BRd 45(1)

CHAPTER4

PROJECTIONS AND GRIDS

Para
0401. Introduction and Scope of Chapter
0402. The Requirement for Projections and Grids

SECTION 1 - PROJECTION CONCEPTS AND PRINCIPLES

0403. Displaying a Three-Dimensional Object in Two-Dimensions - What is a

Projection?
0404. Distortion
04 05. Graticules and Grids
040 6. Spherical Projections
04 07. Spheroidal Projections

SECTION 2 - CYLINDRICAL PROJECTIONS

0408. Principles of Cylindrical Projections

0409. Mercator's Projection (the Mercator Projection)
0410. Equidistant Cylindrical Projection
0411. Mathematics of the Mercator Projection
0412. Meridional Parts of a Mercator Chart
0413. Small Scale Mercator Charts
0414. Large Scale Mercator Charts
0415. Great Circle Tracks on a Cylindrical Projection
0416. Summary of Differences between Mercator and Equidistant Cylindrical
Projection

SECTION 3-TRANSVERSE MERCATOR PROJECTION

0417. Concept
0418. Principles
0419. Uses

SECTION 4 - STEREOGRAPHIC PROJECTION

0420. Concept
0421 . Principles
0422. Uses

SECTION 5 - GNOMONIC PROJECTIONS

0423. Concept
Transfer of a Great Circle Track from a Gnomonic to a Mercator Projection
0424.
Chart ,
0425. Composite Tracks

4-1
August 2018 Version 1
BRd 45(1)

SECTION 6 - GRIDS

0426. Grid Reference Systems

0427. UTM Grid
0428. British (Ordnance Survey) National Grid
0429. Other National Grids
0430. Grivation

4-2
August 2018 Version 1
 BRd 45(1)
CHAPTER4

PROJECTIONS AND GRIDS

0401. Introduction and Scope of Chapter

Having discussed the concept of geodesy in Chapter 3, it is necessary to discuss the
next step which includes 'projections' and 'grids'. The spheroidal Earth model, for example,
is mathematically perfectly suited to navigational calculations, but in order to physically display
a portion of the Earth's surface a projection is required. This chapter describes the conceptual
underpinnings of projections in general, before introducing the projections most likely to be
encountered in the Royal Navy. Closely linked to the projection is the concept of grids, which
are discussed in Section 6.

0402. The Requirement for Projections and Grids

Whether it is a spherical or a spheroidal Earth model which is used, the surface being
represented is something that approximates a 'ball' to use a non-technical term. Unless or
until holographic projectors are developed, parts of this 'ball' need to be displayed on a flat,
two-dimensional surface, whether it be a paper chart historically or, now, a computer display.
Projections are the mathematical technique used to represent these parts of a spherical
surface in two-dimensions. Grids are a separate but related concept, in that they are required
to be able to make proper use of a projection. Grids are the two-dimensional overl,;iy placed
on a chart projection and used to create a co-ordinate system such that positions may be
readily identified. Latitude and longitude, strictly speaking, are not a true grid as they are
Cartesian co-ordinates but rather angular measurements.

Note. It must be remembered that, whilst projections are still critical in that

0 they determine the appearance of the chart on the WECDIS screen, in reality
most if not all of the navigation calculations are in fact carried out on the
reference ~heroid used by WECDIS.

4-3
 BRd45(1)

SECTION 1 - PROJECTION CONCEPTS AND PRINCIPLES

0403. Displaying a Three-Dimensional Object in Two Dimensions - What is a

Projection?

a. The Requirement for Projections. The Earth is a three-dimensional oblate

spheroid, whereas a computer display is a two-dimensional plane. A projection is a
method of representing a spheroidal surface on a plane. It is usually expressed as a
mathematical formula for converting spheroidal geographical co-ordinates to planar
co-ordinates on charts or maps. Suitable projections are available to represent any
portion of the Earth's surface, although not all projections are equally suited to every
region. The generally accepted analogy is to think of a projection as the results of
shining a light from inside the spheroid onto a flat surface, with the resulting shadows
from land masses etc forming the lines on a chart. Different projections arise from
where the hypothetical light source is placed, where it is pointed and so forth.

b. Distortion. It is physically impossible to turn a spheroidal surface into an exact

plane; for the proof, try wrapping a football (soccer ball) in wrapping paper. It will not
be possible to wrap the ball using a single sheet of paper without overlaps, wrinkles,
or gaps appearing. Thus, except over very small areas, whenever a projection is used
it will contain some degree of distortion. When the outline of three identia circular
areas from different latitudes on the Earth's surface are each projected from an origin
at the centre. of the Earth onto a plane chart, they occupy areas of quite different
shapes and sizes on the projection. This is demonstrated in Fig 4-1 below.

