Download iaf2001_paper (doc - 1.8 MB)

IAF-01-Q.1.09
COROT System Requirements for Accurate Stellar Photometry
L. BOISNARD 1
M. AUVERGNE 2
1
2
CENTRE NATIONAL D’ETUDES SPATIALES – Toulouse – France
OBSERVATOIRE DE PARIS-MEUDON – Meudon – France
52nd International Astronautical Congress
1-5 Oct 2001/Toulouse, France
For permission to copy or republish, contact the International Astronautical Federation
3-5 Rue Mario-Nikis, 75015 Paris, France
IAF-01-Q.1.09
COROT SYSTEM REQUIREMENTS FOR ACCURATE STELLAR PHOTOMETRY
L. BOISNARD, COROT System Engineer
CNES
Toulouse, France
[email protected]
M. AUVERGNE, COROT Project Scientist
CNRS
Meudon, France
[email protected]
1.1.1.
Abstract
The COROT satellite, which will be launched late
2004, is dedicated to stellar seismology and search
for extra-solar planets. The mission is led by CNES in
association with French laboratories and several
European countries, contributing to the payload or to
the ground segment. The spacecraft is based on a
PROTEUS low earth orbit recurrent platform.
Stellar seismology
A star is a mass of hot gas, subject to forces of
gravity, pressure and Coriolis inertia if it rotates.
These forces play as the spring forces of an oscillator
with quantified eigen modes. The hydrodynamic
processes make the surface distort and are the
source of photon flux oscillations, whose amplitude is
expected about a few 10-6 (ppm).
The experiment is designed for high accuracy relative
stellar photometry, with long continuous observing
runs. To discriminate its oscillation modes with a
frequency resolution of 0.1Hz, each main target for
seismology will be observed during 150 days. A total
of 100 stars (magnitude less than 9) will be studied
during the 2.5 years mission lifetime.
Central program
Focused on internal hydrodynamic processes, the
COROT seismology central program will measure a
few 10-6 variations of the luminous flux emitted by 5
main bright stars with a resolution better than 0.1 Hz
in the Fourier space. This frequency resolution is
necessary to discriminate a significant number of
modes, to reveal the frequency splittings and to
rebuild the line profiles.
As second objective, COROT will detect the presence
of extra-solar planets if they pass between the
satellite and their parent star. 60000 stars (magnitude
less than 15.5) will be monitored. Between 10 and 40
telluric planets should be detected in the "habitable
zone" (temperature between 200 and 600 K).

