RESULTS FROM TERRASAR-X
GEOMETRIC AND RADIOMETRIC CALIBRATION
B. Bräutigam, M. Schwerdt, M. Bachmann, B. Döring
German Aerospace Center (DLR), Microwaves and Radar Institute, Oberpfaffenhofen, D-82234 Wessling, Germany
Keywords: TerraSAR-X, Calibration, Antenna Model.
As TerraSAR-X, due for launch in June 2007, will be an
operational scientific mission with commercial potential,
product quality is of crucial importance. The success or
failure of the mission essentially depends on the calibration of
the TerraSAR-X system ensuring the product quality and the
correct in-orbit operation of the entire SAR system.
This paper describes the calibration procedures for
TerraSAR-X and the dedicated activities to be performed
during the five months commissioning phase. Results from
on-ground tests are discussed with respect to geometric and
radiometric calibration of the TerraSAR-X system.
The first German SAR satellite for commercial and scientific
applications, TerraSAR-X, will launch in June 2007.
TerraSAR-X is a flexible X-Band SAR built by EADS
Astrium and the German Aerospace Center (DLR) in a
public-private-partnership . The SAR instrument is
designed to cover a wide range of swath positions and to
operate in different operation modes by switching the
instrument over a multitude of antenna beams. For the various
antenna beams, an active phased array antenna electronically
forms antenna radiation patterns. The array consists of 384
radiating sub-arrays for horizontal and vertical polarisation
arranged in a matrix of 12 panels with 32 rows. Active
transmit/receive (T/R) modules individually adjust the array
elements in gain and phase for shaping and steering the
antenna pattern in azimuth and elevation direction.
The essential task of calibrating TerraSAR-X is to estimate
and correct systematic error contributions throughout the
complete SAR system and to tie-down image information
(magnitude and phase) to reference units in geophysical
terms. The quality of this calibration process depends on the
inherent stability of the radar system and the capability to
determine and monitor the radiometric and geometric
Due to the high degree of flexibility of TerraSAR-X
(StripMap, Spotlight, ScanSAR, right/left looking, etc.) and a
tight performance with an absolute radiometric accuracy
better than 1dB, it becomes clear that a conventional
calibration approach is not feasible, i.e. the real measurement
of all antenna beams in all operation modes, as performed for
the ASAR instrument of ENVISAT across the rainforest .
Hence, a new, more efficient, and affordable calibration
concept has been developed [3,4]. The key element of this
calibration concept is a novel antenna model approach [4,6].
The TerraSAR-X antenna model is utilised for generating all
reference antenna patterns and for beam optimisation.
To account for the restricted time of calibration campaigns
performed during commissioning phase of TerraSAR-X, the
number of passes and places of test sites is optimised versus
cost and time effort by calibrating only dedicated beams but
with the same test site.
The paper describes the in-orbit calibration procedure and the
different activities performed before and after launch of
TerraSAR-X. It shows the calibration status until satellite
launch and gives an overview of the planned activities during
the five months commissioning phase.
2 Scope of TerraSAR-X Calibration
Applying the novel antenna model approach for SAR system
calibration, the objective of in-orbit calibration can be sub-
divided into three major tasks:
Geometric Calibration: to assign the SAR system
to the geographic location on the earth surface.
Relative Radiometric Calibration: for radiometric
correction of SAR data within an illuminated scene.
measurement of the SAR system against standard
targets with well known geophysical characteristics.
The calibration steps rely on the antenna model which has
been validated on-ground. In addition to the tasks described
above, the verification of this antenna model has to be
executed in-orbit. This ensures the provision of the antenna
patterns of all operation modes and the gain offset between
The main goal after launch of the TerraSAR-X satellite is to
provide calibrated and verified SAR data products at the latest
by the end of the commissioning phase. Thus, a strategy for
an efficient but robust calibration approach has been
The successive baseline calibration processes are:
1. Geometric Calibration
2. Antenna Pointing Determination
3. Antenna Model Verification
4. Relative Radiometric Calibration
5. Absolute Radiometric Calibration
Characterisation Method [7,8], the actual antenna settings will
be derived in orbit. Individual and simultaneous monitoring
of all T/R modules is possible in-flight under the most
realistic instrument conditions. This saves valuable time
during commissioning and in case of antenna module failure
In the following the process steps of the in-orbit calibration
are described in detail.
by applying the novel T/R Module
3.1 Geometric Calibration
The purpose of the geometric calibration is the geometric
assignment of the SAR system to the earth's surface. Two
effects can influence the correct localisation of the product:
systematic azimuth shifts resulting in wrong data
take start time and
dislocating products in range.
electronic delay of the instrument
For both effects, SAR measurements over corner reflectors
are analysed as passive targets have no additional electronic
delay and consequently no additional source of errors.