Fig 4-1. Distortion of Areas of the Earth's Surface on a Projection

--
The Earth ' ' ,
E

Plane of a Chart

4-4
August 2018 Version 1
 BRd 45(1)
0404. Distortion

a. Properties of Projections. As a spheroidal surface cannot be exactly

represented on a two-dimensional surface, all projections involve some degree of
distortion. The distortion will by no means be the same across the chart, but may vary
considerably from one region on the chart to another. The distortion will affect one,
some or all of the following interrelated properties which are fundamental aspects of
all projections:

SHAPE

BEARING

SCALE

AREA

b. Choice of Projection. When devising a projection it is possible to design it to

optimise, by reducing or eliminating one or more of the distortions, whilst accepting
that other aspects of distortion will remain but remain within reasonable limits (and
thus be usable in a practical sense). The choice of projection for a given purpose is
thus driven largely by that purpose, whilst acknowledging that in reality a modern
navigational system such as WECDIS in fact calculates its navigation solutions on the
reference spheroid and not on the surface of the chart. However, choice of projection
remains a critical human elements decision, as in particular it is essential that a
'glance' at the WECDIS screen provides information that is accurate and not
misleading (as to range or bearing for example).

c. Mariner's Chart Projection Property Requirements. Noting the above

remarks about calculations on the reference spheroid in WECDIS, mariners have a
well understood set of requirements for what they need from a maritime chart. In
particular, for obvious reasons, presentation of the correct area of land on a chart is
not usually of major importance (although other users may have a different view).
Ideally, however, the mariner requires a chart which will provi~e the correct shape of
land, and correct bearings and distances (scale). It is ~ot possible to meet these th~ee
requirements in one single projection, and a compromise must be made, b~ acc~ptmg
· t·ion to all three (shape , bearing and
a very c Iose approx1ma . scale), or satisfaction of
two (usually shape and bearing) at the expense of the third (scale).

• Th k requirements that mariners have, described above,

d Orthomorph1sm. e ey • h" h · rth h" (
· . ,Orthomorphism'; a projection w 1c 1s o omorp 1c or
are usually described as . . ts shape scale and bearing are
. . h" h with certain cavea ' ,
conformal) 1s one in w IC ' erties are critical for mariner because, if distortion
0
correctly represented. These pr P r·ings or compass roses displayed on the
f example range .
of shape occurs, then or Id they have uniform graduations. It would be
screen would not be_circular,_ no~;~~e display in such a case. A non-orthomorphic
very difficult to read mformatio~m ossible to use. Mercator, transverse Me~cator,
st I
paper chart would be almo Prt' nical orthomorphic, skew orthomorph1c and
inverse (oblique) Mercator, Lambe sh~o With an orthomorphic chart projection, the
stereographic charts are aII orthomorp ic.
below characteristics apply.

4-5
Auaust 2018 Version 1
BRd 45(1)
(1) Shape. The shape of the land correctly represents that of the Earth's
surface, at least over small areas.

(2) Scale. At any specific point on that chart the scale is the same in all
directions. However, the same scale may not apply across the whole chart.

(3) [Link]. The parallel of latitude and meridian of longitude at any point are
at right angles to each other. Angles around any point on that chart, and hence
bearings, are correctly represented.

e. Orthomorphism Compromises. Due to the fundamental characteristics of all

projections, no projection meets the mariner's ideal requirement for perfect shape,
bearing and scale over large areas. On a Mercator chart of the world, the area around
Cape Farewell (Greenland) is just as correctly shown for shape as is the Amazon
Estuary (South America), but Greenland as a whole is shown the about the same area
as South America, whereas it is actually about one-tenth . This is because the scale
in the near-polar Greenland area is quite different from the scale being used for
(equatorial) South America on the same chart, and this affects area properties. The
change in scale can also been seen in the expanding intervals towards the poles
between the parallels of latitude. Whilst this does not particularly matter on an
electronic display, when using a paper chart of this scale it is essential that distance
is read off the latitude scale parallel with the point where the distance is being
measured.

Fig 4-2. Chart of the World - Mercator Projection

SO-W ll• rldlanofO-(GrNnwtch) 60"E 120-E 180-E/W 1ZO-W 60'

100
bQ:,?I I rtL I \(j,_ I I I <!4.4..Jd ,I

10'
10'

,.. Irc:;_;::p1:t -..cAPE ilj I~bsL I I / 9f=<'.i ·,<,,r,-.. I (~·w.J

...
·m
FAREWELL ' 80"

~vj -_-I--- =dl

rK·J
50"

:~ -~ ~ (r.~~-
IO"

~ J "
10'

~l..] W 1
v~·
1 · 31
-20"
I
·o-

10'
!O'
10'
1 ..

