PICTURE, a sounding rocket to characterize the debris disk of

Ewan. S. Douglas1 ([email protected]), Chris Mendillo2, Brian Hicks2, Jason Martel2, Timothy A. Cook2, Ron Polidan3, Supriya Chakrabarti2.

1Department of Astronomy, Boston University, Boston, MA. 2Center for Atmospheric Research, Department of Physics, University of Massachusetts Lowell, Lowell, MA. 3Northrop Grumman Corp.

ABSTRACT Mission Properties Observing Plan

The PICTURE (Planetary Imaging Concept Testbed Using a Rocket Experiment) sounding Telescope 0.5 m, f/12.3 Gregorian No. 2, 2009 EPSILON ERIDANI DEBRIS DISK 1533 Observing rocket will demonstrate a nulling interferometer (a “nuller”) to characterize the exozodiacal dust Visible Nulling Coronagraph, TARGET Purpose disk of Epsilon Eridani (K2V, 3.22 pc) in reflected visible light to an inner radius of 1.5 AU (0.5”) Nuller Time from the surface of the . The first launch of PICTURE suffered a telemetry failure and the 600-750 nm, shear of 0.15m Demonstrate nulling, primary mirror was shattered upon landing. A new launch is scheduled and the PICTURE Rigel 60 Sec payload is currently undergoing refurbishment, including receiving a new SiC primary mirror. Science and Wavefront 1024x1024 pixel, back-illuminated record reference PSF PICTURE visible light observations will constrain scattering properties of the Epsilon Eridani Sensing CCDs CCDs from Astro-E2 X-ray mission exozodiacal dust and debris environment. Measuring the dust brightness of the system ϵEri, Roll 1 >105 Sec Characterize Exozodi Inner & constraints the background which must be overcome for future observations. 1.5λ/d (0.5”) & 5” Additionally, PICTURE will demonstrate operation of a MEMS deformable mirror and a visible Outer Working Angles Characterize Exozodi The first PICTURE launch in ϵEri, Roll 2 >105 Sec nulling coronagraph in space. Improved modeling and post-flight measurement of instrument and track speckles October 2011. (Courtesy WSMR.) Tip-Tilt steering mirror, 5 mas Figure 13: Left: PICTURE launch. Right: Telemetry failure. The main science data telemetry transmitter performance allow us to present refined exozodiacal dust sensitivities. Pointing System (TM 2) failed 70 seconds after launch. pointing (Mendillo et al. 2012a) Epsilon Eridani Debris Disk: Reidemeister et al.: ε Eridani ! Interfering the PICTURE wavefronts: Simulated ϵEri observations7. FLIGHT, ANOMALY, RECOVERY & FUTURE WORK Nuller performance is simulated by interfering the complex wavefronts of the sheared nuller arms th 2 0 after subtraction ofPICTURE dustless launched Rigel fromPSF White Sands Missile Range at 4:25 MDT on October 8 , 2011. Approximately 70 IRAS 10 MIPS and generating the resultant point-spread-function (PSF) for a list of discrete points, composed of 1.5 MIPS (SED) seconds after launch, the main science data telemetry transmitter onboard the payload failed. Figure 13 shows SHARC II SCUBA 70% ice + 30% astrosilthe central star and the warm inner ring, with POPPY (Physical Optics Propagation with PYthon) 1 the TM transmitter power during flight. This TM channel transmits the high-bandwidth 4 Mbps data stream 50% ice + 50% astrosil[Perrin et al 2012, github.com/mperrin/poppy]. We computed the polychromatic PSFs in parallel shown in Figure 4, which carries all of the WFS and SCI images and the WCS data products. Due to lack of excess flux [Jy] flux excess PICTURE FOV 100% astrosil 0.5 blow-out limitusing PiCloud.com, which after stacking and addition of noise, produce realistic images. -1 data, the in-flight performance of the nuller and active optics cannot be confirmed. Approximately 20 seconds et 10 n 0 la 10 100 1000 p Simulation Inputs: of flight data has been recovered, unfortunately this data is from very early in the flight, before the FPS had wavelength [µm] r -ratio -5 te β Outer Warm Ring Baseline leakage: 1% amplitude (E0) mismatch (limiting central null to 2.5*10 ) acquired steering lock and thus the nuller had not begun its alignment sequence. This data may be useful in u • o • Measured Flight Deformable Mirror (DM) Surface Figure future analysis for determining some basic functionality of the nuller and telescope. 