The paper recalls the mission and intends to explain
where the critical scientific requirements are for
payload design, on-board treatments and system
engineering.
Seismology measurements will be performed in the
following scientific bandwidth : [0.1 ;10] mHz, covering
both pressure modes of stars of spectral types F and
G (higher frequencies) and gravity modes of stars of
spectral type A (lower frequencies). For stochastically
excited modes, the resolution is a function of the
observation window length, the oscillating mode
lifetime and the signal to noise ratio. Expecting mode
lifetimes about 5 days, the target stars must be
observed during 150 days with a minimum signal to
noise ratio of 15 (in terms of power spectral density).
1. THE COROT MISSION
1.1. COROT scientific objectives
The main target stars are typically F, G or  Scuti (A)
stars with a magnitude less than 6.5. In the vicinity of
a main target, some additional stars, brighter than a
magnitude 9 and belonging to an extended range of
stellar types ( Dor,  Ceph, peculiar metallic stars…),
will also be studied. COROT is designed to acquire up
to 10 stars simultaneously.
The COROT mission has 2 scientific programs, both
requiring long uninterrupted observations with very
high photometric accuracy. They work simultaneously
on adjacent regions of the sky.
Copyright © 2001 by CNES and ALCATEL
SPACE INDUSTRIES. Published by the
American
Institute
of
Aeronautics
and
Astronautics, Inc., with permission. Released to
IAF/IAA/AIAA to publish in all forms.
1
the rotation speed from the center to the surface of
the star. As illustrated for the pressure modes of the
Sun (see Figures 1 and 2) :
The large separation
Exploratory program
The purpose of this program is to observe a wide
variety of stars (from B to K spectral types) up to
magnitude 9, where the Hertzsprung & Russel
diagram is scanned.
   n,l   n1,l
is connected to the sound speed (i.e the density)
across the region of the stationary pressure modes.
The small separation
This will be accomplished by inserting a 20 days
observation between two observations of the central
program. With this shorter time window, the accuracy
on the frequencies falls to 0.6 Hz, but it is sufficient
to produce statistical data about the excitation of the
oscillating modes, as a function of mass, age, rotation
speed and metallicity. Fifty stars should be observed
in five exploratory programs.
   n,l   n1,l 2
is connected to the chemical composition near the
stellar core.
The second order differences
 2   n1,l  2 n,l   n1,l
give constraints on the convective zone of the star.
A total of 100 stars (magnitude less than 9) will be
studied during the 2.5 years mission lifetime, half of
them in the central program.
1.1.2.
Search for extra-solar planets
As second objective, COROT will be able to detect
the presence of extra-solar planets when they transit.
The detectors are 4 CCD 2048x2048 pixels with a
field of view of 8°2. Half is dedicated to the extra-solar
planets program. By adapting both the integration
time and the focus conditions, but without any change
in the mission sizing, luminous flux variations down to
7.10-4 (ground integration time : 1 hour) can be seen
on a large variety of stars whose magnitude is
comprised between 12 and 15.5. That is compatible
with an eclipse detection for a planet slightly bigger
than the Earth. For less bright stars (magnitude 15.5
and over), only giant gaseous planets will be
detected.
Examples of parameters to be measured
DF/F
(ppm)
p
0
-50
-100
t (hr)
-10
-5
0
5
10
Figure 3 : Principle of detection of a planet transit
The flux decrease is given by :
F / F  (Rp / Rs )2
where Rp is the radius of the planet and Rs the radius
of the parent star.
If the impact parameter p is equal to 0, the transit
duration is :
tr  (P / )(R s / a)
where P is the orbit period for the planet and a its
orbit radius (orbit supposed to be circular).
To detect a planet with a complete confidence, the
phase must stay coherent over 3 observed periodic
eclipses. Given the 150 days-long observations
imposed by the seismology central program, this
restricts the mission to planets with a period below 50
days.
The seismology spectrum of a star reveals some
important parameters of its internal structure : the
radius of the core, the Helium content and the mass,
the limits of the convective layers and the profile of
2
Considering that a transit is an achromatic event, the
use of chromatic information is helpful to discriminate
transits against stellar activity (very chromatic events).
So, a 3 colors dispersion device is placed in front of
the exoplanets CCD matrices and will enable, after
analyzing the signal chromatic variations, to widen the
detection domain to cases where the observation
window is not long enough to show a periodicity.
The total mass of the satellite (launch configuration) is
close to 600 kg, with a payload made up of :
In addition to hundreds of Jupiter-like planets,
between 10 and 40 telluric planets should be detected
in the "habitable zone" (temperature between 200 and
600 K), depending on hypotheses about accretion
models and planets existence. 12000 target stars, in a
field of view of 4 square degrees, will be
simultaneously observed during a 150 days-long
period. Five different fields of view, at least, will be
acquired during the whole mission.

an afocal telescope (270 mm entrance pupil)
composed of 2 parabolic mirrors, with a
cylindrical baffle to stop the Earth straylight and
an obturator against Sun blinding in early attitude
acquisition phase ;

a wild-field camera composed of a dioptric
objective (5 lenses) and a focal block equipped
with 4 frame transfer CCD 2048x2048. A bi-prism
is inserted in front of the two CCD matrices
dedicated to exoplanets ;

