STELLAR POPULATIONS

Erasmus Mundus Padova - Italy 2010-2011

Prof. Antonio Bianchini Dipartimento di Astronomia Università di Padova antonio.bianchini@unipd.it

STELLAR POPULATIONS - 2

Erasmus Mundus Padova - Italy 2010-2011 Prof. Antonio Bianchini

A BRIEF HISTORY OF STELLAR POPULATION STUDIES

Understanding the formation and evolution of galaxies through investigation of properties of constituent, luminous parts.

ASSUME:

• Galaxies can be broken down into POPULATIONS - groups of stars, clusters, gas, with shared, or definable distributions of, properties

A basic unit of stellar population studies is the "SIMPLE STELLAR POPULATION" (SSP):

• Coeval • Chemically homogenous (at least at birth) •  Similar orbits/kinematics

One example of an SSP is the typical star cluster (either open or globular).

IN PRINCIPLE:

where here POPULATIONS refers to what I will call a PRINCIPAL COMPONENT POPULATION.

The GALAXY here is a composite population, but so too may be the PRINCIPAL COMPONENT POPULATIONs, which ultimately are likely a superposition of SIMPLE STELLAR POPULATIONS:

“Basis Vectors” Some examples of composite, principal simple component populations

in our Milky Way might be:   halo   disk   bulge

Each of the above represent complex groupings of stars and matter, but with distinct global properties/distributions of chemistry/age/kinematics from one another.

  These differences presumably relate to differing mixtures of SSPs.

  The smaller the basis set (n, mn), the easier is the composite population (and, ultimately, the Galactic history) to unravel

  Identifying individual SSPs may be hard in complex galaxy, but, perhaps SSPs are strung together in somewhat simple patterns. . .

(e.g. a star forming disk with steadily enriching series of SSP "bursts" of increasing rotational velocity about center)

. . . that make up a principal component population of a galaxy

This is one hope of stellar population studies.

We hope to simplify what may be a quite complex problem to one of finding the patterns in Principal Component Populations.

MORE SPECIFICALLY: WE SEEK: Correlations between attributes such as:

  SPATIAL DISTRIBUTIONS, e.g., stellar density laws

  KINEMATICS, velocities, velocity dispersions (i.e. observable dynamical features)

  CHEMISTRY, e.g. mean [Fe/H], chemical abundance patterns ([O/Fe], [Ca/Fe], [Zn/Fe], ...)

  AGES, reflected, e.g., in the types of stars seen (their evolutionary state)

TO IDENTIFY AND DEFINE : Principal component populations that will allow us

TO RECONSTRUCT: A complete, physical, chemodynamical evolutionary model of the Milky Way (or other galactic systems)

The Ultimate Chemodynamical Model for the Evolution of a Galaxy might include as descriptors/variables (all time variable):

  : the evolution of the phase space distribution of stars (and gas, dark matter)

  : the evolution of atomic species Xi as the interstellar gas out of which stars form enriches

  : the STAR FORMATION RATE

  : the instantaneous INITIAL MASS FUNCTION, how the new stars are distributed by mass (which determines how populations evolve chemically and what kinds of remnants are produced)

Brief History of Milky Way Stellar Population Studies

The history of understanding the stellar populations of the Milky Way provides an illustration of the above process.

In the history provided here, the symbol = "A correlation of properties"

1. At the beginning of the 20th century, Milky Way studies was essentially "cosmology", since at the time it was not appreciated that there were galaxies outside our own.

"Sidereal Universe" = Milky Way.

Until 1920: Curtis-Shapley Debate.

2. Early 20th century stellar population studies concerned with taxonomy:   e.g., groups of stars based on:

○  velocities (proper motions, radial velocities) ○  stellar types: e.g., colors, brightness, variability, spectral type ○  location: e.g., in spiral arms, in clusters/associations, in bulge

  Kapteyn, Jeans, Smart, Stromberg, Oort, Lindblad... ○  "Star streams": kinematical subsystems Galactic position/structural shape

Early work by Lindblad (1936, MNRAS, 97, 15) on the understanding of stars of different types in different orbits, characterized by differences in conserved quantities (``integrals of motion"),

energy (I1) and angular momentum (I2).

