An introduction to making maps on a computer
Modern technology allows us to create, edit, view, and print maps on a
computer. In addition to replicating traditional paper maps, computers
open up other possibilities such as “interrogating” a map by clicking on an
object on the map and being shown information about that object. Or
calculating and showing what area will be flooded should water levels rise,
or other more sophisticated analysis. These kinds of applications have
come to be known as “Geographical Information Systems” or GIS.
Some advocates of GIS are so enthusiastic about what can, potentially, be
done that sometimes we loose sight of the benefits to be had from more
elementary computerised mapping.
Until recently the production of maps was in the hands of highly skilled
cartographers and maps were printed by specialist publishers with large
and expensive printing presses. This meant that maps were static images,
frozen in time, often cumbersomely large, usually expensive, and
generally either pinned to the wall or stored in an archive.
Now, in theory, anyone with a PC and a desktop printer can produce
maps. Maps have the potential to become a part of day-to-day work;
sketch maps to indicate where the problems are, maps to show what work
needs to be done where, and maps in reports to share our knowledge and
show where we have done what we do.
When a computer programmer designs any piece of software a balance
needs to be struck between simplicity of use and flexibility. A program to
make maps could be very simple indeed; the base map could be built-in
and the user might be able to label parts of the map and print it out.
However, many potential users of computerised mapping need to do more
than this. They need to bring in data from various sources, they need to
be able to add and edit features and generally customise it to their needs.
To have this flexibility it is essential that the user, without becoming an
expert in GIS, has a grounding in the basics of maps. That is the purpose
of this document. The following topics are discussed:
Scale and map projections
Using the Geographic Positioning System (GPS)
Printing and presenting maps
This document does not describe the use of any particular mapping or GIS
The data that goes into making a computerised map is of two sorts: “Map
data” and “Attribute data”. Map data is that which describes the physical
location, size, and shape of places on the map, while attribute data is
information associated with those places.
Map data is generally of two kinds: “raster” or “vector”.
Raster data is also called bitmap or scanned data. A raster image is like a
photograph in a newspaper. It consists of dots, called pixels, that together
create an impression of an image. Typically a raster image will be a copy
of a paper map, an aerial photograph, or a satellite image.
Vector data is like lengths of string pulled tight between pins on the map.
A vector line is defined by the location – or the co-ordinates - of its
“vertices” (the pins on the map). Vector data is primarily made up of
points, lines, and polygons (enclosed areas).
Typically on a computer-based map a raster image will be used as a
background map while one or more layers of vector data sit on top of the
background image as foreground information.
Unlike vectors, raster images (or “bitmaps” or “scans”) do not store any
information about objects such as lines or points. A raster image consists
of information about lightness, darkness and, generally colour. The three
representations of a polygon shown below are all bitmaps. They are made
up of square dots called pixels. The image on the left has more dots per
inch - each pixel is smaller - so it appears to be a more accurate
representation of a polygon than the others. The image on the right
appears very crude because it contains fewer pixels per inch.
If this polygon were contained in a vector file it would be described in
terms of the geometry of its corners and sides. In a vector file, the
information needed to describe a polygon eight centimetres across
occupies as much space in the file as the description of a polygon of the
same shape but only one centimetre across. In a bitmap file, on the other
hand, the larger the image - the larger the graphics file needed to
describe it. Instead of describing the boundaries of the polygon, the
bitmap file must contain information about the shade and colour of each
dot – each pixel – which makes up the image of the polygon and the
empty space around the polygon.
Take the case of a bitmap containing 300 dots per inch (d.p.i.). An image
which is 4" x 3" would consist of (300 x 4) x (300 x 3) pixels; or
1,080,000 pixels. The amount of data required to describe each pixel is
determined by the number of colours or tones which the image uses.
Computers store data in bytes. Each byte is made up of eight bits. A bit is
like a simple switch. It can be either off (equals zero) or on (equals one).
In a black and white bitmap image, such as an image representing a
simple line drawing, the colour numbers for each pixel range from 0 for
black to 1 for white. This means that the data for each of these pixels can
be stored in just one bit. And the data for eight pixels can be stored in one
At 300 d.p.i. a 4" x 3" image contains 1,080,000 pixels. If this image is in
black and white (a 1-bit image) the file will contain 135,000 bytes – that
is to say the file size will be equal to 1,080,000 bits divided by 8. If the
image uses more colours than simple black and white, more than one bit
will be needed to describe each pixel. Theoretically any number of bits
could be used but, for various reasons, the following common “bit depths”
have become established:
1 bit - 21 = two colours, black and white
4 bits - 24 = 16 colours (no longer commonly used)
8 bits - 28 = 256 colours or 256 grey tones (e.g. aerial photos, etc.)
