sherlock holmes I N T R O D U C T I O N

The topic of these lecture notes of the International School GPS for Geodesy is
the description of the use of the Global Positioning System (GPS) measurements
for geodetic applications. The term geodetic applications is used in the sense that
it
covers the determination of precise coordinates for positions in a well defined
reference system, and the monitoring of temporal changes of these coordinates.
These lecture notes are organized in ten chapters, each of which begins with the
full address of the author(s), and a section introducing the theme of the chapter.
After the main body of text, each chapter is concluded by a summary section and a
list of references. The individual chapters have been written independently, and
they can also be read and studied independently. Their sequence, however, has
been arranged to provide a logical and coherent coverage of the topic of GPS for
G e o d e s y .
Chapter 1 introduces global reference systems for Cartesian and ellipsoidal
coordinates and local reference frames, and their basic relation to the GPS
measurements. Transformations between and motions of the Celestial and
Terrestrial Reference Frames are described. Time systems are introduced to
provide an independent variable for the description of motion and earth
deformation. The concepts and realizations of Conventional Reference Systems are
explained.
The topic of Chapter 2 is the description of the computation of GPS satellite
orbits, and the dissemination of GPS satellite ephemerides. Starting from the
equations of motion for satellites, first the Keplerian orbit is introduced and
then
generalized to include the perturbations resulting from non-central forces. Various
sources of orbital information to GPS end users are described, and the chapter is
concluded with a brief discussion of the effect of unmodelled orbit errors on
positions determined from GPS measurements.
Chapter 3 introduces the GPS signal, its components, and its generation in the
satellites' circuitry. The aspects of signal propagation from the satellite to the
GPS
receiver are described, including the effects of refraction, multipath, and
scattering.
Chapter 4 begins with an introduction to the basic building blocks of a GPS
receiver, and shows how pseudoranges and carrier phases are being measured in
the receiver circuits. The chapter is concluded with a discussion of the
measurement errors in these two observables.
Chapter 5 starts from the complete non-linear observation equations for
pseudoranges and carrier phases and introduces a number of different linear
combinations, in order to eliminate, reduce, and/or emphasize parts of the
equations. Following this exploratory analysis, the observation equations are
linearized with respect to the parameters to be determined. Basic properties of the
linearized equations in the context of single point positioning and relative
positioning are discussed, with particular emphasis on parameter estimability.
Chapter 6 begins with a description of the pseudorange observation in terms of
geocentric coordinates, and proceeds to discuss single site solutions through
linearization of the observation equations. Also included is a presentation of
thedirect solution of pseudorange equations without the requirement of a priori
information. The concept of dilution of precision is introduced. The chapter
concludes with a description of carrier phase and pseudorange combinations for the
reduction of pseudorange noise.
Chapters 7 through 10 present details on the use of GPS measurements for the
spectrum of geodetic applications. There might be a number of ways of structuring
these applications; we have chosen to use the network scale as the criterion.
Accordingly we begin with a chapter on short distance GPS models. In the context
of these Lecture Notes, "short" means that atmospheric and orbital errors do not
significantly affect the accuracy of the positioning result, and do not have to be
included explicitly.
The determination of short baselines and small scale networks with GPS typically
exploits the integer nature of carrier phase ambiguities, and thereby reduces the
required observation time considerably. The process of finding and validating the
correct integer values, often referred to as "ambiguity fixing", is not a trivial
one.
Chapter 8 outlines various strategies to search for and identify integer carrier
phase
ambiguities in the context of least squares estimation algorithms.
In order to retain the full accuracy capability of GPS in networks of larger scale
(typically between 50 km and 1000 km), the atmospheric refraction effects and
inaccuracies of the GPS satellite orbits need to be explicitly included. The
corresponding mathematical models for the GPS measurements and procedures for
the estimation of geodetic parameters are outlined in Chapter 9.
Finally, Chapter 10 discusses the Global Positioning System for geodynamics
applications on a global scale. While primarily discussing the estimation of
various
parameters of interest, this chapter also closes the circle by connecting to
Chapter 1.
The determination and maintenance of global reference systems with GPS is
intrinsically connected to the applications discussed in this last chapter of the
lecture
notes.1. REFERENCE SYSTEMS
Yehuda Bock
Institute of Geophysics and Planetary Physics, Scripps Institution of
Oceanography, University of California, San Diego, 9500 Gilman Drive, La Jolla,
California, 92093-0225 U.S.A.
1.1 INTRODUCTION
Of fundamental importance in space geodetic positioning is the precise definition
and realization of terrestrial and inertial reference systems, It is appropriate
then
that this topic be covered in the first chapter of these notes on the Global
Positioning System (GPS).
As its name implies, the purpose of GPS is to provide a global absolute
positioning capability with respect to a consistent terrestrial reference frame.
The
original intention was instantaneous and global, three-dimensional position with 1-
2
meter precision, for preferred users. Ordinary users would be allowed 1-2 orders
of magnitude less precision. Geodesists realized, at least 15 years ago, that GPS
could be used in a differential mode, much like very long baseline interferometry
(VLBI), to obtain much more accurate relative positions. Relative positioning with
1 mm precision was demonstrated in the early 1980s using single-frequency
geodetic receivers over short distances (100's of meters). Precision decreases,
however, in approximate proportion to intersite distance, about 1-2x 10.6 (1-2 ppm)
circa 1983, primarily due to satellite orbital errors and ionospheric refraction.
