VII SIRGAS School Sto Domingo 2015 P

SIRGAS Scientific Council
Hermann Drewes
International Association
of Geodesy
Munich
Germany
SIRGAS President
Claudio Brunini
Universidad Nacional de La Plata,
Consejo Nacional de Investigaciones
Cientficas y Tcnicas
La Plata
Argentina
SIRGAS Vice-President
Laura Snchez
Deutsches Geodtisches
Forschungsinstitut der
Technishen Universitt
Mnchen
Mnchen, Germany
School on Reference Systems, Crustal

Deformation and Ionosphere Monitoring
1. Contents and introduction
2. Geodetic reference systems and frames
3. Coordinates determination from GNSS
4. Vertical reference systems
5. Crustal deformation, observation and modelling
6. Ionosphere modelling and analysis
7. Reference system and frame for the Americas (SIRGAS)
Aperture
Registration
Monday, November 16th, 2015

07:00-08:30
Adjustment of GPS networks
Tuesday, November 17th, 2015
Vertical reference systems
Coffee break
Crustal deformation observation and

modelling
Lunch
Ionosphere modelling and analysis
SIRGAS: definition, realisation, maintenance
Coffee break
GPS positioning: observation equations and

error analysis
Terrestrial Reference System and Frame

Regional Reference Frames
Types of coordinates
Celestial Reference System and Frame
Rotation and tides of the Earth
08:30-09:00
09:00-10:30
10:30-11:00
11:00-12:30
12:30-14:00
14:00-15:30
15:30-16:00
16:00-17:30
2. Geodetic reference systems and frames
Contents:
2.1 Types of coordinates and transformations
(Cartesian, ellipsoidal, topocentric, plane)
2.2 Definition of reference systems and frames
2.3 Celestial (inertial) reference system and frame (ICRS, ICRF)
2.4 Rotation and tides of the Earth
(nutation, polar motion, length of day, tides)
2.5 Terrestrial reference system and frame (ITRS, ITRF)
2.6 Regional and national reference frames
VII SIRGAS School on Reference Systems, Santo Domingo, Dominican Republic, 16-17 November 2015
2.1 Types of coordinates and transformations

A coordinate system provides the basis for unique determination of
the position of points in lines, surfaces or spaces (1D, 2D, 3D).
To define a coordinate system one has to specify
0. the type of coordinates (rectilinear, curved, plane, spatial);
1. the location of the origin;
2. the orientation of the axes;
3. the unit of measure.
Coordinates cannot be determined directly but only w.r.t. a system!
We shall discuss the following types of coordinate systems:
Global three-dimensional Cartesian coordinates [X, Y, Z]
(OOLSVRLGDOFRRUGLQDWHV>O, h]
Local topocentric Cartesian coordinates [x, y, z], [n, e, u]
Plane coordinates (Mercator, Lambert, azimuthal) [N, E]
Coordinates referring to irregular surfaces [H] (see chapter 4)
Pole
We distinguish right-hand and

left-hand systems. In geodesy we
use the right-hand system.
Cartesian systems have rectilinear

orthogonal axes.
Global Cartesian 3D coordinates
Greenwich
In global systems, the origin is the

centre of mass of the Earth;
the direction of the Z-axis is
towards the conventional pole of
rotation, the direction of the X-axis
is close to the Greenwich meridian;
the unit of measure is the metre.
These parameters are called datum.
Transformation of Cartesian coordinates (1)
T1 0 R3 R2 X
X
X

0 R1 * Y M * Y
Y T2 R3
Z ( P ) T3 R2 R1 0 Z ( P )
Z ( P )
A coordinate transformation is the change of coordinates from

one coordinate system to another one with a different datum.
A coordinate transformation applies parameters which can be
derived empirically from a set of identical points in both systems.
For undeformed coordinate sets in three-dimensional Cartesian
systems it is common to use the 7-parameters similarity
transformation (after Helmert 1893):
X

Y
Z ( D )
R2 X
T1,T2,T3: Translations in X, Y, Z
T1 1 M ' R3

Ri: Rotations around X-,Y-,Z-axes
T2 R3 1 M ' R1 * Y
R1 1 M ' Z ( P ) M: Factor of unit measures (scale)
T3 R2
1 RZ RY X ' X ' p 'X
M RZ 1 RX Y ' Y ' p 'Y
RY RX 1 Z ' Z ' p 'Z
X', Y', Z' = initial system

X, Y, Z = new system
X, Y, Z: origin coordinates
of the initial in the new system.
RX, RY, RZ : Rotations of axes.
X'P, Y'P, Z'P: Average of point
coordinates in the initial system.
M: Factor of measures (scale).
The Helmert transformation

produces correlations of T, R, M
parameters if the centre of points
(i.e. the average of coordinates)
differs from the origin.
To avoid this correlation, the
transformation of MolodenskyBadekas is used:
Transformation of Cartesian coordinates (2)
X

Y

Z
sin X cos O
r sin X sin O
cosX
cos M cos O
r cos M sin O
sin M
Global spherical coordinates can

easily be converted to Cartesian
coordinates X,Y, Z.
The shape of the ellipsoid is given

by semi-mayor axis (a) and polar
flattening (f) or eccentricity (e).
7
The figure of the Earth is better

approximated by an ellipsoid than
by a sphere, because the Earths
rotation produces a flattening with
semi-minor axis (b) about 21 km
shorter than semi-mayor axis (a).
Y
In geodesy we use an ellipsoid that
optimally fits the geoid according
to the Gauss-Listing definition, i.e.
coinciding with the mean sea level
in static equilibrium.
Ellipsoidal coordinates (1)
a b
a
a2 b2
a
A global spherical system is

preferred (and sufficient) for many
representations, e.g. the gravity field,
global geography, kinematics of
tectonic plates, etc.
Y r : radial geocentric distance,
P
P: latitude,
P: polar distance (co-latitude 90-P),
OP: geocentric longitude.
Global spherical and ellipsoidal coordinates
X

Y
Z
a
p
Eccentricity:
Polar flattening:
&
r
X

Y
Z
Zp
Ellipsoidal coordinates (2)
Q
p
Yp
The ellipsoidal coordinates can be

converted to Cartesian coordinates
X,Y,Z using the meridian radius N:
N = a / (1 e2 sin2 0.5
The ellipsoidal surface system is

extended into space by inclusion of
the (ellipsoidal, geometric) height h
of point P above the ellipsoid,
measured along the normal vector.
Y The projection of P upon the
ellipsoid corresponds to point Q.
The coordinates triad (, , h) is
called curvilinear coordinates.
( N h) cos M cos O
( N h) cos M sin O
2
((1 e ) N h) sin M
Q
p
Yp
Zp
A conversion is the change between

two types of coordinates referring to
the same datum.
Conversion of Cartesian coordinates
to ellipsoidal coordinates is done by
an iteration process:
atan (Z+ e2 1VLQ;2+Y2)0.5

atan Y/X
K ;VHFVHF N
7RDYRLGWKHLWHUDWLRQRQHFRPSXWHVE\
atan>=EVLQ3q) / (p e2 a cos3q)]
Conversion of Cartesian to ellipsoidal coordinates
H2 / (1 e2)
b = a(1 f)
p = (X2 + Y2)0.5
q = atan[(Z a) / (p b)]
Transformation of ellipsoidal coordinates

The change of ellipsoidal coordinates from a source system to a
transformed system can directly be handled like a displacement
(offset) using the formulae of Molodensky:
t = s + d; Ot = Os + dO, ht = hs + dh
with:
d" = ( dX sin s cos Os dY sin s sin Os + dZ cos s + (a df + f da) sin2s) / (rs sin 1")
dO" = ( dX sin Os + dY cos Os) / (ns cos s sin 1")
dh = dX cos s cos Os + dY cos s sin Os + dZ sin s + (a df + f da) sin2s da
where:
dX, dY, dZ = parameters of geocentric translation.
da
= difference of semi-mayor axes of the transformed and the source ellipsoid.
df
= difference of the flattening of the ellipsoids.
da = at as,
df = ft fs = 1/(1/ft) 1/(1/fs).
rs and ns = curvature radius of the meridian sections and the first vertical, respectively,
in a given latitude s on the initial ellipsoid:
rs = as (1 es 2) / (1 es2 sin2 s)3/2, ns = as / (1 e s2 sin2 s)1/2
The formulae for dand dO provide the changes in and O in seconds of arc.
Local (topocentric) coordinates (1)

z (u)
10
11
Z x (n)
Local topocentric coordinates refer
y (e)
to a local reference system related to
P
the Earths gravity field: The origin
is the observer (topocentre P), the
orientation is given with respect to
Zp
the local vertical (zenith, plumb line):
Y
p
The
z-axis points to the zenith,
rectangular to the plane x, y;
p Yp
The x-axis points to the north of the
X
meridian;
Topocentric coordinate systems The y-axis points to the east, thus
are typically used in astronomy forming a left-hand system.
(horizontal coordinates, heights x, y, z are equivalent to denotations
north (n), east (e) and up (u).
and azimuths).
Local (topocentric) coordinates (2)
12
u
Point Pi is described w.r.t. the origin P
P
i
E\WKHJHRGHWLFD]LPXWKWKH]HQLWK
DQJOHDQGWKHGLVWDQFHV
7KHJHRGHWLFD]LPXWKLVGHILQHGDV
the angle in the horizon plane between
e
n
the meridian plane of P and the plane
formed by the normals in P and Pi.
7KH]HQLWKDQJOHLVPHDVXUHGLQWKH
P
vertical plane between the ellipsoidal
vertical and the connecting line P- Pi,
counted positive from the zenith.
The topocentric systems are
fundamental for connecting The distance s is only used in local
measurements of different
geodetic systems (not in astronomic
techniques (local ties).
systems).
Meridian of Greenwich
=0
Geocentric origin
=0, =0, h=-a
X=Y=Z=0
E X 0
R 1 * N Y0
U Z 0
with:
R 1
RT
Ecuador
=0
sin O0
cos O0
0
V (north)
cos M 0
sin M 0
13
sin M 0 cos O0 cos M 0 cos O0

sin M 0 sin O0 cos M 0 sin O0
Topocentric origin
=O, = O, h=hO
X=Xo, Y= Yo , Z= Zo
U=V=W=0
U (east)
W (height)
Conversion of topocentric to geocentric coordinates
X

Y
Z
X = XO E sin OO N sin O cos OO + U cos O cos OO

Y = YO + E cos OO N sin O sin OO + U cos O sin OO
Z = ZO + N cos O + U sin O
Transversal
14
15
Cylindrical
Conical
Secant
16
The metric distortions

may be grouped into:
Conformity: does not
present (diferencial)
angular distortion.
Equivalence:
presents coextensive
(equal) areas.
Equidistance: (some)
meridian(s) and
parallel(s) represent
the true (relations of)
length.
Special types of projections and distortions

Tangential
Alemania
3400000
2600000
3500000
4400000
3600000
4500000
5400000
4600000
5500000
17
The Gauss-Krger system is based on

a transversal cylindrical projection. It
is not a geometrical projection but a
mathematical conform transformation
without differential angular distortion.
It is regionally applied and divided in
zones of 3 longitude extension.
The central meridian and the equator
are straight lines. The other meridians
and parallels are complex curves.
The scale is true along the central
meridian and constant along the lines
parallel to the meridian. It increases
with the distance from the meridian.
Gauss-Krger plane conform coordinate system
Azimuthal
Plane coordinates
Plane coordinate systems enable representing the ellipsoidal (or
spherical) surface in a plane through mathematical or geometrical
specifications, e.g. by projections.
As the representation of a curved surface in a plane is not possible
without distortion, one has to decide on a representation distorting
less the angles, the distances or the areas, respectively.
In principle there are three types of projections:
Projections onto a cone,
Projections onto a cylinder,
Projections onto a plane.
The orientation of the surfaces can be
normal (axis of the surface parallel to the
Earths rotation axis),
transversal (axis parallel to the equator),
oblique (axis in an arbitrary direction).
Projections
Oblique
Normal
Cylindrical
Conical
Azimuthal
2500000
Universal Transversal Mercator (UTM) System
18
The UTM system is similar to Gauss-Krger.

It is globally applied and divided in zones of
6 longitude extension.
To reduce the distortion at the zone limits, a
scale factor of 0,9996 is applied in the central
meridian, so that the lines in 137 east and
west distance present the true scale (1,0000).
The Earth from 84N to 80S is divided into 20
stripes of 60 zones of 6 longitude extension.
The border meridians are divisible by 6 and
the zones are numbered from 1 to 60.
The stripes of 8 (12 in the northernmost) are
designated from south to north by letters (C to
X without I and O) beginning with C in 80S.
19
The meridians are straight

lines convergent to the pole.
The angles between the
meridians are true.
The parallels are unequally
separated; they are concentric
circles around the poles.
Two standard parallels are
length-preserving (true scale).
There is symmetry around
each standard parallel.
The pole closest to a standard
parallel is a point, the other
pole is not shown.
Conical conform Lambert system

Standard parallels in 20N
and 60N, central meridian
90W
The distortion is constant along

each parallel. There is no distortion
along the standard parallels.
20
The meridians are straight

lines convergent to the pole.
The angles between the
meridians are true.
The parallels are unequally
separated; they are concentric
circles around the poles.
One standard parallel is
length-preserving (true scale).
There is symmetry around
each standard parallel.
The pole closest to a standard
parallel is a point, the other
pole is not shown.
Polar stereographic system
The distortion is constant along

each parallel. There is no distortion
along the standard parallel.
2.2 Definition of reference systems and frames
Reference System: Definition of standards, parameters, models, etc.

serving as the basis for the representation of the geometry of the
Earth surface and its temporal variation (e.g., speed of light c0,
standard gravitational parameter GM, models of special and general
relativity, models of the atmosphere (ionosphere and troposphere),
three-dimensional orthogonal Cartesian coordinate system with its
temporal variation consistent with the Earths rotation).
Reference Frame: Realisation (materialisation) of a reference

system by a set of physical and mathematical quantities (e.g., a
number of physically marked points at the Earth surface with given
three-dimensional Cartesian coordinates X, Y, Z for a fixed epoch
and its linear variations with time (dX/dt, dY/dt, dZ/dt), i.e. constant
velocities (vX, vY, vZ).
Geodetic datum
Geodetic datum: Parameters fixing the origin, orientation and scale
of a coordinate system w.r.t. the Earth (e.g., in a global 3D Cartesian
system: the origin is realised in the geo-centre, the orientation is
realised by the position of the Earth rotation pole and the direction of
one conventional reference longitude at a defined epoch, the scale is
realised by the metre unit based on the speed of light in vacuum).
Important:
1. Reference systems cannot be determined by measurements, but
they are defined conventionally; i.e., the geodetic coordinates and
directions are not estimable but need a defined coordinate system.
2. Reference frames must realise the reference system strictly
according to its definition (e.g. geocentric and not crust fixed).
3. The geodetic datum must be given unambiguously. 3D-systems
require 7 parameters. There must not be fixed more coordinates.
Hierarchy of reference systems
Denomination
Principal vectors
Application example
Observation
line-of-sight Q
measurement of
system (local)
gravity J
directions & distances
Horizon system gravity J
terrestrial networks
(regional)
(DUWKURWDWLRQ
Equator system Earth URWDWLRQ
satellite geodesy
(global)
celestial origin b
Celestial system celestial origin b
radio astronomy
(extragalactic)
celestial reference pole
Examples of global reference systems
Terrestrial reference systems (TRS)
World Geodetic System 1984 (WGS84): Global terrestrial

reference system, originally established for orbit determination of
TRANSIT Doppler satellites (WGS72), later adopted for the orbit
determination of NAVSTAR GPS satellites (WGS84).
WGS84 adopted the ITRF (see next slide) for its realisation in 2002.
International Terrestrial Reference System (ITRS): Reference

system of the International Earth Rotation and Reference Systems
Service (IERS) established for determining the celestial (ICRS) and
terrestrial (ITRS) reference systems and their interrelation, i.e. the
orientation and rotation of the Earth in space (EOP, ERP).
Examples of reference frames
Global and regional reference frames
International Terrestrial Reference Frame (ITRF):

Materialisation of the ITRS by stations at the Earth surface and given
coordinates for a fixed epoch and its variations with time (velocities).
ITRF2008: more than 900 points in more than 500 sites.
The ITRF serves also for the precise orbit determination of the GPS
satellites by the International GNSS Service (IGS).
Sistema de Referencia Geocntrico para las Amricas (SIRGAS):

Densification of the ITRF, initially established for South America by
a GPS campaign with 58 stations in 1995 and extended to the
Caribbean, Central and North America in 2000 with 184 stations. At
present there are more than 250 permanent stations.
Examples of geodetic datums

Preliminary South American Datum 1956 (PSAD56):
Established by astronomically observed coordinates in La Canoa,
Venezuela, referring to the international ellipsoid 1924 (Hayford).
Deviation from geo-centre: X= -288 m, Y= 175 m, Z= -376 m.
South American Datum 1969 (SAD69):
Established in Cha, Brazil, referring to the GRS67 ellipsoid.
Deviation from geo-centre: X = -57 m, Y = 1 m, Z = -41 m.
North American Datum 1983 (NAD83):
Adjustment of astronomical, Doppler and VLBI measurements
referring to the GRS80 ellipsoid.
Deviation from geo-centre: X = 1,0 m, < -P= P
International Terrestrial Reference Frame (ITRF2008):
Use of Earth gravity field parameters C11=S11=C10=0 in computing
the orbits of Laser satellites for station positioning; orientation
according to a defined meridian; metric scale unit.
Traditional and modern reference systems

Why do we need new reference systems? Why cant we continue
using the traditional systems (PSAD56, SAD69, NAD83)?
Because we are using satellite techniques. We have to guarantee
identical reference systems for terrestrial points and satellites.
Modern reference systems must be defined by a global datum:
The origin is the Earths centre of mass (geo-centre X0=Y0=Z0=0),
because satellites are orbiting around the geo-centre (Keplers law).
The orientation is given by the earth rotation axis (Z-axis) and the
convention of a zero-longitude (X-axis in the Greenwich meridian).
The metric scale is given by the speed of light and (in case of use of
satellite techniques) by the standard gravitational parameter GM.
Modern reference systems are always three-dimensional:
X, Y, Z, or transformed to North, East, Height w.r.t. an ellipsoid.
Modern reference systems:

- global definition
- global and regional realisation
- geocentric datum
One system all over the world
Traditional and modern reference systems
Traditional reference systems:

- regional definition
- national realisation
- local datum
Many systems around the globe
Organisation of modern reference frames
The modern reference systems are installed and maintained by the

services of the International Association of Geodesy (IAG).
There is one service for each of the geodetic observation techniques:
- International GNSS (Global Navigation Satellite Systems) Service
(IGS),
- International Laser Ranging Service (ILRS),
- International VLBI (Very Long Baseline Interferometry) Service
for Geodesy and Astrometry (IVS),
- International DORIS (Doppler Orbitography and Radio-positioning
Integrated by Satellite) Service (IDS).
The International Earth Rotation and Reference Systems Service
(IERS) coordinates the activities for establishing and maintaining
the International Terrestrial Reference Frame (ITRF).
International Earth Rotation

and Reference Systems Service
(http://www.iers.org)
IDS
IVS
ILRS
IGS
Technique Centres
(External Services)
Electr. Interfaces
Central Bureau
Data Centre
IERS Directing Board

Analysis Coordinator
Research Centres
Product Centers
Celestial Reference
Earth Orientation
Terrestrial Reference
Rapid Service / Predicts
Geophysical Fluids
Conventions
Documents
Users
10
International GNSS Service (IGS)
IGS Tracking Network (http://igs.org)
471 stations in total, 407 in operation (2015-10-13)
Int. Satellite Laser Ranging Service (ILRS)
General structure of IAG Services
51 stations in operation (2015-10-13)
ILRS Tracking Network (http://ilrs.gsfc.nasa.gov)
11
- President and elected Directing (or Governing) Board.

- Central Bureau (Coordinating Office).
- Station network for observations maintained by individual
institutions at their own costs with Operation Centre(s).
- Satellites operated by space agencies at their own costs (in case of
satellite observation services).
- Correlation Centres of the observation data (in case of VLBI).
- Data Centres collecting the observation data and the results of the
analysis and combination.
- Analysis Centres for processing the data and generating products
for scientific and practical use.
- Combination Centres for evaluating and combining the products
generated by the different analysis centres.
$OOWKHDFWLYLWLHVDUHYROXQWDU\DQGDWWKHFRVWVRIWKHLQVWLWXWLRQV
12
13
International VLBI Service (IVS)

IVS Tracking Network (http://ivscc.gsfc.nasa.gov)
32 stations in operation (2015-10-13)

International DORIS Service (IDS)
14
15
DORIS stations
co-located
with
other techniques
(http://ids-doris.org)
IDS
Tracking
Network
(http://id
(http://ids.cls.fr)
ds.cl
ls.fr)
59 stations in operation (2015-08)

Internat. Terrestrial
Refer. Frame (ITRF)
realised by the coordinates of

marked stations at the Earths
crust with 3-D coordinates
16
International Celestial
Reference Frame (ICRF)
realised by the coordinates of

Quasi Stellar Radio Sources
(Quasars) observed by VLBI
(www.dgfi.badw.de)
2.3 Celestial (inertial) reference system and frame
Celestial reference system = Conventional inertial system

- Inertial systems are non-accelerated systems without external force.
- The original laws of physics (Newtons laws, celestial mechanics,
relativity theory, ...) refer to inertial systems.
Earth orientation parameters (EOP)

- Rotation of the terrestrial with respect to the celestial system
Terrestrial reference system = Conventional global system

- Terrestrial reference systems are associated with the solid Earth.
- They participate in all the motions of the solid earth and all the
deformations of the Earths crust.
Celestial (inertial) reference system

Ecliptic and
equatorial systems
Coordinates of an
object in space
(star, radio source):
- equatorial system:
right ascension D
declination G
- ecliptic system:
ecliptic longitude O
ecliptic latitude E
(hardly ever used)
Realisation of the celestial reference system by the

International Celestial Reference Frame (ICRF)
The International Celestial Reference System (ICRS) is realised by
the astronomic coordinates (right ascension and declination) of
quasi-stellar astronomic radio sources (Quasars) observed by the
Very Long Baseline Interferometry (VLBI). This realisation is
called International Celestial Reference Frame (ICRF).
The current ICRF (ICRF2 of 2009) contains the coordinates of
extragalactic radio sources for epoch 2000.0 in three categories:
- 295 datum defining sources,
- 3080 additional global sources,
39 instable sources with special treatment,
- 3414 sources in total.
Two telescopes receive signals of

the same radio source (Quasar) at
different times.
Very Long Baseline

Interferometry (VLBI)
Realisation of the ICRF by VLBI
Station OHiggins, Antarctica,

of BKG/DLR, Germany
Principle of VLBI analysis (1)
1. Observation
Simultaneous recording of the
signals from the same Quasars.
2. Correlation
Calculation of the maximum
correlation by delaying the
signal of one receiver: o 't.
3. Corrections of VLBI
observations
- Ionospheric refraction
- Tropospheric refraction (dry
and wet)
- Time (clock) correction
- Telescope calibration
(deformations, eccentricities,
Correlators (IGG Bonn, Germany)
signal travel time within the
(today there are no tapes but files!)
telescope to electronic centre)
One correlator is needed for each
The corrections are similar to
baseline. Today we have broadband
GPS because the frequencies
data transmission (e-VLBI) and near are similar.
real-time computer correlation.
5. Adjustment
Computation of the Quasar
coordinates () relative to
the terrestrial reference frame
by simultaneous adjustment
of all observations.