an equipment bay supporting scientific data
processing
electronics
(camera
controls,
extraction units and data processing units) and
instrument housekeeping electronics (power
distribution, fine thermal control, calibration
sources management, synchronization unit).
1.2. The COROT satellite
The main characteristics and performances of the
spacecraft are summarized hereafter :
The COROT spacecraft is based on a PROTEUS low
earth orbit recurrent platform, developed by CNES
and Alcatel Space Industries.
Mass
Bus dry mass = 270 kg
Propellant mass = 30 kg
Payload mass estimated at 300 kg
Length
4.20 m
Diameter
2.00 m
Power
Bus consumption = 300 W
Payload consumption class = 200 W
Electrical power generated by 2
symmetric wing arrays
NiCd battery (TBC)
AOCS
Gyro-stellar unit with 2 star trackers
and 3 2-axis gyrometers
Magnetometers and sun sensors for
attitude acquisition phase
4 reaction wheels, desaturated by
magneto torquer bars
4 1N thrusters for orbit maneuvers
120 m/s (hydrazine)
V capacity
Pointing
0.05° (3) on each axis
Improved to 0.5 arcsec with payload
ecartometric data
On board
Derived from a GPS receiver
Time
Compatible with 10-6 accuracy
Data handling Centralized architecture
2 MA 31750 Processors
Communication links via
MIL-STD-1553 bus and
discrete point to point lines
Data storage 2 Gbits (Mass memory )
Satellite-toS band QPSK
Ground
CCSDS packet standard protocol
Interface
Down Link
TM frames data rate = 727 kbits/s
TM packets data rate = 550 kbits/s
Up Link
TC frames data rate = 4 kbits/s
Lifetime
At least 3 years
Unavailability 0.88%
XS
Baffle
Telescope
Star Trackers
Camera
Focal Block
Radiator
Equipment
Bay
ZS
PROTEUS
Platform
YS
Figure 4 : COROT satellite overview
(represented with the satellite reference frame)
3
The platform will be used for the first time by the
French-US oceanography JASON satellite, to be
launched late 2001.

Straylight from the Earth : the line of sight must
remain at more than  =20° from the Earth limb,
i.e. the observations are possible when centered
in a direction at less than  = arccos(R/a) - 
from the perpendicular to the orbit plane. The
radius of the observation cone is fixed at  = 10°
(see Figure 6).

Roll domain :  20° on the boresight axis, after
alignment of the solar arrays for the optimum of
power budget. Such a rotation is helpful to
optimize the projection of the target stars onto the
CCD matrices (to get targets out from smearing
columns, for instance)
XS
Data Handling Unit
ZS
Battery
YS
Gyro
electronics
Figure 5 : PROTEUS platform overview
The satellite will be operated from the COROT Control
Center, located in Toulouse and sharing the facilities
of the PROTEUS satellites family. The preparation of
the observation sequences and the pre-processing of
the scientific data will be done by the COROT Mission
Center, located in Toulouse and having interfaces
with the laboratories. A dedicated automatic S band
ground station will be used for communication with
the satellite. Located at Villafranca (Spain), it will offer
4 visibilities per day. The mean volume of data to be
transmitted daily to the ground is limited to 900 Mbits.
Figure 6 : Orientation of the satellite

Platform thermal constraints : with respect to the
PROTEUS normal flight conditions, the satellite
Zs- sidewall (battery) cannot be exposed to high
solar fluxes for a long time. The battery thermal
heat distributor will be adapted to withstand 190
W/m2, which is compatible with a solar incidence
higher than 30°. A rotation on the boresight axis
(Xs) before or after 5 months will fit the central
and exploratory programs observation windows.

Payload thermal constraints : because of the
focal block radiator, the Ys+ satellite wall must be
in the shade as much as possible. Ys+ will be
exposed to the Sun only when the Earth is close
to the Line of Equinoxes (low solar declination,
high solar azimuth in the radiator reference
frame).
1.3. System constraints and mission schedule
The line of sight of COROT is assigned to keep a
same direction during each period of 5 months, with a
90% duty cycle requirement (no occultation by the
Earth). As a result, COROT has an inertial polar
circular orbit, at an altitude between 800 and 900 km.
The lower limit is fixed by the terrestrial straylight and
the upper one by the size of the South Atlantic
Anomaly and the maximum flux of protons acceptable
by the instrument. 826 km is preferred for phase
properties (orbit cycle over 7 days).
Put together, these constraints lead to only one
possible mission schedule, with four rotation
maneuvers per year, separating the different
programs as described on Figure 7.
The orbit parameters are resumed hereafter :
Semi-major axis
Eccentricity
Inclination
Right Ascension of the
Ascending Node
7178 < a < 7278
a = 7203 preferred
e = 0.01°
i = 90°
 = 12.5°
Boresight at 6h50  12h
The right ascension of the ascending node  fixes the
direction of the orbit plane, therefore the mean
direction of the observation cone. It has been chosen
to look at a dense region of the sky where the galactic
plane intersects the equatorial plane (see Figure 8).
The constraints to be taken into account for the
orientation of the spacecraft are :