Between the years 1983 to 1992, geodesists have been able to attain about an order
of magnitude improvement in horizontal precision about every three years (Table
1.1). Vertical precision has also improved and, as a rule of thumb, has always
been about 3-4 times less precise than the horizontal, although with less
dependence
on baseline length at distances greater than several tens of kilometers.
T a b l e 1.1. I m p r o v e m e n t s in horizontal precision and limiting error
sources.
Y e a r b tppm)
- 1983 1
~ 1986 0.1
~ 1989 0.01
~ 1992 0.001
Primary Limiting Error Sources
ionospheric refraction, orbital accuracy
orbital accuracy
orbitalaccuracy
reference systems, station specific errors
o a ~ t a (ram2)= (0.1 - 1.0 mm) 2 + [(2 � b) sli0cm)]2
sij is the distance between sites i and j
One part per billion precision corresponds to 1 cm over a 10,000 km line.
Therefore, GPS can be considered today a global geodetic positioning system
(GGPS) providing nearly instantaneous three-dimensional position at the 1-2 cm1.
ReferenceSystems 4
level for all users, with respect to a consistent global terrestrial reference
system.
These dramatic improvements could not have been achieved without full
implementation of the GPS satellite constellation, expansion and improved global
distribution of the worldwide GPS tracking network, determination of increasingly
accurate positions and velocities for the tracking stations (and in turn improved
satellite orbit determination), and advancements in geodetic GPS receiver
technology.
1 . 1 . 1 Basic GPS Model
The geometric term of the model for GPS carder phase can be expressed in simple
terms as a function of the (scalar) range p~
fik(t) = ~[~(t,t-- tik(t))] = ~ I r k ( t - t i k ( t ) ) - r i ( t ) l (I.I)
where x~(t) is the travel time of the radio signal, r i is the geocentric vector
for
station i at reception time t, r k is the geocentric vector for satellite k at
satellite
transmission time (t - x~(t)), f0 is the nominal signal frequency and c is the
speed
of light. The station position vector is given in a geocentric Cartesian reference
frame as
[Xi(t)]
ri(t) = / Yi(t) /
[ Zi(t) ]
(1.2)
The equations of motion of a satellite can be expressed by six first-order
differential
equations, three for position and three for velocity,
d ( r k) = t k (1.3)
d~t(i -k)= ~ rk + ~kerturbing (1.4)
where G is the universal constant of attraction and M is the mass of the Earth. The
fin-stterm on the right-hand side of (1.4) contains the spherical part of the
Earth's
gravitational field. The second term represents the perturbing accelerations acting
on the satellite (e.g., non-spherical part of the Earth's gravity field, luni-solar
effects and solar radiation pressure).
In order to difference the station vector and satellite vector in (1.1), both must
be
expressed in the same reference frame. The station positions are conveniently
represented in a terrestrial (Earth-fixed) reference frame, one that is rotating in
some
well defined way with the Earth. Solving the equations of motion of the GPS
satellites (i.e., orbit determination) requires a geocentric celestial (inertial)
reference5 YehudaBook
frame. In order to compute (1.1), either the station position vector needs to be
transformed into a celestial frame or the satellite position vector needs to be
transformed into a terrestrial reference frame. Furthermore, fundamental concepts
of time epoch, time interval and frequency must be rigorously described and
fundamental constants (e.g., speed of light and the GM value1) must be defined.
I. 1.2 The Fundamental Polyhedron
As we shall see in this chapter, the orientation of the Earth in space is a
complicated
function of time which can be represented to first order as a combination of time
varying rotation, polar motion, a nutation and a precession. The realization of
celestial and terrestrial reference systems are quite involved because of the
complexity of the Earth's composition, its interaction with the atmosphere, and its
mutual gravitational attraction with the Moon and the Sun. The definition of the
terrestrial reference system is complicated by geophysical processes that make the
Earth's crust deform at global, regional and local scales, at a magnitude greater
than
the precision of present-day space geodetic measurements. The definition of the
celestial reference system is complicated by the fact that stellar objects have
proper
motions and are not truly point sources.
The realization of a reference system is by means of a reference frame, i.e., by a
catalogue of positions which implicitly define a spatial coordinate system. The
celestial reference system is realized by a catalogue of celestial coordinates of
extragalactic radio sources determined from astrometric observations (VLBI).
These coordinates define at an arbitrary fundamental epoch a celestial reference
frame (CRF). The terrestrial reference system is realized through a catalogue of
Cartesian station positions at an arbitrary fundamental epoch, to, i.e.,
[r(t)] o = [X(t), Y(t), X(t)] 0 (1.5)
determined from a combination of space geodetic observations, including satellite
laser ranging (SLR), VLBI, and GPS. These positions define a fundamental
polyhedron. Implicit in the coordinates of its vertices are conventional spatial
Cartesian axes that define the terrestrial reference frame (TRF). Maintaining the
reference frame means relating the rotated, translated and deformed polyhedron at a
later epoch to the fundamental polyhedron. Those motions that are common to all
stations (i.e., rotations and translations) define the relationship between the
polyhedron and the celestial system. The deformations of the polyhedron, by
def'mition those motions that do not contain any rotations or translations, relate
the
polyhedron to the terrestrial system. Earth deformation is accommodated, at least
to
fn'st order, by supplementing the station catalogue with station velocities derived
from global plate models and/or long-term geodetic measurements. The reference
frame does not change (unless a new one is defined, of course). It is fixed to the
station positions (the polyhedron) at to and consists of a set of spatial Cartesian
axes