1. OB390 and the new microlensing planets Christian Coutures Eso Santiago September 2006

2. Planet searches

3. Overview Planets search with microlensing PLANET network OGLE 2005-BLG-390Lb and others

4. In the beginning…

5. Gravitational Lens

6. Home gravitational lens

7. Lens with deflecting angle proportional to 1/u Liebes S., 1969 American Journal of Physics 37, 103-104 Courtesy Phil Yock, Auckland

13. Relative proper motion Source-Lens. S Major image Minor image L

15. Wineglass einstein ring

16. ‘PSPL’ light curveImpact parameter Light Curve

17. Exotics  Parallax: earth movement  Xallarap: mouvement of the (binary) source  Finite size: impact parameter < Rsource (~1  arcsec)  Binary lens (caustic) Distorsions :

18. SINGLE LENS WITH FINITE SOURCE EFFECTS 3 source sizes : R S = 0.05, 0.1 and 0.2 R E (t-t 0 )/t E OB-05-254

19. DOUBLE LENS Different trajectoriesVery different light curves ! M1M2 Caustic: locus of source positions where amplification is ∞ (in the plane of the lens)

20. Wineglass planet

21. Screen light source Camera Inverse ray shooting, caustics

27. Close/wide q=M 1 / M 2 =1

28. Three configurations: - close, intermediate, wide Caustics: - 1 central - 0-2 planetary Séparate regimes (Erdl & Schneider 1993) planetary caustics

29. Jupiter: d - 1/d 1/30 è Jupiter: dégénérescence q Central Caustic degeneracies

30. A planetary companion

31. source star finite size Géante du clump: Turn-off M.S.:

32. Lensing zone

33. Detection probability

34. 1995 - 1998 THE HOLLYWOOD APPROACH “Follow the big stars !” Gould et Loeb 1992 : « Planets in a solar-like system positioned half-way to the Galactic center should leave a noticeable signature (magnification larger than 5 percent) on the light curve of a gravitationally lensed bulge star in about 20 percent of the microlensing events. » Griest & Safizadeh 1998 « We show that by focusing on high-magnification events, the probability of detecting planets of Jupiter mass or greater in the lensing zone [(0.6 1.6) R E ] is nearly 100%, with the probability remaining high down to Saturn masses and substantial even at 10 Earth masses. » Two approaches : Big survey telescope, and network of follow up telescopes (NASA, …) Use existing telescopes, and alerts from surveys (PLANET, GMAN, MPS) - Network of telescopes to do monitoring of on going alerts 24/24. - Online analysis to detect anomalies real time. - Monitoring of Bulge giants (brighter !).

35. PLANET collaboration : Probing Lensing Anomaly NETwork http://planet.iap.fr M. D. Albrow, J.P. Beaulieu, D. Bennett, S. Brillant, J. A. R. Caldwell, H. Calitz, A Cassan, K. Cook, C. Coutures, M. Dominik, J. Donatowicz, D. Dominis, P. Fouqué, J. Greenhill, K. Hill, M. Hoffman, K. Horne, U. Jorgensen, S. Kane, D. Kubas, R. Martin, J. Menzies, P. Meintjes, K. R. Pollard, K. C. Sahu, J. Wambsganss, A. Williams Institut d'Astrophysique de Paris, INSU CNRS, Paris, France Univ. of Canterbury, Dept. of Physics & Astronomy, Christchurch, New Zealand South African Astronomical Observatory, South Africa Boyden Observatory, Bloemfountein, South Africa Canopus observatory, Univ. of Tasmania, Hobart, Australia Niels Bohr Institute, Copenhagen, Denmark Univ. of Potsdam, Potsdam, Germany Space Telescope Science Institute, Baltimore, U.S.A. Perth Observatory, Perth, Australia Boyden 1.5m

36. PLANET Probing Lensing Anomaly NETwork 1995-200? 32 collaborators, 18 institutes, 10 countries 5 telescopes Coopération with RoboNET ( 3 robotic UK telescopes)