4. Analysis
Computation of the terrestrial baseline
length and the direction to the Quasar.
ICRF2
0()
0()
C- mean first
Example of ICRF2 coordinates

Designation Source
last
Nexp Nobs
Interrelation of CRF and TRF
The celestial bodies (stars, radio sources) and artificial satellites

used for space measurements in the Terrestrial Reference System
(TRS) do not participate in the Earth rotation. Therefore we have
to know the Earth Orientation Parameters (EOP) with respect to
the Celestial Reference System (CRS) at any instantaneous time
of measurements at the Earth surface.
The EOP provide the connection between the CRF and the TRF.
They are derived from the continuous observation of the Earths
rotation.
We distinguish between the orientation of the Earth in space
(precession and nutation), the orientation with respect to the body
of the Earth (polar motion), and the variation of the rotational
velocity (variation of Universal Time UT or Length of Day LOD,
respectively).
Connection between CRF and TRF via EOP
Equator
Ecliptic
The reference systems must be consistent: CRF-EOP-TRF

Variation of orientation in space

The cause of the variation of Earth orientation in space (precession
and nutation) is the asymmetric effect of the lunisolar gravitational
forces on the equatorial bulge.
10
11
Precession and nutation
12
The lunisolar forces

produce a wobble of
the rotation axis of the
Earth in space.
Traditionally it was
distinguished between
the frequencies of
precession (lower
frequency) and nutation
(higher frequencies).
This differentiation was
terminated with the
IAU 2000 and IUGG
2003 resolutions.
N = R1(-H-'H) R3(-'\) R1(H)
Precession and nutation in the inertial space
P = R3(-z) R2(T) R3(-])
13
The system 2000 of the International Astronomical Union (IAU2000)

fixes the origin of right ascension at epoch 2000.0 (non rotating).
The new precession-nutation model (, ) includes all the effects.
Z-axis = Celestial
Intermediate Pole
(CIP)
Precession-Nutation Instantaneous (true)

equatorial system
o
Celestial
Intermediate System
Reduction of precession and nutation in

geodetic and astronomical observations
Conventional Inertial
System (CIS)
Realised as ICRF by
Quasar coordinates
Equatorial system at
a fixed epoch
(e.g. t0 = 2000.0)
Z-axis = celestial
pole at a fixed epoch
(e.g. t0 = 2000.0)
2.4 Rotation and tides of the Earth

2. The angular momentum
varies due to displacements
and motions of masses.
Variation of Earth rotation relative to the Earths body

1. The Earth rotation axis does not
coincide with the symmetry axis
(principal axis of inertia)
Symmetry axis
14
ZCT
CTP
GAST
YCT
Instantaneous
(true) pole
Conventional
equator
xp
-yP
zT
UT1 = Universal Time

UTC = Universal Time
Coordinated (TAI)
DUT = UT1 - UTC
= Variation of GAST
dDUT = Variation of LOD
(Length of Day)
CTP = Conventional
Terrestrial Pole
GAST = Greenwich Apparent Sidereal Time
Earth Rotation Parameters

(ERP: XP, YP, DUT)
Polar motion (XP, YP) and variation of the

rotational velocity UT1-UTC
Greenwich
mean time
meridian
XCT
yT
Celestial
intermediate system
Conventional
Terrestrial System
(CTS)
Reduction of ERP variations
Instantaneous (true)
equatorial system
Z-axis = axis of the

coordinate system
for the points in the
conventional system
Polar motion (X , Y ), Global system at a

P
P
and variation of
fixed epoch
angular velocity
(e.g. t = 2000.0)
(DUT = UT1 UTC)
Z-axis = Celestial
Intermediate Pole
(CIP)
Rotation of the Earth as a rigid body

Law of conservation of the angular momentum in an inertial system:
dH/dt = 0
(H = angular momentum)
(I = moment of inertia, Z = rotational velocity)
H =IZ
In an inertial celestial system (Euler equation):
dH/dt = L
(L = lunisolar torque o precession, nutation)
In a terrestrial system (o polar motion):
dH/dt + Z H = 0 (if axis H axis Z there is no motion!)
With principal moments of inertia of the body of the Earth A, B, C:
~A
~ ~ Z1~
B ~ ~ Z2~
H= IZ = ~
~
C ~ ~ Z3~
The solution is a circular motion with the radius
pR Z1 + Z2 = 0,2 and period 1/V = 305 days (Euler)
Rotation of the Earth as a non rigid body

H (t) = I (t) Z
Liouville equation (instead of Euler equation)
time variable moment of inertia due to mass

displacements in the terrestrial system
time variable angular momentum
I (t) = I + 'I (t)
dH/dt + Z H = L
The solution of the Liouville equation results (in the same way as the
Euler equation) in the period and the amplitude of the circular motion
of the Earth rotation pole. This is for an elastic Earth:
1/V = 435 days (Chandler period)
angular momentum due to the motion of masses
In addition to the variation of the moment of inertia by the displaced

masses there is an effect by the motion (velocity v) of the masses:
h (t) = (x v) dM
H (t) = I (t) Z + h(t) forced angular momentum

p + dp
The thermodynamic energy generates an

acceleration of masses in the atmospheric and
oceanic pressure field
dx / dt = dv / dt = -1/U dp/dx
Strongest mass movements in the Earth system

are those of the atmosphere and the oceans.
Temperature variations (dT) change the mass
density. The volume (dv) varies opposing to the
pressure (p) and produces thermodynamic energy
dQ = c dT + p dv
Angular momentum caused by moving masses

p
x
The motion of masses interferes with the Earth

rotation and causes the Coriolis acceleration
dx /dt = dv / dt = -2 : v
Variation of Earth rotation by moving masses
In case of equilibrium between the accelerations caused by pressure

and by the Coriolis force there is a geostrophic motion (Greek:
strephein = to turn)
1/U dp/dx - 2 : v = 0
vg
p + dp
vg
CH
p + dp
This motion has no acceleration and results different in the northern

and southern hemispheres:
p
dv/dt
dv/dt
CH
The high pressure of the atmosphere is on the right of currents in

the northern hemisphere and on the left in the southern hemisphere.
Circulation of the atmosphere (principal winds)
Circulation of the oceans (principal currents)
Humboldt
The deviations of on-going atmospheric and oceanic currents from

the geostrophic currents generate the variations of Earth rotation.
0.6
0.5
0.4
0.3
0.2
0.1
1990
1990
0.004
[sec]
0.003
0.002
0.001
0.000
-0.001
1992
1992
1994
1994
1996
1996
1998
1998
2000
2000
2002
2002
2004
2004
Partitioning of polar motion
1988
1988
2006
2008
2008
IERS C04
LOD
10
11
x annual
y annual
2010
x Chandler
y Chandler
2010
1 ms 46 cm linear motion at the equator
-0.002
1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010
El Nio effect
Variation of rotational velocity (LOD [sec])
Polar motion [] = [arcsec]

1 31 m in latitude and equatorial longitude
Observation of variations of Earth rotation by

space geodetic techniques (VLBI, SLR, GPS)
< ~18 m >
1986
2006
1986
Motion with Chandler period (435 days) [s]
1984
1984
Motion with annual period (365,25 days) [s]
0.0
1982
-0.1
-0.2
-0.3
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1982
-0.1
-0.2
-0.3
YP
0.5
0.4
0.3
0.2
0.1
1986
y annual
y AAM mass
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
Causes of variations of Earth rotation
1984
component and atmospheric angular momentum []
0.0
1982
LOD and angular momentum of atmospheric mass motions [s]

0.004
1990
1992
1994
1996
LOD
AAM motion z
1998
2000
2002
2004
2006
2008
2010
ac
ag
ag
ac
ag
Centrifugal acceleration
ac = Z2* (r* - R cos\)
Gravitational acceleration
ag = GM* / (R2 + r2* - 2Rr* cos\)
Lunisolar tidal acceleration (1)
ac
M* : MEarth
1 : 81,3
332946 : 1
R : r*
1 : 60
1 : 23481
D*
2,628 m/s
1,208 m/s
K1
S0
M0
Symbol
24,07
25,82
23,93
Period [h]
144,6
310,6
436,9
47,7
102,9
S principal wave
L principal wave
L, S inclination of the ecliptic
S constant flattening
L constant flattening
Origin (Lunar/Solar)
O1
S2
M2
12,66
12,00
12,42
71,9
174,8
375,6
L ecliptic orbit
S principal wave
L principal wave
N2
Semidiurnal components
Diurnal periods
Long period components
P1
Amplitude [Pms-2]
Most important partial tides (Torge 1989)
Celestial body *
Moon
Sun
Total acceleration: (Doodsons constant D* = 0,75 GM* R2 / r3*)

- radial: ar = 2/R D* (cos 2\* + 1/3) - tangential: a\ = 2/R D* sin 2\
1988
Effects of Moon and Sun
1986
0.003
1984
2010
12
13
0.002
0.001
0.000
-0.001
1982
Importance of variations of Earth rotation for

reference systems
The variations of Earth rotation with respect to the inertial space
(precession and nutation) affect the coordinates of celestial bodies
(right ascencion and declination). They must be reduced from the
positions in the celestial reference system (Quasars, satellite orbits).
The variations of Earth rotation with respect to the Earths body
(polar motion and UT1/LOD) affect the terrestrial coordinates
(latitude and longitude). They must be reduced from the positions in
the terrestrial reference system (reference frame).
If the variations of Earth rotation are not reduced, they produce
errors in the precise positioning using space techniques (e.g. GPS)
coming up to several metres.
14
15

Inclination [10-3 "]
Variation of amplitudes as a function of latitude

Acceleration [JDO = 10 nms-2]
Deformation caused by Earth tides (1)
W = gravitational potential
Vm = tidal potential
Vd = deformation potential
Deformation of the elastic Earth generated by the potential Vm:

'rel = h 'rm = h Vm/g
(h = Love number)
Deformation of the level surface (W + Vm) generated by the mass
displacements:
(k = Love number)
Vd = k Vm
16
17
Deformation caused by Earth tides (2)
Horizontal surface deformation (l = Shida number):

'yel = l (1/g) (wV/cosM wO)
'xel = l (1/g) (wV/wM)
19
Superconducting Gravimeter
18
Gravimetric amplitude factor:

Inclination amplitude factor:
G = 1 - 1,5 k + h
J=1+k-h
h(f), k(f), l(f), G(f), J(f) are frequency dependent parameters reflecting
the elastic behaviour of the body of the Earth.
Magnitudes
Lunar effect Solar effect
Equipotential surface deformation
~36 cm
~16 cm
h | 0,64 (topography deformation)
~23 cm
~10 cm
k | 0,32 (subsequent level deformation) ~12 cm
~ 5 cm
l | 0,16 (horizontal deformation)
~10-8 = 1 Pm /100 m
| 1,164 (gravimetric amplitude)
-1,1 ... +0,5 -PV-2
J | 0,674 (inclination amplitude)
r 0,017"
r 0,008"
Observed angle
Observation of Earth tides

Plumb line
Horizontal pendulum
Rotation axis
Zllner
principle
Verbaandert-Melchior
Principle of
a tide gauge
rule zero
rule height
zero
mark
Principle of satellite altimetry
Observation of ocean tides
BM
well
Ocean tide model

Example M2-model: Amplitude (dashed lines [cm]) and phase ([h])
20
21
Importance of Earth and ocean tides for

terrestrial reference systems
i 1
(X
Si
X ' Pi ) 2
Satellite (XS, YS, ZS)
Sz
Station
(XP, YP, ZP)
X
Y
22
Tides affect the Earth surface in two ways:

Deformation caused by gravitational forces of Moon and Sun;
Deformation caused by variable ocean and atmosphere loading.
The effects on the terrestrial reference frame must be reduced (e.g.
by the terrestrial model FES2004 and the ocean model Scherneck).
In the (geometric) terrestrial reference system we reduce the entire
effect including the deformation caused by the permanent potential
of Moon and Sun. The result is a system completely free of tidal
effects (tide free system).
In the measurement of terrestrial gravity the effect of the permanent
potential is not reduced but only the temporal effects. The result is a
system of the average tides (mean tide system).
As a consequence positions and gravity values are not consistent.
2.5 Terrestrial reference system and frame

Z
Celestial
Intermediate
System
Z Z
Required consistency of reference systems

Consequences of nonconsistent systems
1. Single (precise) point
positioning (PPP)
Y
Y
X
X
The difference between the reference systems (origin and orientation)
enters completely into the coordinates (e.g. polar motion up to 18 m)
Z Z
Required consistency of reference systems

Consequences of nonconsistent systems
2. Relative (differential)
positioning by GPS
Y
X
X
A scale factor of a 2 ... 3 10-7 per metre difference in the reference
systems enters into the baselines (18 m polar motion: a 5 mm/km)
XPole, YPole, UT1 - UTC
N (, )
Coordinates (angles)
of the Quasars (ICRF)
Consistency of CRS and TRS by transformation
Conventional Inertial System

(CIS)
Precession-Nutation
Polar motion,
DUT variation
Point coordinates (3-D)

at the Earth surface
Celestial Intermediate Pole

(CIP)
Conventional Terrestrial System

(CTS)
International Terrestrial Reference System (ITRS)
The terrestrial reference system must be consistent with the celestial

reference system used for satellite orbits. ITRS is a 3-D Cartesian system.
The origin of the ITRS is defined by the Earths centre of mass
(including atmosphere and oceans), because satellites are orbiting
around the geo-centre according to Keplers laws.
The orientation of the axes is defined conventionally by fixing the
Z-axis close to the Earth rotation axis and the X-axis close to the
Greenwich meridian in its positions at a given epoch. The time
evolution must be given by station motion parameters (velocities)
not generating a global rotation with respect to the Earth rotation.
The scale unit is metric as defined by the speed of light in vacuum
according to the International System of Units (SI) and by the
standard gravitational parameter GM (for satellite techniques).
Standards and models are defined by the IERS Conventions.
Realisation of the datum of the ITRS
The spherical harmonics of

the Earth gravity field are:
C11 = X dm / a M
S11 = Y dm / a M
C10 = Z dm / a M
The datum is realised by combination of geodetic space techniques:

SLR provides the relation to the geo-centre by determining the orbit
in the Earth gravity field with corresponding parameters:
The centre of mass is the integral
over all the masses of the Earth:
X0 = X dm / M
Y0 = Y dm / M
Z0 = Z dm / M
A gravity model with C11=S11=C10 provides geocentric coordinates!
VLBI provides the orientation by the EOP (, , XP, YP, UT1);
SLR and VLBI provide the scale unit by measuring distances based
on the metre definition (speed of light reduced to the atmosphere).
(GPS does not measure distances but differences of distances)
Principle of Satellite Laser Ranging (SLR)

Measurement
of the distance
Earth - satellite
- Earth
Laser-ranging is always global.

The distances from the telescopes
to satellites are measured by the
signals travel time as round trip.
There is only one clock, therefore
there is no need of estimating
clock corrections (like in GPS).
Rsta
dobs
Rsat
True
orbit
Kepler
orbit
Realisation of the origin in the geo-centre by SLR
Laser ranging processing includes

the computation of global orbits.
Employing an Earth gravity field
model with spherical harmonic
coefficients C11= S11= C10= 0
provides ephemerides referring to
the centre of mass (geo-centre).
Subtracting the measured distance
from the geocentric radius vector
provides geocentric coordinates of
the telescope.
This is different in GPS, where we

get only baselines because clock R geoc d = R geocentric
sat
obs
sta
errors have to be estimated.
Realisation of the orientation by convention
S21 <=dm / a2M
One could realise the orientation also by the degree-two spherical

harmonics of the gravity field which provide the symmetry axes
(axes of maximum moment of inertia) of the masses of the Earth:
C21 = ;=dm / a2M,
S22 ;<dm / 2a2M
However, these coefficients cannot be determined with the required

precision. Therefore the orientation is realised by convention, fixing
the Z-axis close to the Earth rotation axis and the X-axis close to the
Greenwich meridian as given by the Bureau International de lHeure
(BIH, a predecessor of the IERS) for the epoch 1984,0.
These parameters have to be extrapolated in time in a way that no
global rotation of the Earth crust (realised by the observation
stations) remains in the coordinates time series (see later slides).
Realisation of the metric scale
Evolution of the ITRS in time (2)
PCFC
SOAM
NOAM
COCO
NAZC
CARB
AFRC
ARAB
INDI
EURA
12 mayor plates with given motion by rotation vectors (see chapter 6)

EURA
AUST
ANTA
Geodetic station motions (velocities) shall refer to this model.
12
13
Geologic-geophysical plate model NNR NUVEL-1A
Besides the varying Earth rotation we have to correct the reference

frame coordinates by its motions w.r.t. the CTS in order to conserve
the consistency with the CIS, i.e. the satellite ephemerides.
Principal terrestrial station motions are caused by plate tectonics
which is a long-time process. Therefore they are modelled by linear
coordinate changes, i.e. by constant velocities dX/dt, dY/dt, dZ/dt.
This modelling requires a kinematic reference system to which the
velocities refer and which is consistent with the Earth rotation, i.e.
it must not generate a global rotation of the Earth crust.
The kinematic reference system is defined by geologic/geophysical
models of tectonic plate motion generating no net rotations (NNR);
- until ITRF91: AM0-2 (Minster and Jordan 1974, 1978);
- until ITRF94: NNR NUVEL-1 (Argus and Gordon 1991);
- until present : NNR NUVEL-1A (De Mets et al. 1994).
10
The scale unit of the terrestrial reference system is the metre as

defined in the Internacional System of Units (SI) by the speed of
light in vacuum. To reduce the speed for atmospheric effects, an
atmosphere model is introduced and its parameters are corrected by
estimation in the common data adjustment.
The estimated parameters are depended on the signal frequency and
correlated with instrumental properties (e.g. electronic delay of the
signal, phase centres, deformations), station heights and others.
Therefore we get different metric scales for different techniques.
In order to avoid distorsions between the techniques, the scale is
defined by one or two techniques (SLR and VLBI in ITRF2008)
and the scales of the others are estimated w.r.t. these.
(The best would be to eliminate all the instrumental effects by an
external calibration resulting in identical scales for all techniques.)
11
Station
(XP, YP, ZP)
Satellite (XS, YS, ZS)

transformed
from CIS to CTS
Evolution of the ITRS in time (1)

Z Z
Geo-centre
The terrestrial stations are moving because of crustal deformations.

