MIRAS In-Orbit Calibration

Abstract

MIRAS has two types of in-orbit calibration: external, in which the instrument makes a maneuver to point to the cold sky, and internal, in which noise is injected to the receivers. This one in turn has two modes, one of short duration carried out periodically interspersed with scene measurements, and other lasting a full orbit and used to obtain the more stable parameters and the sensitivity of all parameters with temperature. These procedures have already been implemented and have been tested at EADS-CASA in the fully integrated instrument. The results show that the in-orbit internal calibration accurately retrieves the needed parameters and will allow to obtain high quality data.

Full text

Page 1

MIRAS In-Orbit Calibration
Ignasi Corbella, Francesc Torres, Nuria Duffo, Adriano Camps, Merc` e Vall-llossera and Ver´ onica Gonz´ alez
Department of Signal Theory and Communications. Universitat Polit` ecnica de Catalunya (UPC)
C/ Jordi Girona 1-3. 08034 Barcelona, Spain. (corbella@tsc.upc.edu)
of designing the testing software and analyzing the resultant data. The testing campaign
will take place in the anechoic chamber of ESTEC during the first quarter of 2007, so
that at the time of the conference it is expected to have already some data available and
processed. A general overview of the MIRAS calibration and results of the tests
campaigns will be presented at the conference.
Abstract—MIRAS has two types of in-orbit calibration: Exter-
nal, in which the instrument makes a maneuver to point to the
cold sky, and Internal, in which noise is injected to the receivers.
This one in turn has two modes, one of short duration carried
out periodically interspersed with scene measurements, and other
lasting a full orbit and used to obtain the more stable parameters
and the sensitivity of all parameters with temperature. These
procedures have already been implemented and have been tested
at EADS-CASA in the fully integrated instrument. The results
show that the in-orbit internal calibration accurately retrieves the
needed parameters and will allow to obtain high quality data.
I. INTRODUCTION
MIRAS (Microwave Interferometric Radiometer by Aper-
ture Synthesis) is the single payload of the second Earth
Explorer mission of the European Space Agency (ESA) Living
Planet Programme: SMOS (Soil Moisture and Ocean Salinity),
designed to observe soil moisture over the Earth’s landmasses
and salinity over the oceans [1]; and scheduled for launch in
early 2008. MIRAS operates at 1.4 GHz (L-band) and it in-
cludes 72 small antennas evenly distributed on the arms of a Y-
shape mechanical structure [2]. Its operation is interferometric
in 2D so that it gets the brightness temperature by mathemat-
ical inversion of the visibility function [3], [4]. The visibility
corresponding to zero spacing is measured independently and
redundantly by three dedicated Noise Injection Radiometers
(NIR) located near the center of the array. The instrument
provides a spatial resolution similar to that of a real aperture
radiometer of similar overall dimensions [5] and gets maps of
brightness temperature at said resolution without the need of
any mechanical scanning.
The procedures designed to obtain the visibility function
out of the raw measurements (correlation counts, detector
voltages, NIR dicke cicle fractions, physical temperatures,...)
are referred to as in-orbit calibration. Table I summarizes
the different modes and types of calibration, along with their
objectives. The proposed calibration time line, including both
internal and external modes, is sketched in Figure 1. In the
present approach, four short internal calibrations per five
orbits, as well as one long internal calibration per month
are scheduled. Frequency of calibration depends on drift and
thermal behavior and therefore will be updated during the
commissioning phase. Calibration time line is based on the
assumption that the parameters having a larger dependency
on temperature are calibrated more often. Other phenomena
causing parameter drift not related to temperature are consid-
ered to be either negligible or with time constants much larger
than those related to temperature drift.
TABLE I
IN-ORBIT CALIBRATION HEADLINES
Table 6: Cold Sky PMS sequence

Time Allocated for PLM Calibration: the total time taken up by the calibration events is the following
(measured per 4-week period):


CalibrationEpochs
Short
9032
Long
4996
External 5668
Total
19696

Table 7: Allocation of calibration time per 4-week period

Thus, the total is 23635 s, which represents 0.98% of the total time, barely within the 1% requirement.

