1. Galactic Understanding and Evolution (LEGUE)
LAMOST Experiment for
Galactic Understanding and Evolution (LEGUE)
The LAMOST Galactic Structure Working Group

2. LEGUE Survey Team
LEGUE Science Working Group: DENG Licai, HOU Jinliang, NEWBERG Heidi, CHRISTLIEB Norbert, LIU Xiaowei, HAN Zhanwen, CHEN Yuqin, ZHU Zi, PAN Kaike, LEE Hsu-Tai, WANG Hongchi
Contributing Authors: LIU Chao, HU Jingyao., XU Yan, ZHANG Yueyang, BEERS Timothy, GRILLMAIR Carl, GUHATHAKURTA Raja, LEPINE Sebastien, O’SHEA Brian, RADDICK Jordan, SELLWOOD Jerry, YANNY Brian, ZHENG Zheng, XIN Yu, CHEN Li, ZHAO Gang, QIAN Shengbang, JIANG Biwei, LI Jinzeng, SHI Jianrong, ZHANG Haotong, SHI Huoming, YANG Fan

3. LEGUE
Licai Deng(NAOC), Yuqin Chen, Jingyao Hu, Huoming Shi, Yan Xu, Haotong Zhang, Gang Zhao, Xu Zhou (NAOC); Zhanwen Han, Shengbang Qian (Yunnan, NAOC); Yaoquan Chu (USTC); Li Chen, Jinliang Hou (SHAO); Xiaowei Liu, Huawei Zhang (PKU); and Biwei Jiang (BNU).
The collaborating group of US astronomers: Heidi Newberg (Rensselaer), Timothy Beers (Michigan State), Carl Grillmair (IPAC), Raja Guhathakurta (Santa Cruz), Sebastien Lepine (AMNH), Brian O’Shea (MSU), Jordan Raddick (education, Johns Hopkins), Jerry Sellwood (Rutgers), Jason Tumlinson (STSci), Brian Yanny (FNAL), and Zheng Zheng (IAS).

4. We hope to convince you that:
A spectroscopic survey of Milky Way stars will provide important new information about how galaxies form, about star formation and the production of heavy elements in the universe, and about the distribution of mass in the Galaxy
LAMOST is the most powerful telescope in the world to carry out this work.
There is no other current or planned spectroscopic survey that will achieve the LEGUE science goals.

5. WMAP Team
The process of galaxy formation is tied to the history of our Universe, involving gravitational attraction, cosmic expansion, gas dynamics, star formation and feedback, and hierarchical merging.

6. 1/2 a degree
To see the Milky Way, we need to survey the entire sky, and build a picture of the sky one star at a time. This has only recently become possible. The Milky Way is the only galaxy for which it is possible to obtain three dimensional position and velocity information, and classifications for every star.

8. SEQUE footprint
Yanny et al., 2009, astroph/ v1

9. Importance of stellar spectra
Stellar spectra provide: luminosity class (distance), radial velocities, chemical composition, maybe ages for stars.
Distances are required to measure physical structure.
Radial velocities distinguish components that formed separately, constraining formation mechanisms.
Chemical composition teaches us about the environment and time period over which stars formed in each component, and about the general processes of metal formation in the Universe

10. Spectroscopic surveys of stars
LEGUE: g0<20, R=2000, δ>-10°, 7.5 M stars, in commissioning
RAVE: I<12, R=7500, δ<0°, |b|>25°, 1 M stars, in progress
SEGUE: g0<20, R=1800, δ>-10°, 400,000 stars, nearing completion
APOGEE: H<13.5, R=20,000, δ>-10°,100,000 stars, infrared, in fabrication
HERMES: V<14, R=30,000, 1.5 M stars, in fabrication
WFMOS: in planning stages
None of these surveys probe the Galactic plane. Except for possibly WFMOS, LAMOST/LEGUE has the largest aperture and the largest number of fibers in the focal plane.

11. Overview Constraints and suitability of the LAMOST Telescope
(2) Scientific justification for a Galactic survey of millions of stars
(3) LEGUE Survey Strategy - a survey of 2.5 million spheroid stars, and 5 million disk stars

12. LAMOST facts Aperture: ~4 m Type: Schmidt, Alt-Az Focal length: 20m
Relative aperture: f/5
Field of view: 5 degree diameter
Size of focal plane: 1.75 m
Sky coverage: Dec>-10 degrees, 1.5 hours around meridian
Wavelength range: 370 nm to 900 nm, R=1000/2000
Number of fibers: 4000, 16 spectrographs with 250 fibers each
>10,000 spectra per night (>2 million spectra/year),
2-3 gigabytes/night
LAMOST: 4000 fibers in 20 square degrees (200 fibers/sq. deg.)
UKST: fibers in 28 square degrees (5 fibers/sq. deg.)
SDSS: fibers in 7 square degrees (90 fibers/sq. deg.)

