1. Confronting Galactic structure with microlensing in an era of large datasets Supachai Awiphan Jodrell Bank Centre for Astrophysics, School of Physics and Astronomy, The University of Manchester, Manchester, United Kingdom In collaborated with Eamonn Kerins and Annie C. Robin Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 1

2. Galactic structure with microlensing Galactic structure with microlensing Microlensing surveys toward the Galactic bulge have provided useful information for the study of Galactic structure. Several microlensing surveys have monitored a large number of stars and detected thousands of events over the bulge. e.g. MOA (Sumi et al. 2013) and OGLE (Wyrzykowski et al. 2015) Detailed theoretical models have been developed in order to predict the microlesning properties toward the bulge. e.g. Han & Gould (2003), Wood & Mao (2005) and Kerins et al. (2009) Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 2

3. Galactic structure with microlensing Galactic structure with microlensing Kerins et al. (2009) Synthetic maps of microlensing properties over the Galactic bulge using Besancon Galactic model. Map for sources I<19 Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 3 Penny et al. (2013) Using Besancon model Optical depths of red clump giants provide 1.8 times lower than observational measurements

4. Besancon Galactic model Besancon Galactic model The Besancon model is a Galactic population synthesis model. Stars are created from gas following IMF and SFR and evolved according to theoretical stellar evolutionary tracks. The model includes a 3D extinction map with 15 arcmin resolution (Marshall et al. 2006). Following work uses a later version of Besancon model (Robin et al. 2014). Kerins et al. (2009) use Besancon version (Robin et al. 2003 + 3D extinction) Penny et al. (2013) use Besancon version (Robin et al. 2012) The model separates stars into 4 populations: Thin disk Bulge/bar Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 4 Thick disk Halo

5. Manchester-Besancon Microlensing Simulator Manchester-Besancon Microlensing Simulator Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 5 http://www.mablus.net Microlensing properties Optical depth Average Einstein radius crossing time Microlensing event rate Filter (Johnson-Cousins) U-band to L-band (9 bands) Magnitude 12 to 23 Area -10.125 < b < 9.875 -9.875 < l < 10.125 Method Resolved or DIA method Lens population Disk, Bulge and All (Disk and Bulge)

6. Average Einstein radius crossing time Microlensing event rate per star Manchester-Besancon Microlensing Simulator Manchester-Besancon Microlensing Simulator Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 6 Maps of all lens population at I<19 Optical depth Microlensing event rate per sky area

7. Manchester-Besancon Microlensing Simulator Manchester-Besancon Microlensing Simulator Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 7 Optical depth maps of all lens population at I<19 Properties mapError map

8. Manchester-Besancon Microlensing Simulator Manchester-Besancon Microlensing Simulator Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 8 Clickable map to obtain event rate histogram of Einstein crossing time or magnitude within 15x15 sq. arcmin Properties mapEvent rate histogram

9. Besancon microlesning of MOA-II Besancon microlesning of MOA-II Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 9 Simulated the MOA-II survey data (Sumi et al. 2013). The simulation catalogues has the same overall areal coverage as the MOA-II survey. The model shows a deficit of short time-scale events (<10 d) and an excess of 10-30 d events which may because by the lack of low-mass stars and brown dwarfs in the model.

10. Besancon microlesning of MOA-II Besancon microlesning of MOA-II Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 10 Add in a brown dwarf with stellar mass function and same kinetic parameters as the original catalogue at the H-burning limit and extended down to 0.001 M sun. The added populations are used for the lens stars only. MF slope of α BD = −0.4 provides the best reduced chi-square value =2.2. Adding low-mass stars and brown dwarfs provides a better match to the MOA-II time-scale. Total mass is increased about 10%.

11. Besancon microlesning of MOA-II Besancon microlesning of MOA-II Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 11 Optical depth The model predicts a significantly lower optical depth compared with Kerins et al. (2009) due to the lower mass of the Galactic bulge. Compatible with the Penny et al. (2013) result which also uses a more recent version.

