1. Introduction 17
1.1 ECDIS – a Complex Navigation Tool 18
1.2 History and Development 18
1.3 ECDIS Display 20
1.4 What is an ECS? 21
1.5 The Potential Benefits of ECDIS 21
1.6 The Potential Disadvantages of ECDIS 22
1.7 The ECDIS Mindset 23
2. Functionality Requirements 25
2.1 “ECDIS” – General Term versus Operation Mode 26
2.2 ECDIS Performance Standards 26
2.2.1 Appendix 6 to Resolution MSC.232 (82): Back-up Requirements 28
2.2.2 Appendix 7 to Resolution MSC.232 (82): RCDS Mode of Operation 29
2.2.3 Power Supply 30
2.3 Upgrading of ECDIS Software 30
2.4 Official Vector Chart Data 31
2.5 Type Approval 32
2.6 Summary Table for “Real” ECDIS Criteria 34
3. Requirements 37
3.1 SOLAS Chapter V Regulation 19 38
3.2 ECDIS Fitted on Board – More than the Equipment 39
3.3 SOLAS Chapter V Regulation 27 41
3.4 Flag State Requirements for ECDIS Carriage 42
4. ECDIS Training 45
4.1 Necessity of ECDIS Training 46
4.2 Legal Training Requirements 48
4.2.1 The STCW Manila Amendments 48
4.2.2 Certification in Accordance with the Manila Amendments 2010 49
4.2.3 International Safety Management – Code (ISM-Code) 50
4.3 Additional Training Requirements 51
4.4 Types of ECDIS Training 52
4.4.1 The IMO Model Course 1.27 (Generic Training) 52
4.4.2 Type Specific Training / Familiarisation with Ship-borne ECDIS 53
4.5 ECDIS Training and Port-State Control 54
4.6 ECDIS Training Providers 55
5. Chart Data 59
5.1 Official and Unofficial Chart Data 60
5.2 Raster Charts 61
5.2.1 Characteristics of a Raster Chart 61
5.2.2 Raster Navigational Chart - RNC 63
5.3 Vector Charts 64
5.3.1 Vector Chart Objects 65
5.3.2 Characteristics of a Vector Chart 69
5.3.3 Chart Production and Fidelity 70
5.4 Electronic Navigational Chart - ENC 71
5.4.1 S-57 Standard: ENC Data Format 72
5.4.2 S-52 Standard: Presentation Library 72
5.4.3 S-100 and S-101 Standard 72
5.4.4 ENC Navigational Purpose 73
5.4.5 ENC Coverage 74
5.4.6 ENC Identification 75
5.5 Active Vector Chart Functions 75
5.5.1 Information Query or Pick Report 75
5.5.2 Continuous Safety Watch or Safety Frame 77
5.6 Mode of Operation 81
6. Chart Presentation
83
6.1 Automatic Chart Loading 84
6.2 Mode of Presentation 85
6.2.1 Chart Orientation 85
6.2.2 Own Ship Movement 85
6.2.3 True or Relative Vector 87
6.2.4 Overscaling and Underscaling 88
6.3 Chart Symbols 90
6.4 Display Filter 91
6.4.1 Base Display 91
6.4.2 Standard Display 91
6.4.3 All Display 92
6.4.4 Custom Display 92
6.5 SCAMIN 93
6.6 Primary / Secondary Display 93
6.7 Depth Shades 94
6.7.1 Safety Contour and Two Depths Shades 94
6.7.2 Four Depth Shades 98
6.8 Safety Depth 100
6.9 Summary on Safety Settings 103
6.10 Colours and Brightness 104
6.11 Presentation of Radar and AIS Information 106
6.11.1 ARPA Targets 106
6.11.2 Radar Overlay 108
6.11.3 AIS Targets 109
6.11.4 Target Data Overview 110
6.12 Presentation of Other Navigational Information 111
6.12.1 Weather Charts 111
6.12.2 Tidal Height 112
6.12.3 Current Flow 113
6.12.4 Ice Charts 115
6.12.5 Temporary and Preliminary Notices to Mariners (T&PNMs) 116
6.12.6 Bathymetric Inlays 117
6.12.7 Piracy Maps 118
7. Own Chart Entries 121
7.1 Electronic Bearing Line & Variable Range Marker 122
7.2 Line of Position / Position Fix 122
7.3 User Charts Additions 124
7.3.1 Objects Associated with Chart 125
7.3.2 Objects Non-Associated with Chart 126
7.4 Event Marker / MOB Marker 126
8. Chart Engines/ System ENC (SENC) 129
8.1 Chart Engines / Kernel 130
8.2 System Electronic Navigational Chart (SENC) 132
9. Chart Logistics, Licensing and Maintenance 137
9.1 WEND / RENC 138
9.2 ENC Distribution Process 138
9.3 IHO Data Protection Scheme 140
9.3.1 Parties Involved in the Protection Scheme 140
9.3.2 Data compression 141
9.3.3 Data Encryption 141
9.3.4 Data Licensing 144
9.3.5 Data Authentication / Integrity Check 146
9.4 ENC Distribution in SENC Format 149
9.5 Chart Licensing Options 149
9.5.1 Direct Licensing 149
9.5.2 Pay as you Sail 151
9.5.3 Licence-free ENCs 152
9.6 Chart Updating 152
9.6.1 Temporary and Preliminary Notices / Navtex Warnings 153
7
9.6.2 Cumulative Versus Sequential Update Strategy 154
9.7 ENC and Update Installation - Summing Up 155
10. Sensor Configuration and Management 159
10.1 Critical Role of Sensors in ECDIS 160
10.2 Sensors Connected 160
10.2.1 Sensor Selection 161
10.2.2 Sensor Alarms 162
10.2.3 ECDIS-Sensor Network 163
10.2.4 Functional Principle, Data Output, and Accuracy of Important Sensors 164
10.3 Sensor References: CCRP and Units 165
