Galactic Archaeology II Conference

Conference (website) in Palm Cove, QLD, Australia.

(Day 1) Tuesday, 27 May 2014

Will transcribe my notes someday.

(Day 2) Wednesday, 28 May 2014

Marino – Multiple GC Populations

  • Observational evidence in various GCs do not match single stellar populations (SSPs).
  • Second parameter: GCs with the same metallicity show different HB morphologies. See Catelan (2009) for a review.
  • Other evidence: multiple/spread main sequence, split subgiant and red giant branches.
  • Prototype of a normal GC is M4 (Marino+ 2008). 2nd generation of stars enriched by AGB (D’Antona+ 2004; Karakas+ 2014), fast rotators (Decressin+ 2007), or massive binaries (de Mink+ 2012).
  • NGC 2808: Multiple HBs – He enhanced HB stars.
  • M22: Prototype of an anomalous GC (Marino+ 2009; Lind+ 2014). Bimodality in s-elements; internal variation in the overall metallicity. Two sub-giant branches in terms of chemical composition (Marino+ 2012).
  • Open issues: Na-O anti-correlation in the normal GC M4 is explained by a two generation scenario.  However, the situation is more complex for anomalous GCs – it is currently unclear how the intra-cluster chemical enrichment proceeds.
  • How massive were these objects in the past?
  • GCs with variations in s-process and metals: ωCen, NGC 1851, M22, M54, M2, NGC 5824, NGC 5286.
  • Open issues (II): Some GCs are remnants of stripped dwarf galaxies, i.e. ωCen, which can retain fast SN ejecta.
  • NGC 1851: Strong observational evidence that its halo is only populated by s-poor stars.

Yong – Galactic Archaeology and GCs

  • There are many systematic and random errors in observations that need to be corrected through “strictly differential analysis” (Meléndez+ 2009; Nissen & Schuster 2010).
  • There are correlations between all elements. 25 of 66 have a 5-sigma correlation. Shows that NGC 6752 is chemically inhomogeneous (and perhaps all GCs). Caused by He variations?
  • Open clusters do not possess the Na-O anti-correlation. There are no observed elemental variations so far.
  • (Liu, Yong, et al., in prep) Some preliminary evidence in the Hyades for variations in nickel. Same GC analysis performed, and they found correlations in all elements, suggesting that the Hyades is chemically inhomogeneous.
  • By comparing the “r+s stars” and “r-only stars”, one can conclude that there is some enrichment by AGB stars.

Shingles – s-process Enrichment of M4 and M22

  • Heavy elements (Z>30) are typically constant in a single GC, but vary from cluster to cluster.
  • (Roederer+ 2011) Subtracting the s-poor from the s-rich abundances gives an empirical s-process distribution so it can be compared with stellar yield models.
  • Rotation in stellar yield models boosts s-process yields in massive stars (Frischknecht+ 2012; Limongi & Chieffi 2012).
  • (Herwig+ 2005) s-process element production in stellar models.
  • In M4, a minimum contributing mass of 3-4 solar masses corresponds to a minimum enrichment timescale of 140-290 Myr. In M22, there is a 200 Myr delay.

Karakas – Nucleosynthesis in He-enriched AGB stars

  • Variations in He up to Y = 0.4. The origin of the helium is unknown.
  • (Johnson & Pilachowski 2010) ωCen MDF enriched by four different events. The maximum He enrichment are in the most metal poor stars.
  • (Karakas+ 2014) Few studies evolve the stars beyond helium burning. Stellar yields of He-enriched stars are significant reduced relative to the primordial abundances. C is reduced up to 60%, F up to 80%. He-enhanced stars evolve much faster than primordial abundance stars, i.e. 6 solar masses with Y=0.4 dies after 24 Myr.
  • What fraction of stars between 8-10 solar masses explode as SNe or leave massive WDs?
  • Stars that proceed through C burning and onto the AGB are known as “super-AGB” stars (see Karakas & Lattanzio 2014 for a discussion).
  • Galactic bulge and early-type galaxies also exhibit He-enhancement.