IN-ORBIT CALIBRATION HEADLINES
Description and objectives
• NIR absolute calibration:
• Deep sky imaging: FLAT TARGET RESPONSE
• One-point calibration for CAS validation/update
• Parametric G-matrix.
• U-noise injection (Visibility offset)
• C-noise injection (FWF shape)
• C-noise injection (same as short calibration)
• C-noise injection:
o PMS calibration (gain and offset) and monitoring (attenuator)
o FWF at the origin (phase and amplitude)
o LICEF receiver temperature monitoring
• U-noise injection (Receivers noise temperature)
• Calculation of normalized complex correlations:
Cal Modes
Sky view
Type
External
Long
Calibration
Internal
Short
Calibration
Internal
Selfcalibration



Internal


Figure 1 Sketch of SMOS in-orbit calibration timeline
Figure 2: Overall SMOS PLM calibration scheme
Fig. 1.Overall SMOS In-Orbit calibration time line


SMOS PLM
No.:
Página/Page: 10
Edición/Issue: 1.0
Fecha/Date:
SO-TN-CASA-PLM-1075
30.11.06



Mod. M-11-E

Registro Mercantil de Madrid - Tomo 18602 - Folio 105 – Sección 8 - Hoja nº M-323546 – Inscripción 1ª fecha 24/3/2003 - Núm. Identificación Fiscal B-83595280
NS
EVEN
NS
ODD
PMS Corr FWF(0)
(V) Del
NIR
Dicke PMS
NIR
CMDValidMkj
5th --OFFL0-04ULICEF-LUOFF
----
9thL0-0--
1st
498 DualLICEF-LA OFFOFF
-0OFFL0 310ULICEF-LUOFF
-9thL1 -0-Dual LICEF-LA2OFF OFF
-5thL1 -0-ULICEF-LUOFFOFF
PMS
Cold
Sky
55
NIR
154
298
Att
LICEF
PMS
DATA
CORRUP
Step
Epochs
LICEF
Cold Sky PMS Validation SEQUENCE

Seconds
10838
5995
6802
23635
%
0.45%
0.25%
0.28%
0.98%


Overall Calibration Scheme: the following diagram shows the overall calibration scheme.

External CalExternal CalExternal CalLong CalLong CalLong Cal
TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4 TMOON/4
1 orbit = 6000 s = 5000 epochs1 orbit = 6000 s = 5000 epochs 1 orbit = 6000 s = 5000 epochs
Short CalShort CalShort CalShort Cal measured
positions positionspositions
Short Cal measured Short Cal measured





II. INTERNAL CALIBRATION
Internal calibration is based on periodically injecting noise
to all receivers via dedicated input switches [6]. Due to the
difficulty of making a large noise distribution network, a
distributed noise approach is used for large baselines [7].
Internal calibration also includes a so-called selfcalibration
mode that is used for automatically correcting samplers offset
and quadrature error [6].
Different calibration parameters are retrieved by means
of dedicated operation modes: The visibility complex gain
(also called Fringe Washing Function -FWF- at the origin)
is measured by injecting correlated noise at the receivers
input using a two-level differential approach to ensure that the
result is independent of the exact value of the noise source
power and of the distribution network physical temperature.
The correlation products are de-normalized by multiplication
by the system temperature, measured by dedicated power
measurement systems (PMS) available at each receiver. The
same correlated noise used for the visibility gain retrieval is
used also to calibrate these PMS in terms of gain and offset
using a two-level, four point approach [8]. The three NIR’s are
1-4244-1212-9/07/$25.00 ©2007 IEEE.3622

