Star formation rate as a function of time

• Techniques
  • Qualitative SF history from existence or lack thereof of important stellar indicators of ages, e.g., upper main sequence, supergiants, blue-loop stars, luminous AGB stars, red-clump stars, giants, horizontal branch stars

  • In principle, quantitative SF histories can be obtained from relative numbers of stars at different locations in CMD.
    • Accurate observations of distribution of stars across upper main sequence constrain SF variations over past few Gyr (feasible entire Local Group).

    • Accurate observations of subgiant branch constrain SF variations over greater lookback times (Milky Way neighbors, now at M31 distances with ACS).

    • In more distant objects, constrain SF over shorter lookback time.

    • In principle, no age-metallicity degeneracy if one observes a population down to main-sequence turnoff

  • Data. HST (WFPC2 & ACS), deep ground-based data now superb, especially for MW neighbors where larger field is huge advantage

    AD: HST archival project (http://ganymede.nmsu.edu/lg) (Holtzman, Afonso, & Dolphin 2003)

    

  • Complications; note, these are mostly systematic
    • Uncertainties in stellar evolutionary tracks, atmospheres

    • Unresolved binaries

    • Uncertainties in photometry, calibration

    • Uncertainties in distance, reddenning

    • Limited time resolution, esp. at larger ages

    • Different techniques, definition of best fits, etc.

  • What to believe? Cynical vs. optimistic opinions, some comparisons of results:
    • Coimbra experiment: multiple groups derive SFH from same dataset (field in LMC), results (Skillman 2002)
      

      ``the disaster of Coimbra I'', or, ''the demonstration of success by the Coimbra experiment''?

    • various sensitivities (Holtzman 2002)
      

    • Carina derived SFHs (from Tolstoy 2001), but note apparent conflicts are not as large as they look if one considers time resolution adopted by individual authors.
      

    • IC1613 derived SFHs (Skillman et al 2003)
      

    • Derived age-metallicity relations (e.g. LMC (Holtzman et al 1999)
      

    • Careful study of sensitivities, uniform analysis warranted and planned

    • Potential of improved SF histories with incorporation of more external information, in particular, abundances and abundance ratios

    • What do you want to/can you know?

• Results: galaxies (more details here)
  • Milky Way bulge and disk.
    • General consensus that bulge/halo is almost exclusively old (e.g. Zoccali et al 2003, van Loon et al 2002, Ortolani et al 1995), but note details of age range depend on metallicity distribution and metallicities of comparison GCs. Potentially strongest evidence comes from enhanced [$\alpha$/Fe] (Rich & McWilliam 2000).

    • Disk studied via Hipparcos CMD (Hernandez et al 2000, derived recent SFR), via chromospheric activity-age relation (Rocha-Pinto et al 2000, derived recent SFR, however, Gizis et al 2002 differs).
      

      Old component ($>$10 Gyr) appears to exist based on Hipparcos data (Binney et al 2000; Jimenez et al 1997)

  • M31 bulge and disk
    • M31 bulge appears different from MW bulge: more metal-rich and broader metallicity distribution, and apparently, significant age dispersion (Brown et al. 2003, ACS CMDs); disk contamination one possible caveat. Not yet able to distinguish two separate components vs. more continuous distribution.
      

    • M31 disk. Old component Ferguson & Johnson (2000) appears to exist (caveat: confirmation of disk nature?). Overall, SF rate appears low now compared to past rate (Williams 2001). Sarajedini (2001) suggests a thick disk component. No obvious evidence of bursts in (young) disk population.

  • M32 Hotly-debated presence of significant intermediate age population

  • M33 Apparent significant age spread in GC system in M33 from HB morphology; several clusters at 5-10 Gyr (Sarajedini et al 2000). Disk populations probed only over last Gyr (Sarajedini et al 2000) or so with no obvious evidence of bursts. Population gradient exists in disk, such that larger contribution of younger stars in inner region.

  • Dwarf irregulars: LMC (Smecker-Hane et al 2001, Dolphin 2000, Holtzman et al 1999, Olsen 1999, others), SMC (Dolphin et al 2001, Harris 2003), IC10 (Hunter 2001), IC1613 (Skillman et al 2003, Cole et al 1999), NGC6822 (Wyder 2003, 2001; Gallart 1997), Sex A (Dolphin et al 2003), WLM (Dolphin 2000), Pegasus (Gallagher et al 1998), Sag DIG, GR8 (Dohm-Palmer et al 1998), Leo A (Dolphin et al 2002, Schulte-Ladbeck et al 2002, Tolstoy et al 1996) , LGS3, Phoenix, Aquarius (latter three ``transition-type'')
    • Recent SF best studied. No strong evidence for strongly episodic SF on scales of $10^7$ Myr, e.g. Dohm-Palmer et al 1998 Possible global events seen in LMC/SMC, but probably not variations of more than factor of a few (but see Harris, this conference); no dramatic differences as a result of proximity to large galaxy?
      

    • In LMC/SMC, SFH can be probed back much further. Again, no strong evidence for extreme burstiness

    • IC10 possibly an exception?

    • All dIrrs in which an ancient population ($>$ 10Gyr) can be seen have such a population

    • Skillman (2003) suggests dominant intermediate age populations in dIrrs, possibly related to effects of photoionization

  • Dwarf ellipticals/spheroidals: NGC205, NGC147, NGC185, Fornax (Saviane et al 2000; Stetson et al 1998, Buonanno et al 1999), And companions (da Costa, Armandroff et al, Sarajedini et al 2002), Leo I (Dolphin 2002, Held et al 2001), Sculptor (Hurley-Keller et al 1999, Majewski et al 1999), Antlia (Piersimoni et al 1999), Leo II (Dolphin 2001), Tucana (Saviane et al 1996), Sextans (Bellazzini et al 2001, Harbeck et al), Carina (Smecker-Hane et al 1996, Hurley-Keller et al 1998), Ursa Minor (Carrerra et al 2002), Draco (Aparicio et al 2001)
    • Wide variety of SFHs. Draco, Ursa Minor, Sculptor, Leo II, Sextans, Tucana predominantly old (but some ``intermediate-old''); Leo I, Carina, Fornax predominantly intermediate age

    • Some dSphs do have apparently episodic star formation, in particular Carina and Sagittarius, Fornax(?)
      
      
      

      Even dSphs with predominantly old stars may show evidence for extended periods of star formation based on chemical evidence (Ikuta & Arimoto 2002)

    • All dSphs in which an ancient population ($>$ 10Gyr) can be seen have such a population

    • Chemical abundance ratios (Sculptor, Fornax, Carina, Leo I) suggesting near solar [$\alpha$/Fe] at all ages (Tolstoy et al 2003) - a bit peculiar for old stars (IMF implications?). Abundance-ratio variations potentially consistent with SFHs derived from CMDs!

• Ancillary issues
  • Population gradients: almost ubiquitous. Generally, younger populations are more centrally concentrated. Metallicity gradients exist in most objects. Unclear whether gradients reflect multi-component (e.g. bulge/disk) population and/or smooth changes in a single population - kinematics needed.

  • Cluster formation vs. field stars: some LG galaxies contain massive compact clusters. Seems unlikely that the age distribution of clusters matches that of field stars of parent galaxies: e.g., MW/M31, LMC, Fornax, IC10.

  • Chemical evolution and single-valued age-metallicity relations: fits to CMDs generally better only when spread in age-metallicity relation is taken into account.