Forthcoming tests to validate the missionNEAT depends on two key technology: (1) formation flying (FF) or mast deployment and (2) high precision centroiding using metrology. We are working on on-fly FF demonstration using the PRISMA satellites with CNES and OHB-Sweden that will mimic NEAT operation and

demonstrate the attitude control performance. For the validation of the performance for micropixel centroïding, two testbeds are being developed in parallel, one at JPL (left figure) and one at IPAG with CNES support (right figure).

1Institut de Planétologie et d’Astrophysique de Grenoble, université J. Fourier/CNRS, Grenoble, France 2 Jet Propulsion Laboratory, California Institute of Technology, Pasadena CA, USA

3 Laboratoire AIM, CEA/Sap, Saclay, France 4 Institut d’Astrophysique Spatiale, Orsay, France

F. Malbet1, A. Crouzier1, R. Goullioud2, P.-O. Lagage3, A. Léger4, M. Shao2, and the NEAT consortium (13 institutes)

Website : http://neat.obs.ujf-grenoble.fr

NEAT: an astrometric mission to detect and characterize nearby habitable planetary systems

Science objectives The prime goal of NEAT is to detect and characterize, in an exhaustive way, planetary systems orbiting bright stars in the solar neighborhood that have a planetary architecture with planets in their Habitable Zone like that of our Solar System or an alternative planetary systems made of Earth mass planets. It will allow the detection around nearby stars of planets equivalent to Venus, Earth, (Mars), Jupiter, and Saturn, with orbits possibly similar to those in our Solar System. It will permit to detect and characterize the orbits and the masses of many alternate configurations, e.g. where the asteroid belt is occupied by another Earth mass planet and no Jupiter. The NEAT mission will answer the following questions:– What are the dynamical interactions between giant and telluric planets in a large variety of systems? – What are the detailed processes involved in planet formation as revealed by their present configuration?– What are the distributions of architectures of planetary systems in our neighborhood up to ≈ 15 pc? – What are the masses, and orbital parameters, of telluric planets that are candidates for future direct detection and spectroscopic characterization missions?Special emphasis is put on planets in the Habitable Zone because this is a region of prime interest for astrobiology. Indeed orbital parameters obtained with NEAT will allow spectroscopic follow-up observations to be scheduled precisely when the configuration is the most favorable.

High precision differential astrometry

The p r i nc i p l e i s t o measu re accurately the offset angles between a target and 6-8 distant reference stars with the aim of differentially detecting the reflex motion of the target star due to the presence of its planets. The output of the analysis is a comprehensive determination of the mass, orbit, and ephemeris of the different planets of the multiplanetary system (namely 7 parameters MP , P , T , e, i, ω, Ω), down to a given limit depending on the star characteristics, e.g. 0.5, 1 or 5M⊕.

Representation of the NEAT targets in the 3D sphere of our neighborhood (D up to ≈ 15pc). They correspond to a volume limited sample of all stars with spectral types between F and K.

Distance (pc)

05

1015

20D

istance (pc)0

510

1520

closest_stars_Aclosest_stars_Fclosest_stars_Gclosest_stars_Kclosest_stars_MA.All Distance (pc)

0 5 1 0 1 5 2 0Distance (pc)

0 5 1 0 1 5 2 0

closest_stars_Aclosest_stars_Fclosest_stars_Gclosest_stars_Kclosest_stars_MA.All

A starsF starsG starsK starsM starsKnown exoplanets

mercredi 21 mars 12

Instrumental conceptThe concept consists of a primary mirror —an off-axis parabolic 1-m mirror— a focal plane located 40 m away to allow enough angular resolution and metrology calibration sources. The large distance between the primary optical surface and the focal plane can be implemented as two spacecraft flying in formation or a long deployed boom. The focal plane with the detectors having a field of view of 0.6° has a geometrical extent of 0.4m×0.4m. It is composed of eight 512 × 512 visible CCDs located each one on an XY translation stage in order to cover the whole field of view, while the central two CCDs are fixed in position. The CCD pixels are 10μm in size.