24 bits - 224 = 16,777,216 colours (sometime called True or RGB colour)
In our example of a 4" x 3" image scanned at 300 d.p.i, 3,240,000 bytes
are required to describe a 24-bit image.
Vector data has two great advantages over raster data:
• Compact. Assuming that two polygons have the same number of
vertices, no more information is needed to record a large polygon than
to record a small polygon. In contrast, the larger the raster image -
the larger the amount of data required to record it - regardless of the
information content of the image. The raster file describes the blank
spaces on the map as well as the objects.
• "Intelligent". While raster data simply record the colour of each pixel
on the screen, vector data describe spatial objects: polygons, lines and
points. These objects have intrinsic attributes such as area and length,
and can also be associated with attribute data. The intrinsic data about
a farm, for example, relate to its size and shape while the attribute
data might contain information about land use, soil quality, crop yield,
etc. One of the great strengths of computerised mapping is its ability
to associate attribute data with vector data.
In addition to representing actual spatial objects (polygons, lines, and
points), vector data can record annotational objects such as text and
arrows. Raster images, on the other hand, offer a simple, intuitive way to
represent a complex visual image. An aerial photograph for example, can
be shown as a raster image but not, realistically speaking, in vector
“Attribute data” can be any data attached (or linked) to an object
(polygon, line, or point) on a map. It might be a name, a number (such as
population or some score), an image – such as a photograph, a document,
or a link to a web page. But typically it will be a record in a database or a
spreadsheet. The object on the map must have a number or name that is
its “ID” (identifier) that links it to one unique record in the database. Data
in that data record can be used to do things like determine the colour that
is used to draw the object on the map. For instance, if the data record
includes a field for “population”, you might set up your computerised map
so that a polygon with a low population is coloured green while a polygon
with a high population is red.
Raster, vector, and attribute data are all stored in computer files. In an
ideal world each of these three types of data would all be stored in a
single consistent way – i.e. all raster data would be stored in one type of
file. However, due to the history of computing a variety of different file
types – or “file formats” – have evolved. Not all formats are supported by
all programs so when you are using data from different sources you will
probably need, at some point, to get involved in choosing between
formats or converting data from one format to another. Each format has
its own advantages and disadvantages:
Raster file formats
The size of a raster file is primarily determined by the number of pixels it
describes (width x height) and the “bit depth” of the image (typically 1,8,
or 24 bits per pixel). However, there are numerous ways in which this
data can be organised in a file, thus each file needs to start with
information which tells the computer program how the data in the rest of
the file is organised. Raster files can be very large. When memory and
disk sizes of computers were smaller than they commonly are today, the
size of such files was a major problem. Today, with the internet, the size
of files is again a significant issue since the smaller the file the easier it is
to download. So it is common to “compress” the data within a raster file.
Any computer file can be “zipped” (i.e. compressed) using widely available
programs like WinZip or Pkzip. However, they need to be unzipped before
they can be used again. But many raster files are compressed internally –
they are like a zipped file but with extra data attached explaining to
computer programs how the compressed data can be read without first
uncompressing the file. Internal compression – or the lack of it – is one of
the major features which distinguishes one format from another.
In addition, some raster files contain information about the “calibration” of
the image – that is to say, it provides the real world co-ordinates of the
extent of the area represented by the image. Common formats are:
BMP files. A BMP file is a “Windows Bitmap”, a format developed by
Microsoft for the first version of Windows.
Advantages: The organization of its data is the same as that used
internally by Microsoft Windows. This means that a BMP image is quickly
loaded and processed by the computer. Other formats have, in effect, to
be converted to the BMP format in the computers memory prior to loading
on the screen.
Disadvantages: In theory data within a BMP file can be compressed but
this feature is almost never used. As a result BMP files are typically larger
than other formats. The BMP format has no provision for calibration.
Typically, any geographic calibration has to be stored in a separate file –
called a “world file” – with a BPW extension (ie a file called test.bmp will
be accompanied by a file called test.bpw).
TIF files. The “Tagged Image File” format pre-dates Windows and was
developed by the Aldus Corporation.
Advantages: The format is widely supported by a wide range of programs,
not only on computers running Windows but also Macintosh, UNIX, and
Linux systems. The word “tagged” conveys the idea of bits of additional
information, identified by numeric tags, being attached to the basic image
data all within the one file. This has the advantage that the format can be
extended in an almost limitless fashion to cope with different kinds of
application. For instance, using “GeoTIFF” tags, a TIF file can be calibrated
without the need for a separate world file. Data within a TIF file can be
compressed resulting in a smaller size than an equivalent BMP file.