Page 2

TABLE II
SHORT CALIBRATION SEQUENCE
NS
EVEN
NS
ODD
Corr
Del
OFF
Att
NIR-R
LICEF-LC HOTOFFL0
Step
LICEF
EpochsNIR
4C
3C1
5C
0
0
0
WARMOFFL0
1C
OFFL1
NIR-RWARMOFFL1
2CNIR-RHOT
0
NIR-RHOTOFFL00
1
1
1
1
6CLICEF-LC01WARMOFFL0
7ULICEF-LU1
0L0HOT
OFFL00
OFFLICEF-LCC1 8
9
10
11C
1
1
1
SHORT_FLIGHT CALIBRATION SEQUENCE
LICEF-LC
0
0
LICEF-LC2
L0WARMOFF
HOTL1
OFF
PMS,
FWF(0) &
Trec
0LICEF-LC2WARML1
OFF
C
C
M η
M η
M η
M η
M
M
M
kM
M
kM
M
M η v M η v M η v M η v M v v M v v M
v M v v M v v M U v U v kj M v v M v v kj M v v M v v M kj M M kj M v v v v M M
U v U v kj M v v M v v kj M v v M v v M kj M M kj M v v v v M M M η
v M η
v M η
v M η
used during calibration as reference receivers, in order to get
absolute reference noise temperature for PMS gain retrievals
[9]. Uncorrelated noise is injected by switching all receivers
to internal matched loads at known physical temperature. This
provides the visibility offset due to cross-coupling and is
also used as a second reference for getting the receiver noise
temperature of receivers [8].
A. Calibration sequences
According to table I, internal calibration is classified in
terms of short and long inter-calibration period. The Short
Calibration is launched once every 1.25 orbits and the fol-
lowing calibration parameters are retrieved: PMS gain and
offset, fringe washing function at the origin and receiver noise
temperature. They are calibrated at changing points along the
orbit in order to obtain data at different parts of the orbital
period. The duration of each short calibration segment is
27 epochs, where one epoch is the duration of the snapshot
integration time: 1.2s. Short calibration segments are normally
interspersed with science measurements. The average time
spent on short calibration is about 10838 s every four weeks
and the sequence of the steps is shown in table II.
The Long Calibration is made of a short calibration se-
quence followed by additional steps. Thus, the same param-
eters are retrieved and also the visibility offset from uncor-
related noise injection and the shape of the fringe washing
function, making use of the three-delay technique described
in [10], [11]. As shown in figure 2, a full orbit (6000 s) is
dedicated to the long calibration procedure. It is subdivided
into 15 segments of 400 s each. The long calibration orbit
is performed once every four weeks and is used also to
get the sensitivity of calibration parameters with respect to
temperature. The sequence of the steps for each individual
segment and for the full sequence is shown in table III.

Mod. M-11-E

Registro Mercantil de Madrid - Tomo 18602 - Folio 105 – Sección 8 - Hoja nº M-323546 –

They are calibrated during a full orbit (6000 s), which is subdivided into 15 segments of 333 epochs
each (400 s). In each segment, a fixed sequence of PLM modes is activated, which allows the
retrieval of all the parameters. The long calibration orbit is performed once every four weeks.

The sequence of the steps is the following:

Long Calibration = 1 orbit ≈ ≈ 6000 s = 5000 epochsLong Calibration = 1 orbit ≈ ≈ 6000 s = 5000 epochs
U-noise
(51 epochs)(51 epochs)
FWF shape
(68 epochs)(68 epochs)
PMS, FWF(0) & Trec
(26 epochs)(26 epochs)
1 Segment = 333 epochs = 400 s1 Segment = 333 epochs = 400 s
Sub-segment
(94 epochs)(94 epochs)
Sub-segment
(94 epochs)(94 epochs)
Sub-segment
(94 epochs)(94 epochs)
Sub-segment = 94 epochsSub-segment = 94 epochs
U-noise
FWF shapePMS, FWF(0) & Trec
Sub-segment Sub-segment Sub-segment