Preliminary Spacecraft Design. The proposed mission architecture relies on the use of two satellites in formation flying (FF). The two satellites are launched in a stacked configuration using a Soyuz ST launcher, and are deployed after launch in order to individually cruise to their operational Lis- sajous orbit. Acquisition sequences will alternate with reconfigurations, during which the Telescope Satellite will use its large hydrazine propulsion system to move around the Focal Plane Satellite and to point at any specified star. At the approach of the correct configuration, the Focal Plane Satellite will use a cold gas μ-propulsion system for fine relative motion acquisition. The Focal Plane Satellite will be considered as the chief satellite regarding command and control, communications and payload handling.The mission relies on a 40 m focal length telescope, for which the preferred solution is to use two satellites in formation flying. The performance to be provided by the two satellites in order to initialize the payload metrology systems are of the order of magnitude of ±2mm in relative motion, and of 3 arcseconds in relative pointing, which are typically compatible with Formation Flying Units and gyroless AOCS6 architecture. The mission aims at a comlete survey of a large number of targets and the maximization of the number of acquisitions will be a main objective of the next mission phases. The mission objectives require a threshold of 20,000 acquisitions. In addition, the time allocation for these reconfiguration maneuvers is quite limited, in order to free more than 85% of mission duration for observations. An alternative mission concept would consist of a single spacecraft with an ADAM-like deployable boom (from ATK-Able engineering) that connects the telescope and the focal plane modules. A possible implementation, EXAM, made by JPL is shown on the left side (see «A scalable concept»).

How well do we know our neighbours (FGK stars,d≤20pc) ?Interestingly, whereas we have detected now several hundreds and even reaching several thousands exoplanets, the number of planetary systems around nearby solar-type stars, a volume limited sample, is less than 10%, mainly because the key detection technique, transit search, can detect only planets with a special orbit (0.5% efficiency for a planet at 1 AU around a Sun-like star).  However, planets around nearby stars are of special interest because one can conduct statistics or precise follow-up. These systems are also interesting because of their proximity that can yield direct spectroscopic studies with coronagraphic or nulling techniques. Therefore, an exhaustive method to detect them is very desirable. 0,2 %

8,3 %

91,5 %No known planets

Known planets

Habitable planets

upsilon And (4.6)

TYC2 2067 (10.4) TYC2 2176

(12.8)

TYC2 1302 (12.0)

TYC2 2188 (11.3)

TYC2 478 (11.7)

TYC2 1396 (12.0)

TYC2 1969 (12.1)

TYC2 1474 (11.6)

TYC2 352 (12.7)TYC2 2010

(10.0)TYC2 707

(12.2)

TYC2 454 (9.8) TYC2 602

(10.1)

TYC2 560 (11.6) TYC2 286

(11.1)

TYC2 1424 (12.0)

TYC2 1182 (12.0)

TYC2 1246 (11.5)TYC2 1542

(10.2)TYC2 1867

(10.8)

TYC2 1880 (11.3)

TYC2 1716 (10.6)

TYC2 1668 (10.0)

0.3°

NE

Typical field of NEAT with a field of view of 0.6°

Simulation of astrometric detection of two type of planets with 50 NEAT measurements (RA and DEC) over 5 yrs. Top row parameters corresponds to an easy detection: MP = 5M⊕, a = 1.8AU, M∗ = 1M⊙, D = 5pc. Bottom row parameters corresponds to a more challenging detection: MP = 1.5M⊕, a = 1.16AU, M∗ = 1M⊙, D = 10pc. Left panels: sky plot showing the astrometric orbit (solid curve) and the NEAT measurements with error bars; middle panels: same data but shown as time series of the RA and DEC astrometric signal (an offset has been put to separate them); right panels: joint Lomb-Scargle periodogram for right ascensions and declinations simultaneously. All measurements corresponds to 1h visits with a 0.8μas/sqrt(h) precision.

A

A

A

A

A

A

A

A

A

A

AA

AA

AA

A

A

AA

A

A

AA

A

A

AA

A

A

A

A

A

A

A

AA

AA

A

A

A

A

A

A

A

A

A

A

A

−2 0 2

−2

0

2

X (µas)

Y (µ

as)

1.5 Earth mass planet at 1.16AU @ 10pc

A

A

AA

AA

A

A

A

AA

A

A

A

A

A

AAA

A

AAA

A

A

A

AA

A

A

A

A

A

A

A

A

A

A

A

AAA

A

AAA

GG

GG

G

G

GG

G

G

GG

GG

GGG

G

GG

GG

GG

G

G

GGG

G

G

G

G

G

G

GG

GG

GG

G

G

G

G

G

0 2 4

−5

0

5

Time since launch (years)