Brief History of Milky Way Stellar Population Studies

3. Stellar evolution theoretical advances in 1950's - understanding of chemistry and ages of stars

 Allows age dating of clusters, CMD's (Sandage & Schwarzschild 1952)

ages CMD types

(Of course, earlier in the 20th century, the Hertsprung-Russell Diagram (HR Diagram) was invented; showed that stars do not occupy random distributions of luminosity and spectral type/temperature/color.)

Sandage & Schwarzschild's (1952, ApJ, 116, 463) development of an understanding of the positions of stars in a cluster color-magnirude diagram on

the basis of stellar evolution.

•  Understanding nucleosynthesis in stars (e.g., Burbidge, Burbidge, Fowler, Hoyle 1957) -- notion of chemical enrichment with time

4. Connections between structure, kinematics, age, chemistry, stellar (CMD) types

A. Chamberlain & Aller (1951) - enrichment levels "written" in spectra

chemistry gives relative age   Note concept of an Age-Metallacity Relation (AMR)   Ultraviolet excess as simple way to get relative metallicities of stars through photometry,

rather than spectroscopy - metal lines are crowded toward UV part of spectrum. (Sandage, Eggen: 1950s, 1960s)

B. Sandage & Walker (1955) - weak-lined Milky Way field stars look like globular cluster stars (via Ultraviolet Excess)

cluster stars calibrate field star age-dating chemistry (e.g., UV excess) absolute ages Develop a general AMR for Milky Way field stars.

C. Baade (1944), Oort (1926)

First sweeping collectivization of ``stellar populations" CMD types structural components

Baade's definition of populations based on CMD type.

structural components kinematical groups   POP I = Disks = "slow-moving" (OB stars, open clusters)   POP II = Bulges, Halos = "fast-moving" (globular clusters, "cluster variables")

Aside: Note potentially confusing, historical nomenclature that "slow-moving" and "fast-moving" (and "low" and "high velocity") here are stated with respect to the Sun, which is itself rotating quickly about the Galaxy. Thus, the true rotational velocities of the stars are actually pretty much the opposite of these labels...

Note also: extragalactic systems analogs to Milky Way populations

D. Nancy Roman (1954), in a spectroscopic study of high and low velocity stars, makes final ties

metal-weak -- high velocity metal-rich -- low velocity age-metallicity groups structure-kinematical groups

E. 1957 Vatican Conference on Stellar Populations:

"Meeting of the world's great astronomical minds" to piece together and organize understanding of Galactic populations.

Subdivide/refine Baade's broad groupings:

Summary tables from the 1957 Vatican Conference proceedings. This book makes great reading, because all of the conversations of participants have been preserved and recorded in the proceedings. Note that the ages listed in the table are based on well outdated stellar evolution models, and are too small by about a factor of two.

F. A “conventional, modern view of the primary Galactic stellar populations and their spatial (density law), chemical, and kinematical properties.

Though it should be kept in mind that this conventional picture is still debated.

Another view of the Milky Way and its populations. From Buser (2000, Science, 287, 5450, 69). His caption: Schematic view of the major components that make up the Galaxy's overall structure, shown in a cross section perpendicular to the plane of rotation and going through the sun and the Galactic center. From the observer's vantage point at the sun's position, the directions to the North (NGP) and South (SGP) Galactic Poles are particularly suitable for studying the layered structure and other properties of the stellar disks and halo, whereas the concentration of gas and dust in the extreme disk severely obstructs observations of the distant bulge at visual-optical wavelengths. The central parts of the Galaxy are better accessible through longer wavelength infrared and radio observations.

Note the difference between the luminous stellar halo, and the dark matter halo postulated to exist and in which the luminous matter is embedded.

