SuperMacho A Next Generation LMC Microlensing Survey

1
SuperMacho- A Next Generation LMC Microlensing
Survey

• K. Cook, S. Nikolaev (LLNL), S. C. Keller, B. P.
  Schmidt (ANU)
• Rest, A. (arest_at_ctio.noao.edu), R. Chris Smith,
  C. Aguilera, K. Olsen, J. L. Prieto (NOAO/CTIO),
  C. Stubbs (Harvard University), A. Becker, R.
  Covarrubias, A. Miceli, G. Miknaitis (U.
  Washington), D. Welch (McMaster), A. Clocchiatti
  , D. Minniti (Pontifica Universidad Catolica)
• The SuperMacho Collaboration

2
Macho Project sought evidence for baryonic dark
matter in the Milky Way Halo

• Macho Project sought evidence for MAssive Compact
  Halo Objects by surveying for the gravitational
  effect of the objects on light microlensing
• Macho Results
• first microlensing detection
• no low mass (lt0.1 Msun) lenses
• the Milky Way is a barred spiral
• 17 events detected toward the Large Magellanic
  Cloud (LMC)
• too many for known populations
• less than 100 of halo
• What and Where are lenses?

MACHO found 20 of halo in 0.5 Msun objects
3
2 Major Dark Matter Results Common to MACHO,
EROS, OGLE
Lack of LMC events of less than 20 days duration
rules out low mass MACHOs
Rate of detected events exceeds that expected
from known stellar backgrounds, and corresponds
to a MACHO fraction of between 8 and 50 of the
standard halo
(Alcock et al ApJ 542, 281 2000)
4
A Next Generation Microlensing Survey was Needed

• Questions which need to be answered
• Where are lenses?
• Halo
• LMC
• Milky Way
• Unknown population/object
• What are lenses?
• Mass
• Composition (baryonic?)

5
Characterizing Microlensing Populations

• Spatial Distribution
• Distribution on the sky
• statistically significant number needed, too few
  from Macho
• Location along the line of sight
• The LMC is dusty
• Can test where source stars are with respect to
  dust weak result

6
Characterizing Microlensing Populations

• Spatial Distribution
• Distribution on the sky
• statistically significant number needed, too few
  from Macho
• Location along the line of sight
• The LMC is dusty
• Can test where source stars are with respect to
  dust weak result

7
Characterization of a Lens for an Exotic Event

• Macho LMC5 event
• Faint source blended with red star
• Foreground dwarf?
• Parallax detected
• Hubble Space Telescope image shows red star in
  direction predicted by fit
• LMC5 system completely solved!
• 0.1 Msun, M4-5 dwarf at 650pc

Alcock et al (2001) Nature
First direct mass measurement of an isolated
system
8
Searching for the Lensing Population

• Direct Detection
• MACHO results suggest white dwarfs (WDs)
• WDs intrinsically faint
• Halo MACHO WDs would be rare with high proper
  motion
• New WD Surveys
• Oppenheimer et al. (2001)
• 2 of halo
• Nelson et al. (2002)
• 7 of halo
• More than expected, but not enough

Nelson et al (2002) likelihood analysis of 5
high proper motion white dwarfs discovered with
the Hubble Space Telescope
9
More Microlensing Events are needed!

• Spatial distribution of population needs
  statistically significant number of events
• Non-standard events needed to characterize
  lenses- binary
• Discriminating lens population

10
Features of a Next Generation Microlensing Survey

• The figure of merit (FOM) for a microlensing
  survey depends upon the number of sources which
  can be monitored in a given amount of time.
  Presented below is a table comparing the original
  MACHO survey with a survey using the Cerro Tololo
  Inter-american Observatory (CTIO) 4-m and prime
  focus mosaic CCD camera.
• MACHO vs. 4 - m MOSAIC

A CTIO survey should detect an order of magnitude
more events
11
A Next Generation Microlensing SurveySuperMacho

• SuperMacho gained NOAO survey status awarded 150
  half nights on CTIO 4m
• NOAO supporting project with hardware, technical
  and scientific staff
• Big aperture (4m)
• Mosaic imager 8kx8k, 0.35 º
• Dark sky and good seeing
• Custom broadband filter (5200A to 7300A)
• 60 fields, 23 sq deg.
• Exposures optimized to maximize of stars
• S/N 0.1 mag at 23rd
• Goal is 50 well-characterized events

12
Secondary Objectives

• Variable stars (instability strip _at_ main sequence
  in LMC)
• Solar System objects at ecliptic pole
• High Proper motion objects
• Supernovae behind LMC
• LMC proper motion w.r.t. quasars
• Develop software for Large-aperture Synoptic
  Survey Telescope
• Planetary transits (OGLE object, Konacki et al
  astro-ph/0301052