3.2 Antenna Pointing Determination
An important task is the determination of beam pointing
errors coming from mechanical and electrical antenna
mispointing as well as from attitude control offsets. These
errors are measured in elevation and in azimuth using
appropriate patterns over rain forest and by ground receivers.
But also a Doppler analysis is performed to evaluate a squint
in flight direction.
3.3 Antenna Model Verification
The total size of the SAR antenna is 4.8 m in length and 0.8 m
in height. Far-field pattern measurement of the complete
antenna was not possible on-ground due to restricted space in
the measurement chamber. Real pattern measurements can
only be performed in orbit.
The characterisation of the antenna is based on a precise
antenna pattern model, which had to be validated against
precisely measured near-field patterns on ground before
launch and will be verified in orbit by in-flight measurements
over homogeneous targets (rain forest) and deployed ground
The antenna model provides a software tool that accurately
determines all beam patterns based on detailed
characterisation of the radar antenna and knowledge of the
antenna control parameters.
It was developed before launch considering different input
geometry of the antenna,
beam excitation coefficients of all 384 T/R modules,
a drift and failure matrix and
embedded radiation patterns measured on ground
from single sub-arrays of the antenna.
The strategy of on-ground model validation was based on the
comparison between simulations and measurements.
For this purpose, the antenna patterns of the sub-arrays were
measured individually while already mounted within the
panel. Mutual coupling is covered by the resulting embedded
sub-array patterns. Based on these embedded patterns, the
radiation pattern of the panel was calculated and compared to
a measurement of the panel. In case of TerraSAR-X the
model was validated on panel level with 32 sub-arrays and on
leaf level with four panels (= 128 sub-arrays).
Figure 1: A) comparison between an exemplary beam measured on
ground (blue/solid) and derived by the antenna model (red/dotted),
B) difference within the main beam between model and
One example of the validation on-ground performed by
Astrium is shown in Figure 1. The solid blue curve is the
measured elevation pattern of one panel and the red curve of
that derived by the model. The deviation of the shape within
the main lobe is less than 0.15 dB. The results of the
validation of the antenna model are described in .
The in-orbit verification will be performed by three selected
beams with low, medium and high incidence angles. For this,
the elevation patterns are measured over rain forest and
compared to the pre-calculated patterns. Furthermore, the
transmitted pulses are recorded on-ground by deployed
ground receivers. With this method the transmit pattern of the
TerraSAR-X antenna can be measured in-flight verifying the
one-way azimuth patterns. One example of really measured
antenna patterns of the ASAR/ENVISAT instrument is shown
in Figure 2. The reduced amplitudes indicate the switching of
the instrument from beam to beam and the envelopes
represent the one-way azimuth pattern of the corresponding
Figure 2: One-way (transmit) azimuth antenna patterns of the
ASAR/ENVISAT instrument in ScanSAR operation measured by a
3.4 Relative and Absolute Radiometric Calibration
After in-orbit verification of the antenna model the beam
patterns required for radiometric correction of SAR data
within an illuminated scene are derived by the antenna model
for all operation modes and all incidence angles. There are
more than ten thousand possible antenna patterns und hence,
the same amount of absolute calibration factors.
The relative gain variation from beam to beam will be
characterised over homogeneous targets in ScanSAR
operation. Thus, absolute radiometric calibration is sufficient
for a few beams instead of measuring TerraSAR-X against
deployed calibration targets for all operation modes. The
effort for relative and absolute radiometric calibration as well
as the duration of the calibration campaigns can be
significantly reduced to achieve a five months commissioning
3.5 Antenna Performance Monitoring
An important input to the model are the actual beam
coefficients of the active phased array which can be
monitored in-orbit using
Characterisation Method . Beyond these capabilities the
antenna model also features a tool for generating optimised
beam coefficients under given constraints. One major
advantage of this beam optimisation is that the reference
the novel T/R Module
setting of the antenna excitation is already calculated before
launch for the best instrument performance. After launch this
optimisation method guarantees dynamic re-calibration in the
event of active antenna module degradations during the
mission. Figure 3 shows results from on-ground tests with the
whole TerraSAR-X antenna for monitoring performance
degradation of individual antenna modules.
Half an antenna panel (16 modules of panel number 10) was
switched off whereas the whole antenna was excited with a
gain taper in elevation (row) direction. The commanded
receive gain taper in elevation applied to all panels is drawn
as a solid line in Figure 3.
The T/R Module Characterisation Method is used to
characterise the individual antenna beam coefficients while
the instrument is operated under nominal conditions. The
actual coefficients of antenna panels are compared to the
commanded ones. The characterised gain of panel number 0
is plotted as a dashed line. It matches the commanded beam
taper very well. Accordingly, the gain characteristics (dotted
line) of antenna panel number 10 correspond to the expected
values, too. The first 16 modules of antenna panel number 10
yield only noise signal as these modules were switched off.
With the help of this method, a fast and dynamic re-
calculation of antenna beams is possible by feeding these
inputs into the antenna model instead of measuring all
antenna patterns again.
Figure 3: Commanded and characterised beam taper coefficients on
antenna panel number 0 and number 10 of TerraSAR-X active
antenna. The first 16 modules of panel 10 were switched off.
4 Calibration Campaigns
To ensure a successful execution of the different calibration
procedures described above, dedicated calibration campaigns
must be performed. This so called external calibration  is
based on SAR data acquisition over test sites with well-
known backscatter characteristics. Basically, these sites can
be homogenous areas like the Amazonian rain forest or sites
with deployed point targets, like corner reflectors, active
transponders or ground receivers. The most important driver
to plan and coordinate these activities is the global coverage
of TerraSAR-X, as the coverage defines the number of
feasible measurements and consequently drives the schedule.
Figure 4: Coverage of TerraSAR-X StripMap beam strip011R over a
complete repeat cycle. Red areas: no coverage, light blue areas:
covered once, dark blue areas: covered twice.
4.1 Global Coverage of TerraSAR-X
The coverage on the earth’s surface needs to be evaluated
with respect to the required number of point target and rain
forest measurements assumed in the radiometric accuracy
budget. As single beams do not cover the complete globe, the
selection of test sites is constrained to the availability of
respective beams on ground. Figure 4 shows the coverage of
the full performance beam 011 over Germany in right looking
mode for a repeat cycle of eleven days. In order to obtain as
many overflights as possible over calibration targets, special
test site configurations have been selected in the crossing
points of ascending and descending passes.
4.2 Test Site Configuration
Driven by the radiometric accuracy budget within a StripMap
swath of 30 km range, two targets each are deployed in near
range, at swath centre, and in far range, see Figure 5.
The targets must be sufficiently separated to avoid
ambiguities in the image. With a slight displacement, most of
the positions (red crosses) can be applied twice for
descending and ascending orbit. Thus, the number of required
target positions within a test site could be extremely reduced.
4.3 DLR Calibration Field
Mainly for logistic reasons test sites are selected near by DLR
Oberpfaffenhofen in South Germany as calibration targets
have to be re-adjusted for each overflight. According to the
previously described test site configuration a total of 30 target
positions have been established in an area of 120 km x 40 km
covering several crossing orbits of TerraSAR-X antenna
Figure 5: Test site configuration for TerraSAR-X. The test site
consists of 7 target positions and encloses an area of 32 km x 32 km.
Implementation of the calibration concept requires a
calibration facility that is well-equipped with ground
calibration hardware as well as software tools for evaluating
the measurements. For external calibration, three different
types of calibration targets are used:
Trihedral and dihedral corner reflectors as passive
targets precisely surveyed (with differential GPS)
and therefore well suited for geometric calibration.
Transponders with high radar cross section
providing accurately defined point targets within the
Ground receivers measuring the one-way azimuth
pattern of the SAR antenna during an overflight.
Different evaluation and analysis software tools have been
tested during pre-launch preparation. The antenna model
provides the validated antenna patterns and the on-board
beam coefficients to be commanded in the SAR antenna. The
T/R Module characterisation software demonstrated its
valuable support of antenna performance monitoring already
during on-ground instrument tests. For point target analysis
the CALIX software features measurements of impulse
response function parameters, absolute calibration factors as
well as geometric analysis.
The existing infrastructure is well prepared for executing
calibration campaigns over large test sites. Accordingly,
reliable and accurate ground equipment as well as evaluation
software is guaranteed during the whole mission life time.
It should be mentioned that additional test sites are deployed
in Northern Germany, Switzerland, and Spain.
In preparation of the TerraSAR-X satellite launch in June
2007 DLR has thoroughly planned the calibration procedures
to be performed during the five months commissioning phase.
By applying the antenna model as key element of the in-orbit
calibration strategy, most of the antenna characterisation
effort has been shifted from in-flight to on-ground activities.
An efficient but accurate way to accomplish a short
commissioning phase has been developed as only a few
beams are measured in orbit.
A multitude of powerful software tools has been established
within on-ground tests to support the evaluation of calibration
measurements. For the first time ever, the innovative T/R
Module characterisation method will be applied in a satellite
environment during the TerraSAR-X mission. Evaluation of
data acquired during the calibration campaigns yields
different calibration and processing parameters like antenna
patterns or absolute calibration
corresponding test site configurations have been developed
and arranged in South Germany, supported by precise and
sophisticated ground equipment
Oberpfaffenhofen calibration facility.
After satellite launch calibration activities consume most of
the time during the five months commissioning phase. The
first measurements begin three weeks after launch. Over 400
acquisitions across the Amazonian rain forest are planned.
The various calibration measurements in Germany require
ground support by professional calibration teams performing
more than 60 campaigns in a time frame of less than five
months. This amount of measurements is sufficient for
calibrating over 10,000 antenna beams.
The results from the in-orbit calibration are a pre-requisite for
successful SAR product verification as they define the
geometric and radiometric performance of TerraSAR-X.
from the DLR
 R. Werninghaus, “The TerraSAR-X Mission,” in 6th
European Conference on Synthetic Aperture Radar,
Dresden, Germany, 2006.
 M. Schwerdt, B. Bräutigam, M. Bachmann, T.
Molkenthin, D. Hounam, and M. Zink, “The Calibration
of TerraSAR-X,” in 6th European Conference on
Synthetic Aperture Radar, Dresden, Germany, 2006.
 M. Schwerdt and D. Hounam, "Quality Control and
Calibration of Future SAR Systems,” in ODAS 2001,
3rd ONERA-DLR Aerospace Symposium, Paris, France,
 M. Schwerdt, D. Hounam, B. Bräutigam, and J. L.
Alvarez-Pérez, “TerraSAR-X: Calibration Concept of a
Multiple Mode High Resolution SAR,” in IEEE: 25th
Symposium, Seoul, Korea, 2005.
 M. Zink and B. Rosich, “Antenna Elevation Pattern
Estimation from Rain Forest Acquisitions,” in
ENVISAT/ASAR Calibration Review (ECR) of ESTEC,
European Space Agency (ESA), Noordwijk, Netherland,
 M. Bachmann, M. Schwerdt, B. Bräutigam, B.
Grafmüller, A. Herschlein, and J. L. Alvarez-Pérez,
“The TerraSAR-X Antenna Model Approach,” in 2nd
International ITG Conference on Antennas, Munich,
 D. Hounam, M. Schwerdt, and M. Zink, “Active
Antenna Module Characterisation by Pseudo-Noise
Gating,” in 25th ESA Antenna Workshop on Satellite
Antenna Technology, Noordwijk, Netherlands, 2002.
 B. Bräutigam, M. Schwerdt, M. Bachmann, and M.
Stangl, “Individual T/R Module Characterisation of the
TerraSAR-X Active Phased Array Antenna by
Calibration Pulse Sequences with Orthogonal Codes,” in
26th International Geoscience And Remote Sensing
Symposium, Barcelona, Spain, 2007.
 J. L. Alvarez-Pérez, M. Schwerdt, D. Hounam, R.
Torres, C. Buck, and M. Zink, “Antenna Pattern
Optimization for the ENVISAT ASAR Antenna with
Failed Elements,” in 25th ESA Antenna Workshop on
Satellite Antenna Technology, Noordwijk, Netherlands,
 B. Bräutigam, M. Schwerdt, and M. Bachmann, “The
External Calibration of TerraSAR-X, a Multiple Mode
SAR-System,” in 6th European Conference on Synthetic
Aperture Radar, Dresden, Germany, 2006.
And Remote Sensing