12. Besancon microlesning of MOA-II Besancon microlesning of MOA-II Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 12 Average event duration Map of the average event duration shows shorter time-scales compared to Kerins et al. (2009) and Penny et al. (2013), due to the addition of low-mass star and brown dwarf lenses. The time scale (21 days) compatibles with MOA-II time scale (All sources: 24.0 days, RGC sources: 19.2 days).

13. Besancon microlesning of MOA-II Besancon microlesning of MOA-II Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 13 The result in each subfield are binned to 0.5 o in Galactic latitude, in similar fashion to the MOA-II survey. For b < 1.8 o, the optical depths decrease due to the high column density of dust. Over lower latitude regions (b < 3 o ), the Besancon DIA optical depth is lower than the MOA-II all-source optical depth by a factor of 2.

14. Besancon microlesning of MOA-II Besancon microlesning of MOA-II Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 14 The Besancon model underestimates (50%) the optical depth compared with the MOA-II data at closer to the Galactic Centre (b < 3 o ) The bulge mass would need to be increased by a factor 2.6 in order to match the overall optical depth distribution. Some recent studies such as Portail et al. (2015) provides more massive bulge mass than Besancon model. The dust map model is likely to be underestimated in the innermost regions due to incompleteness of 2MASS star counts below K=12

15. Besancon microlesning of MOA-II Besancon microlesning of MOA-II Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 15 The MOA-II team parameterize the observed spatial microlensing distribution using a polynomial function. The same function is used to model the Besancon microlensing maps.

16. Conclusion Conclusion Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 16 Besancon model (Robin et al. 2014) is used to calculate microlensing properties. Manchester Besancon Microlensing Simulators http://www.mablus.net Online microlensing simulator to simulate microlensing properties maps (with error maps) toward the bulge. Field-by-field comparison between MOA-II and the Besancon Galactic model Model provides only ∼ 50 per cent of the measured optical depth and event rate per star at low Galactic latitude around the inner bulge (|b| < 3◦) The additional population, along with increased extinction in this region should permit an increased optical depth without violating star count limits. Future works Real time prior - likelihood prediction for future microlensing event. MOA-2011-BLG-262: Sub-Earth-Mass Moon or High velocity planetary system.

17. Thank you for your kind attention Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 17

18. Microlensing sources Microlensing sources Resolved source All sources which are brighter than magnitude limit. DIA source Include fainter sources which may only be detectable during lensing. Less sensitive to blending systematics within crowded fields. Potentially provides a better S/N ratio measurement due to the larger available sample size. The rate of detectable events scales as u (impact parameter). The optical depth weights by The rate weights by Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 18

19. Besancon Galactic model Besancon Galactic model Thin disc Major component in the Galactic central region. Assumed to have an age of 10 Gyr with constant SFR. IMF with two slopes, dN/dm ∝ M −1.6 for M 1M sun The total mass of the thin disc is 9.3 × 10 9 M sun. Modelled with a central hole and the maximum density is located at about 2.5 kpc from the Galactic Centre. The kinematics follow the Hipparcos empirical estimates of Gomez et al. (1997). The populations are divided into seven distinct components with different distribution in age, scale height and velocities (Robin et al. 2012). Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 19

20. Besancon Galactic model Besancon Galactic model Thick disc Lower density than the thin disc. Becomes important at Galactic latitudes above about 8 o -10 o. Recent constraints from SDSS and 2MASS data (Robin et al. 2014) Single thick disc modelled by a 12 Gyr isochrone of metallicity −0.78 dex A density law following a modified exponential (parabola up to z = 658 pc, followed by an exponential with a scaleheight of 533 pc), which is roughly equivalent to a sech2 function of scale height 450 pc. The radial density follows an exponential with a scale length of 2.355 kpc. Kinematics follow the result of Ojha et al. (1996). Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 20

