other datums
(spheroids) are adopted. These may be linked to co-ordinate systems as well such as the
National Grid in the UK.

a. Common Geodetic Datums. Some of the more common geodetic datums,

their associated spheroids, and areas of application are detailed in 'fable 3-1 below.
Some of these datums and/or spheroids are old but remain valid, and even where
they have been superseded they may still be encountered on older navigational
products. The Airy 1830 spheroid remains the best-fit for the UK for example. The
International (1924) Spheroid is used for the calculation of distances in the UKHO
publications NP 350 (Admiralty Distance Tables) and NP 136 (Ocean Passages for
the World). All calculations in this book are based on WGS 84, as are those in
WECDIS and NavPac.

b. Geodetic Datums Worldwide. NATO lists those geodetic datums and

spheroids which may be applicable for mapping and charting products within the
area of NATO interest (and hence is also of interest for naval operations in support
of land-based activities such as naval fire support or littoral manoeuvre). This
information is available through NATO publications, relevant RN publications
including the relevant Gun System Automation (GSA) BRds (which contain details
of gunnery practice ranges worldwide including geodetic information), and if
necessary by reference to NCHQ/UKHO (geodesy@ukho.gov.uk). More generally,
UKHO maintains further details on the matter in ADRS2 (ALRS Vol 2), including the
number of UKHO products based on each datum.

3-13
August 2018 Version 1
JRO 4!>\lJ

Table 3-1. Common Datums, Associates Spheroids, and Areas of Application

Datum Spheroid (a) Equatorial Flattening (f) & Eccentricity (e) Remarks
& (Type) Radius Reciprocal (1/f) & Squared (e2)
(b} Polar Radius f = (a-h)la e2 = (a2-b2)1a2

)SGB 36 Airy (a) 6,377,563.396m (f) 0.003340851 (e) 0.081673374 Some remaining
Regional) 1830 (b) 6,356,256.909m (1/f) 299.324964 (e 2) 0.006670540 UK charts.

)SSN 80 Airy (a) 6,377,563.396m (f) 0.003340851 (e) 0.081673374 Not for charts.
Regional) 1830 (b) 6,356,256.909m (1/f) 299.3249645 (e 2) 0.006670540 OSGB 36 related.

~D 50 International (a) 6,378,388m (f) 0.003367003 (e) 0.081991890 Some remaining

:Regional) 1924 (b) 6,356,911.946m (1/f) 297 (e 2 ) 0.006722670 European charts .

NAD27 Clarke (a) 6,378,206.4m (f) 0.003390075 (e) 0.082271854 Some remaining
(Regional) 1866 (b) 6,356,583.8m (1/f) 294.978698 (e 2 ) 0.006768658 USA charts.

Arc Clarke (a) 6,378,249m (f) 0.003407561 (e) 0.0824834 Some remaining
(Regional) 1880 (b) 6,356,515m (1/f) 293.465 (e 2 ) 0.006803511 S. African charts.

WGS72 WGS72 (a) 6,378,135m (f) 0.003352779 (e) 0.081818811 Being replaced by
(W'ldwide) (1972) (b) 6,356,750.520m (1/f) 298.26 (e 2 ) 0.006694318 WGS84.

NAD83 GRS80 (a) 6,378,137m (f) 0.00335281068 (e) 0.0818191910 Regional WGS 84
(Regional) (1980) (b) 6,356,752.3141m (1/f) 298.25722210 (e 2) 0.0066943800 for N. America
ETRS83 GRS80 (a) 6,378,137m (f) 0.00335281068 (e) 0.0818191910 Regional WGS 84
(Regional) (1980) (b) 6,356,752.3141m (1/f) 298.25722210 (e 2J 0.0066943800 for Europe
WGS84 WGS84 (a) 6,378,137m (f) 0.00335281066 (e) 0.0818191908 WGS 84 is NATO
(W'ldwide) (1984) (b) 6,356,752.3142m (1/f) 298.257223563 (e 2) 0.0066943800 preferred Datum.

0312.
ii Note. The defined parameters are 'a' and '1/f; all others are derived values.

Datum Shifts

a. Datum Shifts. Where two adjacent geodetic datums overlap, the same point
on the Earth's surface will have two distinct sets of geodetic co-ordinates, one in
relation to each geodetic datum. The difference between the two sets of co-
ordinates is called a 'datum shift'. The datum shift may not be consistent in
magnitude and direction (i.e. it is likely to very with geographic location).

b. Datum Shift Information. Datum shift information is encapsulated in ENCS

and RNCs and is applied automatically where known in WECDIS. It can be
displayed manually by displaying the 'Satellite-Derived Positions' and/or
'Overlapping Charts' notes. Some charts in certain parts of the world remain for
which the geodetic datum is unknown; such products cannot be used with GNSS
and relative navigation procedures must be used. WECDIS will issue an alarm
when the ship is being navigated on a product with no known datum .