Cartoon (left) and modeled (right) illustration of the Galactic dark matter halo. In right figure the plot is only of the density of dark matter in a simulated Milky Way halo, with light on a logarithmic scale and 600,000 light years on a side.

From http://archive.ncsa.uiuc.edu/Cyberia/Cosmos/RotationsReckon.html and http://www.mpa-garching.mpg.de/mpa/research/current_research/hl2003-12/hl2003-12-en.html.

5. From groupings of populations by properties to the first physical evolutionary model of the evolution of Milky Way populations

Eggen, Lynden-Bell & Sandage (1962)

 Collected data on two main groups of stars (nearby disk stars, and high velocity stars) under the assumption that these were representative of range of types in Galaxy (preconceived picture?)

 Found trends between population characteristics -- i.e., stellar pops not just disconnected groups but exhibit correlated, continua of properties, evoking an evolutionary sequence of events (where the chronometer adopted is the UV-excess, an indicator of the degree of metal enrichment).

  ELS formulated a physical model to account for continuum of trends.

We shall discuss at greater length later in the semester, but brief outline of model:

  Galaxy starts as ball of unenriched gas.   Self-gravity, spherical collapse begins.   Stars form during collapse, chemical enrichment of gas proceeds as

populations of stars form.   Stars retain chemical and kinematical character of gas at time of their

formation.   Conservation of angular momentum creates ``spin-up" of gas as shrinks.   Collapse proceeds 25X vertically, 10X radially into a disk.   Collapse postulated to be rapid, freefall timescale (107-108 years or so).   Today:

Early-formed stars are metal-poor stars of outer halo on radial, plunging orbits.

Late formed stars are metal-rich stars of disk and on circular orbits.

6. Some later developments (briefly for now and highly selective):

 1970s:

 Searle & Zinn (1978): Discovery of likely age spread in the halo globular clusters -- longer than ``free-fall" timescale.

 Postulate late infall of material from independent ``fragments".

  Isobe (1974), Saio & Yoshii (1979), Mihalas & Binney (1980): Realization of problems with ELS methodology -- selection effects. Model too simplistic to account for some kinds of stellar populations found (e.g., retrograde stars).

 Need slower formation process for halo.

 First large area radio surveys, discovery of Magellanic Stream and other radio evidence for gaseous filagree in the sky.

The Magellanic Stream and other 21-cm features as presented by Mathewson et al. (1974, ApJ, 190, 291).

 1980s:  Codification of computer structure models (i.e., not really

physical evolution models), ostensibly to aid predictions of star counts in HST observations (Bahcall & Soneira 1980, Robin & Creze 1986).

  Influence of BS80 model, with TWO populations, as well as influence of ELS model, which discussed two main populations creates a trend to simpler, Baade-like description of Galaxy (i.e., halo and disk).

 Yoshii (1982) and Gilmore & Reid (1983) starcount studies of structure of Galaxy and (re-)discovery of Vatican Conference Intermediate Population II (``thick disk") component of Galaxy.

 Extensive debate over existence of thick disk ensues.

 Continuing to today, debate over origin of thick disk.

•  1990s:

– Refurbished HST and other large glass leads to revolution in resolution of extragalactic systems and detailed study of deep CMDs of ever more distant galactic and star cluster systems, deeper study of MW clusters, etc..

Lower panels show unprecedentedly deep CMD of the globular cluster NGC6397 by King et al. (1997), showing stars all of the way down the main sequence to the hydrogen burning limit and including the white dwarf sequence. Top panels show HST proper motions used to clean the cluster of contaminating field stars.

HST color-magnitude diagram of the Leo I dwarf spheroidal galaxy at 270 kpc distance, by Gallart et al. (1999, AJ, 118, 2245).

HST CMDs, velocity distributions and derived age-metallicity distributions of stellar populations in fields of M31 via stellar population synthesis modeling. From Brown et al. (2007, ApJ, 658, L95).