~r
IO"

10'
1 . :=± . I ;t I I
50"

60'

75'i 'I
SO- W
/ I
Meridian of 0" (Gl'ftnWlch)
I
60" E
I
120- E
( I
1Ano c • ....
I 10'

75•
120'W 60'W

4-6
August 2018 Version 1
 BRd 45(1)

f. The Flat Earth. Over a limited area (up to about 12 miles radius from a point)
the Earth may be assumed to be flat for all practical purposes, as the errors introduced
by this assumption are less than those resulting from the measurement of angles and
distances. A simple sketch survey plan over a small area may be constructed on a
flat Earth principle may be transferring observed ranges and bearings directly to a
sheet of squared paper or a WECDIS plotting sheet. However, at a distance of 50
miles from a given point, the errors introduced by assuming a flat Earth become
navigationally and operationally significant (and increase rapidly with distance
thereafter).

0405. Graticules and Grids

a. Graticules. graticule is the name given to the network of lines representing

the parallels of latitude and meridians of longitude in a projection . They will probably
not be shown on an electronic chart, but will absolutely be printed on a paper chart.

b. Grids. A grid is name given to the reference system of square or rectangular

Cartesian coordinates obtained when a projection is applied either to part or the whole
of the world . Grids are described in detail in Section 6.

0406. Spherical Projections

If the Earth were indeed a sphere, projections could be created by wrapping a plane
surface around an imaginary sphere, switching on a light at its centre (or any other position)
and projecting features from the Earth's surface onto the plane. Whilst the Earth is in fact
spheroidal , and different methods are used in practice, the above analogy serves to explain
the concepts underpinning the projections described below, and their various properties.

a. Tangent and Secant Projections. Figure 4-3 below shows six common
examples of projections. They can be categorised into two distinct types: tangent and
secant projections.

(1) Tangent Projections. In the left column of Fig 4-3 (Examples 1, 3 and 5),
the surfaces all touch the sphere along a circle or at a point. These are known
as tangent projections.

(2) Secant Projections. In the right column of Fig 4-3 (Examples 2, 4 and 6),
the plane surfaces have been sunk into the sphere. In examples 2 and 4_the! cut
the sphere twice, and in example 6 once. These are known as secant pro1ectIons.

b. A reas of Accuracy and 0 ·1stortion · If the detail on the

. spheres in Fig .4-3 are
• ..,ace from a point on the axis of the one, cylinder or
now proiected onto a p1ane su, ,, . .
• -11 b d"istortion of scale along the lines where contact Is made
plane circle there wI e no . . .
. ' b Id ey-filled dashed lines). Elsewhere there Is d1stort1on
with the sphere (shown by O , gr , d' d 1 ·d
, h" h . remain when the planes are unwrappe an aI
of some sort or another, w ic wi 11
flat.
. . exam Jes 1 through 4 of Fig 4-3, the point from which
c. Points of ProJection._ In ~he centre of the sphere. In examples 5 and 6 it
11
the projection takes place is usua Y x·is at right angles to the plane, but it is usually
nywhere on an a .
may take place f rom a (B) or from the opposite pole A. By convention ,
from either the _cen_tre of the sp~:r;ollowing names:
the projections in Fig 4-3 have t
4-7
A . ,,. 11rof. ?n1 A. \/o.r~inn 1
 ---
BRd 45(1)

(1) Example 1. Example 1 is a conical projection with one standard parallel.

(2) Example 2. Example 2 is a conical projection with two standard parallels.

(3) Example 3. Example 3 is a cylindrical projection with one standard parallel

at the Equator.

(4) Example 4. Example 4 is a cylindrical projection with two standard parallels.

(5) Example 5. Example 5 is a 'zenithal' projection from A, more usually known

as a stereographic projection (or polar stereographic).

(6) Example 6. Strictly, example 6 is a zenithal projection from B, but it is far

more commonly known in the RN as a gnomonic projection.

d. Orientation of Cones, Cylinders and Plane Circles. Although for ease of

understanding the example in Fig 4-3 are all shown as 'North up', there is no
mathematical, trigonometric or cartographical reason why this has to be the case, and
in practice the cones, cylinders and plane circles could equally well be inclined at any
angle to the Earth's axis. Many specialist or regional products are drawn on this basis
for a specific purpose.
 BRd 45(1)

Fig 4-3. Common Spherical Projections

Example 1 {Tangent) Example 2 (Secant)

Conlcal with One Standard Parallel Conical with Two Standard Parallels

----- --~
Example 4 /Secantl
Example 3 !Tangent) Cylindrical with Two Standard Parallels
Cyllndrlcal with One Standard Parallel
at the Equator

--------~-B ---

A
A Example 6 (Secant)
Example s {Tangent) Zenlthal - Projected from A: Stereographic
Zenlthal - Projected from A: Stereographic Zenlthal - Projected from B: Gnonomlc
Zenithal - Projected from B: Gnonomlc

A _O
BRd 45(1)

0407. Spheroidal Projections

None of the so-called 'perspective' projections illustrated in Fig 4-3 are orthomorphic
for the sphere (with the exception of examples 5 and 6 when projected from A), and none are
orthomorphic for the spheroid (and hence the Earth). Whilst they may have value for other
uses, none of the spherical projections are suitable for navigation at sea due to their lack of
one of the fundamental properties required. To correct for this deficiency, a number of
'mathematical' projections have been devised over the centuries, which are analogous to
those described in Para 0406 but have their formulae adjusted to ensure that some have
orthomorphic properties, and others 'equal areas', as required for their intended use. The
main types of spheroidal projections are summarised in the rest of this paragraph and in Table
4-1 below; full explanations are in Section 2 to Section 5 of this Chapter.

