A Survey of Local Group Galaxies Currently Forming Stars


1
A Survey of Local Group GalaxiesCurrently
Forming Stars
Phil Massey Lowell Observatory April 14 2003
2
The Team

• Paul Hodge Univ. of Washington
• Shadrian Holmes Univ. of Texas
• George Jacoby WIYN
• Nichole King Lowell Observatory
• Phil Massey (PI) Lowell Observatory
• Knut Olsen CTIO/NOAO
• Abi Saha KPNO/NOAO
• Chris Smith CTIO/NOAO

3
Overview

• We are imaging all of the galaxies of the Local
  Group that are currently forming stars
• broad-band (UBVRI)
• narrow-band (H OIII SII)
• with the KPNO and CTIO 4-m telescopes and
  Mosaic CCD cameras.

4
Motivation Our Science

• The galaxies of the Local Group serve as our
  laboratories for studying star formation and
  stellar evolution as a function of metallicity
  Z. (Z varies by a factor of 17 from WLM to M31.)

5
Why should the metallicity matter

• Star Formation
• Lower metallicity gas should have a lower cooling
  rate and hence higher temperatures larger
  Jeans mass leading to a top-heavy IMF (Larson
  1998).
• But over the limited metallicity range (3x)
  SMCLMCMW this effect isnt seen!

6
IMF Slope in OB Associations

• From Massey (2003)

Z0.018
Z0.008
Z0.004
7
IMF

• Variations that are seen in the IMF slope are
  statistical not physical (Massey 1998 Kroupa
  2001)
• But what would happen if we extended this to
• one-tenth solar (WLM) to 2x solar (M31)
• The answer is important for understanding the
  integrated properties of galaxies at large
  look-back times.

8
Star Formation/metallicity (cont)

• Some expect that the upper mass limit will vary
  as a function of metallicity
• True only if radiation pressure acting on grains
  is the limiting factor in determining the mass of
  the highest mass star that can form.
• So far we find that the upper mass limits are
  purely statistical and not physical. What ever
  it is that limits the ultimate mass of a star we
  have yet to encounter it in nature (cf. Massey
  Hunter 1998 ApJ 493 180).

9
Why should the metallicity matter (continued)

• Massive Star Atmospheres and Evolution
• Stellar winds are driven by radiation pressure
  through highly ionized metal lines. Mass-loss
  rates will depend upon Z where 0.5-1.0
• This mass-loss has a profound effect on the
  evolution of high-mass stars.

10
Relative number of red supergiants (RSGs) and
Wolf-Rayet stars (W-Rs)
log Number RSGs/WRs
From Massey 2003 ARAA 41 (in press)
log (O/H) 12
11
Relative number of red supergiants (RSGs) and
Wolf-Rayet stars (W-Rs)
log Number RSGs/WRs
From Massey 2003 ARAA 41 (in press)
log (O/H) 12
12
Need good observational database

• New generation of high mass evolutionary models
  are becoming available which include the
  important effects of rotation (mixing introduced
  by meridional circulation and shear
  instabilities).
• Need solid observational database to help guide
  the theorists.

13
Our Science (continued)
Along the way well find The most massive
supergiants. Luminous Blue Variables and other
luminous stars with H emission. Star formation
rates for massive stars. Distribution and
numbers of evolved massive stars (RSGs WRs).
HII regions SNRs PNe and the extent of the
diffuse emission.
14
Your Science

• This survey will provide the source list
  (finding charts) for spectroscopy with 8-10-m
  telescopes for decades to come. Our data
  products include
• Stacked images (UBVRI H OIII SII)
• Individual dithered images (suitable for
  photometry).
• Calibration
• Catalog of UBVRI photometry of roughly 300
  million stars

15
What Were Doing The Sample

• M31 (10 fields) Pegasus Dwarf
• M33 (3 fields) Phoenix
• IC 10 IC 1613
• NGC 6822 Sextans A
• WLM Sextans B

16
How Are We Doing

• M31 (10 fields) Pegasus Dwarf
• M33 (3 fields) Phoenix
• IC 10 IC 1613
• NGC 6822 Sextans A
• WLM Sextans B

17
What Were Doing (continued)

• Aiming for a S/N of 3 at UBVRI25
• in 1 seeing.
• Also imaging in H OIII SII
• Each field 5 ditherings then stacked.