X−20

(µas

) − Y

+20

(µas

)

1.5 Earth mass planet at 1.16AU @ 10pc

1.0 10.00.2 2.0.5 5. 0

5

10

15

20

Period(years)

Pow

er S

pect

ral D

ensi

ty (µ

as2)

1.5 Earth mass planet at 1.16AU @ 10pc

1.5ME @ 1.16AU @ 10pc

1.0 10.00.2 2.0.5 5. 0

10

20

30

40

Period(years)

Pow

er S

pect

ral D

ensi

ty (µ

as2)

5 Earth mass planet at 1.8AU @ 5pc

A

AAA

A

AAA

A

AA

AAA

A

AA

A

AA

A

A A

A

AA

AA

A

A

A

AAAA

A

AA

A

AA

A

AA

A

AA

A

AA

−10 −5 0 5 10−10

−5

0

5

10

X (µas)

Y (µ

as)

5 Earth mass planet at 1.8AU @ 5pc

A

AAA

A

AAAA

AAAA

AAAA

AA

AAAAA

AA

AAAA

AAAA

AAA

AA

AA

AAA

AA

GGGGG

GGGG

GG

GGGG

GGG

GGG

GGG

GG

GGGGG

GGGGG

GG

G

GGG

GGGG

0 2 4−20

−10

0

10

20

Time since launch (years)

X−20

(µas

) − Y

+20

(µas

)

5 Earth mass planet at 1.8AU @ 5pc

5ME @ 1.8AU @ 5pc

8

1 fixed CCD (target star)

Focal plane

1 fixed CCD (telescope axis

tracker)

8 movable CCDs (reference stars)

Telescope spacecraft

Detector spacecraft

Metrology

Telescope axis beam

Dynamical Young’s interference fringes

The principle of the measurement is to point the spacecraft so that the target star, which is usually brighter (R ≤ 6) than the reference stars (R ≤ 11), is located on the axis of the telescope and at the center of the central CCD. Then the 8 other CCDs are moved to center each of the reference stars on one of them. To measure the distance between the stars, we use a metrology calibration system that is launched from the telescope spacecraft and that feeds several optical fibers (4 or more) located at the edge of the mirror. The fibers illuminate the focal plane and form Young’s fringes detected simultaneously by each CCD (right panel). The fringes have their optical wavelengths modulated by acoustic optical modulators (AOMs) that are accurately shifted by 10 Hz, from one fiber to the other so that fringes move over the CCDs. These fringes allow us to solve for the XYZ position of each CCD. An additional benefit from the dynamic fringes on the CCDs is to measure the QE of the pixels.

A scalable conceptOne of the strengths of NEAT is its flexibility, the possibility to adjust the size of the instrument with impacts on the science that are not prohibitive. The size of the NEAT mission could be reduced (or increased) with a direct impact on the accessible number of targets but not in an abrupt way. For instance, for same amount of integration time and number of maneuvers, the options listed in Table 3 are possible, with impacts on the number of stars that can be investigated down to 0.5 and 1 Earth mass. The time necessary to achieve a given precision depends on the mass limit that we want to reach: going from 0.5 M⊕ to 1 M⊕ requires twice less precision and therefore 4 times less observing time a l l o w i n g a s m a l l e r telescope.µNEAT is a proposition submitted to ESA call for S mission in June 2012 based on existing and available technology (current FF performance and state of the art in metrology). The performance has been tuned to study giant planets in the HZ of solar-type stars and a few super Earths.

20 pc 5,0 Earth mass 0,8 uas 0,5 0,15 uas 10 pc 1,8 Earth mass 0,5 uas 1 h 0,11 uas

5 pc 0,5 Earth mass 0,3 uas 3 0,06 uas

SNR6

0,8RSS

0,12 uas 0,61 uas 0,14 uas 0,08 uas 0,26 uas 0,32 uas

6 Mag 11 Mag 32x32 pixels 512x512 pixels1 star 6 stars Tracker 1 CCD

3400 sec 3400 sec 60 sec/axis 3400 sec 5,0E-06 pixel

AstrophysicalErrors

- Stellar aberation - Ref. star geometry - Companions

error1 Targ. CCD vs. 6 Ref. CCDResidual calibration

ErrorsAstrometric errorReference stars

Focal-plane metrologydifferential error

CCD Calibration

3

Astrometric accuracy per 1h observation and per axis (X,Y)

uas

Target star metrology errorAstrometric error Parabola axis Tracker

26

Differential Measurement

Integration time for 1 visit (t)h

h

Total number of visits during the mission lifetime (N)