10.3.1 Consistent Common Reference Point 165
10.3.2 Sensor Units 166
10.4 Own-Ship Presentation 166
10.4.1 Own-Ship Symbol 167
10.4.2 Heading and Stern Line 167
10.4.3 Own-Ship Vector 167
10.4.4 Predicted Vector / Curved Vector 168
10.4.5 Track History / Past Track 169
10.4.6 Past Position Dots 169
11. Voyage Plan Handling 171
11.1 Voyage Plan Principles 172
11.1.1 Guidelines for Voyage Planning 172
11.1.2 Voyage Plan Appearance on ECDIS 173
11.1.3 Waypoint Parameters 174
11.2 Voyage Plan Creation and Modification 177
11.2.1 Constructing a Draft Route 178
11.2.2 Visual Route Check 178
11.2.3 Automated Route Check 180
11.2.4 Geometric Error 182
11.2.5 Saving the Voyage Plan 183
11.3 ETA Calculations 183
11.4 Voyage Plan States 185
11.4.1 Stored State 185
11.4.2 Loaded State 185
11.4.3 Activated State 186
11.5 Voyage Plan Monitoring 187
11.6 Data Exchange for Track and Speed control 189
12. R ecorder Playback 191
13 Alerts and Warnings 195
13.1 Types of Alerts and their Indication 196
13.2 IMO Required Alarms and Indications 197
13.2.1 Chart Alarms and Indications 197
13.2.2 Route Planning Indications 198
13.2.3 Route Monitoring Alarms and Indications 198
13.2.4 Equipment Alarm and Indication 199
14. Effective Navigation with ECDIS 201
14.1 Potential Errors in Displayed Data 202
14.1.1 Errors Due to Quality of Hydrographic Data / Zone of Confidence 202
14.1.2 Errors in Indication of Own-ship Position 204
14.2 Potential Errors of Interpretation 206
14.3 Potential Errors of the System 207
14.3.1 Overview of Potential System Errors 207
14.3.2 Anomalies of Using ECDIS 208
15. Case Studies 211
15.1 Stranding of MV LT Cortesia 212
15.1.1 Preposition 212
15.1.2 Occurrence Summary 212
15.1.3 Analysis of the Use of the Available Electronic Navigational Equipment 214
15.2 Other Case Studies 215
15.2.1 Grounding of MV CFL PERFORMER 215
15.2.2 Grounding of CSL THAMES in the Sound of Mull 215
16. Annexes 217
A. Abbreviations 218
B. Short Reference List of Modules According to the IMO Model Course 1.27
and Sections of this Book 222
C. List of Figures and Tables 232
D. Overview of Flag State Regulations 236
D.1 Australia: Australian Maritime Safety Authority (AMSA) 236
D.2 Antigua & Barbuda: Department of Marine Services & Merchant Shipping 237
D.3 Bahamas: The Bahamas Maritime Authority 238
D.4 Bermuda: Government of Bermuda, Department of maritime Administration 239
D.5 Cayman Islands: Cayman Registry, a Division of Cayman Maritime 240
D.6 Cyprus: Ministry of Communications and Works, Department of
Merchant Shipping 241
D.7 Germany: BSH – Bundesamt für Seeschifffahrt und Hydrographie 241
D.8 India: Directorate General of Shipping 242
D.9 Ireland: Department of Transport, Tourism and Sport 242
D.10 Isle of Man: Department of Economic Development 243
D.11 Marshall Islands: Office of the Maritime Administrator 243
D.12 New Zealand: Maritime New Zealand 244
D.13 Panama: Panama Maritime Authority 244
D.14 Singapore: Maritime and Port Authority of Singapore (MPA) 245

Huibert- Jan Lekkerkerk
Handbook for Surveyors and Survey Engineers
GNSS Survey &
Engineering
Contents
This book is divided into two parts. The first part (Chapters 1 – 4) is concerned
with the general layout of GNSS, its errors and foremost its practical use.
No in-
depth theoretical background is given, as this part is specifically geared
towards the everyday, professional GNSS user.
Introduction 5
Contents 8
Chapter 1 - GNSS – The system 11
1.1 Introduction to satellite positioning 11
1.2 GNSS components 13
1.3 GNSS Position computation 17
1.4 Augmentation - Differential GNSS 25
Chapter 2 - Preparation and planning 33
2.1 Introduction to preparation and planning 33
2.2 Requirements 33
2.3 Project plan 35
2.4 Mission planning 45
Chapter 3 - Operations 49
3.1 Introduction to GNSS operations 49
3.2 Set-up of base station 50
3.3 Rover installation 57
3.4 Rover set-up 60
3.5 Position check 63
3.6 Data acquisition 64
3.7 Monitoring 69
Chapter 4 - Processing 73
4.1 Introduction to data processing 73
4.2 Data validation 73
4.3 Correcting the Data 75
4.4 Post processing 76
4.5 Digital Terrain Models 79
4.6 Data integration 80
4.7 Data delivery 85
The second part (Chapters 5 – 8) discusses specific topics, such as geodesy, satellite
signals and positioning theory. These topics serve as background to the subjects
from the first part, and are geared towards the more specialized, professional user
or student who needs an in-depth knowledge of GNSS positioning.
