601.01 Detecting in the Presence of

The authors are grateful for support from NASA and JPL Exozodiacal Dust Profiles Goddard Space Charley Noecker (BATC), Marc Kuchner (GSFC) Flight Center Abstract: For direct detection mission concepts such as When a direct detection instrument is reduced in size for cost reasons, the Terrestrial Finder or Exoplanet Probe, light from the exozodiacal point spread function (PSF) becomes broader, making it more difficult to R- components of exozodi profiles dust tends to obscure any exoplanets present in the image. Data analysis distinguish a point source from a “blob” of exozodi light. In this case, the shot- -8 -8 10 10 methods to identify point sources against this background have been very noise limited integration time may not be enough; instead we may need an 300 Mean Sin phi -8 -9 Cos phi Sin 2phi 10 simple, traditionally with the simplifying assumption that the exozodi is elevated signal-to-noise ratio and/or later measurements to resolve ambiguities in Cos 2phi -10 Sin 3phi 10 200 -9 Cos 3phi Sin 4phi uniformly distributed, just as our local zodiacal background is uniform over the image data, identify a point source with a calculable and high confidence level, -10 10 Cos 4phi several-arcsec scales. However, the typical size of an exozodi cloud is and isolate the exozodi and exoplanet contributions to the observed light profile. 100 -10 2 2 Exozodi at 10pc, 10× -12 -11 10 10 expected to be comparable to the typical exoplanet orbital radii, or at least We will examine some typical profiles and a few methods of analyzing image data, brighter than Solar zodi,  0 -11 those of greatest interest — the “habitable zone” range from 0.7-1.5 AU. with the goal of structuring an approach to this data analysis problem. -12 no and no , 10 -14 10 -12 -100 =0.5 µm, 60° inclination Jy/mas Coefficients, -13 Jy/mas Coefficients, 10 -13 -16 -200 10 -14 10 Log10 contours, -14 2 -300 jansky/(2.5mas) -15 -18 The Search for Extrasolar Planets 10 10 -300 -200 -100 0 100 200 300 2 10 100 2 10 100 Angle, mas Angle, mas • The Exoplanet Exploration Program (ExEP, exep.jpl..gov) is leading NASA’s Mark a circle on the sky at each Cosines (x): Even harmonics Sines (y): Odd harmonics effort to find and characterize exoplanets, including by direct imaging methods. radius; look for sine and cosine dominate ( symmetric left-right) dominate (asymmetric top-bottom) • For direct imaging detection and spectroscopic characterization of -size harmonics around that circle

exoplanets at Earth-like radii, several mission concepts are being considered, 0 0 10 10 300 – Finder (flagship) Baseline case 1.4 -1 -1 – Exoplanet Probe 10 10 Symmetric exozodi, 200 1.2

No planets 100 1 -2 -2 Same zodi seen through a 10 10 TPF-interferometer 0.8 TPF- 4m diam Lyot coronagraph, 0 -3 -3 2 0.6 10 10 FKSI Coefficients, LZ Coefficients, LZ jinc mask (circular) -100

IWA = 3.1 /D = 80mas 0.4 Mean -4 -4 -200 10 Cos phi  ~zero 10 Sin phi PECO Cos 2phi Sin 2phi  ~zero 0.2 Linear contours Cos 3phi ~zero Sin 3phi -300 Cos 4phi Sin 4phi  ~zero

-300 -200 -100 0 100 200 300 0 50 100 150 200 250 300 0 50 100 150 200 250 300 Angle, mas Angle, mas 0 0 10 10 300 With Exoplanet 1.4 -1 -1 200 10 10 Symmetric exozodi, 1.2 ACCESS Earth-like planet 100 1 -2 -2 10 10 Same zodi with Earth-like Mean planet (on the left side), 0 0.8 Cos phi -3 Cos 2phi -3 10 10

0.6 Coefficients, LZ Cos 3phi Coefficients, LZ EPIC Seen through the same -100 Cos 4phi

coronagraph 0.4 Sin phi -4 -4 -200 10 10 Sin 2phi 0.2 Sin 3phi Sin 4phi Linear contours -300

The Challenge of Exozodiacal Light -300 -200 -100 0 100 200 300 0 50 100 150 200 250 300 0 50 100 150 200 250 300 Angle, mas Angle, mas 0 0 10 10 Pericenter Offset 300 We expect every target star to have zodiacal dust clouds, of unknown brightness Exozodi plus Earth 1.4 Asymmetric exozodi, -1 -1 Exozodi increases integration times 200 1.2 10 10 0.1 No planets  DC background for the planet signal Occulter 100 1 -2 -2 FWHM Same zodi with pericenter 10 10  increases shot noise and integration times 120 mas 0.8 offset of 0.0131 AU, 0