Milone – 2nd parameter in GC HBs

  • Metallicity determines the HB characteristics, i.e. metal-poor have blue HB stars. However, GCs with the same metallicities have different HB properties. Need a 2nd parameter to account for these variations.
  • (Milone+ 2012) The turn-off from MS in NGC xxxx is very narrow but consists of two populations.
  • NGC 2808 is very complex with 5 populations. Delta(Y) ~ 0.13. NGC 6752 (Delta Y ~ 0.03), etc. There are He-variations in all GCs and Y correlates with mass.
  • (Milone+ 2014) Global and non-global parameters of HB morophology in GCs.
  • (Milone+ 2012) Using the properties of the HB to characterize it into metal-poor and metal-rich GCs.

Lind – 3D and NLTE analysis for large stellar surveys

  • (Lind+ 2012) Need to include collisional cross-sections to make accurate predictions for curves of growth with respect to metallicity. NLTE vs. LTE effects can change the metallicity by ~0.2 dex at low metallicities, [Fe/H] < -2.
  • (Keller+ 2014) NLTE corrections to abundance patterns of metal-poor stars, in the paper the [Fe/H] = -7 star. The CNO abundances go down slightly.
  • 3D NLTE calculations takes ~days compared to ~minutes for 1D NTLE.

(Day 3) Thursday, 29 May 2014

Starkenburg – Galactic Archaeology to its Limits

  • See Norris+ (2013) and Carollo+ (2014) for plots that summarize the CEMP stellar observations. Down to [Fe/H] < -7 and [C/Fe] > 4.5 (Keller+ 2014).
  • Two types of CEMP stars: CEMP-s [Ba/Fe] > 1.0 and CEMP-no [Ba/Fe] < 0.0.
  • Why are the stars C-rich?
  • First scenario: They were born with it. Some models predict an over-abundance of carbon in SN [rapid rotators (Chiappini+ 2013) and fallback SN (Tominaga+ 2007)]. Also, C-rich gas can cool more efficiently. Conclusion: Perhaps (CEMP-no).
  • Second scenario: A binary companion dumped metals onto it (Abate+ 2013).  AGB stars make lots of carbon and barium. CEMP-s stars are all contained in binaries. CEMP-no stars are currently being monitored. (Starkenburg+ 2014) Some CEMP-no stars have velocity variations, suggesting that some are contained in binaries. With some careful modeling, they find that nearly all CEMP-s (80-100%) are in binaries with 300 day orbital periods. CEMP-no stars are in a totally different (period vs. binary fraction) phase space than CEMP-s stars. Conclusion: Unlikely for CEMP-no, yes for CEMP-s.
  • Third scenario: Dust-gas winnowing. Elements deplete in the gas because they make dust. Seen in post-AGB & A-type stars. Could other stars show a signature from gas accretion (Venn & Lambert 2008; Venn+ 2014). Also can look at IR-excesses, which is a signature of winnowing. EMP stars do not show this signature. Conclusion: no.
  • (Starkenburg+ 2013) In Sculptor, there are 9 EMP stars but no CEMP stars.
  • C-normal (metallicity floor), CEMP-s (binaries), CEMP-no (2nd generation stars).

Keller – Searching for the oldest stars

  • Overview of Skymapper.
  • Through the use of a metallicity factor (combination of several quantities) on the ANU 2.3m telescope, they identified an EMP for follow-up. With Magellan, they discovered that the iron lines were non-existence. There is detection of Li I 6707A that indicates that envelope had a normal evolution. [Mg/H] = -3.8, [C/H] = -2.6, [Fe/H] < -7.1.
  • The enrichment of this star can be explained by a single SN event. Through abundance patterns, the progenitor is in the range of 10-70 solar masses.
  • Preliminary results of OH lines (UV) shows that O abundance is higher than C.
  • Could compare the simulated MDFs with the observed stars at [Fe/H] < -3.