Sun glare : the observations are possible when
the Sun is at more than 90° of the observed field
4
2. PERFORMANCES AND DESIGN
2.1. Photometric requirements
2.1.1. Seismology program
The accuracy () one can get from a line frequency 
measured in the Fourier space is given by :

( )2   /( 4Tobs ) 1     1  

3
where  is the mode width, Tobs the duration of the
observation and -1 the signal to noise ratio in terms
of power spectral density. To reach 0.1 Hz, a good
compromise is to observe during 150 days and to
have -1 at least equal to 15.
Random noise
Over a 5-day period (expected modes lifetime), it
leads to the following photometric performance :
FT(s)
Figure 7 : Mission schedule and attitude maneuvers
FT(p)
The optimized value =12.5° has been defined by the
COROT scientific committee on the basis of a set of
preparatory observations. The main characteristics of
the observation fields (radius = 10°), centered at 6 h
50 and 18 h 50, are the following :
where :

11 main target stars meeting the criteria of
magnitude and spectral type for the central
program, including 1 star of solar type (same
mass and age) and 2  Scuti candidates. It has
been verified that these stars are not polluted by
faint stars at less than 60 arcsec ;

813 stars as secondary targets, brighter than a
magnitude 8 (excluding the giant ones).
Photometric and spectroscopic ground based
observations are in progress ;

for the exoplanets program, the density of red
dwarfs is higher than 1500 per square degree
(magnitude less than 15.5) in most of the field.