Disadvantages: The very flexibility of TIFF files can also be a
disadvantage; it means that there are numerous versions of TIF files in
circulation. In particular, the data within TIF files can be compressed in a
wide variety of ways. Not all programs can cope with all types of
compression in TIF files, so when using TIF files care needs to be taken to
ensure that they are compatible with the program that you are using.
Similarly the “GeoTIFF” tags for calibration are not recognised by all
mapping programs so it is often necessary to have a separate world file
with a TFW extension.
JPG files. The acronym JPG, or JPEG, stands for the Joint Photography
Experts Group. As the name suggests the format was originally developed
as an appropriate format for storing photographs. So why should one
format be more “appropriate” for photographs than another? The answer
lies in the form of compression. In TIF files when the data is compressed
it is compressed using “lossless” compression. This means that when the
compressed data is uncompressed the resulting data – the image – will be
an exact copy of the original. But JPG images use a “lossy” compression
meaning that the image will give an impression of the original image
rather than an exact copy.
Advantages: The advantage of this is that, thanks to some clever maths,
the size of the compressed image can be considerably smaller than that of
an equivalent compressed TIF file. As with BMP and TIF files, JPG files are
Disadvantages: Where there is a great variety of colours and tones, such
as you may find in an aerial photograph, the JPG “lossy” compression
gives remarkably effective results. But where there are large areas of
continuous colour and sharp edges, such as in a scan of a drawing or a
business graphic, “lossy” images can appear blurred and smudgy. Also, in
these circumstances, a compressed TIF file can actually be smaller. Thus
the use of JPG files should be reserved for the appropriate kind of data –
ECW files. The ECW format was developed by “Earth Resource Mapping”
specifically for large aerial photograph images. As with JPG files they use
a “lossy” compression but even more compressed than in JPG files.
Advantages: Their small size makes them a good format for distribution
over the internet. A disadvantage of an image stored as a JPG file is that
to view part of the image the whole image has to be loaded into memory
and decompressed. The desired part is then extracted for display. This can
be problematic for large images since the computer needs to have
sufficient memory to contain the whole uncompressed image comfortably.
The clever trick of the ECW format is that it is designed so that the
software only needs to load the data that it wants to display. This means
that even very large images can be loaded and displayed rapidly. This
attribute makes ECW a useful format for computerised mapping. Because
the ECW format was specifically designed for mapping, all ECW files are
calibrated. There is no need for a separate “world file”.
Disadvantages. Not all mapping programs yet support ECW files and,
unlike BMP, JPG and TIF files, they are not supported by other image
processing programs that are not specifically designed for mapping (e.g.
PaintShop, Photoshop etc).
Vector file formats
Some formats are considered commercially “secret” and so their
specifications are not published by their authors. Commonly used
published formats are:
Shape files: The Shape file is probably the most widely used vector
format in mapping software. It was developed by Environmental Systems
Research Institute (ESRI) for their ArcView program. The term “shape file”
is a bit misleading since in practice it is a set of a minimum of three files:
the shape file itself, an index file, and a file containing attribute data. The
shape file always has a file extension of .SHP. If the shape file is called
TEST.SHP then the index file is TEST.SHX, and the attribute file is
TEST.DBF. In later versions of ArcView there are more than three files but
the other files are optional.
Advantages: The format is widely supported and well documented by
ESRI. The attributes file format (DBF) is a standard database format
which can be read from most database and spreadsheet programs.
Disadvantages: Each shape file can only contain one object type: i.e. all
polygons, all lines, or all points. The need to have all three files means
that files can get separated. Shape files are in a “binary” format which
means they cannot be read or edited directly by humans via a text editor.
While the DBF file can be read and edited separately, in practice this can
be dangerous since the records in the DBF file must always be in the same
order as the objects in the SHP file, and there must be the same number
MIF files: The MIF file is a format used by the program MapInfo to export
and import data. As with shape files it is not really one file. In this case it
is two files, a MIF file and a MID file. The MID file contains attribute data.
Advantages: Both the MIF and the MID files are text files so they can be
open and examined in a conventional text editor. Unlike shape files, MIF
files can contain any mixture of object types. The start of the MIF file can
contain details of the map projection used.
Dis-advantages: Different versions of MapInfo introduce new object types
and options. As with shape files the MID file can easily become separated
from the MIF file. Being text files they can be quite large but with the
power and capacity of modern computers this is less of a problem than it
used to be. As with shape files the attribute data in the MID file has to be
in the same order as the objects in the MIF file.