Beers – AEGIS Followup of Skymapper Targets

  • [Fe/H] < -1: 56.3%, [Fe/H] < -2: 10.4, [Fe/H] < -3: 1.3% out of 70k stars. Many stars with [C/Fe] > 1.
  • Instead of searching for a lack of Ca H/K lines, search for metal weakness in C lines. Found ~300 CEMP stars in the medium-res search in “bad” weather time for 8m-class telescopes.
  • LAMOST: Chinese survey should capture 5-10 million sources.
  • (Aoki+ in press, Science) Metallicity of [Z/H] ~ -2.5 but has an abundance pattern similar to core collapse pattern but its silicon abundance is low. It isn’t carbon enhanced.  The n-capture elements are quite low compared to other VMP stars. Frequences (~1/500) is similar to the high-mass progenitors (Karlsson+ 2008).
  • Have to be careful about calculating the metallicity of stars because of depletion onto dust grains as the second generation of stars form in the molecular clouds.
  • Kobayashi comment: [Mg/Si] is dependent on progenitor mass.

Kennedy – CEMP stars

  • The fraction of CEMP stars increases with decreasing metallicity. However, they’re not all the same because they have different abundance patterns.
  • 80% of all CEMP stars are CEMP-s stars.
  • (Stancliffe+ 2013) Compared their CEMP observations with theoretical models of mass transfer in binaries.
  • (Kennedy+ 2014) Discovered 8 new CEMP RR Lyrae stars, whereas 2 were known before.

Howes – Discovery of the most metal-poor stars in the bulge

  • (Ness+ 2013) Bulge MDF ([Fe/H]) between -1 and +0.2.
  • (Gonzalez+ 2013) Most of the metal-poor stars exist at high galactic latitudes in the bulge.
  • Observing in the bulge brings the problem of dereddening the colors and spectra, which is very difficult!
  • Over 50% have [Fe/H] < -1 and at least 60 have [Fe/H] < -3, where previously EMP stars were never found in the bulge! Velocity dispersion increases with decreasing metallicity. 60 km/s at solar and 130-140 km/s at [Fe/H] = -2 to -3.5.
  • Some follow-ups have yielded two [Fe/H] = -4 stars.
  • In the [alpha/Fe] vs. [Fe/H] relations, there is some indications that the scatter is similar to the halo stars, but in some elements, the scatter is small like previously discovered metal-poor bulge stars.
  • Followed up the metal-poor stars with an orbital analysis, and found that one star is a halo star that is passing through the bulge.  But the two other stars are truly in the bulge with apocenters of 3-5 kpc.

Feltzing – The bulge seen through micro-lensed dwarf stars

  • (Kormendy & Kennicutt 2004) Review of the bulge
  • To constrain the ages and metallicities of dwarf bulge stars, need spectra, but they are dim (mV = 19-20). Use micro-lensing (µ = 20-1000) to overcome this difficulty.
  • The MDF is double-humped (but small number statistics: 86 stars) between [Fe/H] = -1 and +0.5.
  • Metal-rich stars have an age spread over all cosmic time, but metal-poor stars are all older than 10 Gyr.
  • (Ness+ 2014) Compared these age/metallicity results with an isolated galaxy simulation.

Parker – What is the formation history of the Milky Way?

  • Use planetary nebulae to understand how the MW formed.
  • (Saito+ 2011; Weg & Gerhard 2013) X-shaped bulge.
  • (Fulbright+ 2007; Smith+ 2014) Multiple stellar populations in the bulge – age and metallicity spreads.
  • (Freeman+ 2013; Ness+ 2013) 3 kinematic bulge components. Metal-rich bulge, thick bulge, (and something else…)
  • Use distances and WD masses to determine accurate stellar ages. Have discovered ~1200 PNe in the bulge. PN formation rate is the death rate of intermediate mass stars and are very bright, so they can probe to large distances and high extinction.

(Day 4) Friday, 30 May 2014

Tinney – Mapping the Spectra of the Southern Sky with FunnelWeb

  • Focus on bright stars for detailed abundances, seismology, and exoplanets.
  • FunnelWeb Science: Delivering the HD catalog of the 21st century – [Fe/H], [alpha/Fe], T_eff, log(g) down to m=12.
  • Use Skymapper short-cadence data to provide the input catalog. Commissioning in Q1-Q2 2016. 10 fields per night. Up to I < 12 (2.1 million) and I < 14 (9.8 million).