Satellites do not participate in this movement.
1996
2000
2002
2004
WGS84
ITRFyyyy
2006
2008
Deviations from the geo-centre)
1998
Earthquake
Arequipa,
Peru 2001
Antenna
change
Hfn 2001
Iceland
82.80
82.75
82.70
82.65
82.60
2000
2000
1095
1147
2002
1200
2003
2003
1252
2004
2004
1304
2005
2005
1356
2006
2006
1408
2007
2007
1460
1147
2002
1200
2003 1252
2004
1304
2005 1356
2006 1408
2007
1460
2008
1513
2008 1513
2008
To be considered as a new station after antenna change
2002
Jump in coordinates and different velocities required
2001
2001
2001
A minimum of two years required for velocity estimation
Height
2000
Height
1043
82.55
[m]
1106.04
1106.03
1106.02
1106.01
1106.00
1105.99
Week 1043
Seasonal
variation,
Brasilia,
Brasil
1095
Examples of weekly coordinates analysis
17
16
The (presently two) ITRF combination centres (IGN, DGFI Munich)

- analyse the weekly time series to identify the degree of freedom of
the datum (e.g. eventual constraints by introducing conditions on
station coordinates), gross errors, jumps in coordinates, etc.;
- combine the weekly solutions of each observation technique into
one solution of coordinates for a defined reference epoch and its
OLQHDUYDULDWLRQVYHORFLWLHVintra-technique combination.
Organisation of the ITRF since 2001
1994
Realisation of the ITRS by the ITRF
1992
The International Terrestrial Reference System (ITRS) is realised by

the International Terrestrial Reference Frame (ITRF).
According to resolutions of the International Union of Geodesy and
Geophysics (IUGG) the ITRF is the only terrestrial reference frame
to be used in science and practice. (WGS84 adopted ITRF in 2002.)
1990
14
The IERS Product Centre for the ITRF (IGN Paris) releases a call to
the IAG Services (IVS, ILRS, IGS, IDS) to provide weekly (in case
of VLBI session-wise) solutions of their analysis centres including
3D station position coordinates (X, Y, Z) and daily Earth Orientation
Parameters (EOP) combined by their combination centres to one
technique solution preferably in terms of datum-free (alternatively
ORRVHO\FRQVWUDLQHGLH 1 m) normal equations.
35.0
[cm]
30.0
25.0
20.0
15.0
10.0
5.0
0.0
1988
Realisations of the ITRF

The ITRF is realised by combining individual solutions of positions
(and velocities) provided by analysis centres of different techniques.
until 2000: international call for individual solutions by IGN Paris;
since 2005: combination of solutions of IAG services (IGN, DGFI).
No. of
stations sols: VLBI SLR GPS DORIS Total
ITRF 88
120
5
6
11
ITRF 89
113
6
8
14
ITRF 90
120
4
7
11
ITRF 91
131
5
7
1
13
ITRF 92
155
5
6
6
17
ITRF 93
160
6
4
5
15
ITRF 94
209
6
1
5
3
15
ITRF 96
290
4
2
7
3
16
ITRF 97
309
4
5
6
3
18
21 + 8*
ITRF 2000
477
3
9
6 + 8* 3
ITRF 2005
338
1
1
1
1
4
ITRF 2008
578
1
1
1
1
4
15
regional densifications
Analysis of the normal equations (DGFI)

The normal equations must not have any constraints on the datum
(datum free normal equations) i.e. the datum defect must be 7.
A fixed datum in different weeks refers the coordinates to different
origins, orientations and scales, which then cannot be combined.
All the input data were analysed for the datum defect. The result is:
- GPS (IGS): no defect at all (all parameters are fixed),
/
- SLR (ILRS): defect 3 (3 translations and scale are fixed), - VLBI (IVS) defect 6 (scale fixed),
- DORIS (IDS) no defect at all (all parameters are fixed).
/
If the defect is not appropriate, one has to liberate the datum, i.e.
one has to introduce columns and rows into the normal equations
for those datum parameters which should be free.
(IGN did not use normal equations but solutions + covariance matrices)
Weekly solutions
(LOD)
24 h sessions, free
normal equations
Weekly solutions
(LOD)
Data: epochs,
weekly series
Weekly solutions
(LOD)
1993 - 2008
1980 - 2008
1983 - 2008
1997 - 2008
Interval
Input data for ITRF 2008: X, Y, Z, XP , YP , UT1
ILRS CC
ASI Matera
IVS CC
GIUB Bonn
IDS CC
CLS Toulouse
Service
Technique
Analysis Centre
IGS AC
NRC Ottawa
GPS
SLR
VLBI
DORIS
Total
~1500 occupations ~4500 solutions

~ 920 points
1980 - 2008
with daily EOPs
(UT1 only by VLBI)
578 stations
18
19
GPS
n. eq. week 1
n. eq. week 2
n. eq. week n
...
SLR
n. eq. week 1
n. eq. week 2
n. eq. week n
...
VLBI
n. eq. week 1
n. eq. week 2
n. eq. week n
...
DORIS
Combination procedure (DGFI)

n. eq. week 1
n. eq. week 2
n. eq. week n
...
Multi annual
X, v, EOP
Multi annual
X, v, EOP
Multi annual
X, v, EOP
Accumulation of normal equations of the time series

Multi annual
X, v, EOP
ties datum fixing

Accumulation of normal equations, input of local ties,
ITRF2008: positions, velocities and EOP
20
(IGN did not use normal equations but combined by Helmert transformation)
Local ties at co-location stations
To combine the normal equations of the different techniques we need

connections (coordinate differences) between their reference points.
The definition and realisation of the reference points (phase centres)
is complicated and the ties must be transformed to the global frame.
21
X
The reliability of the local measurements may be evaluated by
- comparing them with the estimated 3-D coordinate differences,
- comparing the velocities obtained from the different techniques,
- comparing EOPs obtained from the different techniques (DGFI only)
Introduction of the datum parameters
23
22
DGFI: The input data are (made to) datum free normal equations.
IGN: The individual solutions are stacked by Helmert-transformation
The datum parameters are introduced in the final combination:
The ITRF2008 datum refers to the epoch 2005.0 by
- 3 translations given by SLR: coordinates origin = geo-centre,
- 3 rotations by the ITRF2005: orientation according to BIH1984,
- 1 scale by SLR and VLBI: speed of light corrected by atmosphere,
- 3 velocities of translations by SLR: velocity origin = geo-centre,
- 3 velocities of rotation (with EOP): by NNR condition:
IGN: geophysical model NNR NUVEL-1A;
DGFI: geodetic model APKIM (see chapter 6);
- 1 drift of the scale by SLR and VLBI (only IGN).
Station velocities of the ITRF2008
Different velocities in a site refer to different periods either due to

seismic events or instrumental changes (e.g. antenna change).
WGS84 in ITRF2008
WGS84 was designed for satellite orbit determination and single point
positioning (PPP). 11 stations are not sufficient for a reference frame.
2.6 Regional and national reference frames
The global ITRF is densified by regional reference frames observed

by GPS in connection with global IGS stations included in the ITRF.
The first network of this type was observed in Europe in 1988.
1988 EUNAV-88 (pre-campaign)
1989 EUREF-89 (densific.)/Turkey
1990 EUREF-NW (Iceld./Greenld.)
24
1991 EUREF-EAST/Hun/Slov/Czech
1992 EUREF-Poland/Baltic/Bulgaria
1993 EUREF-Cyprus/Iceland/D/NL
1994 EUREF-Lux/Slov/Croa/Roman.
1995 EUREF-Ukraine/Slovenia
1996 EUREF-FYROM/Bosnia
1998 EUREF-Serbia/Monte./Albania
1999 EUREF-Moldavia
South American
Reference Frame
Sistema de Referencia
Geocntrico para
Amrica del Sur
(SIRGAS)
58 stations observed for
10 days in May 1995
(see chapter 5)
The Americas
Reference Frame
Sistema de Referencia
Geocntrico para las
Amricas (SIRGAS)
184 stations observed
for 10 days in May 2000
extending the reference
frame to all Americas.
The continental frames
are further densified by
national reference frames
Peru
Argentina
Venezuela
National densifications by GPS campaigns
1998
Colombia
Ecuador
1994
Disadvantages of reference frames by campaigns
When using the reference frame for practical positioning one has to
occupy simultaneously the reference stations in addition to the new
stations by other receivers (more GPS receivers required).
The reference frame coordinates (positions at a defined reference

epoch) serve as reliable values for connecting the new points, but:
The points are moving with the Earths crust
- continuously with the tectonic plates,
- irregularly in deformation zones,
- sporadically by earthquakes.
The coordinates have to be transformed from the definition epoch
to the observation epoch of new (connected) stations. This is
problematic if these motions are not known (observed).
It is much better to establish reference frames by continuously
observing stations. This is done in all continents.
European Permanent Tracking Network (EPN)
African Reference Frame (AFREF)
Asia-Pacific Reference Frame
North American Reference Frame (NAREF)
(6 different reference frames of different institutions)
SIRGAS network of
continuously observing
stations (SIRGAS-CON)
(> 300 stations)
11
10
Covering the area of Mexico,

Central America, The Caribbean,
South America and some parts of
Antarctica (see chapter 5)
Colombia (MAGNA-ECO)
National densifications by continuously

observing stations (examples)
Brazil (RBMC)
Realisation of the regional reference frames (1)
The regional reference frames are densifications of the ITRF. As the

ITRF contains also SLR, VLBI and DORIS stations, and to facilitate
the easy access, the IGS provides an ITRF extract for GNSS users.
The IGS08 core network consists The IGS08 full network consists
of 91 stations and forms the basis of 232 stations and is the basis
for the regional densifications.
for the orbit determinations.
The IGS updates the reference frames by reprocessing and considering

changes in the station occupations. These new frames are called IGb.
The analysis centres for the continental networks compute weekly

solutions of station coordinates referring to the actually valid IGb
frame. Users should apply these frames (SIRGAS) for positioning.
12
13
15
14
All IGS Regional Network

Associate Analysis Centres
(RNAAC) deliver weekly
solutions to the Global
Network Associate Analysis
Centres (GNAAC), where
they are combined to the
global polyhedron (1008
stations at (2013-09-18).
The RNAAC also compute multi-annual solutions including precise
station coordinates at a defined epoch (t0) and its linear variations in
time (constant velocities v). These shall be used to extrapolate the
coordinates from the definition epoch to the observation epoch ti:
X(t) = X(t0) + v (ti t0).
SIRGAS velocities
NAREF velocities
Transformation t0 o ti
X(ti)
Re-transformation ti o t0
Consistent velocities: from

the archive or from a model
= X(t0) + 'T0k + R0kX(t0)

+ dX/dt (ti t0)
Calculation of
coordinates
Y (new)
16
GPS observation
at epoch ti in the
reference system tk
Precise use of the terrestrial reference frame

Definition
of the reference
system
Archive
of coordinates X
for epoch t0
Y(t0) = Y(ti) 'T0k R0kY(tk)
dY/dt (ti t0)
Rem.: A difference of 1 m between reference frames causes relative

errors of ~23 10-7 EDVHOLQHHUURUVFPPPNP
VII SIRGAS School on Reference Systems
November 16 -17, 2015

UNPHU
Santo Domingo, Repblica Dominicana
Sponsored by:
International Union of Geodesy and Geophysics IUGG
International Association of Geodesy IAG
International Association of Geomagnetism and Aeronomy IAGA
International Association of Seismology and Physics of the Earth's Interior IASPEI
Pan-American Institute for Geography and History PAIGH
Geocentric Reference System for the Americas - SIRGAS
School on Reference Systems, Crustal

Deformation and Ionospheric Monitoring
Claudio Brunini
SIRGAS President
Consejo Nacional de Investigaciones Cientficas y Tcnicas
Universidad Nacional de La Plata, Argentina
GNSS positioning
Outlook:
1. Introduction
2. Mathematical foundation
3. Pseudo range measurement
4. Models to reduce error sources
5. Point and differential positioning
6. Network computation
7. Ambiguity resolution
1) introduction
The age of satellite positioning
Began on October 4, 1957 with the launch of the

Soviet satellite Sputnik.
Soon after (1958) William Guier and George
Weiffenbach, at the Applied Physics Lab of the Johns
Hopkins University, computed the satellite orbit using
the Doppler shift measured on the signal from a point of
known position.
The TRANSIT successor
Quickly was recognized that the problem could be reversed: calculate the position
of the measuring point from a satellite of known orbit.
That initiated the Navy Navigation Satellite System (NNSS) program of the US
Army, which spawned the first GNSS known as TRNASIT.
The TRANSIT constellation was composed by 5 satellites and operated from 1960
to 1996, being opened to civil users in 1967.
In 1964-1966 a higher requirement for a new

GNSS was establish by the US AF: determine
the instantaneous position of a point at rest or
moving, at any time and place, by
simultaneously measuring the ranges to 4
satellites of known positions.
That requirement was fulfilled by the GPS whit a 24-satellites constellation.

Satellites have been launched from 1978.
Opened to civil users in 1983.
Initial Operational Capability achieved in December 1993.
Complete 24-satellite constellation in March 1994.

A total of 64 satellites have been launched to date.
Presently there are 31 operational satellites.
Present and planed GPS status
Block II-RM
Block IIIA
All the satellites carry redundant atomic clocks.

Broadcast two carriers named L1 and L2.
Carriers are modulated with different civil and military codes.
Military codes are encrypted.
7 Block II-RM satellites modulate new civil (CA on L2) and military (M) codes.
4 Block II-F satellites broadcast a new carrier (L5) to provide safety-of-life services.
A new generation of satellites (Block III-A) will be launched from 2014.
Block IIR
Present and planed GNSS status

Presently, only the US GPS and the Russian GLONASS are fully operational at
global scale.
China is expanding its regional Beidou into the global Compass .
The UE is deploying its Galileo (presently, only 4 satellites).
France, India and Japan
are developing regional
navigation systems.
There are many
augmentation systems for
improving navigation integrity
and accuracy at local and
regional scale (e.g.: the USA
WAAS or the UE EGNOS).
GPS constelation
6 orbital planes, 55 inclination, 60 in longitude;

31 satellite, 4+ satellites per plane;
~circular orbits, ~26.000 km radii (~20.000 km
above the Earth surface);
12h sidereal (11h 58m UT) revolution period,
orbital speed: 3.9 km/s.
4+ satellites over the horizon at any

place and time.
Same ground track every day 4m after.
Navigation principle
A world-wide infrastructure
composed by tracking stations,
terrestrial and ground-to-satellites
data links and operational
centres, monitor and control the
constellation.
Based on recent measurements, satellites orbits are

computed and extrapolated to the near future.
Extrapolated orbits are broadcasted to users (the
broadcasted ephemerides) by the satellites.
Measuring simultaneously the ranges to 4 satellites
the receiver corrects its clocks and compute its
positioning using the trilateration principle.
The equation of observation
2) mathematical foundation
X R Y S YR Z S Z R X RS
The equation of observation results

from the Pythagoras's Theorem.
The receiver position (unknown) is computed based on the satellite position

(known) and the satellite-to-receiver range (measured).
U RS
S
where X R stands for all the unaccounted errors.
Satellite and receiver coordinates must be

in the same reference system.
The equation of observation

contains 3 unknowns:
U RS
The system of equations of observation
X
Y
Y
Z
Z
Solving the problem requires

simultaneous measurements to,
at least, 3 satellites:
X 1R
X R Y 2 YR Z 2 Z R X R2
X R Y 3 YR Z 3 Z R X R3
1
R
U R2
U R3
Direction cosine of the measured satellite
Linearization of the equation of observation

Receiver approximate position
U RS
X
S
Linearization of the equation of observation

2
corrections to the receiver

approximate coordinates.
direction cosines of the

measured satellite
X R ,0 Y S YR ,0 Z S Z R ,0
receiver approximate coordinates
# U RS ,0 cos D RS ,0, X 'X R cos aRS ,0,Y 'YR cos aRS ,0, Z 'Z R X RS
X R X R ,0
U RS ,0
X S X R ,0
approximate receiver-satellite range

U RS ,0
cos D RS ,0, X
'X R

From the geocentric to the local system
Geocentric coordinates X , Y , Z can be

transformed to local e, n, u coordinates
by means of:
translation of the origin from the
geocenter to the approximate receiver
position;
rotation of the axes by the latitude and
longitude M0 , O0 of the new origin.
The equation of observation in the local system
cos a1R ,0,e cos D 1R ,0, n
cos aR2 ,0,e cos D R2 ,0, n

A n u 3
x 3 u 1
cos aRn ,0,e cos D Rn ,0, n cos aRn ,0,v
v n u 1
XR
cos a1R ,0,v

X 1R
'eR
cos aR2 ,0,v
X2
'nR R

'vR
n
U Rn ,0 cos aRn ,0,e 'eR cos D Rn ,0, n 'nR cos aRn ,0,v 'vR X Rn
U R2 ,0 cos aR2 ,0,e 'eR cos D R2 ,0,n 'nR cos aR2 ,0,v 'vR X R2
U 1R ,0 cos a1R ,0,e 'eR cos D 1R ,0,n 'nR cos a1R ,0,v 'vR X 1R
The system of equations of observation in the local system
E = satellite elevation
A = satellite azimuth
cos aRS ,0,v sin ERS,0
cos D RS ,0,n cos ERS ,0 cos ARS,0
cos aRS ,0,e sin ERS,0 cos ARS,0
U RS ,0 cos aRS ,0,e 'eR cos D RS ,0,n 'nR cos aRS ,0,v 'vR X RS
After applying the transformation the equation of observation reads:

U RS
U 1R
U R2
U Rn
Using matrix notation

U 1R U 1R ,0
U 2 U R2 ,0
R

U Rn U Rn ,0
L n u 1
measured approximated
receiver satellite range
1
2
qnn
V 2 qen
qvn
qne
qee2
qve
1
A AT L
Least Squares solution

supra-index T stands
for transposed matrix
anv
qev variance-covariance of the unknowns.

qvv2
Dilution of the precision factors
geometrical configuration
of the measured satellites
is the unity of weight standard deviation;
are the error estimates;
'nR , 'eR , 'vR are the unknown estimates;

n
i 1
Xi2
n 1
V 2 AT A
v L A x
C
x
measurement errors
eR
n R
vR ,0 'vR
eR ,0 'eR
nR ,0 'n R
V v V qvv
V e V qee
V n V qnn
The receiver position and its error can be computed

from:
vR
The q factors (co-factors) depends only on the

geometrical configuration of the measured satellites
and defined the Dilution of the Precision (DOP ) factors:
3-D
horizontal
V PDOP
V qnn2 qee2
vertical
V P V qnn2 qee2 qvv2

V H
V qvv V VDOP
V HDOP
VV
good (PDOP<3)
Poor (PDOP>6)
Measurement errors: will be discussed later on.

Errors in the satellite position:
Broadcasted ephemerids: 1 m
RT IGS orbits: 5 cm
Final IGS orbits: 2.5 cm
X R ,0 Y S YR ,0 Z S Z R ,0
Error sources
Approximate receiver position (could be improved iteratively).

These errors apply on the approximate receiver-satellite range and, with much less
importance, on the direction cosines:
U RS ,0
cos D RS ,0,n cos ERS ,0 cos ARS,0
cos aRS ,0,e sin ERS,0 cos ARS,0
cos aRS ,0,v sin ERS,0
3) pseudo-range measurement
GPS carriers
Electromagnetic waves are formed by a carrier and different modulations.
The carrier is pure sinusoidal wave mathematically described by 3 parameters:
c
wavelength
299792,458 km/s
speed of light in vacuum
s (t ) a sin(2 S f t I )
a = amplitude
f = frequency
I = phase
GPS codes
f0 =10.23 MHz
GPS satellites broadcast 2 carriers: L1 and L2 (Block II-F also L5), whose
frequencies and wavelength are:
L1: f1 = f0 x 154 = 1575.42 MHz; O1 =19.05 cm
L2: f2 = f0 x 120 = 1227.60 MHz; O2 =24.45 cm
GPS carriers are modulated with civil and

military codes and with the broadcasted
ephemerides.
Codes are a pseudo-random binary sequence
(only two values: 0 or 1) modulated on the
phase of the carrier (code transitions change
180 the phase of the carrier).
Repetition interval
Chip length
L1 (and L2 in the II-RM)
1 ms
~300 m
Anti Spoofing (partially

used by GPS receivers)
L1 and L2
1 week
~30 m
P (Precise)
Modulate on
No
CA (Coarse acquisition)
Encrypted
c W
Measurement of the satellite to receiver range
c W
'T = error of the satellite clock

w.r.t. GPS time;
w.r.t. GPS time.
't = error of the receiver clock
1 ns 30 cm!
W W ' 't 'T
c W ' 't 'T c,

W ' c 't 'T
Measurement of the satellite to receiver pseudo-range
Assuming satellite and receiver clocks synchronized to GPS time:
PRS c 'T S U RS ,0
Equation of observation for pseudo-range

unknown: the receiver clock offset
c 'e
R
X 1R
cos aRS ,0,e 'eR cos D RS ,0, n 'nR cos aRS ,0,v 'vR c 't R X RS
correction: the satellite clock offset
A n u 4
Error sources
x 4 u1 v n u1
cos aR2 ,0,e cos D R2 ,0, n cos aR2 ,0,v c 'nR X R2

'vR
n
cos D Rn ,0,e cos aRn ,0, n cos aRn ,0,v c 't R X R
cos a1R ,0,e cos D 1R ,0, n cos a1R ,0,v
The system of equation of observations for n satellites reads

1
PR1
'T 1 U R ,0
2
PR2
'T 2 U R ,0
c

n
'T n U R ,0
PRn
L n u 1
Behaviour
30 cm Random, increases at low elevation.
1 m
1m
CA
Measurement
3 m
Error source
Multipath
Quasi-random, 12h sidereal time

repetition pattern; increases at low
elevation.
Errors in the satellite clock correction:
Broadcast ephemerid: 5 ns (1.5 m)
RT IGS orbits: 150 ps (4.5 cm)
Final IGS orbits: 7.5 ps (2.2 cm)
Error sources
U ' c 't 'T
Carrier phase measurement
U I,
O N O c 't 'T
an integer number
lR1 U 1R ,0 c 'T 1
lR2 U R2 ,0 c 'T 2
lRn U Rn ,0 c 'T n
In matrix form:
'nR
Equations of observation for carrier phase measurements

cos a1R ,0,e 'eR cos D 1R ,0, n 'nR cos a1R ,0,v 'vR c 't R O N R1 X 1R
cos aR2 ,0,e 'eR cos D R2 ,0, n 'nR cos aR2 ,0,v 'vR c 't R O N R2 X R2
L
't R
Ax v
'vR
n measurements
ambiguities
an integer number
N Rn
0 0
O 0

O
N R1 N R2
cos aRn ,0,e 'eR cos D Rn ,0, n 'nR cos aRn ,0,v 'vR c 't R O N Rn X Rn
'eR
cos D 1R ,0,e cos D 1R ,0,n cos D 1R ,0,v c O
cos D R2 ,0,e cos D R2 ,0,n cos D R2 ,0,v c 0

A

n
n
n
cos D R ,0,e cos D R ,0,n cos D R ,0,v c 0

3+1+n unknowns
Solution of the equation of observation system

Accurate solutions are obtained with the geometrical method:
accumulate measurements until the number of equations gets greater than
the number of unknowns; and
the condition number of the matrix A gets smaller enough to ensure a good
de-correlation of the different unknowns.
direction cosines change slowly
1.5 hours
L1 or L2
30 cm 2 mm
CA
1 m
1m
Error source
Measurement
3 m
10 mm
Multipath
ambiguity
Error sources
Behaviour
carrier phase measurements

are ultra-precise but
ambiguous
cycle slips are produced when the receiver

lost and re-track a satellite signal
4) models to reduce error sources
Error budget (if not corrected)
continental drift
crustal deformation
seismic and volcano activity
Earth tides
Earth orientation and irregular
rotation
post-glacial rebound
ocean, hydrological and
atmospheric loading
Posicionamiento precisoEffects that must be accounted for precise positioning

orbit
clock
relativity
antenna phase centers
wind up
instrumental biases
multipath
ionized atmosphere
(~50-1000 km)
neutral atmosphere
(~0-50 km)
clock
antenna phase center
instrumental biases
multipath
measurement errors
Posicionamiento preciso
Note: an algorithm is needed to

compute the GPS emission time
from the receiver reception time.
1 V
r
2 c
tS tR
tR
Error ionosfrico
UR US
c2
Relativistic correction
Vr = relative velocity
US = satellite potential; UR = receiver potential
tS = satellite time; tR = receiver time
Satellite clock is moving with respectt to

t th
the receiver
i clock; in addition, satellite and
receiver clocks are under different gravity potentials.
tS tR
tR
Antenna phase centers
PE = geocentric gravitational constant
E = eccentric anomaly
a = semi-major axis
e = eccentricity
The major part of these errors is compensate by

changing the frequency of the satellite oscillator
from 10.23 to 10.2299999954 MHz (~38 Ps/day).
The variable part of these errors can be
computed as:
'U REL
2
P E e a sin E
c
f E # 10.2299999954 u106 Hz
The measured pseudo-range is

between antenna phase centers.
Satellite coordinates correspond to

the center of mass
must be translated to antenna

phase center with an eccentricity
vector.
Geodetic coordinates must be

referred to a geodetic mark
the antenna phase center of the

receiver must be related to the
geodetic mark with an eccentricity
vector.
Variable part (Phase Center Variation)
))))))& )))))& ))))))&

'U RPC GU RPC GU RPCV
Constant parts
Antenna phase center of the receiver

)))))))&
'U RECC
PCV in mm (after Steigenberger et al., 2006).
))))))& )))))&
S
S
'U PCM
GU PCV
variable part (Phase Center Variation)
))))))&
S
'U ECC
constant part
Antenna phase center of the satellites
Values are different for L1 and L2 and change if a radome is on the antenna.
The orientation of the x-y-z system changes w.r.t.

the observer as long as satellite moves
Values are different for L1 and L2.
PCV in mm (after Steigenberger et al., 2006).
Carrier phase wind-up
It is due to the circular polarization of the GPS signals.