18
Hasnt All This Been Done Before

• Yes but not with our depth area photometric
  accuracy and resolution!
• Photographic plates had the area coverage and
  (usually) the resolution but neither the
  photometric accuracy nor depth.
• CCD studies had the depth and accuracy but not
  always the resolution and certainly not the area
  coverage.
• Wal Sargent story...

19
Comparison of M31 CCD Surveys
20
Basic Processing

• Generally following the Valdes IRAF pipeline
  but with some enhancements.
• Better flat-fielding techniques.
• Better determination of sky and scaling in the
  stacking process (via scripts using aperture
  photometry).
• Details and software can be found at our web
  site http//www.lowell.edu/massey/lgsurvey

21
Photometry

• For the purposes of photometry we treat each
• Mosaic camera as 8 separate instruments
• PSF variations within a single chip modest
  compared to chip-to-chip variations.
• Different DQE-wavelength dependence for each chip
  means different color terms and even different
  zero-points (despite flat-fielding efforts).

22
U flat divided by I flat
Variations 30
23
Photometry software

• Its a factor of 40 times more work (8 chips x 5
  ditherings) but at least when were done we have
  1 photometry.
• Weve developed a series of IRAF scripts and
  FORTRAN programs that allow us to do the
  photometry automatically chip-by-chip
  dither-by-dither.
• All of this is freely available from our web
  site
• http//www.lowell.edu/massey/lgsurvey

24
How weve solved the calibration problem
Lowells dark-sky site at Anderson Mesa
25
External Calibration using Lowell s 1.2-m Hall
Telescope

• Can use only the most pristine photometric
  nights.
• Select the best calibrated Landolt standards
  covering a complete range of colors
• Investigate gravity effects on the U-band filter

26
U solution always squirrelly near U-B0.
U-B
27
Its a matter of some gravity....
28
Progress Report---How are We Doing

• All images for M31 (10 fields) M33 (3 fields)
  NGC 6822 IC10 WLM Phoenix Sextans A and
  Sextans B are now released and sitting in the
  NOAO NSA archive as well as our own dedicated
  ftp site (which makes bulk downloads easier).
• Poor weather in early September prevented us from
  completing the project still need IC1613 and the
  Pegasus dwarf plus repeat of poor seeing frames.
• Calibration in progress and catalog should be
  complete on schedule release Jan 2004.

29
Did We Achieve our 1.0 seeing goal

• Not really...

30
(No Transcript)
31
1.3
32
0.76
33
1.3
34
0.76
35
Poor seeing matters!

• To redo the images with seeing 1.3 would
  require only a few additional nights.

36
Sadly...

• Weve been told that our time has run out and we
  arent eligible for additional time via the
  survey TAC.
• So weve made our best case to the standard TAC
  and well see what happens.
• (Wal Sargent Cautionary Tale)

37
M31 in 10 fields
38
M31 in 10 fields
39
M31 Fields 2 3
40
M33-North
41
M33-Center
42
NGC 6822
43
Phoenix
44
WLM
45
Whats Next

• Spectroscopy!

46
M31
47
N206 in M31
ob78-231
48
HST/ FUV ob78-231

• Bianchi Hutchings Massey (1996 AJ 111 2303)

49
To take high S/N optical spectra at B19 requires
a really big telescope...
The 6.5-m MMT
50
Optical (blue) spectrum ob78-231
Spectrum in collaboration with Kathy Eastwood
51
OB78-231 at H
Spectrum in collaboration with Kathy Eastwood
52
Meanwhile these data are already being used...

• Ben Williams PhD thesis (Univ Washington)
• Williams MNRAS 340 143 based up a bootstrap
  calibration.
• Forms optical basis for identifying super-soft
  Chandra counterparts (DiStefano et al. in prep)
• Images have been featured in
• APOD 27 Sept 2001
• Astronomy Magazine (Sky Gem feature) Dec 02
• Upcoming Mercury article on super star clusters
  (Hunter Elmegreen Massey 2003 in press)

53
Real Science

• ...will come once the calibrated photometry is
  complete this summer. (Catalog will be released
  at the AAS meeting in Jan 2004).

54
Follow-up Work In Progress

• Well be pushing our studies of stellar winds to
  high metallicities (M31) using Cycle 12 time on
  HST (40 orbits 80 parallels just awarded)
• We also hope to begin extending our studies of
  the IMF to the more distant galaxies of the Local
  Group using DEIMOS on KeckII.