8
.
Chapter 5 - Signals 91
5.1 Types of signals 91
5.2 Signal structure 91
5.3 Differential datalink 101
5.4 Differential messages 104
5.5 Antenna – Receiver 105
5.6 Receiver output 108
Chapter 6 - Geodesy 117
6.1 Geodesy for satellite navigation 117
6.2 Geodetic Coordinate Reference Systems 118
6.3 Coordinate transformation 124
6.4 Vertical Coordinate Reference Systems 129
6.5 Projected Coordinate Reference Systems 133
6.6 Temporal Reference System 138
Chapter 7 - Positioning 143
7.1 Introduction 143
7.2 Determining a range from the satellites 143
7.3 Position computation 148
7.4 Corrections to the measurements 152
7.5 Code phase dGNSS 159
7.6 Carrier phase dGNSS 160
7.7 Carrier phase network adjustment 166
7.8 Precise Point Positioning 168
Chapter 8 - Quality Control and Assurance 171
8.1 Introduction 171
8.2 Error theory 171
8.3 Application of error theory to GNSS 175
8.4 Dilution of Precision (DOP) 181
Finally, the last chapter serves as glossary, index, reference and summary in one.
This chapter should be useful for anybody looking for quick information without
having to go through the entire book. It does not present new information though.
Chapter 9 - Reference section 183
9.1 How to use the reference section 182
9.2 GNSS factsheets 184
9.3 Free-to-air augmentation coverage 192
9.4 Message formats 192
9.5 Glossary 198
9
.
1.1 Introduction to satellite positioning
The basis of any positioning system is the requirement to label ‘objects’ in such a
way that we can either describe their location to others or that we can retrieve them
ourselves. To make this work we need a position reference which can either be a
description, an address locator or a coordinate set. In professional positioning and
surveying the coordinate set is the common method for describing object locations.
Besides a set of coordinates we need a reference frame against which we can
measure our coordinates and a positioning device to a) measure the object
location or b) to return to it. A Global Navigation Satellite System (GNSS) such as
GPS helps us by offering a coordinate set (latitude, longitude, height and time)
against a reference frame (WGS84) with our positioning device (receiver).
Satellite positioning and navigation, the subject of this book, comprises all
the systems developed (or under development) for navigation or positioning
purposes based upon satellites. Satellite navigation dates back to the sixties of the
last century and mainly uses trilateration.
1.1.1 Global Navigation Satellite Systems
At the moment, there are four satellite navigation systems available which can be
classified as being able to give global positions, or of working towards becoming
global. We call these Global Navigation Satellite Systems (GNSS). These GNSS are
GPS, Glonass, Compass / Beidou and Galileo. Except for Galileo, all GNSS have their
basis in a military application; GPS and Glonass evolved from earlier systems, such
as Transit (USA), and both became operational during the first half of the 1990’s.
Glonass then slowly deteriorated due to a lack of funds until around 2000; it had
no practical applications anymore.
Starting mid-2000 Galileo and Beidou / Compass were developed with Compass
getting a fast start in China as a regional system. Galileo had a few financial Chapter
1
GNSS - The System
11
1
set-backs as a civilian system but was brought back with the European Union as
main funder. In Russia in the meantime, Glonass was revived and brought back to
full operational capability in 2011.
India has developed its own regional satellite navigation system (IRNSS) which, will
not be further discussed in this book but functions in a similar way to the regional
segment of Beidou.
1.1.2 GNSS applications
The most common GNSS application is navigation for everyday purposes. Almost
all new cars and smartphones have a GNSS system as standard (and often more
than one). As these users are generally not considered professionals in GNSS
applications they fall outside the scope of this book.
Navigation for aircraft and ships, although similar in nature to other navigation
applications, often involves safety critical navigation. As such, pilots and shipborne
navigators obtain additional GNSS training and are expected to have additional
knowledge.
Those involved in construction and mapping as well as precision farming and
timing applications require the utmost performance of their GNSS receiver.
The output needs to be as accurate as possible in at least three but often four
dimensions (including time). Most examples in this book come from construction
and surveying applications.
1.1.3 GNSS positioning principle
Every positioning system is based on either distance (range) measurements,
angular measurements or a combination thereof. In the early days the angular
method was the preferred positioning method using sextants and theodolites.
With the introduction of electronic navigation in World War II, range measurements
became the preferred method, as they still are in GNSS positioning. The steps
involved in computing a GNSS position are:
.. The distance between a minimum of 4 satellites and the receiver on earth is
measured.
.. As a direct distance measurement is impossible, the travel time between the
Figure 1-1: land, hydrographic survey and construction applications
12
1
satellite and the receiver is measured. For this, the satellites have very accurate
(atomic) clocks. The travel time is converted to a range using the speed of the
signal (speed of light).
.. The satellites function as ‘fixed’ reference points for the range measurement.
Because the satellites do not stay in one place, the satellite positions must be
known at all times.
.. During the signal path from the satellite to the receiver it travels through the
atmosphere of the earth. This causes all sorts of large and small errors that
influence the measured travel time and therefore the measured range.
.. A trilateration computation with a minimum of four ranges leads to a four
dimensional position (latitude, longitude, height and time)
1.2 GNSS components
Though all four GNSS are different in construction and initial purpose they are
designed along the same lines. Each GNSS is based around three distinct system
segments, the space, control and user segment. Each segment will be discussed in
more detail in the next paragraphs.