Values are different for L1 and L2.
Depends on the relative orientation of satellite and
receiver antennas.
Changes in the relative orientation cause a phase
variation that the receiver misunderstands as a range
variation.
'I
d ' d

sign G 'u G arccos
d' d

d x [ u \

d ' x ' [ ' u \ '
Instrumental biases
The satellite and the receiver electronics produce delays (a few

nanoseconds) in the time of emission and reception of the signals.
If not corrected, these delays are misinterpreted as an increase of

the propagation time and hence the measured pseudo-range (a
few meters).
Delays are different for code and carrier measurements.
Delays are frequency dependent.
c
v
refraction index; n=1 in vacuum.
1
1 dr
c
0
refractivity
1
n 1 dr
c
0
106 N dr
Ionospheric error
Errors caused by the atmosphere

Speed of light in the atmosphere, v, differs from the vacuum, c:
n 1 dr
0
W'
The propagation intervals in the atmosphere and in vacuum are:
W W '
dr 1
W
n dr
v c 0
0
(neglecting the bending)
and the difference:
'W
c 'W
which is equivalent to a range error:
'U
NI
40.3 u 106
ED
f2
40.3
ED dr
f 2 0
40.3 u 1016
TEC
f2
L2: 1,22760E+09
L1: 1,57542E+09
frequency
0,267 m
0,162 m
error (1 TECu)
The ionosphere extends from ~50 km above the Erath surface and is characterized by the
presence of free electrons (electrons dissociated from atoms and molecules).
Its mass is lower than 0.1% of the total atmospheric mass, but free electrons interact with
electromagnetic waves changing its speed of propagation.
f = frequency (Hz);
ED = electron density (electrons/m3)
- for carrier / + for code
'U I
TEC Total Electron Content

(TECu 1 TECu = 1016 electrons/m3)
TEC varies with local time,

season, solar activity, latitude and
geomagnetic perturbation.
l2
1.5457
TEC maps computed by SIRGAS
f 22
40.3
TEC
U 'U I ,2 O2 N 2 X2
,
Ionosphere free combination of carrier phase measurements
Ionospheric delay may reach 40

m for a low elevation satellite.
l1
TEC
U 'U I ,1 O1 N1 X1
,
f12
40.3
l3
2.5457
f12
f 22
l
l2
1
f 2 f 22
f12 f 22
1

2.5457 2 V 12 1.5457 2 V 22 # 3 V 1 # 3 V 2
2.5457 N1 O1 1.5457 N 2 O2
~ 3 times greater than the errors in L1 or L2 measurements
2.5457 X 1.5457 X2 V 3
1
Eliminates the ionospheric error:
b3
the ambiguity is not an integer anymore
N D NW
4
hD
Error troposfrico
'U D , z
height variability
'UW , z
11000
hD
hW
Tropospheric error
dry component
hD ~ 40 km
wet component
hW = 11 km
106
N D ,h 0 hD
5
106
NW ,h 0 hW
5
Tropospheric correction
106 NW (h) dh
106 N D (h) dh
hW
40136 148.72 T 273.16
Dry (~90%) + Wet (~20%) components
height variability
p h h
77.64 S
T hD

0
The neutral atmosphere (troposphere + stratosphere) extends ~ 50 km

above the earth surface; according to the Hopfield model:
NT
N ( h)
D
N D ,h
NW ,h
e
e hH h
N ( h)
12.96 3.718 105
2
T

T

hW

hD and hW = scale height (m)
p = atmospheric pressure (mb)
T = temperature (K)
e = partial water vapor pressure (mb)
all reduced to h=0.
'U D , z 0 mD ( z ) 'UW , z 0 mW ( z )
The total tropospheric delay (dry + wet) for a zenith distance z, is given by:
'UT ( z )
mD /W ( z )
a /W
D
1
b
1 D /W
1 c D /W
a D /W
cos( z )
bD /W
cos( z ) cD /W
cos( z )
mD/W(z) = Dry/Wet mapping functions (e.g.: Vienna mapping function);

aD/W, bD/W, cD/W = empirically determined functions dependent on latitude, day of
year and height.
4
Water vapor variability
'UT(z) = ~ 2-2.5 m for z = 0 / ~ 20-28 m for z = 85.
Estimation of the tropospheirc correction

Error ionosfrico
Empirical correction
'U T , z 0 GUT , z
Models are not accurate enough to correct the tropospheric error (specially the wet
component.
Errors affects mostly the estimation of the height.
An empirical correction is added to account for the unmodeled correction:
'UT , z
Zenith delay
Modeled value
An empirical correction is estimated together with the station coordinates, the receiver
clock and the ambiguities.
Typically one correction per hour is estimated.
5) point and differential positioning
Equation of observation for point positioning
Unmodeled
tropospheric
correction
Measurement + multipath errors

on the ionosphere free
combination
Correction to
receiver clock
cos aRS ,0,e 'eR cos D RS ,0,n 'nR cos aRS ,0,v 'vR c 't R
Ionosphere-free combination of
carrier phase measurements
Corrections to the approximate
receiver coordinates
Receiver-satellite range computed
Receiver-satellite
from satellite ephemerides and
direction cosines
approximate receiver coordinates
l3 SR U RS ,0 'U RS
Ionosphere-free bias
(including satellite and
receiver instrumental
biases and wind-up).
Troposphere wet
mapping function
b3 SR mW z RS GUTR X3 SR
Corrections:
satellite clock
relativity
satellite and receiver
antennas phase centers
troposphere (dry + wet)
coordinate variations (solid
tides, ocean loading, EOP, etc.)
Time
dependence
Constant
Varying
none
none
none
Receiver clock and tropospheric corrections

Error ionosfrico
None
relative
Type of
constraint
Receiver coordinates (assuming that

geophysical variations are modeled)
one per epoch
unknown
Ambiguities (assuming that cycle slips

have not occurred)
Updating
interval
Receiver clock
abs. and rel.
Vr
one per hour
Va
Tropospheric correction
V
is
the
standard
deviation
for the
a
absolute value
V
is
the
standard
deviation for the
temporal variation
r
2
l1,2
Single differences
l11 l21
1
1
l12 l22
l21 l12 l22
Double difference
1
2
l1,2
l1,2
Errors in the receiver clocks cancel out;

Measurements errors increase by a
factor 5 .
l1,1,22
Errors in the satellite clocks cancel out;

Measurements errors increase by a
factor 2 .
1
l1,2
Equation of observation for differential positioning
satellite 2
satellite 1
receiver 2
receiver 1
X
2
satellite 2
cos a22
2
1
cos a
receiver 2
cos a12
satellite 1
cos a
1
1
X1
receiver 1
satellite 2
satellite 1
2
1
1
1
l21 l12 l22
Effect on receiver coordinates
1,2
l1,2
1
2
cos a1 cos a1 X 1
1
2
1
1
b12 b22
l21 l12 l22
Effect on bias and troposphere
1,2
l1,2
1
1
1
2
b b
1
m12 GU1 m22 GU 2
1
2
1
2
b1 b1 b2 b2
1
2
1
2
m1 m1 GU1 m2 m2 GU 2
1, 2
b1,2
m GU m GU
1
1
cos a12 cos a22 X 2
cos a12 X 1 cos a22 X 2
cos a11 X 1 cos a12 X 2
b22 ; m22 GU 2
b ; m GU
2
1
b21 ; m12 GU 2
b11 ; m11 GU1
receiver 2
receiver 1
receiver 2
l
l
l
l
l
l
1
2
2
2
1
3
2
3
1
1
2
1
l l
l l
l l
l l
l l
l l
1
1
2
1
1
2
2
2
1
3
2
3
A0 / S
1 0 1 0 0 0 l11
0 1 0 1 0 0 l12
0 0 1 0 1 0 l21
0 0 0 1 0 1 l22
1 0 0 0 1 0 l31
0 1 0 0 0 1 l 2

3
Correlations between baselines
A 0 / S AT0 / S
2
0
1
0
2
0
0
2
0
l11 l21
l12 l22
l21 l31
l22 l32
1
0
2
1 0 1 0
0 1 0 1
2 0 1 0
The variance-covariance matrix is singular
1
1,2
2
1,2
1
2,3
2
2,3
1
3,1
2
3,1
C SD
0 1 0 2
1 0 1 0
0 1 0 1
Correlation between single and double differences
receiver 3
red + green = - blue
Let s assume 3 receivers and 2

satellites which leads to 6 single
differenced combinations possible
receiver 1
1
1,2
2
1,2
1
l2,3
2
l2,3
A S / D A0 / S
l11
l12
1 1 1 1 0 0 l21
0 0 1 1 1 1 l 2

21
A0 / D
l3
l32
l11
l11
1 0 1 0 0 0 l12
0 1 0 1 0 0 l21
0 0 1 0 1 0 l22
0 0 0 1 0 1 l1

32
A0 / S
l3
l
l
C DD
1,2
1,2
1,2
2,3
l
l
l
l
2
1,2
2
2,3
5 0 4 0
0 5 0 4
4 0 5 0
0 4 0 5
1
l1,2
2
1 1 0 0 l1,2
0 0 1 1 l1

2,3
2
AS / D
l2,3
A 0 / D AT0 / D
1
1,2
1
2,3
Lets assume that the red and green independent baselines are chosen (more
measurements, short lengths, other criteria):
1,2
l1,2
1,2
l2,3
l12
l1
22
l2
l31
l32
6) network computation
Average satellites per epoch
Sampling rate
50 x 1/15 x 86400 x 6 = 1728000
24 h (86400s)
1/15s
50
Observations
Observation period
Receivers
Equations of observation
Receiver clocks
50 x 24 = 1200
50 x 1/15 x 86400 = 288000
One per hour; constrained.
One per epoch; constrained.
Just an example
Tropospheric corrections
50 x 31 x 2 = 3100
3 x 50 = 150
unknowns
Biases
292450
Receivers coordinates
Total
Assuming 2 continuous
satellites arcs per day (no
cycle slips).
L = Ax + v
Nx = b
The system of equations of observation
AT L1 /
Normal system
L = variance-covariance matrix of the measurements
N = T L1 $
L Ax ,
v t 6 L1 v
,and the variancenm
The solution provides the unknowns, x = N 1 b , the residuals, v
N 1 .
the standard deviation of the unity of weight, V 0
covariance matrix of the unknowns, X
Weighted solution
Combination of solutions
After L. Snchez
No net rotation & no

net translation solution
The SIRGAS processing centers compute weekly loosely constrained normal

equations (the reference frame is defined by only the satellite coordinates).
The SIRGAS combination centers compute the final solution by stacking the
individual normal equations and imposing the datum to the network.
Fixed solution
6) ambiguity resolution
X5
b5
l1

f12
TEC
l5
f 22
TEC
U 'U I ,2 O2 N 2 X2
,
Wide-lane combination of carrier phase measurements
l2
3.5294
f1
f2
D
D
TEC
TEC
f1 f 2 f12
f1 f 2 f 22
1.3
f O
'U I ,1
f1 D
TEC
f 2 f12
,

c
N1 N 2

f1 f 2

N5
O5 = 0.8619 m
4.52942 V 12 3.52942 V 22 # 6 V 1 # 6 V 2
f1
f2
l
l2
1
f1 f 2
f1 f 2

4.5294
U 'U I ,1 O1 N1 X1
,
f12
f 22
N
N 2 O2
1 O1
f12 f 22
f12 f 22
c
f1 N1 f 2 N 2
f12 f 22
Combination of Ionos-free and wide lane solutions
f1
f2
N
N 2 O2
1 O1
f1 f 2
f1 f 2
'U I ,5
4.5294 X1 3.5294 X2 V 5
b3
2
2
f1 f 2 f1 f 2
O3 = 0.1070 m
1. Calccompute a ionos-free
solution (float biases);
c
f N f N f N f N
1
1
2
1
2
1 2 2

f12 f 22
0
c f2
c
N N
f1 f 2 N1
1
2
f f
f12 f 22

2
1
2
2
c f2
c
N

N1
5
f f
f1 f 2

2
1
2. Use the obtained coordinates

and compute a wide-lane
solution (fixed ambiguities);
3. Use the fixed ambiguities to re-compute a ionos-free solution.
4. Vertical reference systems
L. Snchez (2015): VII SIRGAS School on Reference Systems, Santo Domingo - Dominican Republic, 16-17 November 2015 4-1
Reference systems are needed to describe the Earths geometry and

gravity field as well the Earths orientation in space.
A reference system defines constants, conventions, models, and
parameters required for the mathematical representation of
geometric and physical quantities.
When a reference system is accepted and widely applied, it is called
Conventional Reference System.
A reference frame realises a reference system in two ways:
- physically, by a solid materialisation of points
- mathematically, by the determination of coordinates referring to
that reference system.
The coordinates of the points are computed from the measurements,
but following the definition of the reference system.
The realisation of a conventional reference system is called
conventional reference frame.

The datum fixes univocally the relation between a reference frame
and a reference system.
Reference surfaces are needed:
to handle smaller quantities (e.g. geocentric coordinates are given
in millions of metres, while latitude and longitude in degrees
(max. 360), minutes (max. 60), seconds (max 60))
to distinguish between curvilinear surface coordinates for
horizontal positioning and heights above some zero-height
surface for vertical positioning
to satisfy people's usual sense of orientation.
Any vertical reference system is basically composed by

1) a reference surface, i.e. the zero-height level (vertical datum)
2) a vertical coordinate, i.e. a type of height
Its realisation is given by a vertical network, i.e. a set of points, whose
heights are of the same type considered in (2) and refer to the datum
specified in (1).
If the reference surface and height

type depend on the Earth's gravity
field, we talk about a physical
height system (e.g. orthometric
heights and geoid, or normal
heights and quasi-geoid) if not, it
is a geometrical height system
(e.g. ellipsoidal heights and
reference level ellipsoid).
4.1 Geometrical height system
Vertical coordinate: Ellipsoidal height (h): length, along the ellipsoid

normal, from the ellipsoid to the point P(X).
- today, h is derived from geocentric coordinates [X,Y,Z]
- earlier (for the adjustment of triangulation networks), h = H + N
U X const.
Reference surface: a level (or equipotential) ellipsoid.

U0
GM
a2 b2
2 b2
a
arctan
1 Z 2 a2
3
e' 2
a 2 b2
, m
b2
Z 2 a 2b
GM
GM o geocentric gravitational constant

a o semi-major axis
b o semi-minor axis
Z o (nominal mean Earths) angular
velocity
U0 can be univocally determined as a function

of the ellipsoid parameters:
U0
f
GM
e' 2 n
1
1 1
m
b
2n 1 3
n 1
L. Snchez (2015): VII SIRGAS School on Reference Systems, Santo Domingo - Dominican Republic, 16-17 November 2015 4- 4
Geometrical height system
Dependence of h on
the ellipsoid
parameters (form
and size)
The realisation of a geometrical reference system depends on

- the orientation and position of the ellipsoid whit respect to the ITRS
(reference system for [X,Y,Z]).
- the ellipsoid parameters: if any of those changes, reference surface
and vertical coordinate change.
- the primary geocentric coordinates [X,Y,Z]: if they change, the
vertical coordinate changes.
Dependence of h on the
orientation and position
of the ellipsoid
Source: www.sirgas.org
Geometrical height system

Dependence of h on the changes of [X,Y,Z]
BOGA: Bogot, Colombia

Subsidence
CONZ: Concepcin, Chile

Earthquake
NAUS: Manaus, Brazil

Hydrological loading
UYTA: Tacuarembo, Uruguay

Introduction of ITRF2008
Components of a geometrical vertical reference system
dhX
dt
Vertical coordinate: Ellipsoidal height h referred to a certain epoch t and

its variations with time:
hX, t ;
U X const.
Reference surface: stationary in time and space for a better modelling

of height changes:
U0
Remarks on the geometrical height system
For that, one has to be familiarized with standards, conventions, and

procedures applied for the realisation of the ITRS, i.e. the International
Terrestrial Reference Frame (ITRF) and its regional densifications like
SIRGAS (Latin America and Caribbean), EPN: EUREF Permanent Network
(Europe), etc.
Ellipsoidal heights are not measurable directly; they must be derived

from other geodetic parameters.
The primary coordinates [X,Y,Z] must be computed in the same
reference frame in which the GNSS satellite orbits are given, i.e. ITRF or
its regional densifications must be used as reference frame in GNSS
processing.
For global compatibility, the same ellipsoid must be used for the
conversion [X,Y,Z] o [M,O,h], at present the GRS80.
Ellipsoidal heights in local geodetic datums (e.g. PSDA56, NAD27, etc.)
can only be obtained by means of transformations.
The vertical position in GNSS is 2 3 times less accurate than the
horizontal position, because most of the error sources in GNSS act in
radial direction.
Accurate ellipsoidal heights (at mm-level) require long GNSS positioning
(at least three days) and post-processing following the IERS conventions.
Otherwise, accuracy in cm- to dm-level.
CP
4.2 Physical height systems
W0 WP
(Primary) vertical coordinate: Level difference between a reference

surface W0 and the equipotential surface passing through P(X) i.e. WP
'WP
Reference surface: equipotential surface defined by a potential value

(called W0 ). The realisation (geometrical representation) of this
equipotential surface with respect to a reference ellipsoid is the socalled geoid computation.
Since the primary observable are level differences, the W0 value is
usually selected arbitrarily. In addition, the W0 value per se does not
provide any information about the geometry of the reference surface.
Physical height systems
CP
g
0
W WP
g
Cp is called geopotential number and it is given in [m2s-2]. To facilitate

its use in practice, Cp is converted in a distance (given in [m]) by dividing
it by a gravity value .
HP
represents the mean gravity

value between the reference
surface W0 and the equipotential
surface WP.
The potential difference W0 - WP is constant (because they are
equipotential surfaces) and therefore, the value of HP depends on the
value of . As a consequence, we distinguish dynamic, orthometric and
normal heights.
Physical heights: some glossary
The geoid is the equipotential

surface that would coincide with
the mean sea surface if the
oceans and atmosphere were in
equilibrium; under the influence
of Earth's gravitation and rotation
alone; without other influences
such as winds, tides or
gravitational effects of other
bodies.
The co-geoid is the estimated
geoid, i.e. the real geoid cannot
be known, because the actual
mass distribution and actual
gravity vertical gradient are not
known precisely. The assumption
of hypotheses about that
produces something similar to
the geoid (the co-geoid), but no
the true geoid.
The telluroid is the surface

defined by those points Q, whose
normal potential UQ is identical
with the actual potential WP of
the points P on the Earths
surface, i.e. UQ = WP
The height anomaly ] is the
distance, along the normal plumb
line, between the Earths surface
and the telluroid. When plotted
above the ellipsoid the resulting
surface is called the quasi-geoid.
Telluroid, quasi-geoid and cogeoid are not equipotential
surfaces (the gravity vector is not
perpendicular to them).
Geoid and quasi-geoid are
identical in ocean areas.
CP
g
W0 WP
g

HP
If is the mean real

gravity value, along the
plumb line, between
Earths surface and geoid,
we get orthometric
heights.
If is the mean normal
gravity value between
telluroid and ellipsoid (or
between Earths surface
and the quasi-geoid), we
get normal heights.
If is a constant normal
gravity value, we get
dynamic heights.
Geopotential numbers in practice
g ; dW g dn
The derivative of the gravity potential in a

given direction is equal to the component
of the gravity along this direction. Along
the plumb line (perpendicular to the level
surfaces), it is:
dW

dn
B
gG n
A
g H B H A W A WB
Integrating dn between two points A and B, we have:

B
dW
A
i.e. the level difference dn between two points (A, B), located on the Earths surface
and on two different equipotential surfaces, corresponds to:
W A WB
dn BA H B H A
g
is the mean gravity value between the two equipotential surfaces WA and WB
W0 W B
dn z 0
Spirit levelling measures the geometric height

difference between two points, without
taking care of the effect of the gravity on the
level surfaces (non-parallelism), i.e. in a
closed levelling loop
g G n # g dn
0
If the starting point is on the geoid (W0),

we have the geopotential number of B
CB
g is the average of the gravity along the

levelling line connecting 0 and B.
The height value for B is then
W W
CB
0
B
HB
g
g
Note: g and are different values!
which also means
dn z H B
Levelled height differences are not the same as physical height differences!!
dn k
The levelled height differences must be converted into potential differences or into
physical height differences to satisfy the loop misclosure condition of zero, i.e.
dW
Afterwards they are adjusted by the method of condition equations or by the method
of parameter variation. k is known as the gravity correction (or reduction) to levelling.
gA gB
2
g IA dnIA g AB dn AB g BC dnBC g DII dnDII
In practice, the potential difference from geometric (spirit) levelling between two
bench marks I and II is given by:
'WII , I
g AB
1
sdn
1 / 16
Rkm
Rkm
Rkm
Rkm
C
C
Minimum accuracy of
gravity values for the
determination of
geopotential numbers
Usual availability of gravity for levelling
mg for R = 2 km
[10-5 m s-2]
mg for R = 1 km
[10-5 m s-2]
566
283
189
141
113
400
200
133
100
80
81
57
28
11
10
20
30
40
50
57
40
20
8
8
4
2
Height [m]
70
100
200
500
4
2
1
1000
2000
4000
Required input data:

4) (Real) gravity values
(observed or interpolated) at
levelled points to compute
the geopotential numbers.
pdn
Levelling segments measured in different epochs

3) Measurement epochs: to take into account crustal vertical
movements
r 2 Rkm mm
s
Changes in the levelled height

differences are caused by
vertical movments? Or are
they observation errors?