Figure 1. Long Calibration

Long calibration sequence timelineFig. 2.
TABLE III
LONG CALIBRATION SUB-SEGMENT AND FULL SEQUENCE
NS
EVEN
NS
ODD
Corr
Del
FWF
Shape
4
5
5
5
5
4
4
4
5
+T5
+T
+T
+T
5
OFFL0
22CLICEF-LCHOTOFFL0
21CLICEF-LC WARM
HOTL0
-T
20CLICEF-LCOFFWARML05
19CLICEF-LCOFF
OFFL0 -T
18CLICEF-LCHOTOFFL0
17CLICEF-LC WARM
HOTL0 -T
16CLICEF-LCOFFWARML0 -T
15CLICEF-LCOFF
HOTL00
11CLICEF-LCOFFWARML00
14CLICEF-LCOFF
OFFL00
12CLICEF-LC WARMOFFL00
13CLICEF-LCHOT
Epochs
SHORT CALIBRATION SEGMENT
PMS,
FWF(0)
& Trec
LONG_FLIGHT CALIBRATION SUB-SEGMENT SEQUENCE
Step
LICEF
NIRAtt
M η
M η v M η v M η v M v v M v v M
M
U v U v kj M v v M v v kj M v v M v v
M
M
kM
kM
M
M
M
kM
M
M
M
M U v U v M U v U v M U v U v
NS
EVEN
NS
ODD
Corr
Del
LONG_FLIGHT CALIBRATION SUB-SEGMENT
L00ULICEF-LUOFFOFF
267
450
LONG_FLIGHT CALIBRATION SUB-SEGMENT
Step
LICEF
LONG_FLIGHT CALIBRATION SUB-SEGMENT167
LONG SEGMENT
337
LONG_FLIGHT CALIBRATION SEGMENT SEQUENCE
EpochsNIRAtt
M η v M η v M η v M η v M v v M v v M
U v U v kj M v v M v v kj M v v M v v M v v M v v kj M v v M v v kj M v v M v v M v v M v v kj M v v M v v M v v M v v
M U v U vM U v U v
M
III. EXTERNAL CALIBRATION
Part of an orbit is devoted to observe the sky to complete
the in-orbit calibration, in particular those parameters that
cannot be properly calibrated by internal calibration. They are
the following: NIR units, Deviations in the antenna patterns
and Parametric G-matrix. In one month there will be two
identical external calibration events every two weeks (see Fig.
1). Calibration is carried out by observing an external target
selected in the sky in inertial attitude for part of the orbit. The
inertial attitude is attained by an in-plane tilt maneuver, and
1-4244-1212-9/07/$25.00 ©2007 IEEE.3623