4-10
August 2018 VAr<::inn
 BRd 45(1)

Table 4-1. Spheroidal Projections

tPRC!>.:J1fo,:-,10N TYPE C!IRilil:IQM9RP1:11c ·,i:iGRios~l "1ii'seo 1USEE> · 'REMARKS

liOR "-'.' IFQR
'Sl!JR\f!;Y.S•YGJi:!~~JiS
Lamberts conical I 'Yes 'Very many, Yes Yes but Used for aeronautical
orthomorphic with one or rare charts. Unsuitable at
two standard high latitudes where
parallels modified Lambert
conformal used
instead. Superseded
by UTM for land
ma in .
Mercator I IYes !Not often Yes , on Yes Used for most small
used. small scale BA charts
scales
Transverse IYes Yes Yes Used for large scale
Mercator I charts
National Yes No
Grid of GB
UTM IYes INo Worldwide between
84°N and 80°S
Many others Yes No
Skew !Malaysia Yes No Mainly land surveys,
IYes
orthomorphic Borneo reduces scale errors
I
Madagascar over areas covered
by grids
Inverse Yes Not often No Yes Excellent for air
(oblique) used charts of polar
Mercator regions. Closely
related to transverse
Mercator and skew
orthomor hie
Stereographic !Yes !Universal Yes Yes (polar Good for surveys
I
polar charts) between 80° and
sterographic the poles. Not so
(polar grid) useful for small
scale charts
IYes 1used for small scale
Gnomonic I INo !Not used INo
great circle charts
IPolyconic Yes The so-called
Polyconic I INo IYes
projection gnomonic of old
tables large-scale UKHO
plans. Largely
superseded by
[Link]

Yes No Yes Default projection

Equidistant NO used outside of
cylindrical polar regions in
RN WECDIS (Outfit
4LAA)

4-11
Aun,u:t ?01R V"'rcinn 1
BRd 45(1)

a. Lambert's Conical Orthomorphic Projection. Lambert's conical orthomorphic

projection is a modification of the conical projection with one or two standard parallels.
It has the following properties.

(1) Orthomorphism. As its name suggests, this is an orthomorphic projection . The

parallels of latitude other than the standard parallels appear as arcs, concentric
with the standard parallels, but distances between them are chosen so that the
projection is orthomorphic

(2) Scale Manipulation. To be orthomorphic, the scale along the meridian at any
place must be equal to the scale along the parallel at that place. The scale along
the meridian cannot now be uniform but must be adjusted to the scale along the
parallels. The scale is correct only along the standard parallels; if there are two
of these, the scale is smaller between them and becomes increasingly large
outside. The latitude covered by the projection is limited so that the scale error
does not become unacceptable.

(3) Great Circles. Great circles are very nearly straight lines on this projection.

(4) Uses. This projection is widely adopted for aeronautical charts. It cannot be
used in high latitudes, where a modified Lambert's conformal is more appropriate
instead. It can be used for mapping countries with a large extent in longitude but
not much in latitude. However, it has largely been supersede for this purpose by
the universal transverse Mercator projection.

(5) Derivatives. The Mercator projection is a derivative of Lambert's conical

orthomorphic projection .

(5) Mercator Projection. The Mercator projection is probably the most widely
known projection in use, and is described in full detail in Paras 0409 to Para
0414. The projection's principal properties are described below.

(1) Scale Expansion. The scale units are minutes of longitude measured along
the Equator. The scale expands as latitude increases.

(2) Meridional Parts. The Cartesian coordinate grid is not normally shown,
although_accurate calculations are generally carried out in terms of meridional
parts which form the units of the grid.

.4.1?
 BRd 45(1)

(3) Properties. The Mercator projection is orthomorphic and is in fact a special

case of Lambert's conical orthomorphic projection in which the Equator is used
as the standard parallel. Rhumb lines on the Earth appear as straight lines on
the chat, and angles between rhumb lines are the same on the Earth's surface
and chart. The Equator (which is both rhumb line and great circle) appears as a
straight line. Parallels of latitude appear as straight lines parallel to the Equator,
and meridians appear as straight lines perpendicular to the Equator (use of a
meridional grid to plot Eastings/Northings is not required). A straight line on the
chart joining two points does NOT represent the shortest distance between them,
unless it is also a great circle. A great circle which is not a meridian or the Equator
appears as a curve. Increased spacing of adjacent parallels away from the
Equator leads to scale magnification which increased with latitude. The shape of
charted features is correct for small areas but that of large features is distorted
as latitude increases.