1) Levelled differences (dn) with systematic errors reduced (i.e.
atmospheric refraction, rod graduation errors, temperature effect on
the instrument and rods, etc.). Random errors are compensated by
means of the adjustment.
Accuracy standards in levelling:
s
r 3 Rkm mm
s
mm
mm
mm

r 2 6 Rkm mm
r 2 2
r 2 3
r 2 5

Standard deviation for a level difference

compared with known height differences
(junction with existing vertical networks) :
First-order
s
r 5 Rkm mm
Standard deviation of a level difference

measured in two directions (back-sight
and fore-sigth) :
Second-order:
s
s
Third-order:
Fourth-order:
R: length in [km] of the levelling line associated with the measured height difference.
4mm Rkm

2) Length of the levelling segments: the standard deviation of levelling
increases inversely proportionally with the distance. The levelling
lines shall be adjusted in one block and weighted with the inverse of
this standard deviation. For example in South America:
sdn
DYN
'H AB
DYN
'n AB k AB
Using the dynamic correction:
From geopotential numbers to physical heights

Dynamic heights:
From potential differences:
H DYN
; J
1
HN
HN
J dH
N
B
g J o45
J A J 45
J B J 45
G n m 45 o H AN m 45 o H BN
J o45
Jo
Jo
N
N
'H AB
'n AB k AB
Using the normal correction:
J oM
B
B
g J o45
g J o45
DYN
k AB
A J o45 G n A J o45 dn
J oM Normal gravity for the surface of the level ellipsoid at certain latitude M, normally 45.
C
m
Normal heights:
HN
Jm
N
k AB
A
Z 2 a 2b
GM
N
N 2

H
H
2
N 2
M
2
H J o 1 1 f m 2 f sin M a a 2 [ms ]
o
Mean normal gravity along the normal plumb line between telluroid and ellipsoid (analytically
J m estimable, iterativ)
Jm
1 wJ
1 w 2J
N
Jo
H
2 wH o
2! wH 2
LA MATA
LA MATA
Perfil topogrfico [m]
CORDILLERA
ORIENTAL
BOGOTA
Correccin dinmica (0) [cm]
ALTO DE SANTURBAN
CUCUTA
VALLE DEL RO
CATATUMBO
Courtesy
a semi-major axis, f flattening, M latitude of the point, m
BARRANCABERMEJA
VALLE DER RO MAGDALENA
SABANAS DE CRDOBA Y SUCRE
PUERTA DE HIERRO
CORDILLERA CENTRAL
MEDELLIN
BARRANCABERMEJA
CORDILLERA CENTRAL
Dynamic correction in two Colombian levelling lines

0
LA MATA
VALLE DEL RO MAGDALENA
Dynamic correction vs. topography (levelling loop length: 1364 km)
-450
3000
-900
1000
2000
BOGOTA
VALLE DEL RO
MAGDALENA
CORDILLERA
ORIENTAL
MEDELLIN
Dynamic correction vs. topography (levelling loop length 1820 km)
-600
-1200
4000
2000
50
VALLE DEL RO CAUCA
CALI
VALLE DEL RO
CAUCA
CALI
IBAGUE
MEDELL
CORDILLERA CENTRAL
VALLE DEL RO
MAGDALENA
Molodenki, Vignal, Hirvonen [cm]l
BUENAVENTURA
(Maregrafo)
RO ATRATO / RO SAN JUAN
CORDILLERA ORIENTAL
QUIBDO
ALTO DE LAS DELICIAS
CORDILLERA CENTRAL
CORDILLERA
MEDELLINOCCIDENTAL
CORDILLERA CENTRAL
Molodenki, Vignal, Hirvonen [cm]
MEDELLIN
LADERA DE LA CORDILLERA
CENTRAL
Normal correction vs. topography (levelling loop length 1640 km)
4000
2000
CORDILLERA
OCCIDENTAL
Normal correction vs. topography (levelling loop length 1143 km)

15
-15
1500
2250
BUENAVENTURA
(Maregrafo)
HO
g dH
J o45
g J o45
Gn
O
O
'H AB
'n AB k AB
O
k AB
J o45
g mA J o45
H AO
J o45
Courtesy
g mB J o45
Using the orthometric correction:
; gm
1
HO
H BO
750
BOGOTA
Normal correction in two Colombian levelling lines
C
g
Orthometric heights:
HO
gm
Mean real gravity along the plumb line between Earths surface and geoid. It can only be
estimated by means of hypotheses about the (unknown) Earths internal mass distribution and
the (unknown) vertical gravity gradient. Each different hypothesis produces a different type of
orthometric height.
Some examples of orthometric hypotheses:

gm
gH 2
O
1
g p g 0 g p 3,086 0,83818 U p 10 6 H p
2
2
Helmert:
n
O
1
g p g 0 g 0 g 0AP H O ; H O 1 H iO
2
Hp
ni1

gm

First method of Ramsayer:
HO
1 n
1
g i 3,086 x 10 6 H iO 2 3,086 x 10 6 2p
n i1

gm

Ledersteger:
Units: gp, gm, gH/2, g0 o [m s-2] ; Up o [10-3 kg m-3] ; HO o [m]
Ramsayer (II) [cm]
ALTO DE LAS DELICIAS
QUIBDO
Ramsayer (III) [cm]
VALLE DEL RO
MAGDALENA
CORDILLERA CENTRAL
MEDELLIN
Courtesy
Ledersteger [cm]
BUENAVENTURA
(Maregrafo)
RO ATRATO / RO SAN JUAN
Baranov [cm]
CORDILLERA ORIENTAL
BOGOTA
Ramsayer (III) [cm]
IBAGUE
Ramsayer (II) [cm]
CORDILLERA
MEDELLINOCCIDENTAL
CORDILLERA CENTRAL
Ledersteger [cm]
CALI
Ramsayer (I) [cm]
VALLE DEL RO CAUCA
CORDILLERA CENTRAL
Ramsayer (I) [cm]
VALLE DEL RO
CAUCA
Baranov [cm]
Orhometric corrections in two Colombian levelling lines

100
MEDELLIN
LADERA DE LA CORDILLERA
CENTRAL
Helmert [cm]
Orthometric corrections vs. topography (levelling loop length 1640 km)
50
-50
4000
2000
CORDILLERA
OCCIDENTAL
BUENAVENTURA
(Maregrafo)
CALI
Helmert [cm]
Orthometric corrections vs. topography (levelling loop length 1143 km)

20
-20
2250
1500
750

Comparison of normal (Molodenskii) and orhometric corrections
Definition of
Description
Dynamic heights
to
height
Orthometric heights
Physical heights summary
units
gm
1
HO
point P and the geoid.
C
gm
g dH
Nomal heights
plumb line between the ellipsoid and the telluroid (or
between the quasi-geoid and P).
; Jm
1
HN
J mA J o45 N J mB J o45 N
HA
HB
J o45
J o45
J dH
HN
ellipsoid and the telluroid (or between the quasi-
Jm
geoid and P)
HN
Gn
N
'n AB k AB
J o45
g J o45
Reference surface: the quasi-geoid (close to the
N
'H AB
N
k AB
Magnitude: mm ... dm
O
k AB
Jm is estimable univocally.
Hypotheses are not required
as Jm varies with the latitude.
other cases, heights differ in the same manner
same latitude have the same normal heights. In
Points on the same level surface and at the
h: ellipsoidal height, ]: height anomaly
h ]
HN
geoid but not a level surface)
univocally,
only x
wH are necessary. The value of HO
estimated
H O h N
Heights of points on the same level surface differ
h: ellipsoidal height, N: geoid undulation
Reference surface: the geoid
B
g J o45
g A J 45
g B J 45
G n m 45 o H AO m 45 o H BO
Jo
Jo
J o45
O
O
'H AB
'n AB k AB
Magnitude: mm ... dm
HO
HO
(scaled Distance, along the plumb line, between the surface Distance, along the normal plumb line, between the
between the geoid and P.
J N : constant normal gravity value at an arbitrary gm: Mean real gravity value along the plumb line Jm: Mean normal gravity value along the normal
o
conversion
latitude M (usually M = 45).
Simple
J oM
B
g J 45
J o45 o dn
geopotential numbers)
H DYN
DYN
'H AB
Magnitude: < 20 m
DYN
'n AB k AB
Correction
(for levelling)
DYN
k AB
g J o45
Gn
J o45
No geometrical meaning
Points on the same level surface have the
wg
be
Uniqueness
Heights values shall be univocally determinable, i.e. they shall not
depend on the levelling path.
Orthometric
Normal
Remarks
same height value

Hypotheses are not required
in the same manner as the gm gravity values
Hypotheses about mass density and distribution
as well as about the gravity vertical gradient
cannot
depends on the adopted hypotheses.
approximately.
gm
Zero-height surface
with physical meaning and independent of the heights (i.e. the zeroheight surface shall not change if heights change).
Physical heights summary
Geometric meaning
Heights shall represent the vertical distance between two points (one
on the Earths surface and one on the reference surface)
Height type
Units of length
Heights shall be given en units of length (or distance), i.e. in metres.
Dynamic
The same height value on the same equipotential

If water does not flow between two points, they shall have the same
height value.
Characteristics
Use of hypotheses
The use of hypotheses shall be avoided. If hypotheses are improved, the
height system must be changed totally.
Connection with geometrical heights

Physical heights shall be able to be combined with ellipsoidal heights.
/
/
Small gravity corrections

TTo
ob
be avoided in practical applications of local extension.
4.3 Physical references surfaces
For orthometric
heights: the geoid.
For normal heights:
the quasi-geoid.
Physical references surfaces

The determination of W is, in general, a nonlinear problem and is
further complicated by the fact that W itself is not harmonic
Both limitations can be overcome by linearization and
approximation of the shape, size and gravity field of the Earth by
those of an equipotential ellipsoid of revolution (reference
ellipsoid) with the same mass and same rotational velocity as the
Earth
Approximations:
W U T ; g J Gg ; H h N
Since U and its functionals can be computed analytically for a given
reference ellipsoid, the problem of estimating W now reduces to
determine
T W U V ) u ) V u
T is a harmonic function, i.e. 2T 0
g P0 J P0

Gravity disturbance:
Gg
g P0 J Q0
Gravity anomaly:
'g
Gg
1 wJ
T
J wh
wT 1 wJ

T
wh J wh
The fundamental equation of

physical geodesy:
'g
T
J
Bruns equation:
N
Taken from Torge, 2001
Boundary surface: the geoid
Definition: determine T, harmonic

outside the geoid, from gravity
anomalies on the geoid that satisfy the
fundamental equation of physical
geodesy.
Theory of Stokes (1849)
Approximation surface: the telluroid
Boundary surface: Earths surface
Definition: determine T, harmonic

outside the Earth surface, from gravity
anomalies on the Earth surface that
satisfy the fundamental equation of
physical geodesy.
Theory of Molodenskii (1945)
Approximation surface: the ellipsoid
Hypotheses:
No hypotheses are required.
'g g P J Q
P: on the Earths surface
Q: on the telluroid
Gravity anomalies:

Gravity anomalies:
'g g P J Q g P RFA RB ...

P: on the geoid
Q: on the ellipsoid
Hypotheses:
The geoid enclosed all masses, gravity
reductions for vertical gradient and
inhomogeneous mass density are
required.
1
S \ 'g dV
4SR Earth
T
S \
G1
Tk
0
R2
2S
h hp
R2
3G
] 0 dV |
'g
2R
2S
l 03
l 03
h hp
'g dV
\
\
1
\
6 sin 1 5 cos\ 3 cos\ ln sin sin 2
sin \ / 2
2
2
2
1
S \ ( 'g G1 G2 ...)dV
4SR Earth
Disturbing potential:

T
1
\
\
\
6 sin 1 5 cos\ 3 cos\ ln sin sin 2
sin \ / 2
2
2
2
Disturbing potential:
S \
The Stokes solution includes only the first

term of the Molodenskii solution.
l 0 2 R sin

Geoid vs. Quasi-geoid
N ]
g J
HO
H N HO
On ocean areas, geoid and quasi-geoid are the same, i.e. HO = HN = 0
On land areas, geoid and quasi-geoid differ by:
mean normal gravity value along the normal plumb line between ellipsoid and
telluroid (or between Earths surface and quasi-geoid)
g mean real gravity value along the plumb line between Earths surface and geoid.
g J corresponds, in a very good approximation, to the Bouguer anomaly.
This relationship allows the determination of the geoid from the quasi-geoid (or
viceversa). Its hypotheses correspond to the Helmert theory for the determination
of orthometric heights.
-76
-73
-70
-67
-79
36 [m]
12
32
-67
30
12
24
24
18
16
-70
12
-2
-73
-6
-10
-12
-18
-18
-76
-24
-3
-26
-3
-42 -36 -30
-79
-34

Geoid vs. Quasi-geoid
Differences between geoid undulations

and height anomalies in Colombia.
Courtesy
(Snchez 2003)
http://www.igac.gov.co/
-42
(Quasi-)Geoid computation in practice

f
The determination of the anomalous potential T Tk

S \ ( 'g G1 G2 ...)dV
4SR Earth
0
at any point requires gravity anomalies all over the Earth.
The long wavelength contribution NG
is provided by a set of spherical
harmonic coefficients (geopotential or
global gravity model);
The middle wavelength contribution
NL is estimated from the local
anomalies (terrestrial, marine or
aerial gravity in the area of study)
The short wavelength contribution NT is computed by using a digital terrain model
(topographic heights)
-70
NG
-67
-79
-76
-73
-70
-67
middle wavelength NL
-79
-76
-73
-70
-67
short wavelength NT
The (quasi-)geoid undulation is given by N = NG + NL + NT
Spherical harmonics and the gravity field
Since the gravity anomalies derived from the local terrestrial gravity data contain the
contribution of the global gravity model and of the topography, it is necessary to
remove these two contributions from the local anomalies: 'g= 'gL- 'gG- 'gT . They
are restored again, when the (quasi-)geoid undulations are computed.
-73
http://www.igac.gov.co/
-2
100 200 300 400 500 600 700 800 900 1000 1100 1120 [Pms ]
122
121
121
120
120
119
119
118
118
117
117
116
116
115
115
114
114
113
113
112
112
111
111
110
110
109
109
108
108
107
107
106
106
105
105
104
104
103
103
102
102
101
101
100
100
99
99
98
98
97
97
96
96
95
95
94
94
93
93
92
92
91
91
90
90
89
89
88
88
87
87
86
86
85
85
84
84
83
83
82
82
81
81
80
80
79
79
78
78
77
77
76
76
75
75
74
74
73
73
72
72
71
71
70
70
69
69
68
68
67
67
66
66
65
65
64
64
63
63
62
62
61
61
60
60
59
59
58
58
57
57
56
56
55
55
54
54
53
53
52
52
51
51
50
50
49
49
48
48
47
47
46
46
45
45
44
44
43
43
42
42
41
41
40
40
39
39
38
38
37
37
36
36
35
35
34
34
33
33
32
32
31
31
30
30
29
29
28
28
27
27
26
26
25
25
24
24
23
23
22
22
21
21
20
20
19
19
18
18
17
17
16
16
15
15
14
14
13
13
12
12
11
11
10
10
9
9
8
8
7
7
6
6
5
5
4
4
3
3
2
2
1
1
0
-1
-1
-2
-2
-3
-3
-4
-4
International Centre for Global Earth Models (ICGEM)

http://icgem.gfz-potsdam.de/ICGEM/
EGM2008 coefficients
EGM96 coefficients
Spherical harmonic and the gravity field
-76
long wavelength
-79
12
-67
12
-70
12
-73
12
-76
12
-79
12
36 [m]
-67
30
31
24
23
18
15
-70
12
-2
-73
-6
- 10
-12
-76
- 18
-18
-24
-26
-79
-34
-42
-42 -36 -30
36 [m]
-67
30
30
24
22
18
-70
12
14
-2
-73
-6
-3
-10
-12
-3
- 18
-18
-3
-76
-24
-3
-26
-3
-79
-34
(Snchez 2003)
-42 -36 -30
-3
-42
Courtesy
Taken from GFZ Report 09/02.


Spherical harmonic and the gravity field
Terrestrial and
marine gravity
anomalies of
high-frequency,
topography
Gravity field
satellite missions
Orbit analysis

de
dt
1 e 2 wR
1 e 2 wT

na 2 e wM
na 2 e wZ
da
dt
d:
dt
n
2 wT
na wM
1
wT
na 2 1 e 2 sin i wi
1 e 2 wR
2 wT

na 2 e we na wa
dM
dt
Measurements:
Orbit analysis
Distances, differences of distance, and directions to
satellites (Satellite Laser Ranging - SLR)
Principle:
Comparison of the non-perturbed orbit (normal o
Keppler orbit) with the real orbit described by the
satellites. Discrepancies are caused by gravitational
effects of anomalous masses (non-homogeneous
distribution).
Method:
The variation of the orbit (Keplerian) elements are
represented as a function of the anomalous potential. After
invertion, this potential is estimated:
dZ
dt
cos i
1
wT
wT

na 2 1 e 2 sin i wZ na 2 1 e 2 sin i w:
cos i
1 e 2 wT
wT

na 2 e we
na 2 1 e 2 sin i wi
di
dt
Design:
http://www.gfz-potsdam.de/pb1/op/champ/
Challenging Mini-Satellite Payload for Geophysical Research and Application (CHAMP)

Principle:
Satellite to satellite tracking: gravity field

modelling from variations of the distance
between two satellites (CHAMP GPS).
Modelling of the quasi-stationary and

time-dependent components of the gravity
field.
Resolution:
400 km quasi-stationary component
4000 km monthly variations.
Principle:

Gravity Recovery and Climate Experiment (GRACE)
http://www.gfz-potsdam.de/pb1/op/grace/index_GRACE.html
Modelling of the quasi-stationary and timedependent components of the gravity field.

Resolution:
170 km quasi-stationary component
300 km monthly variations.
Satellite gravity gradiometry:

measurement of the second derivative of
the gravity field (gravitational gradient).
http://www.goce-projektbuero.de/
Gravity Field and Steady-State Ocean Circulation Explorer (GOCE)
Satellite to satellite tracking: distance

between two twin satellites (GRACE) and
between them and GPS satellites.
Principle:
Modelling of the quasi-stationary gravity field

Resolution: ~100 km
GPS receiver
Accelerometer
Star sensors
15.07.2000 19.09.2010
Height: 454 km -> 300 km
Inclination: 87
GPS receiver
Accelerometer
Star sensors
Distances by microwaves
17.03.2002 -
Height : 485 km -> 300 km
Inclination: 89
GPS receiver
Gradiometer
Star sensors
17.03.2009-13.11.2013
Height : 259 km -> 224 km
Inclination: 96,5
Optimal combination: example model EIGEN6C2 (Frste et al. 2012)
Courtesy gren 2004)

Omission error: caused by neglecting degrees larger than lmax
l
The gravity satellite missions provide global gravity models with a

resolution of degree, order 200. This correspond to a spatial
resolution of about 100 km and the omission error in the derived
geoids can reach up to 50 cm! The combination of this models with
terrestrial data is absolutely necessary!

Uncertainties in the geoid EGM2008
Taken from An Earth Gravitational Model to Degree 2160: EGM2008, Pavlis, et al. EGU 2008.
Uncertainties in the geoid EGM2008
Differences between the GOCE combined model (TW01 up to degree 210 and
terrestrial data) and EGM2008 up to degree 2190 (Cortesy M. Sideris 2011)
International Centre for Global Earth Models (ICGEM)

http://icgem.gfz-potsdam.de/ICGEM/

Other global gravity models
4.4 Heights in the time-dependent Earths surface

and gravity field
The geometry and gravity field of the Earth are under the influence of
Secular effects: like plate tectonics, global isostatic adjustment, polar
wandering, subsidence, erosion;
Seasonal effects: like astronomic tides, solid/ocean/atmospheric tides,
ocean/atmospheric/hydrologic loading, variation in the longitude of the
day, polar motion, mass transports, etc.
Sporadic effects: like earthquakes, landslides, volcano eruptions, etc.
Heights in the time-dependent Earths surface and gravity field

In general, plate tectonic movements are represented by means of
constant station velocities, and well-known effects with magnitudes
larger than the precision of the measurements are usually reduced by
using precise models:
Solid Earth tides: deformation of the solid Earth caused by the
gravitational attraction of Sun, Moon, and other celestial bodies;
Ocean loading: elastic response of the Earths crust to ocean tides;
Atmospheric pressure loading: elastic response of the Earths crust to
the changing atmospheric pressure. Two causes are distinguish: tidal
(due to the gravitational attraction of Sun and Moon) and non-tidal
(due to temperature changes, weather conditions, relief, etc.).
Rotational deformation due to polar motion and variations in the
angular velocity (longitude of the day) in solid Earth (pole tide) and in
ocean areas (ocean pole tide).
Heights in the time-dependent Earths surface and gravity field
Treatment of the permanent tide in heights
Earths crust
In general, the time-depending tidal effects caused by Sun and Moon are
reduced from geodetic observations. The treatment of the permanent
deformation defines the so called tide systems, there are two concepts for
Earths geometry (surface, crust or topography) and three concepts for the
Earths gravity field (gravity, geoid, physical heights).
(Quasi-)geoid
W: Earth potential without Sun, Moon, etc.