nrT
2
1
nT
4
FT(p)
nT

1
 6.10 7
nT

FT(s) is the Fourier Transform of a stellar
oscillation to be measured

FT(p) the Fourier Transform of the photon noise

n is the mean flux measured by the detector

r=5.10-6 is the expected relative 0-peak oscillation
of a F, G star.
The photon noise level is fixed at 0.6 ppm and
determines the efficiency of the instrument round the
magnitude 6 : n=6.106 electrons/s. The standard
deviation of the photon noise is about 2500
electrons/s. Note : for a A type star, with greater
oscillations, a photon noise at 2.5 ppm is sufficient.
The objective is to keep every other source of random
noise below 1/10 of this reference noise. The main
performances required are given in the next table :
Straylight
Jitter noise
CCD readout and
electronics noise
Thermal sources
CCD thermal
variations
CCD efficiency
sensitivity
Video electronics
thermal variations
Video electronics
gain sensitivity
Figure 8 : The sky observed by COROT
The field of view is 2.7° x 3.05°, half for seismology
and half for exoplanets. The relative position
(left/right) has been defined as the same time as the
orbit plane, leading to a compromise between the two
programs to set each half-field in the most favorable
orientation inside the mission sky zone.
5
< 15 photons/pixel/s
Including the unavoidable
zodiacal component
< 0,5 arcsec
If the non uniformity of the
detectors is better than 1%
< 14 electrons rms/pixel/s
<  0.015° C peak-peak
< 5 10-3 /° C
<  0.5° C peak-peak
< 0.15 10-3 /° C
The point spread function (PSF) of a magnitude 6 star
stretches over 350 pixels. The elementary integration
time is 1 second.
2.2. Critical elements for payload and satellite
design
For each source of noise identified above, the design
solution is quickly described.
Structured noise and periodic perturbations
Seismology will be performed in the scientific
bandwidth : [0.1 ;10] mHz. Every structured noise
having spectral lines between 1 minute and 3 hours is
likely to be misinterpreted as a component of the star
signal. The orbit period and its first harmonics are in
that band. So, the instrument and the mission are
thought to search out these perturbations, whose
level must be reduced by design or corrected after
calibration. When increasing, a perturbation becomes
considered as unavailability and every interruption in
the data makes the signal to noise decrease and the
spectrum affected by windowing. The mission
scenario must guarantee continuous observations
with a duty cycle higher than 90%.
2.2.1. Straylight and telescope
With a circular orbit at 826 km, the main difficulty is to
trap the straylight coming from the Earth. At 20° from
the limb, the collected flux is around 10 20 photons/s.
To reach a residual flux of 1 photon/pixel/s at focal
plane level, the instrument shall have a 10 -13
attenuation coefficient.
A Three Mirror-Anastigmatic (TMA) solution being
excluded for insufficient straylight rejection capacity, a
2 co-focal parabolic mirrors telescope has been the
concept adopted.
MIRROR M1
0°8
Periodic perturbations, mainly instrument thermal
fluctuations, must be reduced to very low levels,
compatible after correction with a spectral line below
2 ppm.
Entrance
Pupil
Field
Diam. 4°
Line of sight
Xs
At instrument level, to avoid heavy implementation of
an active regulation system, every source of periodic
noise shall be kept at 50 ppm. The sensitive
equipment will be characterized before the launch and
a set of temperature probes will be mounted on the
payload for subsequent light curve corrections.
0°8
MIRROR M2
3°2
Exit
Pupil
3°6
Ys
Dioptric
Objective
Focal
Block
Figure 9 : Telescope concept
For example, the instrument is designed to have a
CCD temperature peak stability about 0.015°
associated with a knowledge of the temperature curve
itself better than 0.005°. The knowledge of every
thermal stability coefficient (CCD efficiency,
electronics gains, offsets) is estimated at 1%.
The focal plane of the two mirrors is shared by the
entrance and exit pupils, and by a square field stop.
The elliptical entrance pupil of equivalent diameter
270 mm is reduced by a factor 3. A dioptric camera
with 5 lens is placed in the collimated exit beam. The
focal length is 1.2 m.
The pupil surface peak stability is required below
3.10-6 and the PSF diameter peak stability at  0.2
pixel over the orbital period.
The straylight rejection coefficient, the highest to have
been required from a space telescope of this class,
associated with the weight and volume allocations of
the mission, needs a high-performance compact
baffle to be achieved. Its main characteristics are the
following :
2.1.2. Extra-solar planets program
The requirement to detect a change of stellar flux
equal to 7.10-4 for a 15.5 magnitude star is fulfilled if
the global random noise is twice the photon noise.
Given the seismology program requirements and the
typical image parameters for the exoplanets channel :

elementary integration time : 32,

PSF over 25 pixels for a reference K0 star,
Length
Entrance diameter
Contamination level
Internal design
Mirror roughness