(4) Uses. The Mercator projection is extensively used for standard UKHO
nautical charts at scales smaller than 1:50,000, up to latitude ao N/80°S. 0

However, for large scale nautical charts (1 :50,000 and greater), the transverse
Mercator projection is generally used instead.

(5) Shape on a Mercator Chart. On the Earth, meridians converge at the

poles, so land masses on a Mercator projection chart will be increasingly distorted
in an East-West direction proportional to their distance from the Equator, until at
the poles their sizes would be infinite. In order to preserve the correct shape, the
parallels of latitude, which are equally spaced on the Earth's surface, must be
spaced increasingly further apart on the Mercator projection chart, until at the
poles the latitude scale becomes infinite. This distortion is governed by the
secant of the latitude. This means that, on a Mercator projection chart of the
world, Greenland appears as broad as Africa at the Equator, although the latter
is three times wider; this fact becomes apparent once the distance is measured
at the latitude scale in the vicinity of the two areas.

Fig 4-4. Spacing of Meridians and Parallels on the Surface of the Earth

N. Pole

.-1eoo~
GeographlCBI
t,\118S Spacing of·Parallels of Latitude
Convergence of Meridians

4-13
A,.~ .. .,+ ?n1R v .. r!::inn 1
BRd 45(1)

(6) Practical Measurement of Distance on Mercator Projection Charts.

Due to the distortions inherent in the Mercator projection, distances must always
be measured using the latitude scale at the latitude of the position concerned.
The longitude scale must never be used for measuring distances on a Mercator
projection chart.

(7) Use of the Rhumb Line as Course. If two places on the Earth's surface
are joined by a rhumn line, and a ship steers along that line, the direction of ship's
head will remain the same throughout the passage. This direction is measured
by the angle from the meridian to the rhumb line, measured clockwise from 000°
to 360°, and is the course. In practice, the difference between the two is
semantic, and the rhumb line itself is often referred to as the course. On the
surface of the Earth, a continuous rhumb line (with the sole exception of
090°/270°) will spiral towards the pole. The Mercator projection chart shows the
ship's track as a straight line between the starting point and destination, and the
measurement of this straight line gives the steady course.

(8) Plotting of Great Circle Tracks. A great circle will appear as a curve on a
Mercator projection chart (with the exception of meridians and the Equator. To
plot a great circle track on a Mercator projection chart it is necessary to use either
a software package or gnomonic chart to generate a convenient number of
waypoints, and then plot those waypoints on the Mercator projection chart.
WECDIS is capable of generating courses as great circles or waypoints, and
NavPac has a convenient function for this purpose as well.

Fig 4-5. Bishop Rock to Bermuda - Great Circle vs Rhumb Line Tracks

so· 70' ... 50• 40"W 30' 20· 10· .

~ £
~ " ' I J "-'i
I I I. =--4---
,.,.,nut'KU\;I\ J...r::;tl
J>' 50'

3..1
IO' 10· IO' so•
1
40"W
1
30•
1
20•
1s
10·
I
o·
I 3·•

c. Transverse Mercator Projection. The transverse Mercator projection is

described in full detail in Section 3, but a brief summary of its principal characteristics
are given below.

(1) Construction. The transverse Mercator projection may be considered as

the same as a standard Mercator projection, with the difference that the 'cylinder'
is rotated through 90°. The central meridian and Equator plot as straight lines,
but all other meridians and parallels plot as curves.

4-14
A •• ____ ,._-- .. - ..
 BRd 45(1)

(2) Properties. The transverse Mercator projection remains orthomorphic. The

scale between meridians expands as longitude increases away from the central
meridian, but due to the large scale used these meridians will appear as straight
lines to the user and, for all practical purposes, straight lines can be used to plot
all bearings and direction lines on the chart. However, in order to enable this to
be the case, and to minimise the effects of scale expansion, the coverage of a
transverse Mercator projection charted is restricted to 3° of longitude either side
of the central meridian . The transverse Mercator projection does not suffer from
the high latitude problems of standard Mercator projections.

(3) Uses. The transverse Mercator projection is in wide use; most land maps,
including UK Ordnance Survey maps and NATO military maps use it. At sea,
most UKHO large scale charts of 1:50,000 or larger (i.e. covering a small area
such as harbour plans) use transverse Mercator. Additionally, some polar charts
and maps use this projection, although the polar stereographic (see below) is
more commonly used for this purpose.

d. Stereographic Projection (or Polar Stereographic Projection). The point of

origin of a stereographic projection can in principal be anywhere. However in practice
this projection is normally only used for polar regions, and s~ only. th~ pol~r
stereographic projection (or the 'universal' polar stereographic proJect1on) Is
considered here.