Vt: Tidal potential
Vd: Further potential caused by the deformation
Deformed crust
Non-deformed crust

System
retained
time-dependent
(periodic)
retained
retained
retained
retained
It reflects the constant effects caused
by Sun and Moon on the Earth
(gravity/potential field and geometry)
It is the reality
Characteristics
instantaneous
removed
Tide effects
time-independent
(permanent)
direct
indirect
mean-tide
retained
It assumes that Sun and Moon do not

exist (or are moved to the infinity)
removed
removed
removed
removed
zero-tide
removed
It affects only the Earth's

gravity/potential field, but not the
Earth's figure; i.e. zero-tide and
mean-tide for the Earth's surface
(crust) are assumed to be identical.
tide-free
Heights of SLR stations in conventional tide free, tide free and mean tide systems
The current resolution (1983) of the International Association of

Geodesy requires the zero tide system for gravity field and related
quantities and zero(=mean) tide system for geometry (ITRF).
0,0602 0,1790 sin 2 M 0,0018 sin 4 M
Ellipsoidal heights from conventional tide-free system to zero tide

system in [m]:
'hT
N mt N nt
0.099(3sin 2 T 1)
Gt N
Conversion of (quasi-)geoid heights between different tide systems

in [m]:
N mt N zt
0.099k2 (3sin 2 T 1)
0.099(1 k2 )(3sin 2 T 1)
N zt N nt

Conversion of levelling heights between different tide systems in [m]:
Mkkinen (2008)
Heights in the time-dependent Earths surface and gravity field: summary
Direct and indirect permanent tide effects remain the most important
issue:
- Ellipsoidal heights referred to the IERS/ITRF are in the conventional tide
free system (nominal Love numbers, anelastic response to semidiurnal
frequencies);
- Sea surface heights are in mean tide system and mean geoids are
needed for the computation of DT;
- Geopotential numbers are in general in mean tide system. If tide
effects are reduced, they can be in tide free (fluid Love numbers) or zero
tide (indirect effect restored);
- Terrestrial gravity is give in:
- Mean tide system for IGSN71 values until 1979
- Tide free system between 1979 and ~ 1988 (International Absolute
Gravity Base-station Network - IAGBN)
- Zero tide system since 1988
Heights in the time-dependent Earths surface and gravity field: summary
- Tide system in geoid determination is dominated by the C20 coefficient,

i.e. the global gravity model used for the long wavelengths.
- Transformation between (conventional) tide free system and zero tide
system is not homogeneous:
k20 = 0,29 IERS89
ks = 0,94 (Lambeck 1980)
k20 = 0,30 IERS1992
ks = 0,933 (Mathews 1999)
k20 = 0,29525 IERS1992
k20 = 0,30190 IERS2003/IERS2010
k20 = 0,30 global gravity models
4.5 Classical height systems

Reference surface:
Definition: the geoid
Realisation: mean sea level measured at a tide gauge and averaged over
different time periods (ideally 18,6 years).
Vertical coordinate:
Definition: orthometric
ometric heights
Realisation: levelling
elling + orthometric correction
Drawbacks of the classical height systems

Problem 1: the mean sea surface is not an equipotential surface and does
not coincide with the geoid.
The dynamic topography varies about r2m globally.
Sea Surface Topography, Courtesy: W. Bosch, DGFI
Problem 2: sea level changes are not homogeneous globally; tide

gauges register different sea level variations depending on the
geographical location.
Mean Dynamic Ocean Topography, Courtesy: W. Bosch, DGFI
Problem 3: secular sea level changes are not taken into account. The
reference level depends on the time covered for averaging.
Tide gauge in La Libertad (Ecuador):
When the vertical datum was established, the mean sea level was 7
cm higher.
1954
1964
1969
CART 5,3 0,1 mm/year
1959
1974
1979
1984
1989
Problem 4: tide gauge registrations describe vertical crustal movements

or sea level changes?
7.35
7.15
7.25
7.05
6.95
6.85
1949

Problem 4: tide gauge registrations describe vertical crustal movements
or sea level changes?
International GNSS Service

Working Group
TIGA: GNSS tide gauge
benchmark monitoring
Problem 5: levelled height differences are corrected using different

gravity reductions (orthometric hypotheses, normal, normal
orthometric, etc.); sometimes no gravity reductions are applied.
Gravity corrections in a levelling loop of

about 1800 km and heights up to 4000 m
Problem 6: (unknown) vertical crustal movements affecting levelling

benchmarks were handled as observation errors within the adjustment.
Time series of the GNSS station BOGA (Bogot, Colombia)
In practice H and N
assumed invariant
Vertical datum discrepancies in South America
Vertical datum discrepancies in Europe
Source: http://www.bkg.bund.de/geodIS/EVRS/
The classical height systems

cannot be thrown away; they
have to be modernised by their
integration in a global unified
vertical reference system!
Classical height systems: Summary
refer to different levels (many

[dm] of discrepancy);
realise different types of heights
(normal, orthometric, etc.);
omit (sea and land) vertical
variations with time,
do not support the precise
combination of h-H-N;
are the base for vertical data
produced in the last 150 years;
cannot be replaced by ellipsoidal
heights (these do not describe
flow of water);
Levelling is (much) more precise
than the existing geoids and
global gravity models.
4.7 Towards a modern vertical reference system
The ITRS/ITRF provides a highly precise geometrical reference frame

(consistent in sub-cm level worldwide);
An equivalent highly precise physical reference frame is missing, it must

be given by realising a unified global vertical reference system;
Main objectives are:
to provide a reliable frame for consistent analysis and modelling

of global phenomena related to the Earths gravity field (e.g. sea
level variations from local to global scales, redistribution of
masses in oceans, continents and the Earths interior, etc.);
to allow the reliable combination of physical and geometric

heights in order to explode at a maximum the advantages of
satellite geodesy (e.g. combination of GNSS with gravity field
models for worldwide unified precise height determination).
Basic idea
Towards a modern vertical reference system
1) To satisfy h-H-N=0 at the cm-level worldwide;

At present: at the dm- to m-level (too imprecise)
Ideal: at the mm-level (still unrealistic)
2) For this combination, the new vertical reference system has to
support geometric and physical heights;
3) Similar realisation to the ITRF, i.e.
a global network with known vertical station positions;
regional and national densifications, i.e. integration
(transformation) of the existing local height systems.
International Height Reference System (IHRS)

Introduced by a Resolution of the International Association of
Geodesy (IAG) during the General Assembly of the International
Union of Geodesy and Geophysics (IUGG) in July 2015 (Prague)
1) IHRS: Geopotential reference

system co-rotating with the
Earth.
2) Coordinates of points
attached to the solid surface
of the Earth are given by
geopotential values W(X)

(and their changes with
time dC(X)/dt), and
geocentric Cartesian
coordinates X (and their
changes with time dX/dt) in
the ITRS.
For practical purposes, potential

values W(X) and geocentric
positions X are to be transformed
into vertical coordinates with
respect to a reference level:
1) geometrical component
h(t0,X); dh(X)/dt
conventional level ellipsoid

U0 = const.
2) physical component
Cp(t0,X); dCp(X)/dt
conventional fixed value

W0 = const.
Remarks on the vertical reference level W0
1
n
i 1
W0(i )
W0 is understood as the potential value of the geoid;

Since there are an infinite number of equipotential surfaces, the
geoid is to be defined arbitrarily by convention;
Usual convention: the geoid is the equipotential surface of the
Earths gravity field that best fits (in a least square sense) the
undisturbed mean sea level;
Since to satisfy this condition is not possible and since the sea level
changes, a convention about mean sea level (time span and area)
is also needed:
(i )
mean value at a local tide gauge W0 W0
mean value a several tide gauges W0
potential value of a best fitting ellipsoid in ocean areas W0 U 0
2
mean value over ocean areas sampled globally W W0 dS min
S
Remarks on the vertical reference level W0
The current (2015) best-estimated value for W0 is

W0 = 62 636 853.4 m2s-2
It was computed following the global approach W W0 2 dS min

S
and considering:
Different estimation methodologies but the same input models (for
redundancy);
Sensitivity of the W0 estimation on the Earth's gravity field model
Dependence of W0 on the omission error of the global gravity model
Influence of the time-dependent Earth's gravity field changes on W0
Sensitivity of the W0 estimation on the mean sea surface model
Influence of time-dependent sea surface changes on W0
Effects of the sea surface topography on the estimation of W0
Dependence of the W0 empirical estimation on the tide system
Rigorous error propagation analysis to estimate the influence of the input
data uncertainties on the W0 estimation.

This value was officially adopted in 2015 by the IAG as the

conventional reference level for the IHRS.
Strategy for the IHRS realisation
Definition:
Physical component
Main requirement: consistent combination of ellipsoidal and physical heights

h - HN - ] | h - H - N o 0 with high accuracy (mm ... cm) worldwide
Geometrical component
Definition:
Realization:
Coordinates: C X, t W0 W X, t ; dC X dt
Reference level: W X W0 const.
Realization:
Coordinates: h X, t ; dh X dt
Reference level:U 0 U X const.
1) referred to the ITRS/ITRF
2) conventional ellipsoid
1) Adoption of a suitable W0 value;

2) Realization of the reference surface
defined by W0 (i.e. geoid modelling);
3) Connection of the local reference
levels with the global one GW0i W0 W0i
(i.e. vertical datum unification based
on geopotential numbers);
4) Conversion into physical heights
(H, HN, ... )
Alignment of standards and

conventions to guarantee the
consistency between physical
and geometrical parameters.
Conventional W0 value
corresponding to a
conventional global geoid
Transformation parameters between local

levels W0i and the global one W0
Solution by
combined
least squares
adjustment.
At geometric reference stations

(ITRF/SIRGAS) and at international
connection points (GNSS + geopotential
numbers from levelling)
Integration of local (regional) physical height systems into the IHRS
On ocean areas around

reference tide gauges
(satellite altimetry + tide
gauge registrations)
At reference tide gauges (GNSS + tide gauge registrations)

Like the ITRF: A global network with regional/national
densifications.
This network shall include:
1) reference tide gauges (local vertical datum points);
2) main nodal points of the levelling networks;
3) geometrical reference stations (ITRF and densifications);
4) fundamental geodetic observatories (connection between
W0 and TAI).
These stations must be:
1) continuously monitored to detect deformations of the
reference frame;
2) referred to the ITRS/ITRF to precisely know their
geometric coordinates;
3) connected by levelling with the local vertical datum to
precisely know their local geopotential numbers.
Coastal approach
(direct at tide gauges)
h GNSS positioning;
Ci = 0;
Ti satellite only GGM +
terrestrial gravity data +
terrain model
Continental approach
(reference stations on land)
h GNSS positioning;
Ci levelling + gravity
reductions;
Ti satellite only GGM +
terrestrial gravity data +
terrain model
Vertical datum unification in practice

1) Based on the reference frame (tide gauges, levelling nodes, ITRF
stations, fundamental geodetic observatories);
2) Combination of free normal equations containing level differences
measured between the reference stations (common adjustment of
geopotential numbers);
3) Estimation of the anomalous potential Tj (after Molodenskii) and
combination with the geometrical reference system (J h) and the
geopotential numbers Ci in three approaches:
Ocean approach
(sea areas around tide
gauges)
h satellite altimetry + tide
gauge registrations;
Ci = ; J;
Ti satellite only GGM
(n=200) + terrestrial gravity
data + terrain model
4) Least squares adjustment of the three approaches.
(after Rummel und Teunissen 1988, Heck and Rummel 1990)
Observation equations for the vertical datum unification
at border points connecting neighbouring vertical datum zones:
H N ,i 1 P H N ,i P qGW0i 1 GW0i
P , Pk
V S \ dV
i
0
indirect effects
(negligible)
j 1
iz j
f P GW P
at tide gauges, levelling nodes, geometric reference stations
1
2SJ
vertical datum
discrepancies
(to be determined)
h P H N ,i P q'W E ] P e i P GW
f 0i P :
'W0 W0 U 0
GNSS positioning on
land and satellite heights from geop.
altimetry on sea
height anomalies
numbers on land
surface around tide
from GBVP
and sea surface
gauges
[GGM (n=200) +
topography around
terrestrial gravity +
tide gauges
terrain models]
1
q : , ei P : q f 0i P ,
q'W0 E ] P
N ,i 1
P H N ,i P @
w\ x wx
Observation equations for the vertical datum unification

Least squares adjustment
P
@ >H
x b v \ x0 A x ; A
Observation equations
N ,i
Observation vector b contains:
>hP H
Elements of the design matrix A are:
GW j ; j 1 J
1, 1 , q, q, e j , f 0i
Unknowns (vector x )
Solution: v
b \ x
A x b \ x0 A x l
x x0 x
l
0
Example: South America
Existing height systems
15 reference tide gauges;

mean sea surface level referred
to a different epochs (some
unknown);
Levelling since ~1940 with
dH/dt = 0;
in general no gravity reductions
applied;
no common adjustment;
First and second order levelling
networks comprise more than
360 000 km and 200 000 bench
marks.
Trend from
satellite altimetry:
2,4 r 0,8 mm/a
mean sea surface heights from

satellite altimetry (OpenADB);
Tide gauge registrations from
PSMSL;
GNSS positioning at tide gauges.
Data standardization (TIGA
objectives):
Geometric heights in sea areas

around tide gauges
Trend from gauge registrations: 0,6 r 0,2 mm/a
Trend from GPS: 2,2 r 2,2 mm/a
o Discrepancy: 2,4 (2,2 + 0,6) = 0,8 mm/a
Determination of vertical trends

from satellite altimetry, tide
gauge registrations, and GPS;
It is assumed that the trends
(dh/dt)Altimetry = (dh/dt)(Gauge + GPS)
Reduction of the reference sea
levels to a common epoch
(2005.0).
OpenADB: Open altimetry database @ DGFI-TUM

PSMSL: Permanent Service for Mean Sea Level
Geometrical heights on
land areas
Reference stations (663):

ITRF stations (10) +
SIRGAS stations (74) +
national densifications (579);
Data standardization:
Transformation of previous ITRF

solutions to the IGb08;
Stations positions given at a
common epoch (2005.0) (with
station velocities or a
kinematics model - VEMOS );
Transformation from
conventional tide-free to zerotide.
Observed height differences
Levelling lines provided by the

countries
Data standardization:

least squares adjustment country

by country to build free normal
equations for each vertical datum
zone;
astronomical correction + indirect
effect (levelling in zero-tide
system);
kinematic adjustment assuming
dH/dt | dh/dt;
combination of free normal
equations for countries with
international levelling
connections
Uncertainty of the input data
Height anomalies
Normal heights
Ellipsoidal heights
with input data standardized
Residuals
with input data as they are

Vertical datum parameters with respect to W0= 62 636 853,4 m2s-2
Uncertainty of about r5 cm in those countries with good data coverage;

Uncertainty of about r20 ... 40 cm in those countries with poor data coverage
(similar uncertainties have been found by other authors in other regions, e.g.
Gruber et al. 2012, Rlke et al. 2014, Gerlach and Rummel 2013)
Definition and realisation of a modern vertical reference system: Summary

A global vertical reference system
To solve the discrepancies
between the existing height
systems and
To support the different
techniques for height
determination.
Implicit characteristics:
One reference level (W0 or geoid) to be used globally;
All existing geopotential numbers (physical heights) referring to one and the
same global level;
Precise combination with geometric heights and geoid models of high
resolution, i.e. h-H-N=0.
5. Crustal deformation: Observation and Modelling

5.1 Earth structure and geodynamic processes
The Earths structure
(radial stratification)
Important geophysical parameters:

0DVVGHQVLW\DVDIXQFWLRQRIGHSWK
(surface: 2.6 g/cm3, average: 5.5 g/cm3)
Gravity g(r) as a function of the density
distribution within the Earth
Pressure p(r) as a function of density
and gravity distribution within the Earth
Elasticity parameters as a function of
depth (e.g. Lam SDUDPHWHUVDQG
Temperature T as a function of depth
Disciplines: seismology, gravimetry, thermodynamics, palaeomagnetism

Epicentre
tF
Focus
(Hypocentre)
tB
A Observatories
Observations: wave arrival times tO

at the observatories
To be adjusted: - travel time tO tF,
- relative positions F,
- event time tF,
- travel path sF-O,
- wave velocities v
along the path sF-O
(a function of depth)
Origination of earthquakes
Earthquakes are caused by the release
of accumulated stress along faults.
The released energies propagate as
elastic waves in the solid Earth.
Principle of seismology
tA
vS
S-waves
< 0.1
18
80
60
44
20
25
80
[109 Nm-2] [109 Nm-2]
- Granite
50
37
Water
1.3
0.3
Examples (Peridotite = 3.2 g/cm3)

Peridotite: vP = 7.3 km/s, vS = 4.3 km/s
Water:
Gum
- Lime rock
Crustal rocks
- Gabbro
Steel
- Peridotite
Material
Example elasticity parameters
Seismic waves
O 2P
U
P-waves
vP
P
U
(Lam parameters and )
vP = 1.4 km/s, vS = 0.0 km/s
Depth
2.0 2.9
3.0 3.9
4.0 4.9
5.0 5.9
6.0 6.9
7.0 7.9

Magnitude
45%
No.
300000
49000
6200
800
120
18
Events/year
12 %
85 %
Energy
<< 1 %
<1%
1%
3%
10 %
35 %
50 %
Energy
3%
30 %
25%
0 70 km
300 720 km
70 300 km
Magnitude and statistics of earthquakes

Magnitude (Richter 1935):
Local magnitude ML is the
base-10 logarithm of the
amplitude of a short-period
seismometer of amplification
2800 in a distance of 100 km.
Moment magnitude scale
(Hanks & Kanamori 1979):
MW = /3 log10 M0 10.7
(subscript w = mechanical work,
M0 = seismic moment [Nm])
Both scales are similar, they
are identical for M = 5.0)
Geographical distribution of earthquakes
Seismic wave propagation
v1
;
v2
v1
v2
sin D
sin E1
r1 sin E1
r2 sin E 2
sin D
sin(180q E1 )
Refraction at spherical boundary surfaces
r1
r2
sin D
sin E 2
const.
The propagation equation fixes the

geometry of the wave path assuming
velocities as a function of the radius
r sin E
v(r )
Fundamental equation of seismology:

The wave path is circular in a radially
stratified Earth.
Inner
no S-waves
in the outer
core: = 0,
i.e. fluid!
Focus
no direct waves
arrive in the
shadow zone
Propagation of waves
Outer
Core
Nomenclature according to propagation path
Mantle
Pressure
Inner
core boundary
Density
Shear module
Compression module
Outer core boundary
Gravity
Derived parameters
Classical Earth stratification from seismology
Discontinuity
Velocity in km/s
Jeffreys 1939, Gutenberg 1958

Crust
Upper mantle
Crust
Lower mantle
Lower
mantle
Outer
core
Inner
core
Principle of gravimetry / Earth gravity field
The Earth gravity field is nowadays determined by a combination of

satellite gravity field missions GRACE (Gravity Recovery and Climate
Experiment) and GOCE (Gravity Field and Steady-State Ocean
Circulation Experiment) with airborne / terrestrial gravimetry results.
Representation of the gravity field by the geoid
Preliminary Earth Reference Model (PREM)
LVZ (LID)
Transition Zone
Lower Mantle
Outer Core
Inner Core
Denomination
6347 ... 6356
6151 ... 6347
5701 ... 6151
3480 ... 5701
1222 ... 3480
Radius
2,60
2,90
3,36 ... 3,38
3,99 ... 3,44
5,57 ... 4,38
12,17 ... 9,90
13,09 ... 12,76
e.g. Gravity field models EIGEN (GFZ Potsdam, Foerste et al.)
Lower Crust
6356 ... 6368
1,02
Density [g/cm3]
Upper Crust
6368 ... 6371
... 1222
Oceans
(Dziewonski/Anderson 1981: Earth mass density from seismology)
Depth in km
10
11
Inversion of the gravity field into mass structures

The in version of the gravity field into the mass distribution inside
the Earth is not possible without constraints (improperly posed problem).
Introducing the structure from seismology (PREM) as constraints,
we may estimate the mass irregularities at the layer boundaries.
12
Lithosphere
part of the
geoid from
an inversion
into 4 layers
of the Earth
Inversion of the gravity field into Earth structure

Earths mantle part
of the geoid from an
inversion into a four
layers model of the
Earth
13
Earths core part of

the geoid from an
inversion into a 4layers model of the
Earth
(Wikipedia)
14
Heat flow:
Q = -JUDG7
[J m-1s-1K-1] = heat conductivity
GLY4 FdT/dt
c [J g-1 K-1] = specific heat
Principle of thermodynamics (Heat flow)
Thermal conduction equation:
/DSODFHRSHUDWRU
WKHUPDOGLIIXVLYLW\
0.14
[106 m2 s-1]

4.2
c [J m-1 K-1]
0.55
0.6
0.8
170
0.8 1.8
[J s-1 m-1 K-1]
1.1
0.23
0.8 1.2
1
2.5
410
24
[106 g m-3]
Glas
10.5
2.2 3.2
Water
Material
Silver
Rocks
Measurement of heat flow
Field measurements: grad T = (Tbelow Tabove) / (hbelow habove)

In the laboratory:
H[SHULPHQWDOO\DWGULOOFRUHVDPSOHV
Spherical harmonics development (degree 12) of measurements
15
[mWm-2]
Thermal definition of the lithosphere

The lithosphere (Greek lithos = rocky) is the solid layer of the upper
Earth. The lower boundary is defined by the melting point of rocks.
Melting curves
Depth [km]
Thickness of the lithosphere
Lithospheric thickness[km]
40
160
40
20
160
80
160
(Pollack and Chapman 1977)
60
40
60
Result from heat flow measurements and conversion via melting

rock temperature into depth [km]
Temperature [C]
17
16
Heat flow [mW m-2]
Principle of palaeomagnetism
The magnetic dipole field
The Earths magnetic field extends

from the inner core to where it meets
the solar wind. It is approximately a
magnetic dipole field tilted at an angle
of 10 with respect to the rotation axis.
Magnetism of rocks
Rocks can be permanently magnetised

by the Earths magnetic field.
The Curie point is where a magnetic
mineral crystal melts enough to loose
the polarity of its magnetism,
e.g. magnetite (Fe3O4) at 580C,
ferric trioxide (Fe2O3) at 680C.
Geomagnetic reversals
The Earths magnetic field inverts occasionally its polarisation.
18
19
Time scale of reversals
Magnetisation of magmatic rocks
21
20
Rising Magma takes the effective polarisation of the Earths magnetic

field when cooling down to the Curie point.
Normal and inverse polarised rocks are situated side by side.

This happens along all the spreading ridges of the Earth.
Magnetisation of magmatic rocks

Pattern of polarised magnetic rocks in oceanic ridges
upper
lower crust
upper mantle
transition
zone
lower mantle
outer core
inner core
continent
lithosphere
(depth
ocean
lithosphere
asthenosphere
asthenosphere
and radius in km)
Summary of the
Earths structure
The nomenclature of the

stratification has changed
due to the modern studies
of composition, dynamics
and physical properties.
The lithosphere includes

the crust and parts of the
upper mantle.
22
The asthenosphere includes

the rest of the lower mantle.
Geodynamic processes
A principal dynamic
process in the Earth
is the convection in
the viscous mantle.
23
Masses heated by
the Earths core are
rising in the mantle
and spreading aside
when cooled beneath
the lithosphere. The
streams separate the
lithosphere forming
diverging ridges.
The lithosphere subducts into the asthenosphere forming trenches.
Trench
Mid Atlantic Ridge
Geodynamic process in the Atlantic
Atlantic
Mantle
Core
24
The Mid Atlantic Ridge is a spreading zone between America and Africa
5.2 Plate tectonics and plate kinematics

Structure of the lithosphere
Dynamics: Oceanic ridge spreading, lithosphere subduction, collison

+ Slab pull
Collision
+ Drag
+ Ridge push
Driving forces of plate tectonics
Resistence
POLE OF
SPREADING
Modelling plate kinematics

POLE OF
ROTATION
Theorem of Euler: A straight motion

on a sphere can be described as a
rotation around a geocentric axis.
To determine the rotation axis and
the rotational velocity we have three
geophysical observation types:
Spreading rates at oceanic ridges
Azimuths of spreading ridges
Azimuths of seismic faults
and four geodetic observation types:
Very Long Baseline Interferometry
Global navigation satellite systems
Satellite laser ranging
Doppler satellite radio-positioning
Spreading velocities at oceanic ridges

Method of spreading rates (velocities) from magnetic reversals
2d
One determines the age of the magnetic stripes (normal or inverse),
measures the distance between those of the same age and computes
the velocity v = d/t.
Azimuths of oceanic ridge spreading
N
A
Compression
Compression
Dilatation
Inverse faulting
Oblique faulting
Strike-slip faulting
Normal faulting
Azimuths of seismic faults: Fault plane solution
Dilatation
Station A: First P-wave move = dilatation
Station B: First P-wave move = compression
The first P-wave move of the seismograph indicates the direction of motion
Fault plane solutions: Example Mediterranean
Inverse faulting along the Eurasian African plate boundary
Modelling of plate kinematics

Theorem of Euler: each differential motion of an entire spherical
cap can be described mathematicaly by one centric rotation:
dX/dt = : u X
Plate name
50.5740
62.9943
45.2329
33.8532
25.0052
24.4856
50.6195
45.5102
-2.4280
47.8005
-63.0451
-25.3483
) >@
286.0407
244.2353
355.5436
33.1708
266.9898
244.2414
247.7258
0.3436
274.1002
259.8728
107.3271
235.5830
/>@
0.2909
0.2383
0.5455
0.6461
0.2143
1.5103
0.2337
0.5453
0.2069
0.7432
0.6409
0.1164
Z>/ Ma@
10
With rotation pole (), /)

and velocity (Z) you can
compute velocities (dx/dt)
of any point on the Earth.