1.2.1 Space segment
The space segment consists of the satellites, or as they are called in GPS, the Space
Vehicles (SVs). Each satellite is built from a number of subsystems amongst which
the most important are the atomic clock(s) and the radio transmitter(s). Additionally,
the satellite has solar panels and batteries for power generation, and rocket
thrusters with fuel for manoeuvring. The crucial elements are usually triplicated to
ensure a continued useful life of the satellite should a single component, such as
an atomic clock, brakes down.
Figure 1-2: Principle of GNSS positioning
13
1
Satellite development is continuous, with new types of satellites with new
capabilities being launched every few years. The main difference between the
satellite generations lies in their ability to send out certain signals, their design life
(extremely long for GPS and rather short for earlier Glonass satellites) and mainly
the orbit they occupy.
For example, GPS has progressed from the first generation (1978 – 1985) with only
two signals, not the most accurate atomic clocks and no autonomous capability
(in case the control segment fails), to a current generation (‘Block IIF’) with three
frequencies, an improved design life, better atomic clocks and 180 days of
autonomous functioning, in case the control segment fails. Just over the horizon
is the next generation of satellites (GPS III), which will add yet other signals and
promises even more accurate positioning.
GNSS satellites occupy either a so-called medium earth orbit (MEO) or an (inclined)
geostationary orbit (I)GSO. The main difference between the two is that a MEO
satellite will circle the earth under an angle at a height of around 20.000 km above
the earth, whereas a GSO stays in one place at a height of around 35700 km (or in
the case of an IGSO, traveling in a figure of eight above a certain place). This makes
the IGSO ideal for a regional system and the MEO mandatory for a global system.
Figure 1-3: The 3 GNSS segments; the space, user and control segment.
Figure 1-4: Galileo [ESA] and GPS [Lockheed Martin] satellites
14
1
As a result of the orbital height of the satellites, they have a certain return period
(time to go through one orbit), in the order of 12 hours for an MEO and 24 hours
for a (I)GSO. The satellites do not fly over the poles but follow an orbit which is at
an angle to the equator (the inclination of the orbit). Depending on the GNSS, this
inclination is between 55° and 65°.
The main effect of the inclination is that a satellite will only fly up to that latitude
above the earth surface. So with an inclination of 55° the maximum latitude that
the satellite will reach equals 55° before returning south / north. As a result, the
closer we get to e.g. the North pole, the larger the number of satellites that will be
Figure 1-5: Left: Parameters of a Medium Earth Orbit; right: track across the earth for Beidou MEO –
red, GSO - green and IGSO satellites – blue
Table 1-1: Details of various (G)NSS
15
1
to the south ánd the closer the visible satellites will fly to the local horizon (lower
elevation), reducing the number of visible and therefore available satellites.
1.2.2 Ground segment
The ground segment consists of the various receivers; both commercial and
the consumer type used in mobile telephones and car navigation systems. Also
included are GPS systems used for timing purposes.
A GPS receiver consists of an antenna, quartz clock, a radio receiver and a
computer for calculating positions. Furthermore, most modern GPS receivers have
a graphical display or controller attached which displays a GPS centred map of
the surroundings. In the case of units created for a mobile telephone all the parts
are self-contained on a single chip. For the commercial types of receiver the parts
are normally larger and separated into different pieces of electronics allowing
replacement and more functionality.
1.2.3 Control segment
The GNSS satellites fly in space and cannot be easily visited for inspection,
maintenance and repair. As a result, they need to be monitored and controlled
from a distance, using earth-based stations. Spread around the world is a series of
monitoring, tracking and control stations. The main function of these stations is to
observe the satellite orbits, monitor the health of the satellite and transmit correct
position information to the satellites.
Without the control segment, the satellites would be nothing but expensive radio
transmitters sending out an unusable signal (or at least after the autonomous
operation period has expired). With the exception of Glonass, all GNSS have their
own control segment with the stations spread around the world. In the case of
Glonass, the entire controls segment is located on Russian soil.
Figure 1-6: Professional GPS receivers (left: series of 5 land survey base stations; right: marine survey
position + heading solution – [Trimble])
16
1
1.3 GNSS Position computation
1.3.1 Trilateration
The process of computing a positing from a set of distance measurements is called
trilateration. Each measured distance will form a circle around the point from
which it is measured. A set of ranges Ra, Rb measured from two known locations
(A, B) will create an intersection and, with a guess (estimated position), this will give
our position P in 2 dimensions (x, y).
Because GNSS satellites travel in space they are at a three-dimensional location.
Therefore a single range does not provide a circle but rather a positioning sphere
with the satellite at the centre and the receiver somewhere on the surface of this
so-called positioning sphere. In a 2D situation a total of two ranges are required;
when a 3D position needs to be computed a total of 3 ranges ánd an estimated
position are required. This estimated position needs to be known to within around
300 kilometres.
Figure 1-7: GPS control segment [gps.gov]
Figure 1-8: Principle of trilateration in 2D
17
1
1.3.2 Range measurement
In a GNSS no direct ranges are measured. Instead the travel time of the signal is
used. The travel time multiplied by the speed at which the signal travels (speed
of light ˜ 300.000 km/s) gives the range used in the trilateration process. With an
average height of around 20.000 km above the earth, this means that the total
travel time is 20/300 of a second, quite a small amount.
The actual range measurement is done using a generated sequence of pulses
called the ranging code. Depending on the GNSS, two or more of these ranging
codes are available at different carrier frequencies. Usually one, less accurate,
code is available to the general public, whereas the other codes are reserved
for governmental use (for example defence purposes). In the early GPS satellites
ranging codes transmitted were the open Coarse Access (C/A) code and the
restricted, military, Precision (P) code.