(1) Construction. With the polar stereographic projection the meridians and
parallels of latitude are projected onto a plane tangential to the pole, the centre
of projection being the opposite pole. Meridians appear as straight lines
originating from the pole, and parallels of latitude appear as concentric circles
radiating outwards from the pole.

(2) Orthomorphic Properties. The polar stereographic projection is

orthomorphic.

(3) Great Circles. Great circles other than meridians are not plotted as straight
lines, although in practice little accuracy is lost by plotting them as such.

(4) Uses. As the name suggests, the polar stereographic projection is used for
polar charts and orthomorphic maps of polar regions.

e. Gnomonic Projection. The gnomonic projection is described in detail in Section

5, but its principle characteristics are summarised below.

(1) c struction. The gnomonic projection projects the Earth's surface from
the Ea~~s centre onto a tangent plane. The gnomonic projection is only applied
to a sphere which represents the Earth.

A_1,;
BRd 45(1)
(2) Properties. The gnomonic projection is NOT orthomorphic. Great circles
are represented by straight lines, and parallels of latitude are curves. The
meridians will not be parallel unless the tangent point is on the Equator. Rhumb
lines are shown as curves, not straight lines. Angles are thus distorted, except at
the tangent point; the further a point on the chart is away from the tangent point,
the greater the distortion. It is therefore impossible to take courses and distances
from a gnomonic projection chart. The gnomonic projection does not have equal
area properties.

(3) Uses. As the gnomonic projection results in a chart which is neither

orthomorphic nor equal area, it is thus unsuitable for general navigation purposes.
Historically, it was used to plot great circles as straight lines in order to generate
great circle waypoints. It was also used to plot long range radio beacon bearings,
or indeed intercept bearings. Both of these usages have almost completely
disappeared as computerised systems (WECDIS, NavPac etc) are readily
available to plot the same information.

f. Skew Orthomorphic Projection. Instead of a central meridian, a central great

circle passing through the axis of the country is used as the line of contact for the
skew ortl:lomorphic projection. It is used mainly for land surveys.

g. Polyconic Projection. The polyconic projection is a modification of the simple

conical projection.

(1) Construction. The central meridian of the area to be display~d- is divided

correctly for intervals of latitude, but each parallel is constructed as 1f 1t we~e the
standard parallel of a simple conical projection. The parallels are arcs of circles,
the radii of which steadily increase as the latitude decreases. The central
meridian is a straight line, although the other meridians are curved .

(2) Properties. The polyconic projection is neither orthomorphic nor equal

area, so it is unsuitable for large areas.

(3) Advantages. The principal advantage of the polyconic projection is a

practical one, which benefits the users of paper products. If small areas are
shown on this projection, with each small area covering the same amount of
longitude, the sheets on which the geographical graticules are drawn fit exactly
along their northern and southern edges. For ordinary purposes, the sheets also
fit along their eastern and western edges, although the join here is a 'rolling fit' as
the meridians are curved.

(4) Modified Form. In a slightly modified form, the meridians can be made to
project as straight lines, and this modified form was used in practice.

(5) Uses. The modified polyconic projection is suitable for topographical maps
which individually covering a small area but combine to cover a large one.
However, today it has been largely superseded by electronic systems and the
transverse Mercator projection. It is not impossible that the reader will encounter
a few old UKHO harbour plan charts may still exist which are described as
'gn~m~nic projection' but in fact almost certainly used the modified polyconic
pr0Ject1on.

4-16
 BRd 45(1)

Fig 4-6. The Polyconic Projection

K2
I
I
I
I
I
I
I
I
I
,' K2
I L2 I
I
II I I
/ ,1' I
I

I I I
I , I
I , M2
I
I
I
,'
I
"' / ,IL2
I I I I I
I I 1, I /
I I ,, ' N2
,' ;'' ,, ' ,,,.,,.,,.,,.,,. I
/
I
,/ ,1M2
I
,' ,' ,,, ,,,-"',-:.,,. ,,. I
,
/
, /
/
1 N2
I

I
I
I
I I
I
,'
I
,' /
,''
I
,'
,, z"/' ,/'
I I
/ ~'·
I ,,' / I , '