13. LAMOST Constraints
(1) Pointing - Light lost with distance from the meridian, declinations away from the optical axis of the telescope.
(2) Fibers - must be uniformly distributed.
(3) Weather – poor summer weather limits view of Galactic center.

14. Optical System
MA is the Schmidt corrector, 5.72m x 4.40m, with 24 hexagonal plane sub-mirrors, each with 1.1m diagonal and 2.5 cm thickness.
MB is the spherical primary, 6.67m x 6.05m, with a radius of curvature of 40m, 37 hexagonal spherical sub-mirrors, each with 1.1m diagonal and 7.4 cm thickness.
Active control for aspheric shape of corrector (34 force actuators plus 3 mount points per submirror). Optimal shape changes with declination and hour angle.
Active control for MB is just 3 mount points plus three actuators per submirror.
Optical axis is 25° from horizontal.
The focal plane has a radius of curvature of 20m.

15. LAMOST effective aperture as a function of declination δ° -10 10 20 3010 20 30 40 50 60 70 80 90
φ(0)
4.88
4.84
4.79
4.71
4.62
4.51
4.37
4.22
4.04
3.83
3.60
φ(0.75h)
4.87
4.83
4.78
4.70
4.61
4.50
4.36
4.21
4.03

16. Spot sizes (80% of light) for central 3° field
δ = 90°
δ = -10°
δ = 40°
δ = -10°
δ = 40°
δ = 90°
with atmosphere
with atmosphere
with atmosphere
Spot sizes (80% of light) for central 3° field

17. 1.5 hours of tracking
atmosphere included
The plots above show the largest linear extent of the spot size containing 80% of the light, as a function of declination. Above declination of 60°, the field of view has been reduced from 5 degrees to 3 degrees in diameter, which is the reason for the apparent sudden reduction in spot size. The spot size at the edge of the field (5°) at δ=60° is the same as the spot size at δ=90° at the edge of the 3° field.

18. In one pointing, fewer than 20 LAMOST fibers can be placed inside the 10’ tidal radius of a globular cluster. Fewer than 50 LAMOST fibers can be placed within 20’ of the center of an open cluster.
3.15
arc minutes
4.7 arc minutes

19. The combination of site weather patterns and the need to look at the meridian puts strong constraints on the footprint of the LAMOST survey.

20. Potential Worries Sky brightness Scattered light Dust/pollution
Temperature
Calibration
But note that we already have a spectrum of similar quality to SDSS.

21. 6660 Å, bandwidth 480 Å

22. Sky brightness measured by BATC in dark photometric nights
New CCD chip

23. Comments on sky brightness
BATC measurements avoid sky emissions (as of KPNO) therefore somewhat fainter than true values;
There is a scatter ~1mag in each night. The scatter has a weak dependence on direction.

25. The LAMOST spectrum (top) is comparable to the SEGUE spectrum (bottom) of the same star, with similar exposure time. The LAMOST sky subtraction and response function still need more work (note O2 line at 6880Å and 7600Å), but it can already achieve simlar S/N as SDSS.

26. Monoceros stream,
Stream in the Galactic Plane,
Galactic Anticenter Stellar Stream,
Canis Major Stream,
Argo Navis Stream
Newberg et al. 2002
Vivas overdensity, or
Virgo Stellar Stream
Pal 5
Stellar Spheroid?
Sagittarius Dwarf Tidal Stream

27. Belokurov et al. 2007 Hercules-Aquila Cloud
Areal density of SDSS stars with and in Galactic coordinates. The color plot is an RGB composite with blue representing , green , and red . Structures visible on the figure include the arch of the Sagittarius stream and the Orphan stream in blue and the multiple wraps of the Monoceros Ring in red. Behind, and offset from, the Galactic bulge, there is a yellow‐colored structure visible in SDSS data in both hemispheres, centered at . This is the Hercules‐Aquila cloud. Overlaid on the plot are contours of constant extinction (black is 0.1 mag and gray is 0.25 mag in the i band). The two red vertical lines show the SEGUE stripes at (runs 5378 and 5421) and 50° (run 4828). The inset is a CMD of the SDSS data, with blue, green, and red bands encompassing the stars used to create the large figure.
Areal density of SDSS stars with 0.1<g-i<0.7 and 20<i<22.5 in Galactic coordinates. The color plot is an RGB composite with colors representing regions of the CMD as shown in the inset. The estimated distance to the cloud is kpc.