optical transmission of the prism : 0.9,
the random noise budget is 1.8 times the photon
noise.
Concerning the periodic perturbations, because of the
windowing of the CCD (see chapter 3), the exoplanets
program asks for an additional requirement of optical
distortion stability : 0.05% over the orbital period.
Coating
6
2700 mm (2 stages)
800 mm
2000 ppm in orbit
Low diffraction chicanes
against grazing incidence
reflections
< 1 nm
(to prevent third diffusion
occurrence)
Low albedo (<4%) black paint
2.2.2. Pointing and AOCS
2.2.3. Focal block and images
A movement of the PSF on the CCD surface changes
the integrated photometry because the set of
impacted detectors is not the same. To comply with a
coupled attitude/photometry jitter noise 10 times lower
than the photon noise, the satellite pointing stability
requirement is stringent : 0.5 arcsec rms and needs to
use the instrument for ecartometry.
The focal plane is equipped with 4 EEV 4280 frame
transfer CCD of 2048x2048 pixels, working in a Multi
Pinned Phase (MPP) mode. This mode, associated
with a temperature regulated at –40°C, reduces the
dark currents to very low levels. The corresponding
noise should be less than 1 electron/pixel/s rms. The
13.5 m detectors are thinned, back illuminated, in
order to have a high quantum efficiency (70% once
integrated) in the bandwidth [370 nm ; 950 nm]. The
images are 16 bits encoded. This technology of
detectors is used by ground observatories, but has
not flown yet. A dedicated space evaluation has been
undertaken by CNES.
As shown by the budget hereafter, the periodic and
random noises due to the sensor are divided per 10
while thermo-elastic biases between star tracker and
payload frames are removed.
Perturbation
f
Thermo-elastic
f0
Sensor errors
f0
Gravity Gradient
2f0
Sensor random noise
Eclipses (transitory)
MTB commands
Amplitude ( line of sight)
PROTEUS
COROT
1"
0"
6"
0,03"
0,08"
1"
< 0.08"
5"
18" (could be reduced)
The light dispersion is performed through the bi-prism
inserted in front of the two exoplanets CCD.
Dispersion is along the detector rows.
Note : f0 is the orbital frequency
The AOCS loop will be modified and a specific
mission mode will be implemented in the PROTEUS
on-board software. At least 2 target stars of the
seismology program will be used by the ecartometric
algorithm, to be performed at payload level (leastsquare method). To avoid angular periodic errors due
to the focal length thermal variations, the focal length
will be estimated in real time. Small gaps of
perturbations
should
remain
during
eclipse
entries/exits, Magneto Torquer Bars (MTB) activations
and solar panels rotations. The control law of MTB will
be adapted to keep these gaps acceptable at system
level.
Figure 10 : Focal block
The limited CCD well capacity (in MPP mode) and the
jitter noise determine the size of the Point Spread
Function (PSF) for the seismology channel. For a
solar type star at the magnitude 6, the defocalized
spot will be 350 pixels large. Although two images per
minute are enough for scientific data reduction, the
ecartometric function asks the CCD to work at the
higher rate of 1 image per second. The windowed
readout is programmed by the camera control.
The transition from the PROTEUS standard mode to
the mission mode is performed in two steps :


The first step uses the star tracker data for the
attitude control loop, while the instrument data
are processed in parallel in another attitude
estimator. The size of the target stars windows is
adapted to the thermo-elastic drift between the
payload and the gyro-stellar unit reference
frames over the useful segment of the orbit (star
tracker's field of view not masked by the Earth). A
comparison
of
both
attitude
estimators
convergence is made on the ground, before
switching to the instrument data in the AOCS
loop.
The exoplanets channel time exposure is 32 s. The
spot size is about 60 pixels at the magnitude 13 for a
K2 type star. For the faint stars of this field, the
readout noise and the background are important
contributors to the noise budget. The readout noise
shall be limited to 5 electrons/pixel/s. Complete
images are transmitted to the on-board extraction
units in charge of soft windowing. It takes 22 s.
The instrument data are used to drive the control
loop, before a second transition to program a
change in the acquisition windows of the CCD
(size, binning). From then on the CCD
sequencers of the seismology channels are in
configuration to perform the scientific on-board
treatments.
7

In order to correlate the photometric signal with
external parameters like temperature or voltage, a
series of calibrations will be done at different levels :
CCD, camera, camera with readout electronics and
complete photometric chain.