10-2 • Secondary/Spider Obscuration & Lyot Mask shape The most important result returned from the PICTURE flight is the successful demonstration of the FPS. planet 'B' Inner Warm Ring • Template source spectra (e.g. Pickles et al. 1998, via pysynphot) & surface In flight, the ACS acquired the calibration star, Rigel according to schedule. Manual ACS uplinks were sent to planet 'A' brightness distribution drive the star into the acquisition mirror pinhole. The functioning FPS telemetry link showed the star appear on 10-1 100 101 Shot Noise and Dark Current i • the live angle tracker GSE display. An additional ACS uplink was sent to drive the star into the capture range of nn ) grain size [µm] er (<10AU the FSM at which point the FPS immediately locked on and stabilized the beam pointing. The FPS operated at Fig. 2. The βadaptedratio from of two Backman silicate et al. – water 2009 ice mixtures, 200 Hz with a 5 Hz bandwidth. Analysis of the raw centroid and FSM data that was stored onboard has shown in ) te U compared to a pure silicate composition, as a function of that the FPS stabilized the 627 mas RMS ACS radial pointing error to 5.1 mas RMS. The FSM position data rm 5A ediate (10-5 size for ε Eri. The bars show the size bins used in the inner also provides a high-precision measurement of the rocket ACS performance. A power spectral density analysis Figure 8. (a) Perpendicular optical depth τ vs. position for the five model disk components. The two unresolved inner belts (dashed lines), outer “icy” ring (dotted Reidemeister et al. 2011 disk simulations⊥ (Sect. 4). The dashed horizontal line shows A dustless simulation showing four regionsof of both the ACS body-pointing and the FPS stability is presented in (Mendillo et. al. (2012) submitted). line), and ring+halo “silicate” dustthe populationsFrom dynamical the (dot-dashed SED, blowout line) Backman are shown. limit (b) et Mass of al.β surface (2009)=0 density.5. vs. radius corresponding to the optical depthprofilesinpanel The predicted 3AU ring can seen above, Epsilon Eridani’s Spectral energy(a), assumingdistribution grain compositions and material densities described in the text. The difference in slopes between the optical depth and mass surface density profiles is scattered light (corners) from the DM at large The operation of the telescope in flight is confirmed through the FPS results. The FPS data suggests that due to radial variation in model particleinfer properties. a system of rings, shown above, the circle is cut-off by the (horizontally (SED) showsou a significant flux) excess at 24um radii and a residual central leakage from colorthe telescope PSF was much larger than anticipated, nearly 10 the diffraction limit. The cause of this is yet ter ring 5-90AU with a thin ring near 3AU, scattering oriented) nuller transmission minima at and beyond, indicating (5 a warm ring within 10AU, 2.3. Dust grain properties mismatch between Epilson Eridani (210 unknown. Unfortunately, the lightweight primary mirror shattered upon landing in the New Mexican desert. The top panel of Figure 8 starlightdisplays the with model an radial expected profile integratedFigure 9 is a two-dimensional representation of the ϵ Eri 0” and 0.9”. proposed by Reidemeister et al.of perpendicular(2011) to be optical depth τ ,thatis,fractional(unitless) debris disk system with a linear spatial scale, showing the second observation) and Rigel (20 second). -4 The remaining hardware was recovered intact and appears to be in working order. Post-flight optical testing has 2 Thebrightness⊥ knee in theof 2*10 IRS2 spectrumF*. at ∼ 20 µm (inset in Fig. 1) Fig. 1. Asustained schematic byview solar of wind-driven the ε Eridanisurface Poynting- system’s density (m archi-of grain cross-sectional area per m of disk sub-mm ring, halo, and inner warm belts. The position of not yet been conducted. surface area; e.g., Backmanis2004(F reminiscent* ),~3e8 and in phot/sec the bottom of a panel, classicalfor PICTURE). the silicatethe innermost feature. warm Since belt the is compared ex- in the inset figure with tecture. TheRobertson outer ring drag. is the region where the dust is pro- Example wavefronts from the DM mirror arm of the nuller. Contrast corresponding mass surfaceact density composition inferred from of the those optical silicatesone is orbit not