•  Big scopes create explosion of high resolution spectroscopic study of multi-element chemical abundance patterns in nearby galaxies.

Provides detailed understanding of star formation and evolution in galaxies at low redshifts.

Plot showing the distribution of [O/Fe] vs. [Fe/H] in Galactic stars. The plot gives information on the relative constibutions of Type Ia and Type II supernovae at different times in the Galaxy's history. From Matteucci & Recchi 2001, ApJ, 558, 351.

 More recently, with access (last decade) to stars in more distant MW satellites, provides some puzzles w.r.t. the prevailing models of hierarchical galaxy formation.

Plot showing the distribution of [O/Fe] vs. [Fe/H] in Galactic stars (small symbols) compared to a MilkY way satellite (the Sculptor dSph, pink symbols). The plot shows that the putative subcomponents used to build large galaxies like the Milky Way have different abundance patterns. From Geisler et al. 2005, 129, 1428.

 Detailed patterns of more "obscure" elements provide promise of "chemical fingerprinting" of stars to their birth sites.

Summary of chemical abundances of a variety of elements in galaxies of the Local Group. From Gibson 2007, IAU Symp. 241: Stellar Populations as Building Blocks of Galaxies, eds. A Vazdekis & R. Peletier.

 HST and big scopes bring new insights into galaxies evolution at high redshifts.

Can image morphology, obtain photometry of presumed high redshift analogues to Local Group galaxies.

•  QSO absorption line studies bring information on chemical and dynamical state of gas in galaxies at high redshift.

Damped Lyman Alpha Cloud at z=2 seen in absorption of a higher redshift QSO (note emission line). The amount of absorption needed to make a trough this deep is consistent with the mass thought to exist in spiral disks. Could these be young versions of the Milky Way?

From zebu.uoregon.edu/~imamura/ 209/may10/baryon.html

The so-called "Madau plot", which shows the evolution of the star formation rate with time, derived for Hubble Deep Field galaxies by Madau (1998). If the plot were made as a function of time instead of redshift, the peak would be more symmetric and centered on 60% of the look-back time from the present age of the universe.

• Growth of realistic N-body codes. Better models of Galaxy formation

and interactions.

N-body model of the collapse of a Milky Way-like galaxy (a la Eggen, Lynden-Bell & Sandage), with gas (yellow), stars (red), and dark matter (purple), by Peter Williams and Alistair Nelson (The Edinburgh Parallel Computing Center).

•  Models of the growth of structure in the universe promote notion of the importance of Cold Dark Matter in the universe, and the possibility of a universal mass profile (e.g., the "Navarro-Frenk-White" or NFW profile) for hierarchical structure growth.

where a is a scale length and ρ0 is a normalizing density.

N-body model of the growth of structure showing gas (left) and dark matter (right) by Navarro, Frenk & White (1995, MNRAS,275,720).

 Growing recognition of importance of local dark matter in the Galaxy. Searches for its source, distribution.

 Growing recognition of possible importance of late infall, accretion of small satellites, in formation of Galaxy (as predicted by Cold Dark Matter theories for growth of structure in the universe).

 Discovery of a potentially disrupting Milky Way dwarf galaxy, the Sagittarius dSph.

Image from Ibata et al. (1995, MNRAS, 277, 781).

•  First quality large-scale, multiwavelength views of Milky Way. E.g., – Infrared. . .

Image of the Milky Way from the COBE satellite.

Image of the Milky Way from the IRAS satellite.

• Radio. . .

Neutral hydrogen, 21-cm map of the Milky Way from Dickey & Lockman (1990).

• Microlensing surveys . . .

Magnification of apparent brightness by microlensing.

 Discovery of brown dwarfs, extrasolar planetary systems as extension of stellar evolution theory and possible dark matter source (but seems less likely now, based on microlensing experiments).