I I I I
I I I
,} I I '<

I
,' I
/,' "•
I I
I I
,1 ''L
I

--..'K K1

Construction Graticule

4-17
.11.11aust 2018 Version 1
BRd 45(1)
SECTION 2 - CYLINDRICAL PROJECTIONS

0408 Principles of Cylindrical Projections . .

· . . • · f h' h the most famous is the Mercator pr0Ject1on, are a
Cylindrical pro1ect1ons, o w 1c . . . t· .
• • h' h 'd ly used in navigation. A cylindrical proJec 10n 1s one
broad class of pro1ect1on w 1c are w1 e .
. • • dt ually spaced vertical lines and parallels of latitude are
in which meridians are mappe O eq . . d h E
· t - It ·s
1 typically visualised as placing a cylinder aroun t e arth,
mapped to horizon a1 1mes. . . • • d'
· · h 1· d d then unwrapping it. In all cylindrical pr0Ject1ons, 1stances
pro1ectmg onto t e cy m er, an . . . .
are distorted east and west; the differences between the different types _of c~lmdric~I p~0Ject1on
are found in the way that north-south distortion occurs. In theory, cylindrical pro1ect1ons are
'perspective' projections (in which the image_ re~ults from t~e s~adows thrown by _the
hypothetical light source), but in practice the cylindrical perspectives m us~ are mathematical
projections, in which the charts are drawn by a mathematical func~ion_ which ~ra~sposes co-
ordinates from the sphere (spheroid) onto the flat plane. Some cylindrical pro1ect1ons can be
used to show the whole world, but the most common form, the Mercator projection, cannot
show the poles as the scale distortion becomes infinite at latitude 90°.

0409. Mercator's Projection (the Mercator Projection)

The Mercator projection was devised by the Flemish cartographer Gerardus Mercator
(he was born Geert de Kremer but adopted the name Mercator whilst studying at a school in
which Latin was the only language), and first published in his world map in 1569. Mercator
used the graticule derived from the projection of his world map, but the graticule was
inaccurately drawn above the parallels of 40° and Mercator left no mathematical description
of the projection. Rather, it is likely that he used a graphical method, in which rhumb lines
were transferred from a globe (with which Mercator was familiar) onto a graticule and the
space between the parallels adjusted until the lines became straight. A mathematical
description was first published by the English mathematician Edward Wright in 1599, and
further enhanced by the derivation of the logarithmic formula for the projection by Thomas
Harriot and Henry Bond independently in about 1600 and 1645 respectively. Mercator's
achievement was to devise a projection specifically for the use of mariners, and by creating a
projection which was conformal and had straight rhumb lines, he achieved it. However, as at
the time it was not possible to accurately determine longitude at sea, and as directions were
magnetic rather than true due to the use of the magnetic compass at the time, it would be
another century and a half until the Mercator projection would be in widespread use at sea.
Nevertheless, Mercator's projection became the standard projection for nautical charts
worl_dwide, _and is still in wide usage today. Indeed, even those who are not professional
ma~ne~s ~111 e~~unter ~he Mercator projection on an almost daily basis as the 'Web Mercator'
proJect1on 1s utilised by internet mapping applications such as Google Maps etc.

a. Perspective Concept vs Mathematical Reality. As discussed in Para 0408

above, th e ~ercator ~roje~tion can be conceptualised as a cylinder wrapped around
the Earth, with the cylinders and Earth's axis coincident, and their surfaces in contact
only at _th e Equator; the Earth's surface is then projected from the Earth's centre onto
th e_ cyl_mder. _W~en unrolled flat the image on the paper cylinder will represent a
cyll nd ncal proJe~tion. To produce a true Mercator projection, a logarithmic formula is
employed t? ad!u st th e spacing of the parallels and to produce a chart which is
ortho_morph1c (different formulae apply for the sphere and the sph 'd) It . t
possible to show the entirety of th I . ero1 . is no
the distance between the II I efplo ~r regions on a Mercator projection chart as
para es o at1tude becomes infinite at latitude goo.

4-18
~n,40 , . ___ !_
 BRd 45(1)

Fig 4-7. Simple Mercator Projection

··- -- - r-·· .-- r - .. _ r - - -· -

-------~Cl.". 60°N

-------~~ 45°N
-------1~ 3(1•1',1
_____ __ _1,li~ tl
15°N
- ~E------~ - o•
~ -- - - - - _j 5~'S 15"5
_______ Jll~ 3o•s
________4.li~ ·~ 45"5

------- 60° ·~ 60°S

b. Properties of the Mercator Projection. See Para 1407 sub para b for a full
description of the properties of the Mercator projection.

0410. Equidistant Cylindrical Projection

The equidistant cylindrical (EC) projection is a cylindrical projection which is not as
commonly used as the Mercator projection, but is now of critical importance to the Royal Navy,
for a simple reason: it is the default projection used for non-polar regions when using RN
WECDIS (Outfit 4LM). The EC projection is an extremely old one, possibly being invented
as long ago as 100AD by Marin us of Tyre (it was certainly known to Ptolemy writing several
centuries later). The EC projection was not much used for navigation prior to the introduction
of ECDIS, as it is in many ways inferior to the Mercator projection as it is not orthomorphic,
and as bearings are not constant across the chart it is not possible to lay off a course (rhumb
lines are curved on the EC projection, however in practice these curves are only apparent on
very small scale charts). However, the EC projection does have some distinct advantages.
Polar regions are less distorted than on a standard Mercator projection for one. Much more
importantly, however, the EC projection is mathematically very straightforward, meaning that
computations are trivial with modern technology. This then makes it very straightforward to
use in WECDIS; the actual navigation calculations take place on the surface of t_he reference
spheroid (WGS 84 ), and the computer is readily able to transform the co-ord~nates _of. the
cursor position on screen (for example) so that they are c~rrect; the on~cr~en 1mag~ 1s 1~ a
sense irrelevant. When zoomed in to practical range scales in_ use for nav1_gat1on, t~e d1stort1on
and other problems of the EC are not practical objections ~o its use, part1cularl~ g1~en as any
· b · tc) are calculated using the reference ellipsoid and not
data b emg used (ranges, eanngs e
the image itself.
D THAT IN WECDIS (OR ANY OTHER ECDIS)
ONCE AGAIN, IT MUST BE REMEMBERE NS COURSES RANGES BEARINGS ETC)
THE ACTUAL NAVIGATION CALCULATIO THE( REFERENCE ELLIPSOID AND NOT ON
ARE CALCULATED ON THE SURFACE OF
THE PROJECTED CHART.