solution known, for we the have candidate planet The 5σ contrast curves at right are derived Future prospects look towards a PICTURE reflight. This would first and foremost accomplish the science duced by parent planetesimals; the intermediatedepth and model zone grain is the properties is displayed. The slopes of the (Section 5.4). chosen astronomical silicate (Laor & Draine 1993) (ρd = from the annular, one pixel wide, standard and technology goals of the first mission. In addition, many lessons have been learned along the 7 road to one whereThe it is transportedPayload inward by dragtwo functionsforces, possibly across the sub-mm3.5gcm ring−3 and). Onhalothe are not other equal hand, by analogy with the sur- because the model includes smooth variation in the particle size 5. DISCUSSION deviation of the expected background (from flight. A number of hardware upgrades have been identified to improve the nuller performance. The two leading interacting with a presumed outer planet; transport contin- face composition of Kuiper belt objects in the solar system ues through the inner region where dustdistribution interacts between with the the fiducial radii, so the ratio of particle 5.1. Some Implications of the Disk Model dustless simulation shown above) versus the contributors to OPD in the nuller, the telescope primary mirror and the DM, should be replaced. A more cross section (and thus optical(e.g., depth) Barucci to mass et surface al. 2008), density we may expect many additional rigid SiC primary would narrow the void between laboratory and space operation, giving near diffraction-limited known inner planet. Two possible orbitsvaries of the slightly inner with planet radius. The optical depths and mass densities star brightness (as in Bailey et al 2013). The species such as ices and organic solids.The derived In particular,surface-density it profile is for the small grain com- performance in both environments with a similar areal density. Higher quality production-run DMs are now are shown. The outer part of the sketch (of> the10 two AU) inner is rings not toin Figurenatural8 are calculated to expect by assuming water ice∆r to beponent present, (Figure especially8)hasanearlyconstantvaluefrom35to given star brightness is taken to be the peak pixel available from BMC. A drop-in replacement could be integrated into the nuller and run with the existing driver scale. Inset: The observed SED. The IRSvalues spectrum of 1 AU at (dots) 3 AU andthat 2 AU the at 20 source AU; they of would dust scale is a ”Kuiper110 AU, belt” consistent located with very dynamics far controlled by P–R radiation intensity of a smoothed, simulated bright inversely with ∆r. drag or its corpuscular wind analog (Section 5.3). In contrast, stems from dust in the inner region and exhibits a charac- from the star (∼ 55–90 AU), and the star itself has a late output. electronics. The nuller calibration system could also be revived by replacing the pinhole mount with a design teristic “plateau” (“shoulder”) at λ ≈ 20–30Theµ totalm.The inferred main mass in detected grains, that is, with sizes the large-grain component has surface density rising from 35 to that does not obscure the calibration reference beam. This would enable full nuller operation and an additional from a few microns to a fewspectral hundred microns, class.Accordingly, carrying most of we also90 AU. tried The homogeneous larger grains are lessmix- subject to drag forces, so their Subtraction of a noisy reference PSF part of the SED with a maximum at λ ≈ 70–80 µm, well 3 avenue for data post-processing. the radiating surface area, istures roughly of 5 astrosilicate10− M 0.4 withMLuna, 50%radial and distribution 70% volume may reflect fraction that of the underlying parent-body removes speckles and increases the effective probed by several photometry points (symbols90% of which with is in error the large (“icy”) grain× component⊕ ∼ of the sub- −3 of water ice (Li & Greenberg 1998)population. (ρd =1.2gcm ). The contrast (solid red line) and enables imaging bars), derives from the outer and intermediatemm ring. regions.The total bolometric fractional (Ld /L )of The void interior to the sub-mm−3 ring is inferred to be partly bulk density4 of these ice-silicate∗ mixtures