The accuracy to which the travel time from satellite to receiver has to be determined
is high, for a range resolution of 15 meters a travel time accuracy of 2 nanoseconds
is needed. For the travel time measurement atomic clocks are used which are
accurate to within nanoseconds per day. There are currently three types of atomic
clocks that are used in GNSS satellites; the caesium clock, the passive hydrogen
maser (both around 0.3 s error / million years) or the rubidium clock (3s error /
million years). As the atomic clock is the most important instrument in a GNSS,
each satellite has multiple atomic clocks on-board which can be of different types.
For each clock the stability and accuracy is monitored by the ground segment.
The clock within the GNSS receiver in the user segment is from a different order
than the atomic clocks in the satellites and usually is a simple quartz crystal clock,
such as the type found in a personal computer. As a result, the receiver clock may
contain a small time offset (clock error). Because the clock error is in the receiver
itself, it will influence all the travel time measurements in the same way. Using an
additional range measurement (so four in total, three for the three dimensional
position and the fourth for the clock error) the clock error can be established (see
the Chapter on position computation).
Figure 1-9: Principle of trilateration in 3D
18
1
1.3.3 Satellite positions
Because the MEO satellites used in a GNSS do not stay in a single place but circle
the earth, their position must be known to compute the correct position. The
satellites travel along a predictable path, so there is no need to monitor their
position at every moment in time. The monitoring / tracking stations follow the
satellites in their orbits and use this information to predict the satellite tracks for
the next couple of days to weeks.
The information to compute the position of a satellite at any time is included in
what is called the almanac and ephemeris and is uploaded to the satellite from
the ground segment. The satellites transmit this information back to the receiver in
the user segment where it is used to compute the most probable satellite location.
Both almanac and ephemeris are included in what is called the navigation message
which is broadcast by every satellite.
Due to small changes in the gravitational field of the earth there will also be small
changes in the satellite orbits which cannot be predicted. As a result, the predicted
orbits are never 100% correct. Using the information from the tracking stations
the actual orbit of the satellite can be reproduced at a later moment. Apart from
the tracking stations in the GNSS itself there are many additional tracking stations
in the so-called International GNSS Service (IGS). With the more accurate orbital
information the GNSS position computation can be done at a later moment using
a process called post processing. It is not possible to re-compute the positions in
real-time although a positioning technique such as Precise Point Positioning (see
further) comes very close.
Until May 2000 the GPS position and timing information transmitted by the
satellites was intentionally degraded. This was called ‘Selective Availability (SA)’ and
degraded the GPS accuracy from around 15 – 40 m to between 50 and 100 m.
SA was implemented to prevent foreign, unauthorised, users from obtaining the
same accuracy as that available to military users using the P-code. In May 2000 –
with many alternatives available to civilian users to get better accuracy such as
dGNSS – by presidential order, the application of SA was terminated. While the
current GPS II satellites still have SA without it being active, their predecessors, GPS
III will not have the option.
1.3.4 Signal path
Whilst traveling from the satellite to a receiver on earth, the GNSS signal needs to
travel through various layers of the earth’s atmosphere.
Coming from space the signal first passes through the ionosphere, which is the
layer receiving / blocking most of the electromagnetic radiation from our sun.
Before entering the receiver the signal passes the troposphere, which is the
bottom layer of the atmosphere where the ‘weather occurs’. 19
1
Finally, the signal passes through that part of the atmosphere where human
activity takes place with buildings, trees and other constructions. Each of these
layers has its own challenges and can potentially influence the path of the signal
between satellite and receiver.
Ionosphere
An ion is a charged particle and potentially harmful to live. These ions are
generated by the sun and ‘thrown’ into space. They are stopped by the ionosphere
where they create a ‘charged’ layer. The equator naturally receives most of the solar
radiation, the poles the least. At noon, with sun intensity at its maximum, the ion
content is highest, at midnight lowest. High sun intensity results in a high ion (or
electron) count and an ionosphere that is more charged. A greater Total Electron
Count (TEC) results in the ranging signal being delayed more, resulting in turn in
signal delays with possible errors of 6-30 m.
Figure 1-10: GNSS signal path
Figure 1-11: Total Electron Content (TEC) of the ionosphere at a given moment [Trimble]
20
1
Each individual GNSS has its own ionospheric model that compensates for the
delays caused by the TEC level. Manufacturers may decide to replace the model
in the GNSS specification with one of their own choice. When a multi-frequency
receiver is used, such as a dual frequency professional survey grade RTK receiver,
the ionospheric delay can be compensated based on differences in the observed
travel time between different frequencies.
Another effect caused by ionization is scintillation where the ions fluctuate in phase
and with similar amplitude as the GNSS signals possibly causing a loss of signal
tracking. These effects cannot be corrected for, not even with a multi-frequency
receiver. The net effect is usually local and may occur in equatorial regions after
sunset and in the Polar Regions at any given time.
When the TEC level becomes too large, no signals are transmitted through the
ionosphere to the receiver. The main source of such a large TEC level is a solar storm.
Solar storms always occur but are usually more intense in periods where large
numbers of sunspots are seen on the sun. The number of sunspots has a harmonic
cycle of approximately 11 years with the most recent peaks in 2001 and 2014.
Troposphere
Due to weather and climatic variations in the troposphere there are variations in
humidity, temperature and pressure. These variations affect the effective speed of
light which is only a constant in a total vacuum. The variations in the speed of light
Figure 1-12: amount of sunspots since January 1984, with a forecast up to 2020 [NOAA]
21
1
result in both delays as well as refraction of the signal along its path.