a focal block radiator, having a radiative surface
of 940 x 280 mm2 and oriented towards the
instrument line of sight ;
two heat distributor radiators, having a radiative
surface of 400 x 100 mm2;
a set of thermistors and heaters.
A MLI system covers all the equipment bay, including
every inner elementary component and Second
Surface Mirror (SSM) sheets are used on the vertical
sides and on the external face of the radiators.
The thermal decoupling of the focal block with its
radiator is guaranteed by a high temperature gradient
(the radiator is at –60° C), filtering the orbital
perturbations. In the same way, simulations (including
eclipses and albedo echelon functions) have shown
that the temperature curves of the electronic units are
close to sine curves of a few 10 ppm, easy to correct
in ground post-treatments, with non phase-coherent
harmonics over 5 days (not likely to be confused with
the scientific signal).
Figure 11 : Examples of PSF for the seismology (left)
and exoplanets (right) fields
2.2.4. Thermal regulation
The thermal stability of the instrument will be carefully
controlled in order not to introduce too high periodic
fluctuations on the light curve induced by eclipses and
orbital variations of terrestrial aspect.
Fine Thermal Control
Radiator
The telescope thermal concept must provide a
temperature stability in order not to exceed a variation
of the star image bigger than 0.2 pixels on the orbital
period (and 2 pixels on the long term). This has led to
the choice of a carbon-cyanate material, associated
with a thermal control range of 20°  4° C. A Multi
Layer Insulation (MLI) system surrounds the
telescope structure.
Camera Control
Housekeeping
Electronics
(analogical telemetry)
Extraction Unit
Heat distributor
Housekeeping
Electronics
The instrument thermal heating is realized by 11
platform-provided heater lines : 7 dedicated to the
telescope and 4 to the equipment bay. Additional
lines are provided at instrument level for fine tuning of
the
focal
plane
and
proximity
electronics
temperatures.
Focal Block
Radiator
Data Processing
Unit
Converter
Figure 12 : Passive thermal regulation of the
equipment bay
The focal plane is at –40° C. This temperature is
driven by active thermal regulation, necessary to
achieve the stringent requirement of stability at CCD
level :  0.015° C. A thermal sensor is fixed under the
Invar block.
It must be noticed that extending the flight domain
w.r.t. solar fluxes implies increasing the rejection
capacity of the radiator, i.e. increasing its temperature
(the equivalent outside temperature is about –80°C).
Because of a lower decoupling efficiency with the
focal block, the orbital periodic perturbations would
get more important.
Concerning the electronics for scientific data
processing and payload housekeeping, the solution is
a passive thermal regulation of the equipment bay,
compatible with two on-board photometric chains
working continuously (stable heat dissipation). The
price is a significant increase in mass.
3. ON-BOARD TREATMENTS
The total scientific telemetry volume is 900 Mbits per
day, received by a dedicated S band antenna
(Villafranca ground station). Complete images cannot
be downloaded : photometry is integrated on-board
within pre-defined masks.
The equipment bay fine thermal regulation subsystem consists of (see Figure 12) :

two aluminum heat distributors 15-30 mm thick,
on which the camera control and housekeeping
electronics cases are mounted ;

an aluminum heat distributor 4 mm thick for less
sensible cases (extraction units, for instance) ;
3.1. Seismology program
The on-board photometric chains are designed to
process, for each CCD :

5 star windows (50x50 pixels2) ;
8


5 sky reference windows (binned in columns) ;
2 offset reference windows.
Figure 13 : Star and sky reference windows simulated
around HD 43318 (image size : 15' x 15')
The image of 2 stars among 5 can be downloaded in
25x25 masks, if accumulated during 32 s.
PSF fitting being ruled out (time consuming), two
aperture photometry methods are under evaluation :

The "threshold" method : a threshold s is defined
in such a way that the signal to noise ratio is
maximum. The set of pixels with intensity larger
than s define the aperture. As the mask is
determined for each new image, this method has
the advantage to be independent of the image
motion inside the CCD window. But, it does not
exclude the faint stars of the window, and if such
a faint star lies on the window edges, it will go in
and out with depointing, increasing the noise.

The "mask" method : first an average image is
acquired and a mask is built by the set of pixels
determined as in the threshold method. Pixels
belonging to another star inside the window are
excluded from the set. This aperture will be the
same for all images of a run. In this case the
photometry is sensitive to large translations,
which shift the spot outside the mask, but is not
polluted by closeby faint stars as long as the two
spot images are not in contact.
Figure 14 : Smearing on-board correction
Two stars of the seismology field are used for the
ecartometry. The barycenter calculation is made
using a threshold method.
3.2. Extra-solar planets program
The exoplanets fields are characterized by a high
density of objects up to the magnitude 23. The useful
pixels represent only 10% of the CCD image.
The on-board photometric chains are designed to
process, for each CCD :