0 -3 -3 Seen through the same 0.6 10 10 -100 Coefficients, LZ Coefficients, LZ Exozodi profile features can mimic small exoplanets Arcseconds coronagraph 0.4 Mean -4 -4

Surface brightness (LZ units) (LZ brightness Surface Sin phi • Exozodi profile traditionally assumed symmetric or even uniform -0.1 -200 10 Cos phi 10 PSF FWHM = 15×30 mas Cos 2phi Sin 2phi 0.2 – Neglects difficulties in fitting exozodi and isolating the planet Linear contours Cos 3phi Sin 3phi -300 Cos 4phi Sin 4phi • Asymmetry in the image can be caused by Arcseconds -300 -200 -100 0 100 200 300 0 50 100 150 200 250 300 0 50 100 150 200 250 300 – Dynamics of the (giant planets  pericenter shift) Angle, mas Angle, mas Exozodi with 0.1AU – Occulter decenter and asymmetry pericenter offset – Requires a more general exozodi profile model for planet detection analysis The differences between these 2 curves are < 1% of the average EZ flux • Broad PSF (small telescope) leaves ambiguity between point sources and exozodi features 0.1 Demands greater measurement sensitivity to distinguish the 2 possibilities

Terrestrial planet detection requires fitting a fully general exozodi model 0 • From panchromatic camera or integral field spectrograph data Arcseconds Modeling uncertainty in exozodi profile • Fit exozodi shape and brightness, coronagraph suppression, and exoplanet position -0.1 • Correct for expected exozodi profile features, based on measurements The example pericenter offset is comparable to • Exozodi modeling could become the dominant • Uncertainty in exozodi profile increases uncertainties in exoplanet position, brightness , and spectrum Arcseconds • Dynamical effect of our on our local zodi uncertainty / ambiguity in finding exoplanets • One Earth signature in the coronagraph

) • Small telescopes (fat PSFs) make it more difficult 2 20 100% • 2% (TBR) skew asymmetry in internal occulter attenuation profile Occulter • 1.3 mas offset of external occulter from star (0.5 m) 15 • Brighter exozodis make it more difficult Causes of exozodi asymmetry transmission 75% • 1.3 mas error in locating the star on the detector Planetary dynamics causes real exozodi asymmetry • 10× smaller pericenter offset for 10× brighter exozodi 10 Exozodi pulled 50% • Gravitational influence of (s) causes exozodi pericenter offset off-center by Exozodi • Offset is locked to the giant planet’s axis of orbital symmetry (periastron-apastron) 5 giant planet 25% flux

• Precise knowledge of giant planet orbits might allow estimates of the asymmetry (µJy/arcsec Exozodi Possible mitigation strategies 0 0% -300 -200 -100 0 100 200 300 ) Occulter decentering causes apparent exozodi asymmetry 2 20 100% • Increased telescope size to sharpen the PSF • Exozodi brightness profile is sharply peaked near center Occulter 15 75% • Decentered occulter (internal/external) asymmetrically blocks the exozodi cloud transmission Occulter • Increased SNR to aid in disti nguishing subtle shape variations aligned • Precise knowledge of the offset vs. the star would allow estimates of the apparent 10 off-center 50% asymmetry • Multiple revisits to same target star to observe exoplanet orbital motion Exozodi 5 25% flux • Advanced exozodi models supported by calibration measurements Exozodi (µJy/arcsec Exozodi 0 0% -300 -200 -100 0 100 200 300 Measurement limitations Angle from star (mas) Conclusions: Broad PSF (small telescope) exacerbates the ambiguity problem (Exozodi asymmetry) ≈ (Fat PSF)  (Faint planet) • Makes point sources (exoplanets) difficult to distinguish from exozodi features • Features of exozodi clouds are ~0.1-1 AU in size;  Faint-planet detection requires careful modeling of exozodi asymmetry • Coronagraph with IWA = 2/D has PSF FWHM ≈ 0.5 AU • A broad PSF washes out details and spreads point sources 1.5m 2.4m 4m From10m P. Oakley  Uncertainty in faint-planet detection can be limited by uncertainty in

Cash et.al., Proc. SPIE 7010, p. 70101Q (Marseille, 2008) modeling the exozodi asymmetry