GNSS receivers employ a model of the earth climate to compensate for climatic
variations. The tropospheric model cannot cope with day to day weather changes
which as a result still affect the measured travel time.
Because the problem is larger for a signal travelling for a longer time through the
troposphere, the signals coming from directly overhead experience the smallest
error. Satellites close to the horizon will have the longest travel time and therefore
the largest errors.
To avoid too large an error from the troposphere, it is common to only use satellites
that are 10° to 15° above the local horizon. This is called setting an elevation mask.
It is good practice to start with an elevation mask of 15°, and only reduce the mask
if not enough satellites can be received. An elevation mask larger than 15° is not
practical, as fewer satellites will be received and the final geometry of the satellites
will not be favourable.
Shielding
A GPS satellite transmits a relatively low amount of power. Combined with the relatively
small reception antennas, this means that the amount of signal reaching the antenna
(the Signal to Noise ratio) is relatively low. Furthermore, the system is ‘line of sight’ i.e.
the signals travel along straight paths and will not curve around objects.
As a result, even a small object blocking the path of the satellite signal will prevent
it from reaching the antenna. Due to this shielding the position calculation will be
based on fewer satellites.
Multipath
Just as light reflects of a shiny surface, radio signals can be reflected by e.g. the
water surface, tanks filled with oil and water but also from cars and ships bridges.
A very common source of multipath, which is not always obvious, is the chain link
Figure 1-13: elevation mask and the path through atmosphere.
22
1
fence where the gaps between the links are smaller than the carrier wavelength of
the GNSS signal (around 0.20 cm -most link fences will qualify). To the GNSS signals,
such a fence has the same effect as a steel plate without holes, and is therefore a
perfect reflector.
The reflected signals may interfere with the signals that are received via a direct
path. If the direct signal is obstructed, the receiver may start using the reflected
signal, which has a longer travel time, instead of the direct signal. As a result the
position will be calculated incorrect, with the position shifting in the direction of
the multipath source.
It is hard to correct for multipath, and as such it is better to prevent it altogether.
The first rule in preventing multipath is to keep the antenna as far away as possible
from any object that can function as a reflector. Changing the antenna height
sometimes helps. As multipath usually comes from below the horizon or at
relatively small elevation angles it is an advantage to block these signals before
they can even reach the antenna.
Applying an elevation mask may help, but better is blocking these signals
physically. For that reason most professional GNSS antennas have a built in ground
plate or choke ring that prevent the reception of reflected signals from under
the antenna horizon. For static measurements the large plate or choke ring is
preferred, whereas on dynamic platforms a small ground plate is to be preferred.
On a dynamic platform a large ground plate may shield too much of the horizon
during platform movement.
Figure 1-14: multipath due to reflection against steel shed. Inset shows the effect of multipath on a
receiver travelling through an urban area (blue path without multipath, green with)
23
1
A multipath error will usually have duration of a few minutes and will disappear as
soon as the signal is no longer reflected towards the antenna. When the satellite is
exactly in the same place (and considering that the receiver and multipath source
are still in the same place) the multipath will reoccur.
1.3.5 Positioning accuracy
As seen in the previous paragraphs, there are a number of potential error sources
in any GNSS. Each of these sources has a contribution to the overall so-called User
Equivalent Range Error (UERE).
.. Clock error and orbital errors in the predicted satellite orbits ˜ 2.3 m
.. Ionospheric error ˜ 0.1 (dual frequency - 7.0 m (single frequency)
.. Tropospheric error ˜ 0.2 m
.. Multipath and shielding ˜ 1.5 m
.. Receiver noise ˜ 0.6 m
The total UERE is between 2.8 m (dual frequency receiver) – 7.5 m (single frequency
receiver). The errors can be mitigated employing a better quality receiver or by
post processing. This is the potential error in an individual measurement, not the
error of the position on the ground.
The accuracy of the position calculation not only depends on the accuracy of
the UERE but also on where the satellites are located relative to the receiver. The
“Dilution of Precision” or DOP describes the strength of the satellite configuration.
It gives an indication on the spreading of satellites around the receiver antenna
and above the local horizon.
Figure 1-16: left: high DOP value (bad geometry); right: low DOP value (good geometry)
Figure 1-15: left, middle: antennas with ground plate; right: with choke ring
24
1
When satellites are not evenly spread around the antenna then the DOP value will
be high. When evenly spread around the antenna the DOP will be low with a value
of 1 in the ideal situation. There are several DOPs used for quality control, the most
important are the Geometric DOP (GDOP – X,Y, Z and time) and the Horizontal
DOP (HDOP – X and Y).
Because the GNSS satellites travel no further North / South than 55° - 65° of latitude,
the satellites cannot be evenly spread. As a result, the accuracy of the GNSS position
is usually slightly worse in North-South direction than in East-West direction. This
effect will increasingly affect the position accuracy in higher latitudes; around 52°
N the observed effect is a factor of 2 for the N-Z precision in GPS.
1.3.6 Integrating multiple GNSS
As shown, a minimum of four satellites are required to perform a position
computation. These satellites should all be of the same GNSS. If more satellites
from different systems are received, a modern receiver can integrate all these
measurements into a single position solution.
At the moment both Glonass and GPS combined receivers can be used anywhere
in the world. Around Asia, Beidou can be used to augment GPS, Glonass or both.
In the near future there will be enough Galileo satellites to allow the possible
integration of four GNSS.