50

m = 5 + 7P

Min. detectable astrometric signature for 3-planet system

Astrometric end-of-missionnoise per axis (X, Y)

sig_X,Y = DMA/sqrt(Nt)

Number of planetsin the system (P)

Minimum detectable1AU planet mass at:

A = DMA*SNR/sqrt((2N-m) t)M = (d/10)*(A/0,3)

# of parameters for a P-planet system (m)

Top-level error budget for NEAT that shows the major contributor to the overall budget. It also shows how the 0.8 μas/sqrt(h) accuracy enables the detection of 0.3 μas (resp. 0.5 μas and 0.8 μas) signatures with a signal to noise of 6 after 50 visits of 3 h (resp. 1 h and 0.5 h) of observation that allows a 0.5 M⊕ planet at 5 pc to be detected (resp. 1.8 M⊕ at 10pc and 5M⊕ at 20pc).

EXAM (NASA)Shao et al.

µNEAT (ESA small mission)Brandeker et al.

(m) (m) (deg) (cm) (R mag) (µas) 0.5M⊕ 1 M⊕ 5 M⊕NEAT plus 1.2 50 0.45 40 11.5 0.7 7 100 200NEAT 1.0 40 0.56 40 11 0.8 5 70 200NEAT light 0.8 30 0.71 35 10.5 1.0 4 50 200

EXAM 0.6 20 0.85 30 10.1 1.4 2 35 200

1M⊕ 10 M⊕ 50 M⊕µNEAT (*) 0.3 12 0.6 15 11 10.2 2 25 200

# targets for a given mass limit

DMA = Differential astrometric Measurement Accuracy (rms); (*) centroiding requirement relaxed to 4e-5

Mission name

Mirror diameter Focal length

Field of view diameter

Focal Plane size

Ref. star mean

magnitudeDMA in 1h

a ULE bench with a collimator parabola glued to it. The parabola was masked to allow a 14 mm ‘primary’ mirror to be defined. A bundle consisting of 7 closely packed fibers was mounted with a defocus in such a way to create an image at an effective focal length of 1.1 m and an F# of about 80. The middle image, top row, shows three of the seven stars created this way. The sampling of the PSF’s, defined as the number of pixels per f / D, was 2.55. Later on in the experiment, the aperture diameter was reduced to allow better sampling. Also mounted on the ULE block was a low expansion ‘metrology block’, on which was mounted six metrology fibers with their tips pointed at the detector. These would be illuminated pairwise to create various fringes.

Figure 6: MCT test setup in the SIM vacuum chamber at JPL. Shown are the chamber (top left), the test setup (bottom), the ULE low-expansion optical bench (top right), and a single frame from the CCD with three stars and metrology fringes (top middle). In the bottom picture the test setup can be seen, with the camera on the left.

Figure 7 shows the metrology block and the imaging parabola during the installation phase. As highlighted by the circles in the picture, 3 horizontal and 3 vertical fibers were mounted on the block. The spacing between the fibers was designed to create a range of fringe spacings. On the right side of the figure can be seen 15 fringe patterns observed as each unique pair of fibers was turned on.

A more detailed look at the metrology system is offered by Figure 8. Acousto-optic Modulators (AOM’s) are used to switch on laser light to a pair of optical fibers, frequency shifting the light to one of the two fibers by a few Hz relative to the other. In general, an optical fiber projects laser light at divergent cone of about 10 degrees (deg) in diameter. On the CCD, the light from the two illuminated fibers interferes and, because of the frequency offset, the fringes “travel” across the CCD surface. The CCD in the experiment could be read at up to 50 frames per second (fps). If the applied (“heterodyne”) frequency offset is chosen at 5 Hz, then the output of any given pixel is a discretely sampled, 5 Hz sine wave with 10 samples per cycle (assuming 50 fps readout). If we compare two adjacent pixels we see two sine waves with a known phase shift. If the fringe spacing is, for example, 4 pixels, then adjacent pixels will have a /4 (i.e. 90 deg) phase shift. In other words, the measured phase difference between any two pixels is directly proportional to the distance between the pixels along the direction of the traveling fringe. By illuminating pairs of fibers with different relative positions, we can produce fringes traveling in different directions on the CCD surface and hence derive the relative

Proc. of SPIE Vol. 8151 81510W-6

Figure 9: MCT test results showing the flat field response (left) and effective pixel location deviations from a regular grid along row (middle) and column (right) directions.