4-19
August 2018 Version 1
BRd 45(1)

0411. Mathematics of the Mercator Projection

a. Mercator Projection - Concept. The Mercator projection is a special case of

the Lambert conical orthomorphic projection in which the Equator is used as the
Standard Parallel. Para 0409 and Fig 4-7 above provide an outline of the principles
of the Mercator Projection. The Mercator projection is widely used in paper charts but
is generally not used in ECDIS chart displays.

b. Mercator Projection - Governing Criteria. The requirements for the Mercator

projection are g0verned by two criteria:

(1) In a cylindrical projection based on the Lambert conical orthomorphic

projection, the constant of the cone must be zero (see Para 0411 sub para d).

(2) The Projection must retain orthomorphic properties (see Para 0411 sub para
e) .

c. Mercator Projection - Detailed Mathematical Analysis. Amplifying Para 0409

and Fig 4-7 above, Fig 4-8 (below) takes the Equator as the latitude of the origin $0
The following can be observed/deduced. •·

• RO is a central meridian and is equal in length to Vo cot$, where Vo is the

radius of curvature at right angles to the meridian at O for the spheroid of
the Earth in use, and$ is the latitude of 0.

• As the cotangent of 0° is infinity, R recedes northwards (or southwards) to

infinity (see Fig 4-8); this can also be seen from Fig 4-7 if latitude is
extrapolated to 90°. Thus the complete polar regions can never be shown
on Mercator projection charts as expansion between parallels of latitude
expands to infinity at latitude 90°.

• The angle between true North and grid North becomes zero for this
projection, thus there is no convergence.

• OPo coincides with grid East, all parallels become straight lines parallel to
OPo and, since convergence is zero, all the meridians are parallel with grid
North.

• One minute of longitude measured along the Equator (or standard parallel)
as the unit of the grid makes this projection suitable for navigation use.

d. Constant of the Cone. The quantity Sin <?o is the 'constant of the cone' (where
$ is the latitude of 0) and it is a constant for any given latitude of the point of origin .
When the Equator is the point of origin:

sin ¢0 = sin 0° = 0

e. Orthomorphism and Rhumb Lines. The scale along a meridian in the

neighbourhood of a point in latitude$ stretched by the same amount (sec $) as the
scale along the parallel through that point.

4-20
 BRd 45(1)

(1) Orthomorphism. In view of this, and that the meridians and parallels on
the Mercator projection are at right angles, the projection must be orthomorphic.

(2) Rhumb Lines. The parallels are spaced at increasing intervals as they
approach the poles (see Fig 4-7), while the meridians are spaced equally. Thus,
any straight lines (including diagonals) make a constant angle with the meridians
with no distortion; they are thus rhumb lines.

Fig 4-8. Detailed Analysis of Mercator Projection

ToR ToR
(at Infinity) {at Infinity)

Grid North

NI ',' ','IP
i I
I I
I I
I I
I I
I I
I I
I I
I I
m m
I I
I I
I I
I I
I I
I I
I I
I I

~I
I
I Grid East
)I(

0 Po

f. Use of Longitude Scale When Constructing a Mercator Chart. As meridians

of the mercator graticule are straight lines at right angles to the Equator, the longitude
scale is the same everywhere and provides a means of measurement in the
graticule.

(1) Let the scale of any Mercator projection chart be x millimetres to 1' of [Link] .

(2) Departure "' [Link] cos Lat (see Note below), thus departure on the chart
represented by 'x millimetres' approximates to 1' _cos lat (ie on~ sea mile in that
particular latitude is represented by x sec lat millimetres on the chart).

(3) Therefore, the scale of latitude and distance at any part of a Mer,cator
projection chart is proportional to the secant of the latitude of that part;
thus the amount of distortion in any latitude is also governed by the secant

of that latitude (see example below).

(4) The latitude scale cannot be used to draw the parallel in its correct position
on the graticule because it is continually being stretched as the latitude increases.
Thus, for this purpose, the spacing of any parallel from the Equator must
be calculated in units of the longitude scale.

4-21
A11nust 2018 Version 1