microlensing in wide-field surveys

Transcript

1. Identifying Microlenses in Wide-field,
   Non-uniformly Sampled Surveys
   Adrian Price-Whelan
   Marcel Agüeros
   Amanda Fournier (UCSB)
   Rachel Street (LCOGT)
   May 2012
   1
   Tuesday, May 29, 12
   

   View Slide

2. May 2012
   Will wide-field, time-domain surveys
   aid microlensing event discovery?
   2
   Tuesday, May 29, 12
   

   View Slide

3. What is microlensing, and why is it
   interesting?
   Can we model the expected microlensing
   signal in synoptic surveys?
   How should we identify microlensing
   events in single-color, photometric data?
   May 2012
   3
   Tuesday, May 29, 12
   

   View Slide

4. What is microlensing, and why is it
   interesting?
   May 2012
   4
   Tuesday, May 29, 12
   

   View Slide

5. May 2012
   ✓E =
   q
   GM
   c2
   DS DL
   DS DL
   =
   q
   GM
   c2
   Drel
   5
   Tuesday, May 29, 12
   

   View Slide

6. Microlensing
   May 2012
   ✓E
   6
   Tuesday, May 29, 12
   

   View Slide

7. Microlensing
   May 2012
   ✓E
   tE =
   ✓E
   µrel
   /
   p
   M
   p
   Drel
   6
   Tuesday, May 29, 12
   

   View Slide

8. 7
   Tuesday, May 29, 12
   

   View Slide

9. Why is it interesting?
   Exoplanets
   Galactic structure
   Stellar abundances
   Compact objects / dwarfs
   May 2012
   8
   Tuesday, May 29, 12
   

   View Slide

10. What is microlensing, and why is it
    interesting?
    Can we model the expected microlensing
    signal in synoptic surveys?
    How should we identify microlensing
    events in single-color, photometric data?
    May 2012
    9
    Tuesday, May 29, 12
    

    View Slide

11. Modelling
    To predict the microlensing signal, we need a
    Galactic model and some measure of the
    detection efficiency
    May 2012
    10
    Tuesday, May 29, 12
    

    View Slide

12. Galactic Model
    Lensing rate / nSD2
    S
    nLD2
    L
    2✓E µrel
    Pre-2008 (Peale et al. 1998):
    Uniform densities, uniform velocity
    distribution
    Han et al. 2008:
    More realistic model with disk-like
    distributions
    11
    Tuesday, May 29, 12
    

    View Slide

13. Galactic Model
    Our model (Amanda Fournier):
    - Galactic disk (no bulge)
    - Main sequence sources (0.07 to 12 )
    - Scale heights dependent on age
    - SFD dust maps
    - Flat rotation velocity + 3D dispersion
    M
    12
    Tuesday, May 29, 12
    

    View Slide

14. Galactic Model
    Next steps:
    - Apply to a single cluster (Rup 147)
    - Update velocity distribution
    - Add bulge
    - Add dwarfs / stellar remnants
    13
    Tuesday, May 29, 12
    

    View Slide

15. May 2012
    Will wide-field, time-domain surveys
    aid microlensing event discovery?
    14
    Tuesday, May 29, 12
    

    View Slide

16. Detection Efficiency
    OGLE, MACHO, MOA (1990’s - present)
    Limited coverage: LMC, SMC, Bulge
    PTF, Pan-STARRs, LSST (present / future)
    Wide-field cameras
    Extremely irregular sampling
    May 2012
    15
    Tuesday, May 29, 12
    

    View Slide

17. Keck NIR WFC3
    OGLE IV
    May 2012
    16
    Tuesday, May 29, 12
    

    View Slide

18. May 2012
    17
    Tuesday, May 29, 12
    

    View Slide

19. May 2012
    18
    Tuesday, May 29, 12
    

    View Slide

20. Detection Efficiency
    Depends on survey sampling
    baseline
    photometric errors
    detection algorithm
    Previous studies used matched filters
    We aim to combine this with statistical
    variability indices
    May 2012
    19
    Tuesday, May 29, 12
    

    View Slide

21. What is microlensing, and why is it
    interesting?
    Can we model the expected microlensing
    signal in synoptic surveys?
    How should we identify microlensing
    events in single-color, photometric data?
    May 2012
    20
    Tuesday, May 29, 12
    

    View Slide

22. Variability Indices
    σ / μ : root variance / mean magnitude
    η : mean square successive difference /
    sample variance
    J : large for potentially variable sources
    K : kurtosis
    Δχ2 : difference in chi-square for Gaussian
    vs. straight line fit
    May 2012
    21
    Tuesday, May 29, 12
    

    View Slide

23. May 2012
    22
    Tuesday, May 29, 12
    

    View Slide

24. 23
    Tuesday, May 29, 12
    

    View Slide

25. 24
    Tuesday, May 29, 12
    

    View Slide

26. May 2012
    25
    Tuesday, May 29, 12
    

    View Slide

27. Issues: Blending
    May 2012
    26
    Tuesday, May 29, 12
    

    View Slide

28. Issues: Blending
    May 2012
    26
    Tuesday, May 29, 12
    

    View Slide

29. Issues: Blending
    May 2012
    26
    Tuesday, May 29, 12
    

    View Slide

30. Issues: Pipeline
    May 2012
    27
    Tuesday, May 29, 12
    

    View Slide

31. Future
    Develop selection criteria on all variability
    indices to pick out microlensing events
    Run Amanda’s code for a single cluster,
    compare to observations
    Predict microlensing signal in PTF, then run
    our code on the entire archive
    Real-time identifier?
    LSST?
    May 2012
    28
    Tuesday, May 29, 12
    

    View Slide