Africa
Antartica
Arabia
Australia
Caribbean
Cocos
Eurasia
India
N. America
Nazca
Pacific
S. America
NAZC
CARB
COCO
NOAM
SOAM
11
General motion and plate rotation poles in the SIRGAS region

( is positive in counter-clockwise rotation).
Observations of the plate model NUVEL-1A (De Mets et al. 1990)
277 extensions, 121 fault azimutes, 724 spreading rates = 1122 in total
Problem of the NUVEL-1A velocities for geodesy

Global rotation with respect to the Earth crust (no net rotation)
The ITRF refers to the geologic-geophysical model NNR NUVEL-1A.
The integration (interpolation) of the ITRF velocities over the entire
Earth crust yields a global rotation (1 mas = 31 mm at the equator):
ZX = -0,04 mas/a, ZY = 0,03 mas/a, ZZ = -0,03 mas/a (~ 2 mm/a)
This is due to the NNR NUVEL-1A model, which does not provide
the present rotation, but the average over ~3 Million years. In addition
it does not include plate (crustal) deformations.
12
Its possible to compute a kinematic plate model including crustal

deformations from geodetically observed velocities (e.g. ITRF2008)
fulfilling the NNR condition and providing consistency with the EOP.
The input data are all the observed station velocities of the ITRF.
Geodetic modelling of a no net rotation plate model
13
Integration: 6 6 v1u1 o 0
no net rotation compatible with EOP
Modelling of deformations of the

Earth crust (non rigid plates):
(dv/dt)pred = cT C-1 (dv/dt)obs
v = vector of velocities (M, O)
obs = observed
pred = predicted
(dO/dt)k
(dM/dt)k = Zi cos )i sen(Ok - /i)

= Zi (sen )i - cos(Ok - /i)
tan Mk cos )i)
Modelling of plate rotation vectors from geodetic point velocities:

Geographic North pole
Plate rotation pole
Geophysical plate model PB2002 (Bird 2003)
14
Velocity
0.2909
0.3079
0.2384
0.5929
1.2102
0.6462
0.2912
0.2337
0.5454
1.1971
0.2069
0.7432
0.2359
0.6408
0.1164
0.3478
0.4755
0.3928
Actual Plate Kinematik Model (APKIM)
PB2002
Longitude
-73.9481
-98.3803
-115.7365
-6.1940
27.2489
33.1689
-87.5775
-112.2750
0.3448
139.5455
-85.8946
-100.1303
-97.7137
107.3246
-124.4300
-93.2661
-73.1976
-150.4806
Latitude
50.5643
44.3353
62.9970
46.6691
40.9338
33.8554
34.0376
50.6311
45.5120
15.1572
-2.4098
47.8041
31.8492
-63.0448
-25.3247
49.7622
45.1695
66.8484
15
APKIM ITRF2008 NNR

Plate
Latitude
Longitude
Velocity
(PB2002)
AF Africa
48.80 0.16
-77.11 0.53
0.2808 0.0010
AM Amur
60.59 8.21 -104.91 16.9
0.2851 0.0240
AN Antarctic
59.51 0.24 -120.15 0.46
0.2293 0.0027
AR Arabia
49.42 0.41
2.94 1.74
0.5880 0.0145
AT Anatolia
40.14 0.12
28.11 0.22
2.0153 0.0790
AU Australia
32.89 0.09
35.86 0.18
0.6381 0.0007
CA Caribbean 31.11 1.92 -107.81 5.12
0.1968 0.0206
-94.40 0.36
54.74 0.22
0.2674 0.0009
EU Eurasia
49.84 0.93
IN India
9.87 6.77
0.5599 0.0203
12.32 0.49 144.12 0.27
MA Mariana
3.6205 1.2470
NA N.America -4.39 0.55
-84.11 0.13
0.1951 0.0010
45.07 0.55 -100.95 0.24
0.6519 0.0041
NZ Nazca
39.17 1.34 138.00 1.59
1.7743 0.3439
ON Okinawa
PA Pacific
-63.27 0.06 110.79 0.32
0.6705 0.0008
SA S. America -15.33 0.64 -122.22 1.72
0.1185 0.0011
SO Somalia
50.16 0.69
-88.42 1.52
0.3212 0.0056
SU Sumatra
-79.98 1.29
33.35 3.65
0.5001 0.0578
YA Yangtze
56.80 6.02
0.3164 0.0043
-99.58 8.49
16
Geophysical and geodetic plate kinematic models
(The model PB2002 is identical to NNR NUVEL-1A in the mayor plates)

5.3 Intra-plate and inter-plate crustal deformation
- Conservative (transverse faults)

- Constructive (spreading ridges)
- Destructive (subduction, collision)
Crustal deformations mostly are generated by colliding/diverging plates

Three types:
From en.wikipedia.org
Plate motions and deformation types
Long-term deformations
(a) Plate rotations
(b) Intra-plate deformation
(c) Inter-plate deformation
(c1) Sea floor spreading
(c2) Ocean plate subduction
(c3) Continents collision
(c4) Shear strain (strike-slip)
(d) Other deformations
(d1) Isostatic adjustment
(d2) Sediment basins
There are other short-term (periodic, seasonal, episodic) deformations,

e.g. from environmental loading, seismic events, which are not treated
Deformation zones (Argus and Gordon 1996)
Peter Bird 2003)
(cf. also PB2002,
Diagram
V-H or H-t
V0
Relative strain
at V = const.
Examples for geodyn. applications
Mantle convection,
long-term deformation of lithosphere
plates flexure under

loading
models of rigid
plates
(e.g. NUVEL)
H = V/E
H = (V/K) t
H = 0 for V<V0 processes in the

H = indefinite fault plane during
for V >V0 seismic events
5
Modelling as
a deformable
continuum
Defining the
used material
property is the
most important
prerequisite for
an adequate
approximation
of reality.
How to model crustal deformations?
H=0
for all V
Rock rheology
Mechanical
analogon
Basic physical materials

Denomination
rigid
(Euklid
material)
H
V
elastic
(Hook
material)
viscous
(Newton
material)
plastic
(St. VenantMaterial)
Diagram
V-H or H-t
V0
where t1 = K/E1
H = V/E+(V/K )t
for V > V0
H = V/E for V< V0

H = indefinite
V/E for V > V0
H = V/E for V< V0

H = (V-V0 )t/ K +
processes of
mountain building
flexure of plates in
subduction zones
general lithosphere
(incl. mantle)
H V / E (1 e t / t ) Earth tides,
Chandler wobble
of polar motion
Relative strain Examples for geodyn. applications

at V = const.
Rock rheology
visco-elastic
(Kelvin-Voigt
material)
Combined model materials
elasto-viscous
(Maxwellmaterial)
Mechanical
analogon
elasto-plastic
(Prandtlmaterial)
De-nomination
elasto-viscoplastic
(Bingham-mat.)
1 F
; H
EQ
1
p
E
F / Q)
G = shear modulus
W G J (W
'r 'l
V
:
(0 d V d 0,5)
r l
V = Poissons ratio (number)
Elastic deformations: elasticity parameters

Linear
deformation:
'l
l
E = elasticity module (Youngs-modulus)
Volume
deformation:
'V
p K
V
K = compressive modulus
Interrelation of elasticity parameters
E
3 (1 2V )
2
K G
3
E
2 (1 V )
E
2 (1 V )
1/ 2 K 1/ 3 G
K 1/ 3 G
(in any case only two are independent of each other)
3 K G
K 1/ 3 G
E V
(1 V ) (1 2V )
O, P = Lam parameters (mainly used in seismology)
w2u
wt 2
O G ij H kk 2 P H ij
P 'u (O P ) grad div u f
(p = stress-, H = deformation tensor,

O, P = Lam elasticity parameters,
G = Kronecker delta, i, j, k = 1 ... 3)
pij
Modelling of an elastic continuum

Hookes law for isotropic elastic material:
Equation of motion of
the elastic deformation:
U = rock density, ' = Laplace operator

Solution of the differential equation by integration with finite elements
u = S-1 f
u = point velocity
f = force function
S = stiffness matrix
S = S(x, O, P)
Shear : J
1 wv ww

2 wz wy
1 wu wv

2 wy wx
1 wu ww

2 wz wx
1 wv ww

2 wz wy
Extension : H
wv
wy
wu
wx
ww
wz
Example for three-dimensional models:
Result of the modelling using the finite

element method
Strain-Tensor e = H + J
Solution of the modelling
0
1 wu wv

2 wy wx
1 wu ww

2 wz wx
Example for South America
10
11
Alps
2008
M=8.0
2012
M=8.6
2011
M=9.0
(DUWKTXDNHV0IURPWR
2010
M=8.8
2009
M=8.1
Persia-Tibet-Burma
Global modelling of deformation zones

Alaska-Yukon
GordaCaliforniaNevada
1RUWKHUQ$QGHV
Peru
Altiplano
Puna-Sierra Pampeanas
Displacements observed
by the GPS GEONET of
the Geospatial Information
Authority (GSI) of Japan
before and after the M=9.0
earthquake, 2011-03-11.
There are more than 1000
continuously observing
GPS sites in a spacing of
about 25 km.
Data are processed and
interpreted in nearly realtime in order to allow a
prognostic alert.
Deformations of the 7KRNX earthquake 2011
The seismic deformations cannot be modelled like plate motions and

inter-plate deformations. They must be observed by geodetic methods.
12
(names of PB2002 orogens)

5.4 Monitoring seismic deformations by GNSS
Probability M > 8 is
one per year
Besides long-term plate tectonic motions and intra-plate deformations

there are episodic deformations, mainly generated by earthquakes.
10 largest earthquakes
since 1900
(from USGS)
Campaign observations of the Cariaco earthquake
SIRGAS
AS provides the continuous time series showing clearly the jump.
GPS weekly time series in Arequipa
After
shock
series
Full
time
series
The M=7.0 earthquake in Cariaco, Venezuela, 9 July 1997 was one of

the first large earthquakes monitored by GPS (CASA) in Latin-America.
1996.18 1997.73
1997.73 1999.20
1999.20 2002.18
El Salvador M=7.7 earthquake 2001
Continuous observations in Arequipa
Main shock (15 Jan.) and aftershock (13 Feb.) clearly identified
The M=8.4 earthquake in Arequipa, Peru, 23 June 2001 was the first
large earthquake monitored by SIRGAS continuously observing stations.
SLR observes the displacement after one week

Mexicali M=7.2
earthquake
4 April 2010
Maule M=8.8 earthquake, 27 February 2010

Deformations
up to 1600 km
from the focus
Large vertical
displacements
close to focus
Time series of deformations in Concepcin
Velocities have not yet recovered to the normal status
Earthquake in Guatemala, 2012-11-08
10
11
No earthquake in Manzanillo, Mexico, 2002-11-15
12
In Manzanillo
there was no
earthquake on
15 Nov. 2002,
but a collision
of a ship with
the pier where
the GPS antenna
was installed.
There were some
earthquakes on
22 Jan. 2003!
Caution with
interpretations!
Ionospheric modelling and analysis
Discovery of the ionosphere
In 1839, the German scientist Carl Friedrich Gauss

speculated about the existence of electrically
charged layers on the upper atmosphere, as the
responsible of the perturbations of the magnetic
field measured on the Earth.
In 1901, the Italian scientist Guillelmo Marconi

confirmed the existence of a electrically charged
layer in the upper atmosphere, which allowed
radio-waves emitted from USA arrive to Europe.
The Earths ionosphere
The ionosphere the region of

the high altitude atmosphere
where free electrons exist in
sufficient amount to disturb the
propagation of electromagnetic
waves.
Free electrons are produced

mostly by the UV and X
radiation of the Sun, which
causes the photo-ionization of
the atmospheres gases.
Chemical composition of the ionosphere

The ionosphere contains less than 0.1% of the atmospheres
mass, and less than 1% of that mass becomes ionised.
In spite of that, the charged particles make the gas
electrically conducting, which completely changes its
characteristics.
The most abundant gas molecules in the ionosphere are:

below 200 km: molecular oxygen (O2) and nitrogen (N2);
Between 200 and 600 km: atomic oxygen (O);
above 600 km: hydrogen (H) and Helium (He).
The geomagnetic field
The geomagnetic field can be approximated by a

magnetic bar in the Earth, tilted at an angle of 11
with respect to the rotational axis.
The differences between the actual

geomagnetic field and the dipole
approximation are called geomagnetic
anomalies; the most noticeable of them is the
South Atlantic Magnetic Anomaly (SAMA), which
is characterized by a depletion of the field intensity.
Solar radiation, solar wind,

interplanetary magnetic field
Electrodynamics of the ionosphere
Due to the Lorentz Force, an electrically

charged particle cannot move across the
magnetic field lines.
A typical example are electrons and ions

coming from the solar wind, which are trapped
in the van Allen radiation belts.
This particles are forced to rotate around the

magnetic field lines, drifting (north-south) along
the magnetic field lines, and preceding (eastwest) around the magnetic dipole axis.
The solar-terrestrial environment
Earth atmosphere, Earths magnetic field,

magnetosphere
Solar Wind interaction
The magnetospheric tail flaps in space much like a flag in the wind.
Solar cycle
The activity of the Sun (magnetic field, ejection of solar
material, radiation intensity) changes with a period of
approximately 11 years.
These changes are highly correlated with the number of dark
regions in the solar corona (sunspot) that have been observed
for several centuries.
Magnetic and ionspheric storms
Massive explosions in the sun's corona

(specially during high solar activity periods)
result in Coronal Mass Ejections (CME).
If the ejected mass reaches the Earth (in 1-2

days), it may produce a great distortion of
the whole magnetic field, thus creating a
magnetic storm.
In most cases, magnetic storms are

associated to great perturbations in the
ionosphere, which are known as
ionospheric storms.
Physical process in the ionosphere
Solar flares are sudden brightening

observed over the Suns atmosphere
associated to energetic emission of X
and UV radiation.
They are mainly followed by a coronal

mass ejection.
Both together radiation and mass

ejection can severely disturb the
onosphere and disrupt radio
communications, radars, GNSS and
other devices.
The ionosphere is stratified in different layers

according to the chemical and physical processes
that dominate at different heights.
Edward Appleton named the E (electrical) layer in
1927 after Guglielmo Marconi showed, during
communication experiments in 1901, that radio
waves between Europe and America had to be
bouncing off an electrically conducting layer at
around 100 150 km altitude.
Subsequently, other layers were discovered which
simply received the names D- and F-layers.
Vertical structure of the ionosphere
Parameters used to characterize the ionosphere
Electron density (ED)

Height of the F2 peak (hmF2)
Electron density of the F2 peak (NmF2)
Total Electron Content (TEC)
Quiet ionosphere
Disturbed ionosphere
Galileo
GLONASS
Observation of the ionosphere

In 1957 the spatial era began allowing the exploration of the ionosphere using
the satellites as dual frequency beacons (e.g.: TRANSIT system).
Presently, more than 60 satellites beacons are available including GPS,
GLONASS and Galileo.
Large networks of
continuously operational
receivers (such as
SIRGAS and IGS) collect
high quality measurements,
s,
which are useful for
ionospheric studies.
GPS
Observation of the ionosphere

Several dual frequency GPS receivers are flying onboard low Earth orbiter (LEO) satellites at
600800 km above the Earth surface.
These receivers provide ~2500 radio ocultations per day, during which the LEO 'sees' the GPS
satellite set or rise behind the Earth's limb, while the signal line-of-sight slices almost horizontally
through the Earths atmosphere.
GPS receivers onboard LEOs provide dual frequency
CHAMP
observations
distributed almost uniformly around the globe,
describing the TEC distribution at different heights.
GRACE
SAC-C
FORMOSAT-3/COSMIC
Dual frequency GNSS measurements (either

ground- or space-bases) provide information on
the total electron content (TEC).
L2 L1
m /TECu
S
S
0.105
, TEC E R E E R XTEC
TEC is usually measured in TEC unities (TECu),

1 TECU being 1016 electrons per square meter.
L5
ER and ES are the differential hardware delays

for the receiver and the satellite;
ERS=N2.O2 N1.O1 is the combination of the

carrier phase ambiguities.
GNSS-based ionospheric observable
SIRGAS ionospheric model
Combination of ground-based and

space-based GNSS measurements
allows a tomography reconstruction of
the electron distribution in the
ionosphere (like the technique used in
medicine to study the human brain).
Presently, the SIRGAS model is capable

of describing the 4-D (latitude, longitude,
height, and time) distribution of the
electron density based on the
assimilation of ground- and space-based
GNSS measurements.
It is based on the La Plata Ionospherc Model (LPIM) developed at the GESA laboratory of the
La Plata University; its development started in 1994 and the first operational version was
released on 1998.
height
latitude
longitude
TEC 1016 N e ds
SIRGAS TEC model

S is the emitter to receiver ray path (from the GNSS satellites
to either, the ground-based or the space-based receivers).
F1 region
resulting profile
F2 region
height of the
layer peak
Ne ( h )
N e,max e 2
Scale eight of
the layer
maximum
electron density
of the layer
1 z e z
Each layer is characterized by 3 parameters:
LPIM models the vertical profile of electron density by means of a superposition of 4 alphaChapman layers which represent the E, F1, F2 and topside.
topside
E region
h hmax
H
SIRGAS model for the F2 layer parameters
The F2 peak parameters:

NmF2 (maximum electron density);
(height of the F2 peak); and
HF2 (scale height of the layer),
hmF2
are separately modeled with spherical harmonics expansions dependent on the modip latitude
and the geographic longitude, and time dependent coefficients that are the main parameters to be
estimated by the Kalman filter:
L
l
2S
2S
: P , O , t a0 t alm t cos m
O blm t cos m
O Plm sin P
24
24
l 1 m 1
GNSS
Satellite Altimetry
CHAMP
COSMIC
SAC-C
SIRGAS model for the vertical profile
The vertical profile is modeled as a

superposition of three alphaChapman layers for the E, F1, F2
and topside layers
Data assimilation scheme
The main parameters of the E, F1 and topside

layers are anchored to the parameters of the F2
layer, and modeled in accordance to the ITU-R
recommendations.
robust and adaptive

Kalman filter
Quasi-real-time data from different

sources (ground- and space-based
GNSS, and satellite altimetry)
GRACE
Global maps of NmF2, hmF2

and HF2 and electron density
profiles
hmF2 for Diciembre, 2011, 19 LT
NmF2 for Diciembre, 2011, 19 LT
Some results
Some results
Vertical TEC for Diciembre, 2011, 19 LT
Some results
7. Reference system and frame for the Americas (SIRGAS)

Objective
To provide a reliable reference frame for:

1) Earth System research;
2) Scientific and practical applications based on high-precise

positioning.
This implies a reference system realisation with
1) A significantly higher accuracy than the magnitude of the

phenomena we want to study;
2) Homogeneous reliability and global consistency (the same accuracy

everywhere);
3) Long-term stability (the same accuracy at any time).
VII SIRGAS School on Reference Systems, Santo Domingo - Dominican Republic, 16-17 November 2015 7-1
SIRGAS comprises:
Standardisation of the gravity

field-related heights.
Full consistency with the

reference system of the
(GNSS) satellite orbits.
Accessibility to the global

reference system at regional,
national, and local levels.
To guarantee:
7.1 SIRGAS components

1) A regional densification of
the ITRF (the realisation of
the ITRS), as continental
reference frame;
2) National densifications of
the continental reference
frame;
3) A unified vertical reference
system.
Precise combination of
physical and geometrical
parameters.
7.2 SIRGAS structure

French Guyana
Guatemala
Honduras
Mxico
Nicaragua
Panam
Paraguay
Peru
Uruguay
Venezuela
Members
Argentina
Bolivia
Brazil
Chile
Colombia
Costa Rica
Dominican Rep.
Ecuador
El Salvador
Guyana
SIRGAS is the Sub-commission

1.3b (Regional Reference Frame
for South- and Central America)
of the Commission 1 (Reference
Frames) of the Internatinal
Association of Geodesy (IAG).
SIRGAS is a Working Group of the
Cartography Commission of
the Pan-American Institute for
Geography and History
(PAIGH).
7.3 SIRGAS definition and realisation
First SIRGAS realisation: SIRGAS95
SIRGAS as geometrical reference system is defined to be identical

with the ITRS (International Terrestrial Reference Frame).
SIRGAS as reference frame is a regional densification (continental
network) of the global ITRF (International Terrestrial Reference
Frame).
At present, there are three SIRGAS realisations:
by means of GPS campaigns:
1) SIRGAS95 (ITRF94, 1995.4)
(58 stations over South America)
2) SIRGAS2000 (ITRF2000, 2000.4)
(184 stations over North, Central, and South America)
by means of continuously operating stations:
3) SIRGAS-CON
Obs.: 1995-05-26 ... 1995-06-04;

58 stations in South America;
Processing:
- DGFI (Bernese V. 3.4)
- NIMA (GIPSY/OASIS II)
Discrepancies between the two
solutions: max. 3,5 cm with RMS
1,0 cm in X, 1,4 cm in Y and
0,7 cm in Z;
Causes: different satellite orbits
and non-dependence of the phase
centre variations on the vertical
angle (in the NIMA solution);
Final combined solution:
SIRGAS95: ITRF94, epoch 1995.4.
Second SIRGAS realisation: SIRGAS2000
Re-measuring campaign to
Estimate constant velocities for
SIRGAS95
First approach for a SIRGAS
vertical reference frame;
Obs.: 2000-05-10/19;
184 stations in North, Central and
South America;
Processing:
DGFI (Bernese, v. 4.0)
IBGE (Bernese, v. 4.0)
BEK (GIPSY/OASIS II);
Final solution: combination of the
individual normal equations.
SIRGAS2000: ITRF2000, poca 2000.4;
Precision: 3 6 mm.
From South America to the Americas