5000 stars within chromatic masks ;

1000 stars within monochromatic masks ;

9 small sky reference images (10x15 pixels 2) ;

2 offset reference windows.
Both the methods will be implemented in the payload
software. The threshold method will be preferred for
isolated bright stars (magnitude less than 7).
The on-board software also corrects the photometry
from a series of undesirable components : offset
variations, dark currents, flat-field (TBC), smearing
due to detector frame transfer and false pixels when
crossing the South Atlantic Anomaly (high probability
of proton events).
Figure 15 : Simulated exoplanets field up to mv=16.5
around HD 49434 (image diameter : 20')
The masks have to be programmed among a predefined list of uploaded patterns. The monochromatic
masks are used for faint and cold stars analyzed in
9
white-light, as well as for background photometric
references.
36 stars can be oversampled to 32 s, in case a transit
event is detected. Every other correction of the light
curve will be done on the ground.
The shape of the mask is a function of several
parameters :

the magnitude of the star ;

the temperature of the star (or its color) ;

the position on the CCD, because of the space
variations of the PSF ;

the contamination of the target by closeby stars.
Conclusion
The COROT experiment addresses major scientific
topics. Designed to probe the internal structure of
stars and bring a new understanding of the stellar
evolution processes, it has been adapted to the
search for extra-solar planets. The photometric
detection of telluric planets is the next challenge after
the discovery of Jupiter-like objects. Besides, the
COROT stellar microvariability database, containing
the light curves of 100 000 stars, will provide the
astronomers community with wide research
opportunities (stellar magnetism, binary systems…)
It has been shown by simulation that 256 different
patterns allow to fit every possible target. A system
tool is under development to determine the optimal
programming of an exoplanets observing run, taking
into account the mask definition criteria, the behavior
of the instrument (PSF variations, local noise or CCD
defaults) and a number of effects such as spot
overlapping and smearing.
The project activity is currently focused on the
instrument and system engineering. Straylight
rejection, pointing, thermal stability and photometry
on-board processing are the main critical points of the
mission, for which cost-effective compromises have
been found. The satellite Preliminary Design Review
will be held in July 2002 while the instrument is
already in development phase for a delivery of the
flight model in 2004. The launch is scheduled late
2004, by a ROCKOT or SOYUZ launcher.
Simulations of representative fields will be done
before the launch, using the catalogues of targets
obtained from space or ground based observations.
The on-board software calculates the photometry in
one or three colors, depending on the selected
pattern, and accumulates the data during 32 cycles
(17 minutes). The movements of the line of sight
make the image spot move on the surface of the CCD
and the dispersed colors blend together over the
mask. To correct this depointing noise, the fluxes are
real-time interpolated before accumulation by the onboard software. The attitude data come from the
seismology channel.
Project Team
The mission is led by CNES and the following French
laboratories :

Observatoire de Paris-Meudon, involved in the
development of the camera, the on-board
software, the equipment bay and the ground data
processing for seismology ;

Laboratoire d'Astrophysique Spatiale (Marseille),
in charge of the telescope and the ground data
processing for exoplanets ;

Institut d'Astrophysique Spatiale (Orsay), in
charge of the calibrations and the prism delivery ;

Observatoire Midi-Pyrénées (Toulouse), which
conducts the important program of preparatory
observations.
The scientific committee, with members of the seven
European partners, is led by Annie Baglin, PI.
References
R. Samadi, G. Houdek, M.-J. Goupil, Y. Lebreton, A.
Baglin. Oscillation power across the HR diagram :
sensitivity to the convection treatment. Eddington
Workshop, 2001. ESA Special Publications series.
Contact
For more information and links towards other
sites, one can contact the internet server :
http://corot-mission.cnes.fr
Figure 16 : Optimal mask at mv=14 as a function of
the position (left) and the star temperature (right)
10
Acknowledgements
This paper presents results of the work of the COROT
project team, gathering people working in CNES as
well as in the French laboratories. We want to give
credit to their skill and their contribution.
11