Page 3

TABLE IV
EXTERNAL CALIBRATION
Atttitude
Nominal
Mode
Epochs Minutes
1
LICEF
Dual
NIR
NIR-ANominal Measurement
0,02
First Slew
Ext-Measurement
U-Noise
534
50
26
26
26
29
27
10,7
1,0
0,5
0,5
0,5
0,6
0,5
Dual
U
Dual
Dual
Dual
Various
Various
NIR-A
LICEF-LU
NIR-AR
NIR-A
LICEF-LA
Various
Various
Cold Sky PMS
IC4EC
Ext-Measurement (FTR /
G-matrix)
IC4EC
Cold Sky PMS
74014,8DualNIR-A
27
29
26
26
26
50
30
1162
0,5
0,6
0,5
0,5
0,5
1,0
0,6
23,2
Various
Various
Dual
Dual
Dual
U
Dual
Dual
Various
Various
NIR-AR
NIR-A
LICEF-LA
LICEF-LU
NIR-A
NIR-A
U-Noise
Corr Test
Ext-Measurement
Return
Slew
Time
Inertial
(ETO)
NIR Calibration
NIR Calibration
the return to nominal attitude is achieved by performing a turn
in the same sense as the first one (that is, performing a back
flip). The duration of external target observation (ETO) is of
22 minutes, with special events of internal calibration placed
within ETO. The target and timing are selected to minimize
perturbations by having the Sun and Moon in the back lobes
of antennas and avoiding equatorial targets. The total monthly
duration of the external calibration events is 6802 seconds,
including maneuver and settling time. A dedicated internal
calibration event is devoted to process the external calibration
measurements. It is called Internal Calibration for External
Calibration (IC4EC). In addition, a sequence is performed to
measure the PMS by looking at an external scene. It is called
Cold Sky PMS calibration. This allows to validate the PMS
internal calibration, and the stability of the noise distribution
network. The sequence of the steps for external calibration is
shown in table IV
IV. MIRAS INTEGRATION AND TESTING
The whole instrument has been successfully assembled at
EADS-CASA premises. Figure 3 shows a picture just after
placing all the components and before being covered by the
protecting layer. Once integrated, it has been subjected to
an extensive testing process aimed at checking the hardware
and software operation, as well as validating the calibration
procedures and algorithms.
A. Calibration sequences evaluation
The first tests were carried out just after integration in the
clean room of EADS-CASA. During these tests the payload
module was set to acquire several calibration sequences pro-
ducing the corresponding raw data files. A software package
developed by UPC was used to read these files and compute
the calibration parameters according to the programmed se-
quences and the procedures defined in the previous sections.
The following tests were performed
1) Short-flight + short extended 100 times cyclically
2) Short-extended 100 times
3) Short, short-extended, long and external one time
Fig. 3. MIRAS fully assambled at CASA-EADS premises
Fig. 4.Power measurement system gain and offset comparison
The results of these tests were highly successful and the
instrument showed a very consistent behaviour with respect
to the on-ground characterization of the components. Figure
4 shows the gain and offset of all the PMS included in the
receivers. Two values are given for each, one corresponding
to the on-ground characterization carried out during receivers
manufacturing, and the other to the values retrieved using the
in-orbit internal calibration sequences. A high consistency is
seen between both data sets both for gain and offset.
Figure 5 shows the fringe washing function at the origin
computed using the in-orbit internal calibration sequences and
1-4244-1212-9/07/$25.00 ©2007 IEEE.3624