The SIRGAS2000 station positions
were transformed to the ITRF94 to be
compared with the SIRGAS95
positions and to determine constant
velocities.
The United Nations Organization,
through its 7th Cartographic
Conference for The Americas (New
York, January 22 27, 2001),
recommend to adopt SIRGAS as
official reference system in all
American countries;
The original acronym of SIRGAS
(Geocentric Reference System for
South America) was changed in 2001
to Geocentric Reference System for
the Americas.
Third SIRGAS realisation: SIRGAS continuously

uously operating network SIRGAS-CON
One core network (SIRGAS-C) as the

primary densification of ITRF in Latin
America;
National reference networks
(SIRGAS-N) improving the
densification of the core network
and providing accessibility to the
reference frame at national and
local levels.
The core network and the national
networks satisfy the same
characteristics and quality; and each
station is processed by three
analysis centres.
Third SIRGAS realisation: SIRGAS continuously

uously operating network SIRGAS-CON
389 stations;
75 common stations with the IGS (to
be included in the next ITRF
solution);
278 stations with GLONASS;
20 stations with GALILEO;
7 stations with BEIDEU;
Weekly processing.
Third SIRGAS realisation: SIRGAS continuously operating network SIRGAS-CON
In the early years of SIRGAS the CON stations were installed mostly by the IGS.
By 2003 the national organizations committed to a policy aimed at
improving the geodetic infrastructure of the Americas and the number of
stations began to grow quickly.
The national reference stations are integrated into the continental
reference frame (SIRGAS-CON) for common processing and to guarantee
consistency with the ITRF.
VII SIRGAS School on Reference Systems, Santo Domingo - Dominican Republic, 16-17 November 2015 7- 10
7.4 SIRGAS National Reference Networks

No. stations
pillars/CON
Densification
SIRGAS/ITRF
269 / 9
1903 / 118
County
SIRGAS2000, 2000.4
125 / 7
SIRGAS2000, 2002.0
178 / 45
Brazil
34 /
SIRGAS95, 1995.4
Chile
ITRF2000, 2005.8
SIRGAS95, 1995.4
135 / 39
ITRF2005, 2006.6
Colombia
SIRGAS95, 1995.4
Argentina
Costa Rica
IGS05, 2007.8
Bolivia
Ecuador
ITRF93,
7/
160 / 15
34 /
13
70 / 47
El Salvador
ITRF2005, 2009,6
1995.0
Guatemala
F. Guiana
ITRF92, 1988.0
0 / 19
Mexico
21
47 /
17 /
17 /
156 /
SIRGAS95, 1995.4
SIRGAS95, 1995.4
ITRF2000, 2000.0
SIRGAS95, 1995.4
Panama
Uruguay
Peru
Venezuela
7.5 SIRGAS contribution to the International GNSS Service (IGS)
SIRGAS
SIRGAS contribution to the International GNSS Service (IGS)

IGS tracking network
IGS polyhedron
IGS Regional Network Associate Anaylsis Centre for SIRGAS

IGS RNAAC SIR
It is under the responsibility of DGFI since June 1996, when it was created;
It has to deliver loosely constrained weekly solutions of the SIRGAS-CON
network to the IGS. These solutions are combined together with those
generated by the other IGS Global and Regional Analysis Centres to form the
IGS polyhedron.
The processing of the SIRGAS-CON network in the frame of the IGS RNAAC
SIR also includes the computation (since 2001) of weekly coordinate
solutions aligned to the ITRF and cumulative (multi-year) position and
velocity solutions for estimating the kinematics of the network.
Until 31 August 2008 (GPS week 1495), DGFI processed the entire SIRGASCON network in one block. Now, it is responsible for processing the SIRGASC core network, for the weekly combination of the individual solutions, and
for the computation of the multi-year solutions.
7.6 Weekly processing of the SIRGAS Reference Frame
INEGI-Mx
SGM-Uy
CPAGS-Ve
IGN-Ar
IGM-Cl
10 processing centres;
2 combination centres;
Each station is included in three
individual solutions;
Unified processing standards
following IGS and IERS conventions.
CEPGE-Ec
IGAC-Co
DGFI-De
IBGE-Br
CNPDG-Cr
Weekly processing of the SIRGAS Reference Frame
Basic observable: ionosphere-free linear combination;

Sampling rate: 30 sec;
Elevation cut-off angle: 3 deg;
Elevation-dependent weighting of observations: 1/cos2z, with z being the
zenith distance;
Satellite orbits, satellite clock offsets, and EOP fixed to the combined IGS
weekly solutions (Dow et al. 2009, www.igs.org/components/prods.html);
Application of antenna phase centre offsets and direction-dependent phase
centre variation values for both transmitting and receiving antennas
according to igs08.atx (Schmid et al. 2007, http://igs.org/igscb/station/
general/pcv_archive/);
Antenna radome calibrations applied if given in the model igs08.atx.
Otherwise, the radome effect is neglected and the standard antenna model
(radome NONE) is used;
Solution of phase ambiguities for L1 and L2 as far as possible (strategy
depends on the analysis software);
The tropospheric zenith delay is modelled using the Vienna Mapping

Function 1 (VMF1, Bhm et al. 2006). The a priori values (~ dry part) from
the climate numerical models of ECMWF (http://ggosatm.hg.tuwien.ac.at/
DELAY/GRID/VMFG/). These a priori values are refined by computing partial
derivatives of the troposphere zenith delay parameters (~ wet part) with 2 h
intervals (using also VMF1) within the network adjustment;
Tidal corrections for solid Earth tide, permanent tide, and solid Earth pole
tide are applied as described in Petit and Luzum (2010).
Ocean tide loading: based on FES2004 (Letellier 2004)
Atmospheric tidal loading (semidiurnal constituents S1/S2) based on
the model of van Dam and Ray (2010).
Non-tidal loadings as atmospheric pressure, ocean bottom pressure, or
surface hydrology are not reduced;
(mean values GPS weeks 1700 1758)
Quality control of the individual solutions
Removing
constraints
Free Normal
Equations (NEQ)
Accumulation of NEQ
- outliers removed
- variance factors included
Solution of individual NEQ

wrt IGS Reference Frame
Residual analysis
- Detection of outliers
- Thresholds: r 10 mm in N or E,
r 20 mm in Up
Validation of the individual

stochastic models for relative
weighting (variance factors)
Loosely constrained solution

(all coordinates constrained to r 1 m)
for the IGS global polyhedron and
for multi-year solutions
Quality control of the weekly coordinates

- formal errors
- time series analysis
- comparison with IGS weekly coordinates
and with IBGE weekly combinations
Weekly solution aligned to IGS Frame

- IGS weekly positions for IGS ref. stations
- Constraints to reference positions
for datum realisation
Loosely
constrained
SINEX files
Combination of the individual solutions in a unified solution for the entire

network
Reliability of the
individual solutions
(comparison with IGS).
Consistency of the weekly

station positions in the
individual solutions.
Formal errors in [mm] of

the individual solutions.
Processing centres compute daily free normal equations that are be

combined for providing a loosely constrained weekly solution for station
positions (all station coordinates are loosely constrained to r1 m).
Stations with large residuals in the weekly combination (more than 20 mm
in the N-E component, and more than 30 mm in the height component)
are reduced from the normal equations.
The individual loosely constrained solutions are made available to be
combined with the corresponding solutions delivered by the other SIRGAS
Processing Centres.
Quality control of the individual solutions

Objective: Determination of variance factors for relative weighting of
the individual solutions to compensate possible differences in the
stochastic models.
Criteria:
Mean standard deviations of coordinates based on minimum
datum conditions (NNR and NNT) with respect to IGS Reference
Frame (formal errors of the individual solutions);
Weekly repeatability of station coordinates for each Processing
Centre (individual precision of the weekly coordinate solutions);
Comparison with the IGS weekly coordinates for common stations
(reliability of the individual solutions).
Quality control of the final weekly station positions

Objective: to ascertain the accuracy and reliability of the weekly
solutions for the entire SIRGAS-CON network (combination of the
core network with the national reference networks).
(mean values GPS weeks 1700 1758)
Quality control of the final weekly station positions
Criteria:
Mean standard deviation for station positions after aligning the
network to the IGS05 reference frame (formal error of the final
combination);
Residual analysis after combining the individual solutions
(internal consistency of the combined network);
Time series analysis for station coordinates (compatibility of the
combined solutions from week to week);
Comparison with the weekly IGS coordinates (consistency with
the IGS global network);
Comparison with the IBGE weekly combination (required
redundancy to generate the final SIRGAS products).
The mean standard deviation of the combined solutions agrees quite well
with those computed for the individual contributions, i.e. the quality of the
individual solutions is maintained and their combination does not deform
or damage the internal accuracy of the entire SIRGAS network.
Internal consistency: ~1,0 mm in N-E and ~3,6 mm in h.
Reliability: ~2,0 mm in N-E and 4,5 mm in h.
The differences between both processing centres are within the expected
level (less than 1,0 mm).
7.7 Multi-year (cumulative) solutions
Objective: to know the kinematics of the reference frames (station

positions in a certain epoch + constant velocities).
These solutions include those models, standards, and strategies widely

applied at the time in which they were computed and cover different
time spans depending on the availability of the weekly solutions:
Comparison of the different SIRGAS-CON multi-year solutions

with the ITRF2008
Multi-year (cumulative) solutions: computation
Multi-year (cumulative) solutions: time series analysis
Identification and reduction of outliers

Identification of jumps
Identification of trend changes.
Setting up discontinuities
Setting up discontinuities
Latest multi-year
ear solution SIR15P01
Recomputed weekly normal

equations applying new
IGS/IERS standards;
Time span: 2010.2 - 2015.2;
303 stations;
Frame: IGb08, 2013.0
Accuracy:
Pos.: N - E = r1,8 mm
h = r3,5 mm
Vel.: N - E = r1,0 mm/a
h = r1,2 mm/a
VEMOS: Velocity model for SIRGAS
Reliable station velocities can

be computed after two years
(at least).
Applications requiring timedependent coordinates and
relying on stations with less
than two years of operation
may use a model representing
the annual mean station
position changes.
7.8 Present challenges: modelling station seasonal movements

Most of the SIRGAS stations present significant seasonal position variations
(mainly in the vertical component). These variations are ignored (as in any
reference frame) when constant velocities (linear position changes) are
computed.
El Maule, Chile, Mexico,

February 27, 2010
Guatemala,
November 7, 2012
Present challenges: modelling reference frame

deformations due to seismic events
Arequipa, Peru, June 23, 2001
Baja California, Mexico,

April 4, 2010
Present challenges: modelling reference frame

deformations due to seismic events
Strong earthquakes produce not only discontinuities in the

station position time series, but also in the usual lineal
movement of the stations;
The caused deformations are not homogeneous along the
reference frame and, therefore, the transformation between the
pre-seismic and the post-seismic coordinates can not be carried
out with the usual network transformations (like similarity or
affine);
Seismic deformations in the SIRGAS region make the existing ITRF
solutions unusable and ITRF updates (re-computations) take too
long;
Geospatial information referring to the pre-seismic coordinates
has to be transformed (recovered) to the post-seismic reference
frame to continue being valid (specially in official maters like
legal borders, cadastre, land management, etc.).
On-going activities to keep undated the SIRGAS reference frame
1) Second reprocessing of the entire SIRGAS reference frame:
New computation of daily normal equations between January

1, 1997 until December 31, 2012
Including the new geodetic standards outlined by the IERS

(International Earth Rotation and Reference Systems Service)
and the IGS (International GNSS Service);
Inclusion of GLONASS measurements;
2) Modelling of seasonal movements at the combination level of the

weekly solutions;
3) Computation of deformation models derived from discrete (weekly)

station positions to incorporate seismic discontinuities in the
computation of the reference frame.
7.9 Availability and applicability of SIRGAS
SIRGAS PRODUCTS
from the SIRGAS reference frame
- weekly station positions
- multi-year solutions (position + velocities)
- ionospheric maps
- troposphere estimates
- station position time series
- deformation/velocity models
-
from supporting SIRGAS activities at national
level
- redundancy in the computation of national
reference frames
- transformation parameters between old
national datums and SIRGAS
- alignment of the cartography projection
parameters to SIRGAS
-
Availability and applicability of SIRGAS

SIRGAS PRODUCTS
from the SIRGAS vertical datum unification (not reliable yet)
- standardised physical heights
- connection parameters
to a global
vertical datum
p
g
-
the SIRGAS products are freely available for everyone but depend
on the participation of the countries in SIRGAS. If national reference
frames are not integrated into SIRGAS, there are not SIRGAS
coordinates for those reference frames. If levelling networks are not
integrated into SIRGAS, there are not connection parameters to the
global vertical datum, and so on
http://www.esa.int/esaLP/LPgoce.html
http://geodesy.hartrao.ac.za/
SIRGAS provides the core

data for any discipline
requiring geospatial
data in the Region.

SIRGAS MEANS GEODESY...
As the science of accurately measure and
understand three fundamental properties of
Earth: its geometric shape, its orientation in
space, and its gravity field; and the changes
of these properties with time (Precise
Geodetic Infrastructure: National
Requirements for a Shared Resource).
The science for measuring changes in the
Earth System.
http://www.igp.ethz.ch/geometh/
http://ggos.gfz-potsdam.de/
SIRGAS MEANS INFRASTRUCTURE...
As the set of human and technical

resources devoted to the long-term
realisation, maintenance and
modernisation of a multipurpose
continental network, which is a
regional densification (realisation) of
the global International Terrestrial
Reference Frame (ITRF). They are
components of the infrastructure,
oriented to the monitoring and study
of different phenomena occurring in
the Earth System with special
influence in the SIRGAS region.
http://www.agu.org/

SIRGAS MEANS EARTH SCIENCE...
As the contribution of geodetic science and
techniques to the family of Earth sciences
by sharing data, providing services and
generating information that combined with
those provided by different sources lead to
a better comprehension
comprrehension of Earth.
Cartography
Visualization
Social
Sciences
GIS
Geodatabases
Business
SIRGAS
SSI
IRGAS
AS
AS
Geodesy
In situ data
collection
Remote sensing
Photogrammetry
Physical
Sciences

SIRGAS MEANS SOCIAL BENEFITS...
As a practical application
focused on solving problems
derived from natural hazards,
global change and the social
evolution itself. It is related to
all the elements, variables
Biological
Sciences
and processes that can be
located by geopositioning.
This covers, by far, the most
of the human activities and
their relation with the
environment.
Adapted from: http://cast.uark.edu/home/research/geomatics.html
2009
1997 - 2009
SIRGAS MEANS CAPACITY BUILDING AND DEVELOPMENT...

One of the most successful international
initiatives. A growing number of people
and institutions contributing in a
voluntary basis for the development of
Geodesy in the Region.
1993 - 1997
SIRGAS DATA ARE

The most basic layer in the spatial
data infrastructures of the
Americas
The basis for spatial data
standardization
The space-time link among data
sets and information
The common language for data
sharing, interoperability and
compatibility
http://www.fgdc.gov/library/whitepapers-reports/
7.10 How to joint SIRGAS?
AS A MEMBER COUNTRY
Addressing a letter to the SIRGAS President making clear the
interest on joining SIRGAS, mentioning what it is expected to find in
SIRGAS, and indicating the names of two national representatives to
integrate the Directing Council.
The national representatives are appointed by the member countries,
through the entities that congregate the scientists and technicians in
Geodesy and related subjects. The election of the national
representatives follows the procedures established in each country.
Each American and Caribbean country is very welcome as a SIRGAS
member.
Being a SIRGAS member does not have any cost, but many
advantages: establishment, adoption and use of a modern reference
frame, direct access to the biggest geodetic forum in the region,
participation in outlining coming activities and making decisions,
exchange of experiences related to geodetic topics, capacity building
possibilities, etc.
How to joint SIRGAS?

The most important component of SIRGAS
are all those people contributing to the
adequate functioning and improvement of
this infrastructure:
Station operators
Data centre responsibles
Members of the analysis centres
Working groups members
Researchers developing modelling
strategies
Entities and decision makers supporting
the SIRGAS initiatives
Users asking for better results and more
comprehensible ways of application
Young people and experts intending to join
SIRGAS for a reciprocal benefit

AS A SIRGAS REFERENCE STATION or AS A SIRGAS NATIONAL

DENSIFICATION
Following the SIRGAS guidelines for the installation and registration
(in the ITRF) described in the document Procedure for becoming a
SIRGAS-CON station (available at www.sirgas.org)
Guaranteeing that the stations are reference stations in your country
and not private stations (therefore, an agreement of the national
representative is required);
Addressing a letter to the President of the SIRGAS Working Group I
describing stations (log files) and availability of the data (FTP access).
Advantages: the stations will be processed by three analysis centres
applying the most modern standards and conventions; full
consistency with the ITRF, availability of weekly coordinates, also
velocities after two years, continuously monitoring of the national
reference frame, etc. All of this for free.
Data policy within SIRGAS

The SIRGAS member countries provide free-of-charge observational
data needed to reach the SIRGAS objectives, e.g. GNSS
measurements, levelling differences, tide gauge registrations, etc.
The SIRGAS components working with those data are not able to
provide them to third parties. If someone asks for any kind of data,
he is oriented to contact the direct responsible of the data.
The geodetic data given to SIRGAS is only applied to realise the
SIRGAS objectives; other application requires the authorisation of
the data owner.
The SIRGAS products, like weekly solutions, ionospheric maps, time
series, etc., are free-of-charge and available for everyone.
Each governmental, academic, scientific, or private institution
joining SIRGAS has to agree with this data policy.

AS A SIRGAS PROCESSING CENTRE
Following the SIRGAS guidelines described in the document
Guidelines for SIRGAS Analysis Centers (available at
www.sirgas.org)
Addressing a letter to the President of the SIRGAS Working Group I
describing software to be used, stations to be processed, members of
the processing centre, etc.
Advantages: support in capacity building for precise GNSS processing,
updated know-how to handle and to analyse geodetic reference
networks, equilibrium between the present precise technology and
capacities of the human resources, etc.

AS A MEMBER OF A SIRGAS WORKING GROUP OR PROJECT
Contacting the chair of the Working Group or Project and
mentioning your intention of contributing with the present activities
of facing new challenges.
To start and keep working. Inactive members are withdrawn from
the working groups after two years.
7.11 How to know SIRGAS?
Specialized courses for the establishment of SIRGAS analysis centres

Instituto Geogrfico Agustn Codazzi, July 2006
Instituto Geogrfico Militar of Ecuador, December 2008 and February 2011
Servicio Geogrfico Militar of Uruguay, March 2009
Instituto Geogrfico Militar of Chile, September 2011
Universidad Nacional, Costa Rica, December 2012
Instituto Geogrfico Militar of Bolivia, May 2013
Visiting the web site www.sirgas.org

Participating in the SIRGAS annual meetings (in symposium format since
2008)
Participating in the capacity building activities
SIRGAS Schools on Reference Systems

Bogot, July 2009, IGAC, 120 participants, 12 countries.
Lima, November 2010, IGN, 122 participants, 13 countries.
Heredia, August 2011, ETCG-UNA, 116 participants, 18 countries
Concepcin, October 2012, UdeC, IGM, 50 participants, 16 countries, dedicated to real
time applications
Panam, October 2013, IGNTG, 140 participants, 26 countries, extended to crustal
deformation and ionospheric monitoring
La Paz, November 2014, IGM, EMI, 34 participants, 15 countries, vertical reference
systems
Curitiba, May 2015, UFPR, 23 participants, 9 countries, adjustment of vertical networks
-

VII SIRGAS School Sto Domingo 2015 P

Enviado por

Dados do documento

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

VII SIRGAS School Sto Domingo 2015 P

Enviado por

Direitos autorais:

Formatos disponíveis

SIRGAS Scientific Council

School on Reference Systems, Crustal

Monday, November 16th, 2015

Adjustment of GPS networks

Tuesday, November 17th, 2015

Vertical reference systems

Crustal deformation observation and

Ionosphere modelling and analysis

SIRGAS: definition, realisation, maintenance

GPS positioning: observation equations and

Terrestrial Reference System and Frame

2. Geodetic reference systems and frames

2.1 Types of coordinates and transformations

We distinguish right-hand and

Cartesian systems have rectilinear

Global Cartesian 3D coordinates

In global systems, the origin is the

Transformation of Cartesian coordinates (1)

A coordinate transformation is the change of coordinates from

1  RZ  RY X ' X ' p 'X

M  RZ 1  RX Y '  Y ' p  'Y

 RY  RX 1 Z '  Z ' p 'Z

X', Y', Z' = initial system

The Helmert transformation

Transformation of Cartesian coordinates (2)

Global spherical coordinates can

The shape of the ellipsoid is given

The figure of the Earth is better

Ellipsoidal coordinates (1)

A global spherical system is

Global spherical and ellipsoidal coordinates

Ellipsoidal coordinates (2)

The ellipsoidal coordinates can be

The ellipsoidal surface system is

A conversion is the change between

atan (Z+ e2 1VLQ  ;2+Y2)0.5

Conversion of Cartesian to ellipsoidal coordinates

Transformation of ellipsoidal coordinates

Local (topocentric) coordinates (1)

Local (topocentric) coordinates (2)

 sin M 0 cos O0 cos M 0 cos O0

Conversion of topocentric to geocentric coordinates

X = XO E sin OO N sin O cos OO + U cos O cos OO

The metric distortions

Special types of projections and distortions

The Gauss-Krger system is based on

Gauss-Krger plane conform coordinate system

Universal Transversal Mercator (UTM) System

The UTM system is similar to Gauss-Krger.

The meridians are straight

Conical conform Lambert system

The distortion is constant along

The meridians are straight

Polar stereographic system

The distortion is constant along

2.2 Definition of reference systems and frames

Reference System: Definition of standards, parameters, models, etc.

Reference Frame: Realisation (materialisation) of a reference

Hierarchy of reference systems

Examples of global reference systems

Terrestrial reference systems (TRS)

World Geodetic System 1984 (WGS84): Global terrestrial

International Terrestrial Reference System (ITRS): Reference

Examples of reference frames

1 RZ RY X ' X ' p 'X

M RZ 1 RX Y ' Y ' p 'Y

RY RX 1 Z ' Z ' p 'Z

atan (Z+ e2 1VLQ;2+Y2)0.5

sin M 0 cos O0 cos M 0 cos O0