., 4.
 BRd 45(1)

Table 3-2. Examples of Common Datum Shifts

DATUM DATUM DEVON PORT ROSYTH

FROM TO LAT LONG LAT LONG
WGS84 WGS72 -0.10N -0.55E
-0 .55E -0.09N
WGS84 ED 50 +3.40N +5.16E +2.75N +5.82E
WGS84 OSGB 36 -2 .22N +4.22E +0.24N +5.06E
OSGB 36 ED 50 +5.62N +0.94E +2.51N +0.76E

0
0313.
Note. See also ADRS which provides further detail on this subject, including
the number of UKHO products currently based on each datum.

Datum Transformations
Several techniques exist to transform positional variety across a large area from one
geodetic datum to another (most commonly in the navigational context into WGS 84). These
are different from shifts, which apply to points or small areas, in that they operate across a
wide range of space and time. These are complicated operations which will not be
conducted manually, but can be carried out using appropriate software. It is rare that the
practical navigator will have to carry out such a transformation, but it is conceivable that the
NO might need to understand that such a transformation is required and request it from the
appropriate source.

a. Non-Homogenous Geodetic Datums. Some earlier local or regional geodetic

datums, such as OSGB 36, contain internal inconsistencies in scale and azimuth,
and are thus known as 'non-homogenous geodetic datums'. These inconsistencies
may be significant for situations where precise navigation is required if calculating
position transformations to other geodetic datums, especially if using automatic
systems.

b. Published Geographical Shifts. When the datum shift has been determined
by comparison of positions common to both datums, a simple block shift in
geographical coordinate values may be applied. As a result, many charts referred to
datums other than WGS 84 now show the datum shift between WGS 84 and the
horizontal datum in which the chart is published. These datum shifts are quoted to
an accuracy commensurate with the scale of that particular chart and should not
produce an error capable of being plotted on that chart at that scale.

3-15
August 2018 Version 1
BRd 45(1)

c. Datum Conversions. Full details may be found in NATO STANAG 2211 which
mandates that AGeoP-21 is the authoritative NATO document. STANAG 2211/
AGeoP-21 is available in the public domain via the NATO Standardisation Office
(NSO) website (https://nso.nato.inUnso/). To transform positions between different
geodetic datums, the geodetic datum coordinates of latitude and longitude are first
converted into three-dimensional Cartesian coordinates. Using the Molodensky
equations, translations along the X, Y and Z axes, rotations round these axes and a
scale factor are applied to the Cartesian coordinates. The modified Cartesian
coordinates are then converted into the second geodetic datum coordinates of
latitude and longitude.

(1) Homogenous Geodetic Datums. For a homogenous geodetic datum, an

average set of values for X, Y and Z axis rotation will provide sufficient accuracy
in datum transformation. This process is easily automated.

(2) Non-Homogenous Geodetic Datums. For non-homogenous geodetic

datums, average values will not always be adequate. Two principal methods
are commonly available to overcome thus. One method is to used localised
values for X, Y and Z axis rotations. This method suffers from the
disadvantage that no single value can be used in automated systems, and that
it must interrogate a digital record of the contours. These values should never
be extrapolated. Multiple regression equations (MRE) are appropriate for
determining datum shifts for land-based applications, but are not a valid
technique for maritime applications since MREs become rapidly unstable
outside the area for which they were defined.

3-16
August 2018 Version 1
 BRd 45(1)

SECTION 3 - EARTH MODELS FOR NAVIGATION SYSTEMS

0314. Types of Earth Model

a. Principal Types of Earth Model. Naval systems concerned with what could be
termed 'navigational matters' (so WECDIS, command systems, fire control systems
etc)_ aU require an Earth model of some sort to be used in order to carry out
navigational calculations (whether of own ship movement, aircraft movement,
missile engagements etc). These fall into three broad categories; one could also
consider 'inertial space' as well although movements in inertial space would
inevitably have to then be converted into an Earth-referenced model in order to have
meaning for the navigator.

(1) A Spheroidal Earth. As discussed above, a spheroidal Earth model

assumes that the Earth has the characteristics of an oblate spheroid, as defined
by the various parameters.

(2) A Spherical Earth. As discussed above, a spherical Earth model

assumed that the Earth is a true sphere.

(3) Flat Earth or Tangential Plane. In a 'flat Earth' or 'tangential plane'

model, the Earth is assumed to be flat and all calculations are carried out on
that plane. Whilst clearly not suitable for modelling the globe as a whole, for
applications involving short distances from the ship a flat Earth may be an
acceptable generalisation, especially if available computing power is limited.

b. Choice of Earth Models. The choice of which Earth model to use is ultimately
driven by the purpose of the system, and hence the required accuracy, as well as
limits on the computational power available (although in practice this constraint
probably no longer exists today, but was a factor when legacy systems were
designed). As discussed at length above the spheroidal Earth model is the closest
approximation to the true Earth shape, and hence it is the most suitable surface on
which to base precise calculations for navigation, especially if they concern locations
at long ranges from the ship. Other Earth models are less accurate approximations
of the 'ground truth' and will give progressively less accurate results, especially
when dealing with long-range contacts beyond the horizon. The size of these errors
should be compared with those of the spheroidal Earth model.