Spectrum of the brown dwarf Gl229B compared to the very similar spectrum of Titan, the gaseous satellite of Saturn. From Gordon Walker, 1998 C.S. Beals Lecture, www.astro.ubc.ca/E-Cass/1998-JS/ thetalk/thetalk.html.

•  First clear signature of a dwarf galaxy stellar tidal stream:

Map of the distribution of M giant stars seen in the 2MASS survey, in two apparent magnitude bins. The Sagittarius stream is evident. From Majewski et al. (2003, ApJ, 599, 1082).

The distribution of M giant stars seen in the 2MASS survey seen in projection onto a plane perpendicular to the Galactic plane. The Sagittarius stream is seen wrapping around the Milky Way in a nearly polar orbit. From Majewski et al. (2003, ApJ, 599, 1082).

•  The substructure of the outer Galaxy becomes clear:

The "field of streams" from the SDSS survey. From Belokurov et al. (2006, ApJ, 642, L137).

• Missing satellites problem.

Current CDM models predict that the Milky Way should have hundreds of satellites. Model from Ben Moore's Zurich group. From http://cfcpwork.uchicago.edu/seminars/talks/040206/slideshow/3.html

From http://cfcpwork.uchicago.edu/seminars/talks/040206/slideshow/4.html

•  Expectation of luminous halo substructure from late infall of satellites.

CDM models mated with tidal disruption models make predictions of halos filled with streams. Models from Kathrn Johnston and James Bullock.

 Discovery and mapping of actual stellar streams around the Milky Way and Andromeda, and convergence of observation and theory of halo substructure.

(Left) Known Milky Way streams as mapped by Majewski & Law. (Right) Model of Milky Way from Bullock & Johnston model.

 In past few years (!) a substantial increase in the number of known Local Group dwarf galaxies (mostly satellites of Milky Way and Andromeda).

About a dozen new Milky Way dwarf galaxy satellites have been found in the SDSS data (doubling known number).

Some Challenges of Stellar Population Studies Physical Limitations   Difference between OBSERVED and BIRTH properties

Some information on the past can be erased and be unrecoverable. For example:   Dynamics:

○  Stochastic ``scrambling" of stellar motions by dynamical relaxation, scattering off of Giant Molecular Clouds, etc.

○  Tidal/evaporative processes act to winnow small star systems (e.g., dwarf galaxies, star clusters) down; change their appearance or remove them completely from sample.

  Chemistry: ○  Convection in stars alters chemical abundance patterns present in atmospheres --

dredges up/introduces new elements, destroys others. ○  Astration: Stars pick up material from ISM (small effect).

  Problems of perspective. For example:   Some aspects of MW difficult to study. For example:

○  No clear view of bulge. However, Skrutskie/2MASS working in the infrared (less extinction): barred MW.

2MASS image showing relative densities of 30,000 carbon stars in the disk of the Milky Way. Our position marked with green cross. The carbon stars appear to show the Galactic Center to have a bar shape, tilted to the line of sight.

 Compare to external views of other galaxies, e.g., M31, for which a central bar has just been discovered by Beaton, Majewski, Skrutskie et al. (2006):

 We know rotation curves of external galaxies much better than that of MW. Cannot do outside of solar circle very well.

Summary of measures of the Galactic rotation curve. From Heather Morrison's web page: http://smaug.astr.cwru.edu/heather/222/mar1.html.

 What is on other side of Galaxy or beyond disk? (e.g., recent Sloan ``ring"; Sgr dwarf galaxy)

Schematic representation of the Sagittarius dwarf behind the Galactic bulge (R. Wyse, JHU.).

 Connecting Milky Way to extragalactic context without similar, external view.

For example:   Until Morgan

(1950's) recognized nearby spiral arms in MW, still uncertain whether we were a spiral galaxy.

Discovery of the Galactic spiral arms by Morgan, Whitford & Code's (1953, ApJ, 118, 318) observations of blue supergiant stars.

 Now some understanding of spiral structure of the Milky Way, though still incomplete due to obscuration problems.

(Left) Our best guess at the spiral arm structure of the Milky Way. Note that the one arm should be called "Scutum-Crux" (not "Scuturn-Crux"). From wikimedia. (Right) Schematic view of our current notion of Milky Way structure. From http://www.daviddarling.info/images/Milky_Way_schematic.jpg.

  No rotation curve, can't accurately put MW on Tully-Fisher relation.   Current best evidence puts Milky Way 2-sigma off the T-F relation.   As may be seen by some examples given above, there is hope that

problems of perspective can eventually be solved.

The Tully-Fisher Relation correlates the mass of a galaxy, as measured by the rotation curve, typically measured in HI, and the optical or NIR luminosity of the galaxy. The relation is a useful rung on the distance scale ladder. From www.astro.columbia.edu/ ~bureau/astronomy.html.

Technical Limitations  Difficult to study stars in other galaxies to level in MW, make

comparisons.

  HST, CCDs, large telescopes, adaptive optics making/will make huge advances here possible.

 Outer reaches of Milky Way challenging, particularly for dynamics and parallaxes (distance scale problem).

  SIM, GAIA will bring astrometry to microarcsecond levels, enabling motions of stars in outer Galaxy to be measured well.

  Accretion means many subsystems (subpopulations) in Galaxy. Systematic mapping needed.

  Huge databases, like Sloan, 2MASS, Hipparcos, and faster computers are allowing great inroads into mapping Galactic phase space.

Representations of Complex Populations in Multivariate Space

SOME EXAMPLES YOU WILL SEE IN THIS CLASS   Most commonly, will use color-magnitude diagrams.   We will learn how to interpret such diagrams to understand

something about the star formation history of a system.   For example, decomposing the Carina Galaxy into SSPs:

Color-magnitude diagram of the Carina dwarf spheroidal galaxy by Smecker-Hane et al. (1994, AJ, 108, 507) showing evidence of the superposition of multiple simple stellar populations.

Evidence for three bursts of star formation but with similar

abundance

• We may also represent the chemical and star formation history of a galaxy by way of the HODGE POPULATION BOX (Hodge 1989).

– AGE vs. [Fe/H] vs. STAR FORMATION RATE

– Note definition of metallicity [Fe/H]:

• Note that the Hodge box variables are derived, not observed.

Hodge Population Box for Carina rather simple:

 Hodge Population Box for a more complex (fictional) galaxy:

 More or less continuous star formation could be one "principal population component" made up of numerous SSPs (e.g. an enriching disk) ... or multiple pops -- need dynamics

From Grebel (199x).

 Current estimation of the Milky Way's Hodge Population Box, as viewed from above

An estimation of the Hodge Population Box for the Milky Way, using a variety of available data. See Majewski (1999, in Globular Clusters, eds. C. Martinez Roger et al., Cambridge Univ. Press).

• Chemodynamical models require dynamical constraints/ observations as well

– We can construct, analogously to Hodge box, a DYNAMICAL POPULATION BOX, with one axis a measure of the ratio of ORDERED MOTION (Vrotation) to DISORDERED MOTION (σ = velocity dispersion):

Unfortunately, hard to construct this information for external galaxies:   Line-of-sight, Doppler velocities hard enough, integrated information, v sin

i problems.

  Proper motions? Even SIM (launch sometime this decade??) will measure only very nearest galaxies

  Milky Way is one "laboratory galaxy" where we can get detailed info on chemistry, age, r, and kinematics of stellar populations (we assume M.W. is "typical").

e.g. Eggen, Lynden-Bell and Sandage Formation Model:

An estimation of the Dynamical Population Box for the Milky Way, using a variety of available data. See Majewski (1999, in Globular Clusters, eds. C. Martinez Roger et al., Cambridge Univ. Press).

READING ASSIGNMENTS:

•  Baade (1944) articles.

• Majewski ARAA article, Introduction and Section 1.

• Majewski Canary Island article, Section 1.1.

•  Binney & Merrifield, Chapter 1.