Zucker – Other Galactic Archaeology with HERMES (OGALAHLA)

  • Most of the HERMES time has been allocated to GALAH, whose main science drivers are chemical tagging and galactic archaeology.
  • Abundance groups: Light elements (Li, C, O, Na, Mg, Al), other alpha elements (Ca, Si, Ti), Fe and Fe-peak elements, Light s-processes (Sr, Zr), Heavy s-process (Ba), r-process (Eu).
  • Some examples of other science with HERMES: fainter specific targets, observations in support of other space missions, direct time domain (astroseismology, exoplanets, etc)
  • Halo stars: Up to V ~ 17, can reach the TRGB out to ~100 kpc or HB out to ~25 kpc. Identifying and studying streams in the halos.
  • HERMES will make it possible to study the chemical and dynamical evolution of the Magellanic Clouds.
  • Astroseismology can yield very accurate measurements of radii and masses, which, in turn, yields the gravity and distances.

Da Costa – Can GCs with Fe-spreads be nuclei of dwarf galaxies?

  • There are at least 5 GCs associated with the Sagittarius dwarf / stream, which will be incorporated into the MW’s GC population.
  • NGC 2419: No Fe and Ca spread, but there’s an anti-correlation with Mg and K.
  • M15 and M92: Intrinsic spread in r-process elements, which isn’t seen in other GCs.
  • There are stars in the solar neighborhood that has the same unusual chemical signature as ωCen (Wylie de Boer 2010).
  • (Carretta+ 2010) M54 is at the center of the Sgr dwarf, which in a few Gyr will be just another GC.
  • Other GCs with Fe-spreads (M2, NGC 5824, NGC 6864, NGC 1851, NGC 3201).
  • (Grillmair+ 1995) M2 has a very extended blue HB morphology out to 300 pc with a power-law density profile outside of the main GC.
  • (Marino+ 2014) The halo stars of NGC 1851 have the same signature as its 1st generation of stars.
  • (Norris+ 2010) MDFs for 5 dwarf galaxies, including ωCen.  All of the dwarf galaxy MDF, but ωCen rises very rapidly from low metallicities (rapid enrichment).
  • (Georgiev+ 2009) Shows that nuclear star clusters have similar properties (luminosity and radius).
  • Interesting back of the envelope calculation (see slides) of ~8 former dwarf nuclear star clusters in the halo. But if more GCs are found with substantial Fe-spreads, then this estimate must be re-considered.

Martell – GCs as probes as halo formation

  • GC-halo connections: donors of stars, self-contained objects with cosmological histories.
  • (Odenkirchen+ 2003) Palomar 5 GC is associated to a tidal stream.
  • (Martell+ 2011) 1/2 to 2/3 of original GCs have been completely disrupted. Can find the donor stars that are chemically similar to GC to find their origin.
  • (Gnedin & Ostriker 1997) Stability of GCs in mass-radius phase space.
  • Global properties: Is there a mass threshold for self-enrichment? But it should depend on environment.
  • Detailed properties: anticorrelations, presence / absence of s-process, extent of photometric complexity.
  • In-situ vs. accreted clusters: depends on timing, which affects everything else. Ages shown in Marin-Franch (2009). Bifurcation of formation time based on radius (i.e. blue vs. red GCs), which is seen in other galaxies.
  • (Mackey+ 2010) In M31, the GCs are associated with the streams and can be correlated with recent mergers.
  • It’s hard to compare with local SSCs because they can evolve tremendously after 12 Gyr. Also, the physical conditions at those earlier times are totally different.

Anguiano – Exploring stellar orbits from blind chemical tagging

  • Can the stellar kinematics inform us about stellar ages?  Then, it can be used in conjunction with abundances.
  • (Mitschang+ 2014) Age-metallicity relationship in [Ti/Fe] and [Fe/H].
  • Explored any evolution in orbital parameters and stellar group ages (see slides).

Zwitter – Studies of circumstellar and ISM with spectroscopic surveys

  • There are peculiar spectra from stars in transient modes with emission lines and in binaries.
  • (Roweis & Saul 2000; Daniel+ 2011) Local linear embedding (LLE) makes a projection of data to a low-D space that optimally preserves the local similarity of observed spectra.
  • GALAH can give 6-D mapping of potentially young stars, and 4-D mapping of gas, dust, and DIB carriers in the Galaxy.
  • (Khoperskov+ 2014) Smaller dust grains are driven farther than the Galactic plane.

McMonigal – From the Milky Way to Andromeda: A PAndAS View of Galactic Halos

  • PAndAS: imaging the surrounding 150 kpc around M31 and a little around M33. Difficult but possible to remove all of the foreground contamination.
  • (Martin+ 2013) Source identification of (faint) dwarf galaxies and GCs. Detected And XII (MV = -6.5) at 7-sigma.
  • PAndAS-31: new extended cluster never detected before.
  • (Ibata+ 2014) Analyzing the structure of the halo and dividing the sources into metallicity bins. 10% of the baryonic mass of M31 is contained in the halo stars – 8 x 109 solar masses. The metal poor stellar projected density profile has a slope of -2.3 (-2.5 < [Fe/H] < -1.7; unmasked), -2.08 (masked), -2.7 (-1.7 < [Fe/H] < -1.1; unmasked), -3.7 ([Fe/H]> -1.1; masked).  Check paper for details.
  • (Martin+ 2014) New MW stream detected.

Gilbert – galactic Archaeology in Andromeda

  • SPLASH: spectroscopy and photometry of line-of-sights toward M31, using KPNO and Keck.
  • Many different discoveries with SPLASH (many references; see slides)
  • (Guhathakurta+ 2005; Gilbert+ 2012) Surface brightness profile: 1/R2.2 power law in the halo outside of ~20 kpc. Use spectroscopy to obtain velocities and then the velocities to remove substructures.
  • (Kalirai+ 2006; Gilbert+ 2014) Radial metallicity gradient: [Fe/H] ~ -1.4 at 50-90 kpc with ~150 stars. Newer work uses ~1500 stars out to ~150 kpc. Outside of 90 kpc, there are very little stars with [Fe/H] > -0.5.
  • (Dorman+ 2012, 2013) Found evidence of rotation in the halo and lower velocity dispersion at large radius. Some evidence of disk stars that might have been perturbed into the halo. Interaction with M33?
  • (Fardal+ 2012) From the velocity amplitudes and gradients, one can determine the halo formation history through mergers and subsequent tidal stripping.
  • (Ho+ 2012) Rotational analysis of And II. Its kinematical major axis is misaligned with the isophotal major axis by 67 degrees!
  • (Gilbert+ 2009, 2014) Comparing observations and simulations of surface brightness and [Fe/H]. The correlation between [Fe/H] and surface brightness originates from the Z-L relation of dwarf galaxies.

Bate – A plethora of substructure: the view from the PAndAS

  • (Collins+ 2013, 2014) 16 new dwarfs found. Brighter M31 dSphs are more extended than MW ones (also see Tollerud+ 2012, McConnachie 2012). M31 dwarfs tend to be kinematically colder. If they (XIX, XXI, XXV) are removed from the analysis then the mass profiles are the same as the MW.
  • (Huxor+ 2008, 2011, 2014; Veljanoski+ 2013) Discovered 98 GCs outside 15 kpc, of which 79 lie outside of 30 kpc. Found 3 new GCs in NGC 147. M31 population larger and brighter, and there is a secondary peak around M_V = -6. Clear preference for GCs to be associated with substructure.
  • The outer M31 GC system is rotating at ~100 km/s, which is similar to the inner GC and disk. What does this tell us about the accretion history? (Veljanoski+ 2014)
  • (Mackey+ 2013; Bate+ 2014) There are three GCs possibly associated with the SW cloud. Ages ~6 Gyr. SW cloud has an average [Fe/H] = -1.3. Velocity dispersion 14 ± 3 km/s.

Conn – 3D PAndAS and a Disc of Dancing Dwarfs

  • With a depth, further constrains orbits, metallicities, and link various objects with structures. Use the TRGB stars to measure distances. Using density-matched filters (MCMC) to detect the TRGB to an accuracy of ~0.02 mag.
  • Several of the dwarfs are in the same plane, which could have been accreted through the local sheet or filaments; however, there is a scatter of 60 kpc, where this planar detection is only ~2-sigma. If only the 15 most significant dwarfs are taken, then the plane is ~15 kpc.

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.

*