3-17
August 2018 Version 1
BRd 45(1)

0315. Spheroidal Earth Models .

Within the general class of spheroidal Earth models, as has been explained above,
a number of different spheroids have been defined, each for a distinct purpose. It is
therefore essential to know not just that the navigation system uses a spheroidal Earth, but
which spheroid.

a. WGS 84. By default, the most commonly encountered spheroidal Earth model
in the Royal Navy will be WGS 84. This spheroid, based on a modified GRS 80
spheroid with an integrated gravity model, is a fully realised terrestrial reference
frame (TRF) and is the TRF on which GPS is based. In addition to being the basis
of GPS, WGS 84 is the NATO-preferred geodetic datum and spheroid for mapping
and charting products in NATO maritime areas, and it was adopted for all RN
applications in 1991. Furthermore, WGS 84 has been adopted by the IMO for
ECDIS, ECS, and AIS, and all electronic navigation charts (ENCs) and raster
navigation charts (RNCs) used with them.

b. Differences between Spheroids. The correct spheroid associated with the

geodetic datum to which geodetic coordinates are referred must always be used in
all calculations, otherwise large and unexpected errors can be introduced. A
common case in which this will be encountered is during naval gunfire support,
during which geodetic latitude and longitude coordinates may require to be
converted to grid coordinates in order to be compatible with land maps used by
spotters ashore. The example below gives an indication of the sorts of errors which
can occur if incorrect datum/spheroid information is applied to geodetic/grid
conversions.

Example. During a medium range gunnery exercise, a target position on the North East
corner of Garvie Island at 58° 37.09'N 004° 52.20'W (WGS 84) translates into a British
National Grid position of 233418E 973594N when using the correct Airy (1830) spheroid and
OSGB 36 datum. Fleet ARCS products are available with overprinted grids for certain
areas, but otherwise the transformation must be carried out mathematically. Some
transformation facilities are found within WECDIS (and may be available in other ECDIS
types), and are also available via the Fleet gunnery calculator. If the wrong spheroid/datum
combination is selected for the transformation then a significant error can occur; for example,
using the example of the target on Garvie Island, if the International Spheroid (1924) is used
instead of Airy (1830) the target's British National Grid position will be plotted 19 metres East
and 288 metres North of the correct position. Clearly, this sort of error is not acceptable
during gunnery, and so the utmost caution must be exercised when carrying out these sorts
of conversions (whether using WECDIS, the command system/gun system or the gunnery
calculator).

3-18
 BRd 45(1)

0316. Spherical Earth Models

Spherical Earth models, in which the Earth is assumed to be perfectly spherical ,
enable positional calculations to be made based upon straightforward (for a computer)
spherical trigonometry. The errors introduced by this assumption and associated calculation
methods are principally, but not exclusively, dependent upon the adopted value of the radius
of the spherical Earth, the latitude of own ship, and the distance and direction to the contact.
Table 3-3 below indicates the likely maximum size of error introduced by the use of a
spherical Earth model, by comparison with WGS 84, when dealing with contacts at long
range (other sources of error such as gyro error, radar range error etc are ignored).

Table 3-3. Errors Arising from a Spherical Earth Calculation compared with WGS 84

RANGE OF CONTACT POSITIONAL ERROR POSITIONAL ERR©R

(n. miles) AT LATITUDE 0° AT LATITUDE 75°
(n. miles\ (n. miles)
50 0.1 0.3
100 0.1 0.7
250 0.3 1.7
500 0.5 3.4
1000 1.1 6.5
2000 1.9 11.3

a
Note. In the above calculations and results, no account has been taken of
other sources of error. In particular, no account has been taken of the height
above (or below) the reference surface of either own ship or the contact.
This in itself will lead to further range errors, in addition to those directly due
to the use of approximate Earth mod~ls.

0317. Flat Earth Models

a. Flat Earth Assumptions. Flat Earth models assume a plane Earth, which then
enables the use of a two-dimensional system of Cartesian coordinates. These
Cartesian coordinates have axes arranged to lie along true (or theoretically
magnetic) North and due East from an arbitrarily chosen grid origin. This grid origin
will be chosen based on a number of factors, which may not always coincide with
own ship's position.

b. Errors Associated with Flat Earth Assumption. The errors introduced by the
flat Earth assumption are dependent upon the specific algorithms implemented in
the particular system. However, in general it will be deduced that the size of the
error will be a function of both the distance from the (arbitrarily selected) grid origin,
and the latitude of the grid origin. Thus errors in bearing and distance increase with
distance from the grid origin and with increase in the latitude of the grid origin. For
close ranges from own ship, the errors will either be negligible or insignificant for
most purposes.

3-19
Auam;t 2018 Version 1