28. Belokurov et al. 2007 Hercules-Aquila Cloud
Areal density of SDSS stars with and in Galactic coordinates. The color plot is an RGB composite with blue representing , green , and red . Structures visible on the figure include the arch of the Sagittarius stream and the Orphan stream in blue and the multiple wraps of the Monoceros Ring in red. Behind, and offset from, the Galactic bulge, there is a yellow‐colored structure visible in SDSS data in both hemispheres, centered at . This is the Hercules‐Aquila cloud. Overlaid on the plot are contours of constant extinction (black is 0.1 mag and gray is 0.25 mag in the i band). The two red vertical lines show the SEGUE stripes at (runs 5378 and 5421) and 50° (run 4828). The inset is a CMD of the SDSS data, with blue, green, and red bands encompassing the stars used to create the large figure.
The large scale lumpiness of the stellar halo density has made it difficult to determine whether the outer parts of the Galaxy are axisymmetric (Xu, Deng & Hu 2006, 2007).

29. Tidal Stream in the Plane of the Milky Way
Blue – model Milky Way
Pink – model planar stream
TriAnd,TriAnd2
Explanations:
One or more pieces of tidal debris; could have puffed up, or have become the thick disk.
Disk warp or flare
Dark matter caustic deflects orbits into ring
Monoceros, stream
in the Galactic plane,
Galactic Anti-center
Stellar Stream (GASS)
Sun
Canis Major or
Argo Navis
Tidal Stream in the Plane of the Milky Way
If it’s within 30° of the Galactic plane, it is tentatively assigned to this structure
Press release, November 4, 2003

30. Summary of Spheroid Substructure
Dwarf galaxy streams:
(1) Sagittarius: Ibata et al. 2001a, Ibata et al. 2001b, Yanny et al. 2000
(2) Canis Major/Argo Navis? Monoceros (Newberg et al. 2002, Yanny et al. 2003), GASS (Frinchaboy et al. 2004), TriAnd (Majewski et al. 2004), TriAnd2 (Martin, Ibata & Irwin 2007), tributaries (Grillmair 2006)
(3) ?? Orphan stream, Grillmair 2006, Belokurov et al. 2006
(4) ?? Virgo Stellar Stream, Vivas et al. 2001, Newberg et al. 2002, Zinn et al. 2004, Juric et al. 2005, Duffau et al. 2006, Newberg et al. 2007
(5) Styx, Grillmair 2009
(6) Cetus Polar Stream, Newberg, Yanny & Willett 2009
Globular cluster streams:
Pal 5: Odenkirchen et al. 2003
(2) GD-1 Grillmair & Dionatos 2006
(3) NGC 5466: Grillmair & Johnson 2006
(4-6) Acheron, Cocytos, and Lethe:
Grillmair 2009
Sino-Western collaborations on spheroid substructure:
Xue, X.X., Rix, H.-W., Zhao, G., et al. 2008, ApJ, 684, 1143
Liu, C., Hu, J.Y. Newberg, H.J., & Zhao, Y.H. 2008, A&A, 477, 139
Other:
Hercules-Aquila Cloud
Virgo Overdensity?

31. density
selected
sky density of SDSS spectra
velocity
selected
velocity selected Sgr stream
The Cetus Polar Stream

32. Stars within 1 kpc of the Sun, with Hipparcos proper motions
Following Helmi et al. 1999
Tidal streams separate in angular momentum
– need 3D position and velocity through space.

33. Smith et al. 2009 Klements et al. 2009
Moving groups found from 22,321 low metallicity ([Fe/H]<-0.5) SDSS/SEGUE stars within 2 kpc of the Sun, from SEGUE data (above). Moving groups from SDSS /Segue stars within 5 kpc of the Sun (right). Both analyses show significant velocity substructure in the Solar neighborhood.