The main advantage of such integration is that receivers will be faster with their
first position (called the Time to First Fix), have improved accuracy and greater
reliability. The disadvantage is that more expensive receivers are required
(although all modern professional receivers are capable of receiving at least GPS,
Glonass and Galileo and usually also Beidou) and that the noise levels and possible
radio interference will increase with so many satellites transmitting in the same
frequency bands.
1.4 Augmentation - Differential GNSS
Stand-alone GNSS is usually not accurate enough for professional use which is why
it is commonly augmented with other information / sensors. The most common
augmentation technique is differential GNSS (dGNSS). In dGNSS a separate
antenna is placed over a benchmark or control point (base station) to measure
any GNSS errors and transmit that information to the user in the field (rover). All
augmentation systems aim to reduce or mitigate the common GNSS errors. In
addition, they usually give extra insight into the accuracy of the position solution
and are a means to warn the user for any outages / badly functioning satellites. In
this chapter the following augmentation techniques are discussed.
25
1
Differentially augmented GNSS (dGNSS).
Uses a (network of ) base stations to correct
any GNSS errors. Comes in two types:
.. Code phase dGNSS
.. Carrier phase dGNSS
The main difference between code phase
dGNSS and carrier phase dGNSS is the
part of the GNSS signal they use and the
accuracies achieved.
Absolute Precise Point Positioning (PPP),
which is an augmentation technique
where the satellite orbits and clock errors
are observed in detail at a number of base
stations. Using this very accurate orbit and
timing information the rover receiver then
computes an accurate standalone position
Aided GNSS, where a GNSS or dGNSS
receiver is combined with heading and
speed (or acceleration and angular) sensors.
Depending on the type of aiding the
sensors are fully integrated into a so-called
Inertial Navigation Solution or are used to
verify the correct positioning of the GNSS.
1.4.1 Differential GNSS
The most common form of differential augmentation is that of code phase dGNSS
(for example dGPS). Code phase dGNSS was developed to counteract Selective
Availability in e.g. surveying applications, and uses for example, the C/A ranging
code to determine the position of the receiver, whereas carrier phase dGNSS
uses the carrier frequency of the signal (see the chapter on Signals). Carrier phase
dGNSS has accuracy in the order of mm to cm(s), whereas code phase dGNSS has
an accuracy of dm(s) to m(s). Other classifications of dGNSS systems are generally
based on the area over which the corrections are valid:
.. Wide Area Augmentation Systems (WAAS) cover an area the size of for example
a continent or an ocean.
.. Local Area Augmentation Systems (LAAS) cover a small area, for example a
construction site or an airfield.
Or dGNSS is classified based on the location of the station transmitting the
corrections:
.. Space Based Augmentation Systems (SBAS) employ satellites to transmit the
corrections towards the mobile receivers. Usually SBAS systems are also WAAS
systems but not necessarily the other way around!
26
1
.. Ground Based Augmentation Systems (GBAS) employ transmitters at ground
level. These transmitters can either transmit over radio or (mobile) telephone line.
The main implication of a classification, other than by the part of the signal
used (code or carrier), is only relevant when discussing the means of getting the
corrections to the rover receiver, and as such is discussed further on.
Code phase dGNSS
Every dGNSS system uses a base station which is an additional receiver whose
antenna is located as accurately as possible above a ‘known’ point; usually a
benchmark or control point. In code phase dGNSS the base station observes the
range towards the visible satellites as described for regular GNSS positioning. Due
to its placement over a known point the true range can be computed as well.
The difference between the computed and the observed range signifies the
measurement error or range correction (.Rx). This correction can be transmitted
to one or more mobile receivers (‘rover’). At the rover receiver the range correction
for a specific satellite is applied to the measured range from that satellite. Because
a correction is applied to a measured range, the base and rover need to measure
towards the same set of satellites. As a result at least four common satellites must
be visible at both the base and rover location.
Although in theory the accuracy of a code phase dGNSS is only depending on the
accuracy of the control point, in practice other factors also play a role:
.. Noise in the base station (and the rover) receiver
.. Different atmospheric conditions on base and rover location
.. Different satellites in view on both locations (due to shielding or a large distance
apart).
Figure 1-17: principle of code phase dGNSS
27
1
Carrier phase dGPS
As implied by its name, carrier phase dGNSS uses the carrier frequency of the GNSS
signal rather than the ranging code superimposed on it. For example with GPS C/A
code is put on a carrier frequency with a wavelength of around 0.19 m.
Carrier phase dGNSS uses this carrier signal to measure the range rather than
the ranging code superimposed on it. By counting the number of cycles and the
phase of the carrier wave, the range can be determined. Due to the long range
between the satellite and receiver there are many waves of 0.19 m length. The
determination of the total number of waves, a process called integer ambiguity
resolution, is a tedious process requiring much computation power.
In order to resolve the integer ambiguity (and to determine the whole number of
cycles between a receiver and the satellite, an intricate process called differencing
is performed. Measuring towards at least 5 common satellites between base and
rover, the receivers solve for a number of errors / unknown factors (satellite clock
and orbit, receiver clock, atmospheric error and integer ambiguity). The more
satellites are being received and the less noise there is in the environment, the
faster the integer ambiguity resolution process will go.
When both base and rover are capable of receiving multiple GNSS then satellites
from other GNSS can be combined into a single solution. For each additional GNSS
one additional satellite is required as a minimum. So if GPS and Glonass are used
together, 5 GPS satellites and 1 Glonass satellite are required (or vice versa).
