+ Current efforts for modeling exozodiacal disks Jean-Charles Augereau & Olivier Absil LAOG, Grenoble, France & U. Liège, Belgium Barcelona, September 2009

+ Various kind of models Radiative transfer To reproduce the scarce exozodi measurements, in particular IR interferometric data (CHARA/FLUOR, KIN) To reproduce extreme Spitzer spectra showing unusually large amount of warm dust and clear spectral features (e.g. Lisse et al. 2009) Dynamical modeling Sculpting of an asteroid/dust belt by a planet: capture in mean motion resonances (e.g. Stark et al. 2008) Larger scale models, with outward planet migration toward a Kuiper belt Sudden event, e.g. Nice LHB model (Booth et al. 2009) More progressive (Vandeportal et al., in prep.) Collisional models to evaluate lifetime of asteroid belts due to collisions (works by Krivov, Löhne, Thébault)

+ ISSI working group on exozodis ISSI international team assembled in order to Interpret the pioneering observations of exozodis with (near-)IR interferometers Identify the dominant source of dust (dust transport, collisions, comet evaporation) Predict the structure and amount of dust that can be expected around nearby stars Help to prepare future missions aiming at finding planets in the habitable zone Two one-week meetings were held (August 2007, April 2009) Next and last meeting in 2010  Anyone interested to attend should contact Jean-Charles Augereau ISSI team: Exozodiacal dust diks and Darwin

+ Exozodi emission spectrum Exozodis: are they zodi analogs? Fit of the 2.2 μ m excess with the Solar system zodiacal model : Vega : 3000 zodis to reproduce the CHARA K-band excess Fomalhaut : 5000 zodis to reproduce the VLTI/VINCI K-band excess Problem: it predicts too much flux in the mid-IR (by ~ an order of magnitude) Our zodi is not representative : detected exozodis have dust much closer to the star Flux Wavelength 2.2 μ m10 μ m ZODIPIC package [Kuchner]

+ Development of an optically thin exozodiacal disk model, based on debris disk model by Augereau et al. (1999) Scattered light and thermal emission Parametric surface density profiles and size distribution Variable dust composition (silicates, carbonaceous grains, porosity, …) Specific treatment of the exozodi disk inner edge to account for the size-dependent sublimation radius Outcome of the model : Peak position of surface density Grain properties (size, composition) Mass, optical thickness Typical dynamical time-scales (vs collisions, radiation pressure) Exozodis : radiative transfer - Solid line : 50% silicates + 50% carbons - Dashed line: 100% carbons

+ Result for Vega (Absil et al. 2006) All parameters go in the same direction: shifting the exozodi spectrum to shorter wavelengths compared to our zodi Small grains (mostly < 1 µm) at distances ~ 0.2 – 0.5 AU Highly refractive grains, no silicate feature  carbons > 50% Steep density profile:  (r) ~ r -4 (or steeper) Thermal emission only Total Exozodis : radiative transfer MMT/BLINC upper limit

+ Exozodis : radiative transfer Tau Ceti (Di Folco et al. 2007) Solar-type star (G8V at 3.65pc) Spitzer upper limits

+ Exozodis : radiative transfer Zeta Aql (Absil et al. 2008): More and more similar cases found with CHARA/FLUOR ~20% of stars? Depends on spectral type M sun 5.5-8AU

+ Other types of detections Spitzer/IRS (see Chas’ talk) Mid-IR imaging on 10-m telescopes VISIR, T-ReCS, Michelle, etc Only in favourable cases (~5 so far) due to limited resolution/sensitivity JWST/MIRI will do much better Mid-IR interferometry (see Chas’ talk) BLINC, KIN, (MIDI) All detections are showing large amounts of warm/hot dust HD (Moerchen et al. 2007)

+ Vega’s exozodi : origin of the dust? Exozodiacal dust grains have very short lifetime  high dust production rate, ~10 -8 M earth /yr for Vega Equivalent to 1 medium-sized asteroid every year Equivalent to a dozen of Hale-Bopp-like comets passing every day (400 tons of dust per second) Equivalent to the mass of the Vega Kuiper Belt per Myr If we exclude the case of a dramatic event, the question is: where does all this dust come from?

+ Vega’s exozodi : origin of the dust? Inward migration of grains due to Poynting-Robertson drag is excluded : too slow compared to other dynamical timescales (collisions, radiation pressure) Steady-state evolution of an asteroid belt excluded : km-sized bodies in a M Earth belt at around AU do survive ~10 5 years (Löhne & Krivov, priv. communication – see also Wyatt et al. 2007) Reservoir of mass in Vega’s Kuiper Belt (~85AU, ~ 10M Earth ). How to extract of this mass every Myr and transport it well inside 1AU? Planets can help … Schematic representation of the Vega system

+ Vega: planet migration Structures in Vega’s Kuiper belt : migrating planet trapping the parent bodies of the dust grains in MMR Wyatt (2003): Neptune-mass planet, migrating from 40 AU to 65 AU, at a rate of 0.5 AU/Myr. Circular orbits. Reche et al. (2008): Saturn mass planet, with e < 0.1) 2:1 MMR3:2 MMR

+ Vega: the comet factory A two planet system … Migrating Saturn mass planet [Wyatt 2003, Reche et al. (2008)] 0.5 – 2 Jupiter mass planet inside … and a Kuiper Belt … … to produce star-grazing comets. Numerical simulations with SWIFT (symplectic integrator). Counting of the test particles entering the 1AU zone

+ Vega: the comet factory A two planet system : Migrating Saturn mass planet [Wyatt 2003, Reche et al. (2008)] A 0.5 Jupiter mass planet at ~20 – 25 AU does the job well: Sufficient rate of comets in the 1AU zone for 40Myr Resonant structures at 85AU preserved

+ More extrasolar LHBs? See M. Booth’s poster Constrain: MIPS statistics ~12% of debris disk systems affected by LHB event? Mid-IR excess short-lived Expect (very) few detections

+ How does all this help? Asymmetries Most problematic exozodi feature (but also interesting on their own) Could they be predicted? Dust sources need to be known Planetary system architecture matters Extrapolate the Bryden exo-Kuiper belt histogram? Also valid for exozodis? How deep should we survey to make sure that XX % of exozodis are < 10 zodi? Can we rely on statistics only? Do we need to observe the exozodis of actual candidate targets? (  wait for SIM’s input) ~100 zodi