21. Besancon Galactic model Besancon Galactic model Bulge/bar In Robin et al. (2012), the sum of two ellipsoids: A standard boxy bulge (bar), the most massive component which dominates the stellar content of latitudes below about 5 o. Ellipsoid (thick bulge) with longer and thicker structure Robin et al. (2014) showed that the ‘thick bulge’ population was in fact the inner part of the thick disc which short scale length makes a large contribution in the bulge region. The angle of the bar to the Sun-Galactic Centre direction is 13 o. The bar kinematics are taken from the model of Fux (1999). The stellar density and luminosity function are assumed from the result of Picaud & Robin (2004) with a single burst population of 10 Gyr age. The IMF below and above 0.7 M sun are assumed to be dN/dm ∝ m −1.5 and a Salpeter slope, dN/dm ∝ m −2.35, respectively (Picaud & Robin 2004). The total bar mass is 5.9 × 10 9 M sun. The model mass to light ratio is 2.0 at (l = 1.25 o, b=−2.65 o ) in Johnson I band which is compatible with result of Calamida et al. (2015) in F814W filter (wide I). Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 21

22. Besancon Galactic model Besancon Galactic model Halo Older than the thick disc (14 Gyr) and metal poor ([Fe/H] = −1.78). A single burst population with an IMF, dN/dm ∝ m −0.5. Total mass of 4.0 × 10 10 M sun are assumed (Robin et al. 2003). The density law has been revised in the study of SDSS+2MASS star counts with a power- law density with an exponent of 3.39 and an axis ratio of 0.768 (Robin et al. 2014). The kinematics is modelled with Gaussian distributions of velocities of dispersion (131, 106, 85) in km s -1 in the (U,V,W) plane, and no rotation. Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 22

23. Lens/source star catalogues Lens/source star catalogues Lens/source star catalogues spanning four H-band magnitude ranges. −10 ≤ H < 15 15 ≤ H < 19 19 ≤ H < 23 H > 23. Using R- and I-band magnitudes of the sources which correspondence to MOA-II filter. The solid angle in each catalogue is chosen to contain ∼ 6000 stars in each range towards Baade’s Window (l = 1◦, b = −4◦) Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 23

24. Besancon populations Besancon populations Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 24 Sumi et al. (2011) find a favoured mass function index in the brown dwarf regime, 0.01M sun ≤ M ≤ 0.08M sun, for the 2006–2007 MOA-II data is α BD = −0.49. From our simulation, an MF slope of α BD = −0.4 provides the best reduced chi-square value. This result is consistent with MOA-II results, but disagrees with the result from some field surveys for young brown dwarfs which suggest a power-law MF with slope α BD > 0.0 (Jeffries 2012; Kirkpatrick et al. 2012).

25. Mean Einstein crossing time Mean Einstein crossing time Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 25 MOA-II All sources24.0days RGC sources19.2days Besancon Resolved sources25.5days DIA sources26.3days Besancon with low-mass stars and brown dwarfs Resolved sources20.3days DIA sources20.9days OGLE-III survey Positive longitude (l>2 o ) 22.0 days Central(-2 o < l < 2 o ) 20.5 days Negative longitude (l < −2 o ) 24.2 days

26. Number of microlensing events Number of microlensing events Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 26 Number of events from the Besancon resolved sources is 0.83 N MOA. Number of events from the Besancon DIA sources is 2.17 N MOA. In the absence of significant blending effects, resolved and DIA predictions to bracket the true result. The effects of blending are complex and a more detailed comparison would require modelling both the source selection criteria and the source blend characteristics of the MOA-II data. 12 per cent of faint stars which can only be detected by the DIA method are observed.

27. Microlensing properties Microlensing properties Confronting Galactic structure with microlensing in an era of large datasets 20th mircrolensing workshop Paris, France 15 January 2016 27 Optical depth Average crossing time Event rate