Figure 1-18: problem of integer ambiguity resolution
28
1
The result of the carrier phase initialisation process is not a series of range
corrections but rather the difference between the base and rover location (.X, .Y,
.Z). This is called relative positioning and requires that the base position (Xb, Yb,
Zb) is known at the rover position to find the final position of the rover receiver (Xr
= Xb + .X etc). For land surveyors this is a not uncommon method of positioning
as the Total Station uses a similar method. An advantage over Total Station is,
however, that carrier phase dGNSS does not require line of sight and can cover
much larger distances. The Total Station, if used correctly, is however much more
accurate (mm level rather than cm level for carrier phase dGNSS).
The concept of carrier phase dGNSS can be applied in different ways, depending
on how initialisation is achieved and whether one can get a measurement ‘on the
move’. The most accurate results are generally obtained with static measurements
where raw data is collected in the field and then post-processed in the office.
On the other end of the spectrum is Real Time Kinematic (RTK) On The Fly (OTF)
where the initialisation and measurement is done in real time whilst the antenna
is allowed to move during initialisation (making the integer ambiguity solution
much harder). Where static measurements can be used to obtain sub centimetre
level accuracies, RTK OTF usually generates results at the centimetre level.
The vector between the base and rover antenna also allows for different types
of application such as GNSS heading computation. In this case, the base and
rover antenna are mounted close together on the same platform or even in the
same sensor housing. Using an accurately measured baseline between the two
antennas it is possible to compute the orientation of the baseline between base
Figure 1-19: principle of carrier phase dGNSS
29
1
and rover relative to true North, without requiring the actual position of the base
antenna. The achieved accuracy is in the order of 0.1° – 0.7°.
Another use is to position two objects relative to each other. An example is
offshore seismic surveying, where the base station is mounted on the seismic
vessel which is towing so-called seismic streamers of a few kilometres long. At the
end of the streamer there is a tail buoy with a roving receiver mounted on it. The
position of the tail buoy can now be measured relative to the base antenna on the
towing seismic vessel. When the position of the base station is made known using
e.g. a code phase dGNSS system, all sensors can be positioned accurately to the
decimetre level while maintaining a relative positioning accuracy of centimetres.
1.4.2 Precise Point Positioning
The concept behind Precise Point Positioning (PPP) is that a regular GNSS receiver
would be much more accurate when more accurate orbital information is
available. In the case of PPP, accurate orbital and atmospheric data is distributed
from a network of monitoring stations. Using the more accurate data the receiver
computes a stand-alone position. A common source of orbital information is the
International Geodetic Service (IGS) comprising a number of scientific stations
around the world. An alternative are commercial services such as those hosted by
Fugro, Veripos and CNav.
If, in addition to the accurate data, the carrier phase information is used as well, we
speak of PPP Ambiguity Resolution (PPP-AR). In the future it may even be possible
to combine regular carrier phase measurements with PPP as PPP-RTK.
PPP can give very accurate (dm level) positions after an initial solution has been
acquired. The time to the first solution is generally over 20 minutes for a high
accuracy fix; when the satellite signal is lost a reacquisition takes about the same
amount of time. If a PPP-AR solution is used the time to first fix stays the same, but
the reacquisition can be reduced to seconds.
1.4.3 Aided GNSS
Aided GNSS is an integration of GNSS with auxiliary sensors, such as compasses,
acceleration and rate of turn sensors. Using data from these auxiliary sensors the
GNSS is improved upon, and in the case of high quality sensors, it is even possible
Figure 1-20: left: heading GNSS; right: seismic survey set-up [govt.nz]
30
1
that the position computation is continued during (small) periods of GNSS outage
using Inertial Navigation System (INS) capabilities of the other sensors.
In order to make this work, the GNSS positions are integrated with the other sensors
through a so-called coupling model. Depending on the type of model, the GNSS
positions are taken as ‘true’ and the sensors are used to assist in the computation
between GNSS fixes (loose coupling), or the solution is based on a mix of all the
data (tight coupling).
1.4.4 Supplying augmentation signals
In all augmented GNSS solutions, with the exception of Aided GNSS, a base station
or a number of base or monitoring stations are required for determining the
correction data. The data then needs to be transmitted to the rover for use. For
small projects, the single base station is usually the most cost-effective solution.
Such a single base station can be owned and set-up by the organisation executing
the project, or alternatively the signal can be obtained from a base station provided
by third party. Third party systems can be Free to Air or commercially operated.
Free to Air systems can be either carrier or code phase but are commonly code
phase augmentation systems. Existing Free to Air systems are the land-based IALA
beacons located around the world at major ports, and the government operated
Space Based Augmentation Systems, such as the European Egnos or the American
WAAS. Commercial suppliers, such as Veripos and Fugro, supply both code phase
corrections as well as PPP solutions.
A single base station consists of the following elements:
.. High quality GNSS receiver with antenna
.. Well know control point / benchmark
.. Telemetry to transfer corrections from the station to the user(s)
.. Power source
The antenna is set-up over the known control point or benchmark. The receiver
Figure 1-21: Single base stations with radio telemetry set-up
31
1
is connected to both the power source and the antenna. The telemetry system is
connected to the output of the receiver.
For larger areas the corrections can be improved upon by using a network
solution. Network solutions are usually found in commercial code phase and
carrier phase networks. In a network solution the corrections from multiple
stations are combined into a single correction that can be tailored to the specific
location of the receiver or is a better average correction than could be obtained
from a single station. In a network solution, either the receiver itself combines the
corrections from multiple base stations or the rover position is transmitted back to
the network, which in return transmits the corrections in such a way as if a local,
virtual, reference station (VRS) is available.
Figure 1-22: Concept of a virtual reference station (background: Fugro correction station network
– [Fugro])
32
1