Figure 10: Metrology of pixel locations for a 10x10 zone achieves 20 pix accuracy in 25 sec (left). Also, the measured distance between the centroids of two artificial stars dips down to 30 pix after 200 seconds (right) under conditions where the individual centroids are moving by about 10X as much.

5. SUMMARY AND CONCLUSION Micro-pixel centroiding holds promise as the enabling technology for low cost micro-arcsecond astrometry and particularly the search for Earth-mass planets in the habitable zones of nearby sun-like stars. We have developed an approach that uses precision metrology to calibrate the otherwise intractable focal plane systematic errors that would be encountered in getting down to micro-arcseconds in astrometric accuracy. We have checked the algorithmic aspects of our approach using simulations and begun testing the remaining aspects using a small testbed called MCT.

Using simulation we found that our image position sensing algorithm is capable of 4 pix accuracy in the presence of wavefront errors and displacements up to half a pixel. With the MCT testbed, we were able to calibrate the focal plane with a pixel-to-pixel differential measurement precision of less than 20 pix after 25 seconds of integration. Also, we demonstrated star-to-star differential measurement precision of less than 30 pix after 200 seconds of integration. The next set of tests will aim at systematic errors and aim to show accuracy at the few pix level by measuring the post-calibration inter-star distance repeatability under conditions where the actual distance is effectively constant.

Proc. of SPIE Vol. 8151 81510W-8

4 10-5

JPL testbedNemati et al. 2012, SPIE 8442, 57

Pseudo stellar sources

E2V - CCD 39

Mirror bloc

Zerodur bench

Translation deck

Metrology fibres

Invar bench

Liquid core fibre

Reading electronics

Heat-pipe to Peltier cooler

mardi 26 juin 12

CNES testbed

Crouzier et al. 2012, SPIE 8445, 59

Crouzier, Malbet, Preis et al. 2012, SPIE 8445 (arXiv:1208.0099)Malbet, Léger, Shao et al. 2011, Experimental Astronomy, in press (arXiv:1107.3643)Malbet, Léger, Goullioud et al. 2011, Proc. of SFFMT conference (arXiv:1108.4784 )Malbet, Goullioud, Lagage et al. 2012, SPIE 8442 (arXiv:1207.6511)

Nemati, Shao, Zhai, et al. 2011 SPIE 8151, E28 Nemati, Shao, Zhai et al. 2012, SPIE 8442, in pressShao, Nemati, Zhai et al. 2011, SPIE 8151, E27Zhai, Shao, Goullioud, Nemati, 2011, RSPSA 467, 3550

Acknowledgments:

0

5

10

15

20

25

30

35

40

45

50

55

60

65

0 2 4 6 8 10 12 14 16 18 20star distance

Cou

nt

Nearby FGK w known exoplanetsNearby FGK from Hipparcos catalog

Distance (pc)2 4 6 8 10 12 14 16 18 200

0

10

20

30

40

50

60

Only 10% of nearby stars have known exoplanets so far !

Nearby FGK stars 0-10 pc 10-20 pc Total I FGK stars with known exoplanets 6 36 42 I FGK listed in the Hipparcos catalog 68 426 494

02

4

6

8

10

12

14

16

18

20

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0eccentricity

Count

0

5

10

15

20

25

30

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0Planets

Count

01

2

3

4

5

6

7

8

9

10

5.05.0 1e01 5.0 1e02 5.0 1e03 5.0mass*300

Cou

nt

nearby FGK

01

2

3

4

5

6

7

8

9

10

5.05.0 1e-01 5.0 1e00 5.0 1e01

axis

Count

Planet mass (ME) Semi-major axis (AU)

Number of planets Eccentricity

0

2

4

6

8

10

0

10

20

30

0

2

4

6

8

10

0

4

8

12

16

20

2

6

10

14

18

1 2 3 4 5 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.00.0

0.05 0.1 0.5 1 5 105 10 50 100 50010001

N JSE

Statistics on nearby exoplanets (d