is ρ =2.35 g cm all the disk components is 1.1 10− . − filled by two additional belts of particles with mean sizes the expected inner debris disk. and× ρ =1.89 g cm 3, respectively. The optical constants of the mixtures were calculated by effective medium theory Acknowledgements: with the Bruggeman mixing rule. In all three cases (pure (Mendillo et al. 2012b) 2. Model setup astrosilicate, two ice-silicate mixtures) the dust grains were PICTURE is funded by the National Aeronautics and Space Administration, Proc. of SPIE Vol. 8442 84420E-13 assumed to be compact spheres. Grant NNX13AD50G. Boston Micromachines Corporation provided DM surface New Primary Mirror New Deformable Mirror Downloaded From: http://proceedings.spiedigitallibrary.org/ on 12/14/2012 Terms of Use: http://spiedl.org/terms 2.1. Method measurement data used for this work. A new lightweight Silicon Carbide primary2.4. mirror Radiation is pressure Thanks to D. Hammerle and T. Bruno at AOA Xinetics for their help and Our model includes gravitational forces from the star and A new MEMS an inner planet,currently radiation being and manufactured stellar wind pressure, by Northrop drag Using Grumman the optical AOA constants Xinetics. and adopting Mie theory, assistance in refurbishing the PICTURE telescope. forces induced by both stellar photons and stellar wind par- the radiation pressure efficiency Q ,averagedoverthedeformable rp References: ticles (Burns et al. 1979), as well as collisions. Dust produc- stellar spectrum, was calculated as a function ofmirror size sfrom Boston tion in the outer region and dust transport through the in- Backman, D., M., et al. 2009..ApJ 690,2. (Burns et al. 1979; Gustafson 1994). We then computedMicromachines the is Bailey, V., et al. 2013. ApJ 767, 31. termediate region are modeled with a statistical collisional radiation pressure to gravity ratio, β (Fig. 2). The resulting code. To study the dust in the inner region, we need to han- being prepared for Mendillo, C. B et al. 2012a. Applied Optics 51, 29. β(s) was utilized to compute the direct radiation pressure Mendillo, C. B., et al. 2012b. In Proc. SPIE, 84420E–84420E. dle dust interactions with the inner planet. These cannot and Poynting-Robertson forces. integration. be treated by the collisional code, so we model the inner Perrin, et al. 2012. In Proc. SPIE, 8442:84423D–84423D–11. region by collisionless numerical integration. The figure at right Pickles, A. J. 1998. PASP, 110, 749. 2.5. Stellar wind shows the Reidemeister, et al. 2011. Astronomy & Astrophysics 527, A57. expected flattened ! 2.2. Stellar properties The stellar wind was included by a factor βsw/β, which is the ratio of stellar wind pressure to radiation pressure:surface figure. Example 20” SiC mirror, We assumed a stellar mass of M⋆ =0.83NewM⊙ primary mirror undergoing initial figuring. (Benedicthttp://www.northropgrumman.com et al. 2006) and a luminosity of L. =0.32L⊙ βsw Fsw M˙ ⋆vswc Qsw ⋆ = = , (1) (Di Folco et al. 2007). For the stellar spectrum we used a β Frp L⋆ Qrp NextGen model (Hauschildt et al. 1999) with an effective temperature of 5200K, log g =4.5, solar , and where Qsw is the efficiency factor for stellar wind pressure stellar radius R⋆ =0.735R⊙ (Di Folco et al. 2007). (Burns et al. 1979; Gustafson 1994). We adopted Qsw =1.

2 Printing Instructions

Completing Your Poster Design!

! 1. After you have completed your entire layout, delete all slides except for your poster.!

! Exporting Your Poster Template for Print!

! 1. Go to File / Print! 2. In Print Menu, select button PDF! 3. Select “Save as PDF”! 4. This will now create a PDF document on your Desktop or chosen area!

! Pre-Press Information!

! The PDF that Keynote will produce is at a reduced resolution and will need resizing by your printer to print as a full 4 feet by 3 feet poster. Have them resize the file in Adobe Illustrator to 7,200 pixels by 5,400 pixels in size before printing. This will ensure the best quality print for you to display.!

! The instructions below are pre-press instructions for your printer.!

! 1. Open PDF document in Adobe Illustrator! 2. Select desired page! 3. Select all text and objects! 4. Group all text and objects! 5. Resize grouped area to fit in 7,200 x 5,400 space