37. SITES PLANET/RoboNet ESO Danish 1.54m 2003-2006+ Sutherland, SAAO 1m 1996-2006+ Boyden, 1.5m, CCD 2006+ Perth 0.6m 1996-2006+ Hobart 1m, 1996-2006+ RoboNet/Liverpool 2m, Canary 2005- RoboNet Faulkes North 2m, Hawaii 2005- RoboNet/Faulkes South 2m, Australia 2006 ? Goals : - 1 % photometry, - Sampling 1 point/hour- - Online analysis. Boyden 1.5m

38. Followed Alerts 1993: MACHO (†1999), OGLE, EROS (†2003) 1995: MOA Followed Alerts

39. PLANET DATA PROCESSING At each site : - Relative photometry for all stars real time -“Planet Plotter”: real time fit & display Data from all sites are uploaded to Paris: Every day, homebase checks : data, light curve fits, BAP, StAndrews priorities algo, Choose strategy, sampling, … Alert the community if anomalies SAAO Boyden ESO LS Hobart Perth RoboNet

40. First successfull online prediction of anomalous event, rotating binary with parallax. 9 updates of PLANET alert.

41. time scale 60 days, u0=0.128, mass ratio 0.53, d0=0.718 Rotation parameters : Effective motion (-0.062 +0.052), PiE=0.14 Dominik et al., 2006 in preparation

42. Other planets hunters GMAN (Global Microlensing Alert Network): 1995 – 1999, CTIO 0.9m MPS (Microlensing Planet Search): 1997 - 1999, 1.9m Mount Stromlo MicroFUN (Microlensing Follow-Up Network): 2003, CTIO 1.3m, CTIO-Yale 1.0m, Wise 1.0m, Palomar 1.5m, MDM 2.4m, LOAO 1.0m + 2 t é lescopes amateur RoboNET: 2004, Faulkes-N 2m, Liverpool 2m, (Faulkes-S 2m) OGLE-III EEWS: 2003

43. HOW about Planets ?

44. MB9947 At first sight, it looks like a planet… But it is a binary 

45. THE PLANET THAT NEVER WAS ? Jupiter at 4 AUs ??? A cool model A variable source ? Investigation on the way

46. MicroFUN PLANET 262-2003

47. THE ZOO OF PLANET DATA ANOMALIES PLANET 1995 — 2005 About 40 high magnification events sensitive to planets Identified planets: 1 (+1) Multiple anomalies: 3 (2 rotating binary lenses, 1 binary- lens/binary-source or triple lens) Finite source, single lens: 4 Caustic passages of finite source: 28 (of which 4 over cusps) -- measurements of proper motion, stellar brightness profile Weak, suspected, unclear anomalies: 27 (this pool harbours potential and likely contains some other jewels)

48. FINALLY A PLANET DETECTED ! MOA 2003-BLG-53/OGLE 2003-BLG-235 Best fit lens distance = 5.2 kpc 90% c.l. range is 2.3-5.4 kpc Best fit separation = 3.0 AU 90% c.l. range is 1.3-3.1 AU Best fit stellar mass = 0.36 M  90% c.l. range is 0.08-0.39 M  Best fit planet mass = 1.5 M jup 90% c.l. range is 0.3-1.6 M jup If lens star is a 0.6 M  white dwarf Dlens = 6.1 kpc ap = 1.8 AU Mp = 2.5 M jup Bond et al., 2004, ApJ 606, L155

49. 2003: MOA 2003-BLG-53/OGLE 2003-BLG-235 BEST FIT (Bennett et al. 2006): Lens distance = 5.8 kpc Proj. separation = 3.6 AU Stellar mass = 0.63 M sun Planet mass = 2.6 M JUP HST images discard that the lens is a 0.6 M sun white dwarf Bond et al., 2004, ApJ 606, L155

51. 2005 (1): OGLE 2005-BLG-071 Two models: close binary : d=0.758 q=6.7 10 -3 wide binary : d=1.294 q=7.1 10 -3 M * ~ 0.13 M sun D OL = 2.9 ± 1.2 kpc 0.9 M JUP a ~ 2.3 AU P ~ 10 yr Wide slightly favoured:

52. OGLE Field HST Image E N 1” Subo Dong ©

53. 2005 (2): OGLE 2005-BL53G-390

54. 390 story (1) OGLE alerts on july 11, 2005 PLANET observes from july 25

55. 390 story(2) Maximum on july 31st : A = 2.9

56. 390 story(3) another boring event …

57. 390 story(4) Should we drop it?