13
(No Transcript)
14
Discriminating self-lensing vs screen-lensingvia
spatial event distribution

• Red screen-lensing
• Black self-lensing,

Differential approach reduces sensitivity to LF,
efficiencies, etc
15
Discriminating self-lensing vs screen-lensingDif
ferential rate exclusion plot, lower limit
16
Discriminating self-lensing vs screen-lensingDif
ferential rate exclusion plot, upper limit

• Nml(set 5)0
• Screen-lensing can be excluded if Nml(total)gt42

17
How we do it

• Reduce a lot of data in a short time

18
SuperMacho Reduction Pipeline
Make Flats and Bias Frames Cross-talk correct
and WCS register Carve MultiFits 8K x 8K MOSAIC
image into 1K x 4K amplifier images Flatten
Images DoPhot reduction to obtain PSF and noise
array Determine appropriate template by field
and filter Register, convolve and subtract
images Fixed-PSF, bipolar DoPhot on difference
image Filter detections to reject artifacts,
cosmic rays, etc. Upload observation and its
detections to database SN microlensing fits
with visual interface to high-confidence
detections for interactive classification by
astronomers
Modular Flexible Directory-based
19
Parallelization

• Initial steps sub-units not independent
• Crosstalk instrumental interdependencies
• WCS mapping need enough stars
• Produces 16 independent units
• Remaining data reduction and transient analysis
  done in parallel
• Use Condor for load balancing
• 18 CPUs available, 16 units to be processed
• Robust to CPU problems
• Scaleable

20
Hardware

• Computers
• Linux PCs running RedHat 7.x
• 10 x dual 1.2Ghz Athlon CPU processing cluster
• 2 x dual 2.2Ghz Intel CPU storage servers
• Storage
• 1TB SCSI RAID array for mission-critical
  transient detection system
• 2TB/yr IDE RAID array for review and online
  storage
• Infrastructure
• 77Mbps link from Tololo to La Serena
• 1Gbps link between all components in La Serena
• 10Mbps link to U.S.

21
Pipeline Layout
22
SuperMacho Status Pipeline

• Present version is robust, flexible and
  efficient.
• Real-time detection
• Data pipeline is treated as an LSST prototype
• Current focus dB, automatic classification,
  fitting, and visualization of events

23
False Positive Filtering

• Life is not always perfect
• Example near center of LMC bar

24
Database

• Using PostgreSQL fast, relational, open-source
• Combined Operations and Science DB
• Operations
• Observations, processing, and templates used
• Transient analysis
• Detections in any given observation
• Detection based, not object based
• Detections to Objects OPTICS clustering
  algorithm
• Additional tables
• Classification (SN, variable star, moving object)
• Spectroscopic information (IDs, etc.)
• Relations to external DBs (previously known
  variables)

25
Data Products

• Transient announcements
• Email list (bidirectional!) IAU Circulars
• Finder charts, fits and spectra on web
• Images
• Raw Reduced (flat fielded) data
• No proprietary period

26
What have we found?

• Lots of variables
• Lots of Supernovae 10/month
• Lots of Microlensing 11/month

27
SN 1987a
28
2003-LMC-1 Microlensing chi2 2.66               
        T0 52980.58                       That
52.16                       U_min
1.51                       Template flux
90825.11                       Supernova chi2
4.71                       Supernova dilation
3.11 Stellar spectrum
29
2003-LMC-58 Microlensing chi2 5.94 T0
52984.88 That 219.02 U_min 1.07 Template flux
163983.09                       Supernova chi2
37.36                      Supernova dilation
4.00 Stellar spectrum      
30
2003-LMC-80 Microlensing chi2
2.10                       T0 52999.31            
           That 77.38                      U_mi
n 0.23                      Template flux
336.96                      Supernova chi2
0.64                      Supernova dilation
2.07 SN spectrum

31
2003-LMC-11Microlensing chi2
44.71                      T0 52956.93           
           That 322.80                     
U_min 0.06                     Template flux
648.84                     Supernova chi2
1.94                     Supernova dilation
1.22 ?? spetrum
32
Status

• Follow-up photometry (Boyden)
• Spectroscopy (contaminant SN)
• Automatic fitting, classification alert
• Alert web page
• http//darkstar.astro.washington.edu
• Images publicly available within days

33
Future Tasks

• More follow-up
• Efficiency Calculations
• Database Search Tools
• Deep variable catalogs