Page 4

Fig. 5.
from fringe washing function estimation
Fringe washing function at origin (Gkj) and bandwidth computed
equation (15) of [8]. Also shown in the figure is the receivers
bandwidth retrieved from the three-delay fringe washing func-
tion estimation [10]. The result is consistent with the nominal
bandwidth of 19 MHz in all receivers.
B. Image validation test
After these tests, the payload module was sent to the
European Space Research & Technology Centre (ESTEC) to
being further evaluated. An image validation test is scheduled
to take place in the anechoic chamber of ESTEC during the
first weeks of May 2007 including the following specific tests:
• Instrument check
• Polarization check
• Recovery of relative phase
• Fringe Washing function test
• Stability Test
The first two of them are devoted to check the overall instru-
ment operation, but the most critical one is the recovery of rel-
ative phase. This is devoted to computing the phase difference
between all pairs of antennas [12], which is a parameter that
cannot be measured during standard antenna characterization
and is used in the visibility inversion algorithm. The method
consists of measuring the visibility produced by an active point
source and compare with theoretical predictions. To increase
robustness of the retrievals, several active probes will be used,
one at a time, at different relative locations with respect to the
instrument.
V. CONCLUSION
After a successful manufacturing and integration of MIRAS,
the instrument is being tested in terms of hardware and soft-
ware operation. One of the most critical aspects is the in-orbit
calibration procedures. A number of tests have been designed
to infer the quality of this calibration and consolidate the
operations. The first results obtained have been very successful
and consistent calibration parameters have been retrieved.
ACKNOWLEDGMENT
This work was supported by the European Space Agency
and EADS-CASA Space Division under ESTEC contract
17950/03/NL/FF-SMOS; and by the Spanish Ministry of Sci-
ence and Technology (MCYT) under project TEC2005-06863-
C02-01/TCM.
REFERENCES
[1] P. Silvestrin, M. Berger, Y. Kerr, and J. Font, “ESA’s second Earth
Explorer Opportunity Mission: The Soil Moisture and Ocean Salinity
mission - SMOS,” IEEE Geoscience and Remote Sensing Newsletter,
no. 118, pp. 11–14, March 2001.
[2] M. Mart´ ın-Neira and J. M. Goutoule, “A two-dimensional aperture-
synthesis radiometer for soil moisture and ocean salinity observations,”
ESA Bulletin, no. 92, pp. 95–104, November 1997.
[3] I. Corbella, N. Duffo, M. Vall-llossera, A. Camps, and F. Torres,
“The visibility function in interferometric aperture synthesis radiometry,”
IEEE Transactions on Geoscience and Remote Sensing, vol. 42, no. 8,
pp. 1677–1682, August 2004.
[4] P. Moreno-Galbis, J. Kainulainen, and M. Mart´ ın-Neira, “Experimental
demonstration of the Corbella equation for aperture synthesis microwave
radiometry,” IEEE Transactions on Geoscience and Remote Sensing,
vol. 45, no. 4, pp. 945–957, April 2007.
[5] J. Bar´ a, A. Camps, F. Torres, and I. Corbella, “Angular resolution of two-
dimensional hexagonally sampled interferometric radiometers,” Radio
Science, vol. 33, no. 5, pp. 1459–1473, September-October 1998.
[6] I. Corbella, F. Torres, A. Camps, A. Colliander, M. Mart´ ın-Neira,
S. Rib´ o, K. Rautiainen, N. Duffo, and M. Vall-llossera, “MIRAS end-to-
end calibration. Application to SMOS L1 processor,” IEEE Transactions
on Geoscience and Remote Sensing, vol. 43, no. 5, pp. 1126–1134, May
2005.
[7] F. Torres, A. Camps, J. Bar´ a, I. Corbella, and R. Ferrero, “On-board
phase and modulus calibration of large aperture synthesis radiometers:
Study applied to MIRAS,” IEEE Transactions on Geoscience and
Remote Sensing, vol. GRS-34, no. 4, pp. 1000–1009, July 1996.
[8] F. Torres, I. Corbella, A. Camps, N. Duffo, M. Vall-llossera, S. Beraza,
C. Gutierrez, and M. Mart´ ın-Neira, “Denormalization of visibilities for
in-orbit calibration of interferometric radiometers,” IEEE Transactions
on Geoscience and Remote Sensing, vol. 44, no. 10, pp. 2679–2686,
October 2006.
[9] A. Colliander, S. Tauriainen, T. Auer, J. Kainulainen, J. Uusitalo,
M. Toikka, and M. Hallikainen, “Calibration of fully polarimetric
noise injection radiometer of MIRAS,” in 8th Specialist Meeting on
Microwave Radiometry and Remote Sensing Applications - µRad 2004.
Rome: Univerity La Sapienza, 24-27 February 2004, p. 80.
[10] A. Camps, F. Torres, I. Corbella, J. Bar´ a, and F. Monz´ on, “Automatic
calibration of channels frequency response in interferometric radiome-
ters,” Electronics Letters, vol. 35, no. 2, pp. 115–116, January 1999.
[11] R. Butora, M. Mart´ ın-Neira, and A. L. Rivada-Antich, “Fringe-washing
function calibration in aperture synthesis microwave radiometry,” Radio
Science, vol. 38, no. 2,1032, pp. 15–1,15–15, April 24 2003, doi:
10.1029/2002RS002695.
[12] I. Corbella, F. Torres, S. Blanch, S. Rib´ o, S. Beraza, N. Duffo, A. Camps,
M. Vall-llossera, and M. Mart´ ın-Neira, “Inter-element phase calibration
in interferometric radiometers,” in International Geoscience and Remote
Sensing Symposium, IGARSS 2006, vol. Draft, no. ISBN DVD-ROM:
0-7803-9510-7. Denver, Colorado USA: IEEE, 31 July - 4 August
2006, pp. –.
1-4244-1212-9/07/$25.00 ©2007 IEEE.3625