34. GAIA Astrometric Satellite
Magnitude limit: 20
1 billion Galactic stars
Astrometry and radial velocities
Will only get precise radial velocities
for stars brighter than 15th magnitude!
With LAMOST, radial velocities
can be obtained for the most
interesting magnitude range
of 15<V<20
Other large spectroscopic surveys of stars include RAVE (I<12), SDSS III/ SEGUE II (400,000 stars), APOGEE (infrared bulge), HERMES (V<14, in fabrication), and WFMOS (in planning stages).

35. Metal-poor stars t Norbert Christlieb See also:
This slide summarizes Galactic chemical evolution by means of high-resolution spectra of stars. The spectrum of the Population III star is of course a simulation. The slide also illustrates that there is a metal-poor star “lookback time”.
See also:
Zhao, G., Butler, K., Gehren, T. 1998, A&A, 333, 219
Zhao, G. et al. 2006, ChJAA, 6, 265
Norbert Christlieb
MPIA, January 09
35/37 35

36. Number of contributing SNe
Karlsson & Gustafsson (2005, IAU 238)
These histograms show the distribution of the number of supernovae which contributed to the chemical enrichment of metal-poor stars, according to the stochastic chemical evolution model of Torgny Karlsson. At [Fe/H] < -3.5, only very few or even only a single SN have contributed, hence we can study the properties of the progenitor stars (e.g., mass, rotation) by means of the abundance patterns of the next generation of stars.
Norbert Christlieb
MPIA, January 09
MPIA, January 09
36/49 36

37. The halo metallicity distribution function
This slide shows how rapidly the halo MDF drops towards low [Fe/H], and it shows how weakly the tail at [Fe/H] < -4.0 is populated right now: Only three stars are known there, and these are the most interesting ones (see previous slide). Because LAM
Norbert Christlieb
MPIA, January 09
MPIA, January 09
37/49 37

38. EMP and HMP stars expected to be found
Survey
Effective sky coverage
Effective mag limit
N < 3.0 (EMP)
N < 5.0 (HMP)
HES
6400 deg2
B < 16.5
200 2
SEGUE
1000 deg2
B < 19
1000 10
LAMOST
12,200 deg2
B < 18.0
3000 30
SSS
20,000 deg2
B < 17.5
2500 25
V1/V2 = A1/A2 * 10**((m1-m2)*3/5)
Norbert Christlieb
Heidelberg, April 2009
MPIA, January 09
38/48 38

39. Outstanding Problems
Describe the chemical evolution of the Galactic disk(s), and especially the first generation of stars.
What is the detailed structure of the Milky Way’s disk? How is it related to Monoceros/ Canis Major?
What does the dark matter potential of the Milky way look like? We have yet to successfully extract information about the Galactic potential from tidal streams.
How many stellar components are there in the Milky Way, and how do we describe them?
Galactic stellar data in all Galactic components is more complex than the models in structure and in dynamics. How do we compare them?
How many small galaxies merged to create the Milky Way, and when?
So far, advances have primarily come from reducing the data size to analyze very clean samples. How do we utilize all of the partial chemical, kinematic, and spatial information at the same time?

40. The Future of Galactic structure
In the Milky Way, we have the opportunity to learn the whole history of one galaxy instead of comparing snapshots of many. It is only now that we have large surveys of the whole sky that we are able to comprehend the Milky Way as a whole. Unlike external galaxies, the picture we are building is in three dimensions of position and velocity, with much higher accuracy information for each star.
Many surveys currently in progress will provide multi-color imaging of the sky. However, there is a great need for spectroscopic surveys of millions of stars.
Twenty years ago, when the idea for the SDSS was born, large scale structures of galaxies had just been discovered. But there was structure on all scales of the largest surveys of the day. There was a pressing need for a larger spectroscopic survey.
We are at the same place now in the study of the Milky Way. Spatial substructure and moving groups are found in every spectroscopic sample of spheroid stars that is well constrained in position and stellar type. It is guaranteed that a larger survey will reveal more substructure.