58. 390 story(5) ? August 10

59. 390 story(6)  ???

60. 390 story(7) Aug. 11th, OGLE confirms !

61. 390 story(8) Yes !

62. 390 story(9) Yessssssssss ! indeed

63. 390 story(10) YESSSSSSSSSS!

64. A planet ? A binary lens ? A bump in the night…

65. OGLE-2005-BLG-390 Coopération : PLANET/RoboNET, OGLE, MOA-II

68. Data in the anomaly from : PLANET-D154, OGLE, MOA-II, PLANET-Perth Data outside the anomaly from : PLANET/Robonet, PLANET-Hobart (Gould & Loeb 1992, Bennett & Rhie 1996) A TEXTBOOK MICROLENSING EVENT

69. PROBABILITY DENSITIES OF THE STAR AND ITS PLANET

70. Final Parameters Source distance D S = 1.05 ± 0.25 R GC (R GC =7.62 ± 0.32 kpc) Source star: a Galactic Bulge giant G4 III, R S = 9.6 ± 1.3 R SUN Primary lens mass: Primary lens distance: D L = 6.6 ± 1.0 kpc Planet mass: Planet separation: Orbital period:

71. Planet temperature and radius Star properties(MS): Planet properties: Model L é ger et al. (2004):

72. Detecting the lens 390L

73. contrast: ~4000 en J 5 ans? 35 mas

74. 2005 (3): OGLE 2005-BLG-169 New planet of 13 Earth masses detected by microFUN et OGLE in a very high amplification event (800, i.e. 7.3 mag) Source is a dwarf, moving ~ 8.6 mas/year Close-wide ambiguity (central caustic).

75. Comparaison fit PSPL et ESBL

76. Comparing the 4 planets (1) Event qb caustic Max Amp. source OB03235 = KB03053 0.00391.120unique7.6MS OB05071 close0.00670.758central44MS OB05071 wide0.00711.294unique d w =1.297 42MS OB05169 close0.0000830.9819unique d c =0.969 800MS OB05169 wide0.0000861.0198unique d w =1.067 806MS OB053900.00007571.6097planétary2.9RG

77. Comparing the 4 planets (2) t E days u0REu0RE  *  as  E mas  mas yr -1 D L kpc M * M sun M p M earth a AU K m s -1 OB0323561.50.1330.530.553.35.80.638304.344 OB05071 wide 70.90.02360.48 2.52.90.133002.357 OB05169 wide 42.30.00120.441.008.62.70.49133.41.1 OB0539011.00.3595.250.206.86.50.225.52.70.65

78. Summary Exoplanets

79. The core accretion model

80. Events caracteristic times Do giant planets prefer long lasting events (i.e. more massive lenses)?

81. PLANET detection efficiency

82. 2006 Season May1 - Aug.31 : 7 telescopes 96 alerts followed 12 high amplification, 16 medium (10-40) 5 stellar binairies, 9 anomalous 2 with variable source 52 « PSPL » Promising events: 207, 238, 245, 265 OGLE 2006-BLG-245: The planetary model has

83. 2006 (1): OGLE 2006-BLG-245

84. Conclusion(s) µlensing still explores ranges complementary to other methods (R.V.) It is able to find light planets around 1 - 4 AU Much less easy than it looked: Detection ≠ Caracterization Photometry, sampling, fitting model(s), ambiguities 4 detections with caracterisation 3 central caustics: 2 with high amplification, 1 with low amplification 1 planetary caustic, with low amplification => High Amp events does not work: no simple single strategy! 2 Jupiters, 1 x 13.Earth, 1 x 5.Earth: Good indication that little planets may be more common than gaz giants around M stars (30% vs 0.6%). This confirms the core accretion model. 3 long events, 1 short: Giant planets may be more common around bigger mass stars Imprecision on the star-planet parameters: Better to wait and try to detect the lens …

85. END

90. Finding the source size Fitting the curve: We can estimate the apparentangular radius of the source star from measured derredened mags and colours of the source in a CMD: Using the red giant clump in the CMD: position calibrated with the local red giant from Hipparcos: Transform for the distance from the clump, as a function of (l,b) because of the Galactic bar: