Microlensing planet surveys:
the second generation
Dan Maoz Tel-Aviv University
with Yossi Shvartzvald, OGLE, MOA, microFUN
Conceived problems with microlensing:
1. Seems complicated…
2. and hence results suspect…
3. No “follow up” of planets possible
4. Statistically useless due to haphazard survey
strategies
5. Planet yield so small -- not worth trouble?
ë =
a
4G M
c 2R
R
( ü 1)
ë =
4G M
c 2R
( ü 1)
I
a
R
E
S
DLS
DOL
DOS
“microlensing”
(in our Galaxy):
In distant galaxies:
òE = 10 à 6 asec
“macrolensing”,
“galaxy lensing”:
cluster lensing:
òE ù 1 asec
òE ù 10 asec
ë =
4G M
c 2R
( ü 1)
I
a
R
E
S
DLS
DOL
DOS
ë =
I+
4G M
c 2R
( ü 1)
a
S
R
b
-
A
I-
DLS
DOL
DOS
+
I+S + SA = I+A
~milliarcsec
Magnification=image area / source area :
magnification ~ 1/ (impact parameter)
S. Gaudi
Einstein-ring crossing timescale:
t =E DOL / v ~ M1/2
For DOL=8 kpc,
v=20 km/s
t(1Msun) = 2 months
t(1MJ)=2 days
The first
microlensing
lightcurves
(LMC)
Alcock et al. 1993
Nowadays, ~1000 microlensing events/yr
detected toward Galactic bulge
Yee+ 09
Bond et al. 2004
Beaulieu et al. 2006
Udalski
et al.
2005
Gould et
al. 2006
Gaudi et al. 2008
“Jupiter”+”Saturn” system: 1+2+3+5=“Saturn”, 4=“Jupiter”
Our solar system:
1 Msun
5.2 AU
9.5 AU
1 Mjup
1 Msat
Msat/Mjup = 0.30
Rjup/Rsat = 0.55
OGLE-2006-BLG-109L,b,c:
Mc/Mb = 0.37
0.50 Msun
Rb / Rc
2.3 AU
4.6 AU
0.71 Mjup
0.90 Msat
= 0.50
Second 2-planet system discovered:
0.7MJ (4.6 AU) and 0.1MJ (3.8 AU)
Han+2012,
OGLE-2012-BLG-0026
Simulation by S. Gaudi
Simulation by S. Gaudi
q = Mp / Mhost
Simulation by S. Gaudi
Caustics:
points in the source plane which get infinite magnification.
For a point lens, caustic is a single point behind the lens.
(source there gets magnified into Einstein ring)
Caustic cusps
Magnification still ~ 1/(distance to caustic)
Source passage on or near central caustics:
high mag  almost full Einstein ring  ~100% detection
efficiency for planets near Einstein radius (lensing region).
planetary caustics: low mag  Lower planet detection
efficiency per event, but much more common.
A. Cassan
Microlensing probes a unique region of planetary
parameter space…
Gould et al.
2006, 2009
…near the Einstein radii of stars ~ their snow lines.
Gould et al.
2006, 2009
Snowline scaling
with mass:
R  M
Lens
star

S
Snowline-region planet frequency based on
microlensing discovery statistics:
Gould et al. (2010, based on 6 planets):
~1/3 of stars have snowline-region planets;
~1/6 of stars have solar-like planetary systems;
Cassan et al. (2012, based on 2 (!) planets):
~1/6 host jupiters
~1/2 host neptunes
~2/3 host super-earths
To date, only ~20 microlensing planets.
Why so few?
“1st Generation” survey strategy (Gould & Loeb 1992)
focused on bright, high-magnification (mag>100) events.
Udalski et al. 2005
Gould et al. 2006
Gaudi et al. 2008
1st Generation microlensing
OGLE, Chile, 1.3m
• low cadence (~ once a night)
MOA, NZ, 1.8m
1st Generation microlensing
1st Generation microlensing
~ 650 events/year
1st Generation Microlensing
Follow-up search for planetary perturbations with global
network on bright, high-magnification events:
High-magnification (mag >100) events are:
Good: ~100% sensitivity to planets projected near
Einstein radius,
+ high S/N light curves even with small and amateur
telescopes.
Bad: Rare events (~1%)  ~7 events/year  1-2
planets/year.
As opposed to high-mag (central caustic) events,
Low-magnification (planetary caustic) events:
Lower planet detection efficiency, but much more common:
Potential for tens of microlensing planets/year.
A. Cassan
Beaulieu et al. 2006
Need network of 1-2m class telescopes with degree-scale
imagers for continuous monitoring of many low-mag events in
search of planetary perturbations:
“Generation II microlensing”
Since 2011: A generation-II microlensing experiment:
Wise Obs., Israel, D=1m, 1 deg2
Yossi Shvartzvald is there
OGLE IV, Chile, D=1.3m, 1.4 deg2 MOA-II, NZ, D=1.8m, 2.3 deg2
The generation-II network
The generation-II network
Group
OGLE
The generation-II network
Group
OGLE
MOA
The generation-II network
Group
OGLE
MOA
WISE
8 deg2 of bulge with highest lensing rate covered quasicontinuously by all 3 telescopes, cadences 20-40 min
Gen II
2011 season: some typical low-mag event
light curves (no anomalies):
2011 Generation-II planetary events:
I-band (mag)
MOA-293
Yee, Shvartzvald et al. 2012
OGLE
MOA
Wise
I-band (mag)
Survey
data only:
All data:
HJD-2450000
2011 Generation-II planetary events
OGLE
MOA
Wise
What to expect from
Generation II?
a simulation:
Monte-Carlo of many SolarSystem-like planetary
systems, host star properties
matching those of bulge
microlensing population,
random inclinations.
Shvartzvald & Maoz 2012
• Various scalings of orbital
radius with host mass
R  M
Lens

S
S-
Shvartzvald & Maoz 2012
Ray trace through systems……
…add real sampling sequences,
photometry errors…
…search for planetary-type anomalies with same detection criteria as real data
Simulation results: can detect
~10-20% of planets around
microlensed stars;
~100 stars in Gen-II footprint, so
(10 to 20)*f planets per season.
S-
Shvartzvald & Maoz 2012
Conceived problems with microlensing:
1. Seems complicated…
but calculable. An elegant geometric method.
2. Light curve complexity
 uniqueness of models
3. No “follow up” possible
Not quite valid/true.
4. Planet yield so small -- not worth trouble?
Untrue! Unique probe of normal planetary
systems near snow line, beyond Solar
neighborhood, free-floating planets, yield
growing thx to Generation II (plus controlled
experiment)
Some calendar numerology:
Today, 18 Dec 2012 is:
12 / 2 days since 12.12.12 (just married?);
3 days until 21.12.12 (end of the world);
24,377 days since May 2, 1946
=0.66667 century
Happy 2/3 centennial Birthday, Tsevi !!