41. LAMOST Experiment for Galactic Understanding and Exploration (LEGUE)
Science Goals
Discovery of spheroid substructure
Constrain Galactic potential
Disk/spheroid interface near Galactic anticenter
Search for extremely metal poor stars
Identify smooth component of spheroid
Structure of thin/thick disks, including chemical abundance and kinematics
Search for hypervelocity stars
Survey OB stars and 3D extinction in Galaxy
Globular cluster environments
Properties of open clusters
Complete census of young stellar objects across the Galactic plane

42. Survey Strategy (five years)
LAMOST Experiment for Galactic Understanding and Exploration (LEGUE)
Survey Strategy (five years)
Three subsets:
Spheroid (|b|>20°) portion will survey at least 2.5 million objects at R=2000, with 90 minute exposures, during dark/grey time, reaching g0=20 with S/N=10.
Anticenter (|b|<30°, 150°<l<210°) portion will survey about 3 million objects at R=2000 with 40 minute exposures, during bright time (and some dark/grey time), reaching J=15.8 with S/N=20.
(3) Disk (|b|<20°, 20°<l<230°) and will survey about 3 million objects at R=2000 and R=5000, with 10 and 30 minute exposures, respectively, during bright time, reaching g0=16 with S/N=20

43. Survey footprint, shown as an Aitoff projection in Galactic coordinates. The region with open circles will be observed at R= The region with filled circles at low Galactic latitude will be surveyed with shorter, bright time exposures including R=5000 and R= There is a small region that is part of both the anticenter and speroid surveys. The Celestial Equator is shown as a solid line. This illustrates the footprint, but not the exact placement of the survey plates.

44. Spheroid Survey
Use SDSS, PanSTARRS, or SuperCOSMOS photometry, in that order, as available. A u-band survey is planned on the 2.3 meter Bok telescope of Arizona’s Steward Observatory; a camera is currently being fabricated. This survey can be used in conjunction with PanSTARRS or SuperCOSMOS
Select as many 0.1<(g-r)0<1.0, g0<17 stars as possible (a nearly complete sample where surveyed, except below b=40°, randomly sample to g0<18
Randomly sample stars with (g-r)0<0.4 in the magnitude range 17<g0<20
If u-band photometry is available, we will deselect QSOs. The subsampling will be about one in two or one in three at higher latitudes.

45. Additional Criteria for spheroid
We will observe a sample of high proper motion stars with colors of M dwarfs in the magnitude range 16<g0<20 (local spheroid stars)
If u-band available, subsample K and M giant candidates with 17<g0<20
Within 3 tidal radii of 40 selected GCs, we will use a completely different selection algorithm to select stars with color/magnitude of the GC stars
We will include bright (V<12) K and M stars from the Tycho-2 catalog, without regard to their proper motion.

46. Deng Licai, Liu Chao
Simulation of LAMOST stellar spectral density in a five year survey, using an Aitoff projection in Galactic coordinates, using *all* of the clear weather. The result is 7.5 million halo objects, 5 million anticenter objects, and 3 million disk objects.

47. 7.5 M halo objects
5 M anticenter objects
3 M disk objects
Deng Licai, Liu Chao
Sample survey coverage in fibers per square degree, shown as an Aitoff projection in Equatorial coordinates (Galactic coordinates shown in blue). The survey simulation was done assuming all of the time for a five year period, including moon and likely weather conditions as a function of season.

48. 2.5 M halo objects
3 M anticenter objects
3 M disk objects
Deng Licai, Liu Chao
Sample survey coverage in fibers per square degree, shown as an Aitoff projection in Equatorial coordinates (Galactic coordinates shown in blue). The survey simulation was done assuming 1/3 of the dark/grey time and all of the bright time for a five year period, including moon and likely weather conditions as a function of season.

49. Deng Licai, Liu Chao
Number of spectra
winter
summer
Survey footprint covered
Accumulation of data as a function of time assuming all of the dark and grey time for a five year survey. After the first year, we begin to re-observe parts of the sky that have already been covered.

50. Deng Licai, Liu Chao
Number of spectra
winter
summer
Survey footprint covered
Accumulation of data as a function of time assuming 1/3 of the dark/grey time and all of the bright time for a five year survey. After the first year, we begin to re-observe parts of the sky that have already been covered.

51. 2.5 M halo objects
3 M anticenter objects
3 M disk objects
LEGUE Science Working Group: DENG Licai, HOU Jinliang, NEWBERG Heidi, CHRISTLIEB Norbert, LIU Xiaowei, HAN Zhanwen, CHEN Yuqin, ZHU Zi., PAN Kaike, LEE Hsu-Tai, WANG Hongchi

52. Recent development
We learnt last week that the telescope can point 2 hour away from the meridian without loosing much efficiency;
The survey plan may change a lot.
The expected data size for LEGUE will not change.
Site quality decays quickly, if we want to do the survey, we should act fast!