DXF files: DXF files (Data Exchange Files) were developed by AutoDesk
to import and export data to and from their AutoCAD program. A DXF file
does not need any other files with it.
Advantages: DXF files are widely supported not only in mapping software
but also in CAD (Computer Aided Design) and other graphics programs.
Being a single file they do not have the potential problem of shape and
MIF files of becoming separated from other necessary files. Like MIF files
any DXF file can contain any mixture of object types. They are usually text
files (though some are binary) so they can be viewed and, in theory,
edited in a text editor though in practice they are not easy to understand.
Disadvantages: DXF files were designed for CAD, not for mapping. This
means that they can contain many object types that relate to engineering
rather than maps. With every new release of AutoCAD there is a new
variant on DXF files so there are many different versions of DXF that can
be encountered. DXF files can be very large. Unlike other formats
discussed here there is no simple and standard way of associating
attribute data with points, lines, and polygons. Where possible DXF files
should be avoided,
DRA files: Drawing files (DRA) were developed by Map Maker Ltd. As
with DXF a DRA file does not require other files with it.
Advantages: A DRA file can contain any mix of objects. It is compact.
Unlike shape or MIF files each object can have an name without the need
of a separate database, this makes it easier to associate attributes with
objects at a later date.
Disadvantages: They are not supported by other programs and are much
less common that the previous three formats. The binary format means
they cannot be read in a text editor.
Attribute data file formats
Attribute data is stored in tables of columns and rows. There needs to be
one row for each object. Potentially, attribute data could be held in a wide
variety of database formats. In practice, four formats are commonly used:
DBF files: DBF (DataBase Files) is a format originally developed for a
program called “dBase” which pre-dates Windows.
Advantages: Because DBF files have been in circulation for many years,
the format is in the public domain, and so they are widely supported. Also
the format is relatively simple and limited which means that DBF files are
usually reliable – i.e. a DBF produced in one program will usually be
readable in another program. Their simple format also means that they
tend to be compact.
Disadvantages: Their simple format means that they do not have all the
options that more modern formats have. A particularly irritating limitation
is that in DBF files column names can only have a maximum of 11
characters and none of those characters can be a space. This means that
often column names in DBF files are rather cryptic.
XLS Excel files: The XLS file is used to store “spreadsheet” data from
Microsoft’s Excel program. It is important to remember that Excel is not a
database. It is a spreadsheet. This means that Excel is good at
manipulating data – i.e. applying simple formula – but an XLS file is not
designed to be a repository of data to be used by other applications. That
said, Excel is now such a ubiquitous program that it is commonly used to
store data as though it were a database.
Advantages: XLS files are very flexible and can store a wide variety of
data and, unlike DBF files, they have no restrictions on column names.
Since so many people use Excel, XLS are a popular format for exchanging
Disadvantages: The flexibility of XLS files is also their weakness. Because
users have total freedom on how to structure and lay out their data there
is no guarantee that data in an Excel spreadsheet can be read as though it
were a data table – i.e. a systematic grid of columns and rows. The XLS
file format is considered a secret by Microsoft and so it is not directly
readable by other software. Other programs have to read and write data
from an XLS files via the intermediary of a Microsoft program.
MDB Access files: “Access” is Microsoft’s main database program.
Advantages: Unlike Excel, Access saves data in formally structured tables.
This means that other programs can exchange data with Access tables
with full knowledge of the type of data being exchanged: i.e. is the data in
a given column a number or a piece of text? If a number is it a whole
integer or a decimal number? If the latter how many decimal places of
accuracy are being used? Access is what is known as a “relational
database” which means that data tables can be linked to each other so
that a change in one table can generate a change in other tables. With the
appropriate software the same result can be achieved with other data
table formats, but with Access it is built-in.
Disadvantages: Like Excel the internal format of Access tables is a secret
so Microsoft software needs to be used to read and write the data. Excel is
part of Microsoft’s “Office” suite and so is available on most computers
running Windows. “Access” is part of “Office Professional” and so only
available to users with more resources.
CSV files: A Comma Separated Variable (CSV) file is simply a text file.
Each line represents one record. Within each record, each item (each
“column”) is separated from the next by a comma. Usually the first line
contains the names of the columns.
Advantages: Great simplicity. Users with little computer skills can edit the
data with any text editor.
Disadvantages: When storing large data tables they can be slow to
access. They cannot store “binary” data such as photographs or other
bitmaps. Though relatively easy to edit in a text editor, to edit them in an
more conventional data form you will still need software designed to
manipulate the CSV file.
Other formats: There are other databases and database file formats.
Some are free programs, such as “SQLite” and “MySQL”, which aim to
offer some of the database functionality of Access but with without either
the cost or the need for memory-hungry software. Their limitation is
simply that compared to the above formats they are not widely used.
Any point that appears on a map has, like a point on a graph, an X and a
Y ordinate. On maps these are also commonly known as Eastings and
Northings. The values describe how far the point is east and north of a
“point of origin”. A point to the west of the origin has a negative X
ordinate (negative Easting), a point to the south has a negative Y ordinate
(negative Northing). These X and Y ordinates are on a square grid, like
graph paper – one unit in the X direction is the same size as one unit in
the Y direction. An understanding of this simple system is all that most
map users in the past required.
However, with the advent of low-cost GPS (Global Positioning System)
there is a greater need for map users to understand the relationship
between XY co-ordinates on a map and latitude and longitude locations on
the globe of the earth.
The latitude and longitude grid
X and Y values describe a location on a flat plane. Necessarily a flat plane
only approximates to the curved surface of the earth. Putting aside for the
moment the question of altitude, the location of any point on the globe
can be described by a latitude and a longitude – lines of latitude are
circles that run around the world parallel to the equator, lines of longitude
are half-circles that start at the south pole and end at the north pole.
The latitude and longitude grid is not a square grid. Lines of latitude are
equally spaced but lines of longitude converge towards the poles. Only on
the equator is one unit of latitude measured on the ground the same
length as one unit of longitude. Elsewhere one degree of latitude is larger
than a degree of longitude. In Oslo, for example, at the latitude of 60
degrees north, one unit of latitude is double the length of one unit of
If you attempt to plot the world on a square grid (one degree of latitude is
drawn equal to one degree of longitude) the resulting map flattens and
severely distorts the shapes and areas of countries that are further away
from the equator. There are many other ways to represent the latitude
and longitude values of the earth’s curved surface on a flat plane and
these methods are called map projection systems. All of them involve
distortion of some kind. The challenge is to minimise these distortions.
World map drawn on a simple square latitude and longitude grid
Think of the world as a glass globe with coastlines and other geographic
features drawn as lines on the surface of the globe. At the centre of the
globe there is a small, powerful light bulb. Take a sheet of tracing paper
and wrap it around the globe to form a cylinder, touching the globe all the
way around at the equator. The shadows of the coastlines etc are cast
onto the tracing paper. If you were to take a pencil and trace those
shadows on the paper, and then unwrap the paper and lay it flat you
would have a “Mercator” projection of the world.
The idea of a light bulb at the centre of the world sending out straight
light rays that cast shadows is where the word “projection” comes from. A
traditional map projection is conceived in terms of straight lines being
“projected” from the curved surface of the globe onto either a flat surface
or a surface that is curved in one direction – i.e. curved in the same way
that a piece of paper can be curved; normally a cylinder or a cone.
If you imagine a point at a far northerly latitude, such as the Norway or
Alaska, then the straight line projected through it from the centre of the
globe will have to travel a long way until it meets the cylinder of paper. A
line projected from the north pole will never reach the cylinder. This
means that when you unwrap the cylinder and lay it out as a map places
far away from the equator – north and south – will be badly distorted;
they will be stretched out. This is why on familiar maps of the world
Greenland can look so large, when it fact it is a quarter of the size of
Brazil. Areas close to the equator have negligible distortion. Ideally we
would like maps in which the distortions are always small. In an attempt
to achieve this map makers have devised a system called the “Transverse
Mercator” system. As with the regular Mercator projection it is conceived
of as a cylinder wrapped around a globe. But this time rather than being
wrapped around the equator the cylinder is wrapped so that it touches
both the north and south pole. In fact, it is not a cylinder but a half
cylinder, starting at the south pole wrapping around the globe so that it
touches the globe along one line of longitude and finishes at the north
pole. You can choose the line of longitude closest to the area of the earth
that you are interested in mapping. On the finished map the distortions
close to that line of longitude will be small, getting bigger the further that
you move from that line, east or west.
For a Transverse Mercator projection you can choose any line of longitude
for its centre – its “central meridian”. The problem with this is that we can
end up with many maps produced using the Transverse Mercator
projection but with a different central meridian meaning that the maps
can not be readily used together. In an attempt to bring some order the
“Universal Transverse Mercator” projection, or “UTM” was introduced. The
UTM projection is in reality a family of 60 projections, each one a
Transverse Mercator projection optimised for a different region of the
earth. Each UTM “zone” is designed for an area of 3 degrees each side of
a specified line of longitude. For instance UTM zone 31 is designed for the
slice of the earth centred on the 3 degrees East line of longitude – i.e.
stretching between longitude zero (the Greenwich meridian) and longitude
6 degrees east.
Locations at the centre of their UTM zone are shown in their true
proportions. As you move further away – east or west - from the centre so
the distortions grow. In practice, many countries are more than six
degrees of longitude wide, or else they do not fall neatly within one UTM
zone. In these circumstance it is common to choose the UTM zone that
sits nearest the centre of the country and accept that longitudes towards
the easterly and westerly extremes of the country will be more distorted
than they ideally would be. In large countries, such as India, Brazil, or
Russia several different UTM zones need to be used.
The Mercator projection is just one projection. There are very many map
projections. However, for detailed mapping the UTM system is the most
The shape of the world
The “datum”. Latitude and longitude is an internationally agreed system
which always uses 360 degrees of longitude, and 180 degrees of latitude.
But this simplicity is deceptive. The process of measuring the exact size
and shape of the earth and agreeing on the precise position of the centre
of the earth, its axis, and the location of the zero meridian (zero line of
longitude) has gone on for centuries and is still evolving. We have not yet
arrived at a universally accepted definition of latitude and longitude
values. The latitude and longitude values used in the creation of historic
maps and long established National Grids usually differ from the values we
use today. Geographers and cartographers have imposed their definitions
of the earth’s shape and the latitude and longitude system. Such a
definition is known as a “datum”.
To determine the XY co-ordinates on a map for a given latitude and longitude
requires a knowledge of both the datum used and the map projection system.
The earth can be described as a slightly flattened sphere – or an ellipsoid.
The height of the earth from pole to pole is about 0.3% less than its
diameter at the equator. Over the years there have been many different
attempts to define the size of the earth in terms of its radius at the
equator and the extent to which the sphere has been flattened. Today
about twenty different definitions of the earth’s ellipsoid are in more or
less regular use.
While it is convenient to think of the world as an ellipsoid, in practice -
even if we ignore mountains and valleys - the underlying shape of the
world undulates. Many ellipsoids, were defined to make the best sense of
a given locality - typically a particular country. To achieve the best fit, the
centre of the ellipsoid is often offset from the “real” centre of the earth.
Using different offsets and slight rotations to achieve the best fit, the
same ellipsoid can be used in different parts of the world. So while there
are 20 or more ellipsoid shapes, more than 200 different combinations of
ellipsoid shape and ellipsoid position are in common use. Each of these
combinations is known as a “datum”.
In more recent times people have promoted datums that can be used all
over the world. The WGS84 is the one used by the GPS satellites and is
the most important of these “geo-centric” datums. Unless the WGS84
datum is also being used for the local maps, then data from a GPS
receiver must be converted from WGS84 into the locally used datum. In
other words WGS84 latitude and longitude values must be converted to
other slightly different values of latitude and longitude to suit the local
datum. The modified latitude and longitude values then need to be
converted into the locally used map projection system.
The Geographical Positioning System is a set of satellites that broadcast
signals enabling ground-based devices to determine their position. If you
place a GPS receiver in a fixed position and monitor the readings you will
find that the readings are constantly wandering.
This illustration shows the variation of the recorded position of a fixed point over a 24
hour period. The grid is a 10 metre grid.
The more expensive and elaborate GPS devices use “Differential GPS”
(DGPS) in which radio signals from base stations of precisely known
locations give details of the GPS error at any moment. The errors are
automatically subtracted from the GPS values being recorded. An
alternative for people on a tighter budget is “post processing” in which
data is recorded in the field and the GPS device records the precise time
of each reading. Subsequently, data recorded at the same moment at a
known base station can be used to subtract the errors from the data.
Note that without the use of Differential GPS a low cost GPS device will
generally be just as accurate as an expensive device. The difference in
cost being in the additional features such as the number of points that can
The concept of “scale” is one of the most fundamental aspects of
mapping, yet it can be a source of confusion and misunderstanding. If a
paper map has a scale of 1 to 10,000 it simply means that representations
of objects on the paper map are 10,000 times smaller than the real
objects. People new to maps and the concept of scale commonly ask
questions such as “1 to 10,000 what?”, the answer being “units”: one
metre on the map represent 10,000 metres on the ground, just as one
inch represents 10,000 inches. While this response is true it does not
always help to clarify the situation; in reality one is unlikely to measure a
whole metre on a paper map, just as one is unlikely to conceive of a
distance in the real world as 10,000 inches.
In older maps, particularly, it is common to express “scale” as, for
instance “one inch to one mile”, or “1 centimetre represents 100 metres”.
These expressions can be easier to grasp than their numerical equivalents
of 1 to 63,360 and 1 to 10,000 respectively. Numerical scales are
generally written using a colon, i.e. “1:10,000” or as a fraction,
“1/10,000”. Sometimes, maps are described by the numerical scale, so
that a map drawn at a scale of 1 to 25,000 will be called a “25,000 map”
or a “25k map”.
Another common confusion is the difference between “smaller” and
“larger” scale maps. If you have a 1:10,000 map and a 1:5,000 map, the
1:5,000 map is correctly said to be at a larger scale than the 1:10,000
map, despite the fact that 5,000 is a smaller number than 10,000. This is
because objects of the same size appear larger on a 1:5,000 map than on
a 1:10,000 map. When calling for “a larger scale map” we are asking for a
more detailed map, not a map covering a larger area.
One of the big differences between a map on paper and a map on a
computer screen is the ability to zoom in and out – in other words to
change the scale. This means that digital map data, unlike a paper map,
cannot be said to have a scale. The scale of a computerised map is the
scale at which it is printed out or at which it is currently viewed on the
screen. The same data may be viewed at different scales. However, digital
mapping data available on web sites or from data suppliers is often
referred to as being of a particular scale (e.g. “1:10,000 raster data”).
This “scale” should be thought of as a “serving suggestion”; the data
supplier is simply suggesting that the data has been designed to look at
its best at the stated scale, or that the data originated by scanning or
tracing a paper map which was of that scale and so should not be treated
as accurate if you zoom in to a greater extent than the scale of the
original source material.
As mentioned above, with the growing use of GPS it is becoming more
common for people to want to plot latitude and longitude locations on a
simple square grid representing latitude and longitude coordinates. These
maps, where one degree of longitude is shown as the same size as one
degree of latitude, cannot be considered to have any meaningful scale.
Such maps should be considered more as schematic diagrams than scale
maps. Similarly, because they have no scale they cannot be used to
measure distances and areas. Where you want latitude and longitude
locations “drawn to scale” the latitude and longitude coordinates should
first be translated into X an Y coordinates using a map projection as
Many map users only need to view ready-made maps but others need to
be able to record real world locations and place them on the map. To do
this requires some knowledge of surveying. Clearly modern accurate
surveying is a complex and extensive area of expertise requiring
sophisticated equipment. However, the basic principles of surveying are
Low-cost GPS devices make it a simple task to determine the location of
the spot where you are standing. Using a GPS aerial attached to a car,
GPS can be used to plot a road network. But there are many situations
where GPS is not appropriate. These can be overshadowed locations, such
as under trees or in a dense urban context. Or they can be locations that
you cannot physically reach, such as the other side of a river. In these
circumstances more traditional surveying techniques are required. In
essence, there are three techniques for determining locations in plan.
These require a basic knowledge of trigonometry:
The classic surveying technique is triangulation. If you know the location
of two points – the two ends of a straight line “AB” – and so the distance
between them, “c”, then you can determine the location of a third point
“C” in one of three ways:
If the distances “a” and “b”, from
A to C and B to C, are measured
a then the position of C can be
If the angles BAC and CBA are
known then the position of C can
If the compass bearings of point C
from North are known from both A
and B then the position of C can
Where you have a meandering line, such as the path of a stream, the
easiest way to survey the line is often to create a straight line (AB)
approximately parallel to the line and then measure a series of “offsets” at
known distances along the straight line where you know the location of
the two ends of the straight line. The “offset” is the distance from the
straight line to the line being surveyed measured at right angles to the
straight line at a known distance (x) from the start of the straight line.
By measuring a series of offsets the path of the line being surveyed can
Bearings and distances
In the above two cases it is assumed that you can see the object to be
surveyed from some other known location. In some situations this will not
be possible. For instance, if you wanted to survey the course of a path
through dense forest you could neither use GPS or view the path from
known locations. In this case you would survey the path as a series of
straight segments – or legs. For each leg you measure both its length and
its compass bearing from the end of the previous leg.
More often than not, after producing a map on a computer you will want
to present the map as a finished product. This might be as a printed map,
as an image on a web site, in a document, or in some other computer
context, such as a Power-point presentation.
The requirements for a printed map vary between software packages, but
with all printing it is important to recognise that the resolution (i.e. dots
per inch) of the printed output will always be significantly higher than that
of the screen. Typically a printed page will have 300 dots per inch, though
it is often up to 1200 dots per inch, while a normal screen will have
around 100 dots per inch. An image which looks fine on the screen may
look unacceptably coarse when printed. This is particularly true when your
map has a raster layer, such as an aerial photograph. So when designing
your map, if you will eventually need printed output, you should keep in
mind the resolution of the final output.
