Microlensing Planets from the Ground and Space

1. Microlensing Planets from the Ground and Space David Bennett University of Notre Dame

2. The Physics of Microlensing • Foreground “lens” star + planet bend light of “source” star • Multiple distorted images • Total brightness change is observable • Sensitive to planetary mass • Low mass planet signals are rare – not weak • Peak sensitivity is at 2-3 AU: the Einstein ring radius, RE • 1st Discovery from Ground-based observations announced already

3. Target Field = Galactic Bulge A high density of source and lens stars is required for a high microlensing rate.

4. Lensed images at arcsec resolution A planet can be discovered when one of the lensed images approaches its projected position.

5. MOA/OGLE Planetary Event Best fit light curve simulated on an OGLE image

6. 1st Exoplanet Discovery by lensing The OGLE 2003-BLG-235/MOA 2003-BLG-53 light curve (Bond et al, 2004). The right hand panel shows a close-up of the region of the planetary caustic. The theoretical light curves shown are the best fit planetary microlensing light curve (solid black curve indicating a mass ratio of q = 0.0039), another planetary mass binary lens light curve (green curve with q = 0.0069), and the best fit non-planetary binary lens light curve (magenta dashed curve), which has q > 0.03.

7. 2nd Exoplanet Discovery by lensing OGLE 2005-BLG-71 (Udalski, Jaroszynski, et al - OGLE & FUN. Addl’ data from MOA & PLANET). Central caustic light curve perturbation (d = 1.3 or 1/1.3): Additional planet discoveries by PLANET, MOA & OGLE, also in preparation

8. Planetary Parameters from Microlensing • Mass ratio & planetary separation in Einstein radius units • Radial velocity planets only give mass ratio  sin(I) • But the properties of the source star are well known for radial velocities! • High resolution observations can reveal source star • Light curve fit gives source star brightness • HST observations may reveal a source apparently brighter than required by the fit - due to light from the lens • Pending HST DD proposal by Gould, Bennett & Udalski • Favorable case due to long timescale event and indications of blending in ground-based photometry • 30-50% of events have detectable sources • Future JWST or AO observations will confirm the lens star ID and determine the lens-source proper motion (~10 years later) • Measurement of microlensing parallax plus finite source effect gives planetary mass directly • Weak parallax detection for OGLE-235/MOA-53 gives mass between ~0.06 and ~0.7 M • MOA upgrade from 0.6m to 1.8m telescope should improve data for future events

9. Extend the search from Jovian to Terrestrial Planets: Escape ground-based confusion with space-based resolution • Systematic photometry errors for unresolved main sequence stars cannot be overcome with deeper exposures (i.e. a large ground-based telescope). • high magnification events reveal low mass planets at Einstein radius separation • High Resolution + large field + 24hr duty cycle => space mission • Microlensing Planet Finder

10. The Microlensing Planet Finder(MPF) D. Bennett, PI E. Cheng, Deputy PI GSFC Management • MPF scored highly in the 2004 Discovery Science review • MPF is complementary to Kepler, with sensitivity at larger separations • MPF + Kepler yield statistics on planets < 1 M at separations 0–∞ • MPF • MPF finds mass ratios and separations, plus masses for > 30% of planets. • MPF can see 100 smaller masses than current methods, and smaller masses than any other technique. • If the probability for Earth analogs is less than 10%, MPF may be the only way to detect them prior to TPF.

11. Planet Detection Sensitivity Comparison • Sensitivity to all Solar System-like planets • Except for Mercury & Pluto • most sensitive technique for a  1 AU • Good sensitivity to “outer” habitable zone (Mars-like orbits) where detection by TPF is easiest • Mass sensitivity is 1000  better than vrad • Assumes 12 detection threshold • Can find moons and free planets Updated from Bennett & Rhie (2002) ApJ 574, 985

12. Exoplanets via Gravitational Microlensing • Planetary signal strength independent of mass • if Mplanet/M* 310-7 • low-mass planet signals are brief and rare • ~10% photometric variations • required photometric accuracy demonstrated • Mplanet/M*, separation (w/ a factor of 2 accuracy) • Mplanet and M* measured separately in > 30% of cases • follow-up observations measure Mplanet , M*, separation for most G, K, and some M star lenses • finds free-floating planets, too

13. Simulated Planetary Light Curves • Planetary signals can be very strong • There are a variety of light curve features to indicate the planetary mass ratio and separation • Exposures every 10-15 minutes • The small deviation at day –42.75 is due to a moon of 1.6 lunar masses.

14. Technical Summary • 1.1 m TMA telescope, ~ 1.5 deg FoV, at room temperature, based on existing Kodak designs and test hardware • 35-70 2Kx2K HgCdTe and Si PIN detector chips at 140 K, based on JWST technology • 0.24 arcsec pixels, and focal plane guiding • 6  62 sec exposures per pointing • SIDECAR ASICs run detectors, based on JWST work • No shutter • 1% photometry required at I=21.5 • 28.7 inclined geosynchronous orbit • Continuous viewing of Galactic bulge target (except when Sun passes across it) • Alternating between 2 view directions in 15 minute cycle • Continuous data link, Ka band, 42 Mbits/sec • Launch Feb. 2012

15. W Sun Ecliptic Plane Equatorial Plane Orbit 23.5 W 28.7 Vernal Orbit Plane MPF field Equinox MPF in Geosynchronous Orbit

16. Focal Plane Concept • 35-70 2Kx2K detectors: 56 near IR HgCdTe, 14 visible Si PIN arrays from Rockwell • Divided into two banks of 35 each, separated by width of one bank • Sidecar ASIC – Reduces wire count, produces clock signals, provides 16-bit ADC’s, and digital signal processing (Fowler sampling) • One ASIC per 5 detectors • Each detector can watch a guide star in a sub-window while taking long exposures

17. Detector Sensitivity The spectrum of a typical reddened source star is compared to the QE curves of CCDs and Si-PIN detector arrays. The HgCdTe detectors developed for HST’s WFC3 instrument can detect twice as many photons as the most IR sensitive Si detectors (CCDs or CMOS). MPF will employ 56 HgCdTe and 14 Si-PIN detectors.

18. 2 Interleaved MPF Fields • MPF alternates between 2 fields every 7.5 minutes • Fields are oriented parallel to the Galactic plane to maximize the microlensing rate.

19. Operations Requirements • Stare at Galactic Bulge as long as possible • Avoid Sun passages for 3-4 months around Dec. 21 - point elsewhere • Toggle between 2 pointings on 15 min cycle • Occasionally (at least daily), subpixel raster (dither) to confirm photometry and get best angular resolution • Weekly, sweep visible detectors across field to get star colors • Occasional orbit maintenance • Commanding, health and safety monitoring (routine after checkout) • Data collection, archiving, processing into light curves • Rotate spacecraft 180 deg around LOS around June 21 • Potential Moon avoidance, Earth shadow actions

20. Science Data Products • ~ 4,000 light curves of candidate lens systems, with photometry errors limited by photon noise plus at most 0.3% systematic errors, in 2 colors • Interpreted light curves with models of star and planet masses, locations, and velocities • ~ 50,000 transit light curves • Archives of 100 million stellar light curves • Raw data for further study

21. Similar Designs for Planet Finding & Dark Energy MPF SNAP DESTINY Wide-FOV near-IR optimized telescopes: Joint Mission?

22. MPF Summary • Only MPF can complete the census of extra-solar planets in the Milky Way • Only MPF can determine the frequency of extra-solar planets like those in our own Solar System • MPF can be built and flown with current, well established technology • MPF should be selected in the 2005 Discovery Competition