The same applies where you are preparing a map to insert into a
document which itself is likely to be printed. If the map is a raster image
there is a danger of it looking crude when printed. Most mapping or GIS
programs will include an option to simply save the image of the map as a
raster file. This is quick and easy but will not necessarily give the best
result. If your map includes vector layers then you should try and find an
option to export the image in a vector format. Vectors will stay sharp
when viewed in another context with a higher resolution, such as printed
If your computer is running with Microsoft Windows then a useful format
for exchanging images between programs is the “Enhanced Metafile”
(*.EMF). An EMF file is simply a Windows “macro”, which means it is like a
recording of a sequence of Windows graphics commands. This sequence
can be saved to a file in one program and then “played back” in another
program where it appears as an image. Two great advantages of EMF files
are that they record all the visual attributes, such as line colour etc, and
that you can mix both raster and vector data in the one file. So if your
map includes a raster background map with vector layers on top then all
this can be included in the single image and the vectors will remain sharp.
If your mapping program can export an image of the map as an EMF file
then the image can be included in Word documents, Power-point slides,
and other programs.
The disadvantage of an EMF file is that because it is a recording of
Windows graphic commands it is tied to the version of Windows in which
the file was recorded. This means that an EMF file is not “backwardly
compatible” – e.g. if the EMF file is made on a computer running Windows
XP it will not necessarily display correctly on another computer running an
earlier version of Windows, such as Windows 98.
If you want to produce a map to put on a web page then EMF files are not
suitable since they depend on Windows. Unless the web browser has a
proprietary “plug-in” (add-on) then the map will need to be a raster image
in either a GIF or JPG format. In practice, the JPG format will almost
always be better. GIF files are good where there are large areas of plain
colour (typically business charts) but where there is a lot of variety or
relatively fine detail then JPG images will generally be superior.
Of the relevant proprietary plug-ins the most important is the “Scaleable
Vector Graphics” (SVG) viewer from Adobe. It is widely and freely
available since it now comes as part of the ubiquitous Adobe Acrobat
Reader. The advantage of SVG is that it is possible to display maps which
can zoom in and out while remaining sharp and which can be panned
(moved up, down, left, and right). The disadvantages are that they do not
readily support raster layers and they are not as flexible in terms of
graphic effects as a simple JPG file. If your mapping program can export
images and data in the form of an SVG image then SVG does open up
possibilities for creating an interactive web map, but to do so does require
fairly advanced computing skills.
Most modern computers have adequate memory to cope with mapping
software. However if you expect to be using large raster images, such as
aerial photography, a computer with more memory will perform better. If
you are buying a new computer it is better to spend money on more
memory (RAM) than on a fast processor, dual processor, or a superior
Mouse or Pointing device
Many people now use a laptop computer as their main, or only, computer.
Most laptops use a small built-in pad as the pointing device. While such
pads are adequate for normal text and menu-based programs they are
hard to use for any fine work such as that involved for drawing or graphic
editing. If you use a laptop I strongly recommend that you invest in either
a mouse or a pen tablet. If using a mouse it is advisable to get an
“optical” mouse since it will operate more smoothly than a conventional
Most printers are one of two types: laser or ink-jet. If you want to
produce maps in colour, and particularly if you may be using aerial
photographs, then ink-jet printers produce the best results. However, ink-
jet printers are expensive to use because of the ink cartridges.
When a computer program wants to print a page it does not send the data
direct to the printer; it sends the data as a series of drawing instructions
to an intermediary program called a “printer driver”. In some cases the
printer driver does all the hard work of converting the drawing instructions
into an image – a raster image – that is then sent to the printer line by
line. In other cases the printer driver does little more than pass the
instructions on to the printer which then does the processing to convert
the drawing instructions into the page image. The advantage of this latter
approach is that it rapidly frees the computer to get on with other tasks.
Where the processing is done in the computer the user may be left waiting
for some time watching an hour-glass on the screen. However, the
disadvantage is that the printer must have the capacity to process the
image. When dealing with simple text based programs, such as a word
processor or spreadsheet, this is seldom a problem. But when dealing with
a complex map the complexity and size of the data files involved in the
map may exceed the memory capacity of the printer.
As a general rule, ink-jet printers tend to take the first approach of
processing the page image on the computer, while laser printers will
process the image in the printer. Laser printers, and large format ink-jet
printers, often have the option to add additional memory. If you are
buying a new printer and you are planning to produce maps it is advisable
to buy as much additional memory for the printer as you can.