NEARBY YOUNG STELLAR CLUSTERS: AN OVERVIEW OF IMF S AND AGE SPREADS LYNNE A. HILLENBRAND (CALTECH)
I am only the messenger – please no imf baggage or weapons! Abstract: This talk will summarize methods for measuring the masses and ages of young stars, including the techniques and their challenges, then review recent literature results concerning mass and age distributions in clusters. Some longstanding problems apropos distances, cluster membership, spatial resolution, and sensitivity have been improving, while othersincluding systematics in assumed intrinsic photospheric propertiesand theoretical pre-main sequence evolutionary tracks, remain.
BR BROAD SUMMARIES OF THE IM IMF CO CONCE CEPT AND OB OBSERVE VED OR OR PROP OPOS OSED FUNCTION ONAL FO FORMS (E (ESPECIALLY APRO ROPOS YOUNG-ST STAR-SP SPECIFIC ISSU SSUES) S) • Yorke, Zinnecker – 2009 Salpeter – 1955 • • Elmegreen – 2009 Miller and Scalo – 1979 • • Bastian – 2010 Scalo – 1986 • • Parravano – 2011 Kennicutt - 1998 • • Jeffries – 2012 Larson – 1998 • • Luhman - 2012 Scalo – 1998 • • Kroupa – 2013 Kroupa – 2002 • • Offner – 2014 Chabrier – 2003 • • Maschberger – 2013 de Marchi – 2005 • • Krumholz – 2014 McKee, Ostriker – 2007 •
WH WHAT ARE WE WE TRYI YING TO MEASURE? dN(M)/dM α M - α dN(logM)/dlogM α M Γ Kouwenhoven et al. Hopkins, 2018
SOME REALITIES SO S OF ANY IM IMF ST STUDY We want to infer stellar mass , but we actually measure light ! • Usually we measure as single point sources what are actually multiple star systems - • - combined light that we naively convert to mass. It is not possible to distinguish every individual object given the wide range of binary parameters. For young clusters, inferring an IMF is completely dependent on assumptions about • the cluster age (and any age spread). Metallicity effects are not relevant. Producing the IMF over the full mass spectrum requires a populous enough cluster • to have formed massive stars, and a young enough cluster to still have them. The dynamic range from 20 Msun to 0.1 Msun is ~12 magnitudes in the red optical, • and from 0.1 Msun to 0.01 Msun is another ~7 magnitudes è 4x10 7 in total! There is a wide range of measurement techniques (monochromatic, polychromatic, • spectroscopic) and a wide range of analysis techniques. There is likely some stochasticity in the star formation process. Thus, due to finite • sampling, we might not expect the inferred IMF to be exactly the same everywhere – though it is clearly quite similar in many different types of star-forming environments.
ST STARS S >100-500 500 MY MYR AR ARE Meneses-Goytia et el., 2015 RELATIVELY STRA RE RAIGHTFORW RWARD RD WHEN LOCATED IN CLUSTERS WH • Evolved stars are bright! • Isochrones are readily traced by accurate/precise photometry. • Free parameters of distance and reddening are uniform across the population, and can be solved for. • Some nuisance parameters like metallicity and binaries. • Challenges: upper IMF depleted by stellar evolution and lower IMF can suffer from dynamical evolution.
ST STARS S <10 MY MYR OL OLD, BY CON ONTRAST, ARE A MESS, EV EVEN EN WHEN EN LOCATED ED IN CLUSTER ERS Amard et el., 2019 • Young stars are also relatively bright. • Empirical sequences are much more diffuse due to several effects: Some regions close enough that there is a • distance spread è enhanced dispersion in brightness / luminosity. • Extinction is differential and can be quite large. Reddening can be non-standard. • Activity, especially accretion • Still “nuisance” parameters ** such as binaries. But no metallicity effects. ** well-determined binary parameters are of course an important constraint on s.f. physics
OB OBSERVABLE CON ONSEQUENCES OF OF ACCRETION ON Increasing accretion rate Barensten et al. 2013 - Line emission complicates spectral typing. - Continuum excess distorts colors. Hartmann, Herczeg, Calvet2016
BU BUILDING HR DIAG AGRAM AMS FOR YOUNG STAR ARS [Robinson et al. 2019] • Spectral type from optical or infrared spectra used to estimate photospheric temperature. • Accretion effects should be taken into account during the spectral typing process. • De-reddened photometry plus bolometric correction allows 10 um 1 um luminosity estimate.
A A CONTINUUM OF AC ACCRETION BU BURST BE BEHAVIOR [Cody et al. 2017] At 5-8 Myr, 14% of the objects with disks exhibit with these types of lightcurves
AL ALSO A A CO CONTINUUM OF OF FA FADING/DIPPING BE BEHAVIOR Quasi-periodic Examples Aperiodic Examples Δ mag = 7! Cody and Hillenbrand (2018) [see also Ansdell2016 and Hedges 2018]
VA VARIABILITY TIME SC SCALES S AND AM AMPLITUDES ( AT AT RELAT ATIVELY OLD AGES OF 5-8 8 MY MYR ) 300% Long = Bursters, Stochastics Aperiodic Dippers Int = Quasi-periodic Dippers Quasi-periodic Symm. 10% Short = Periodic Multi-Periodic Amplitude ranges of most ..disk categories are similar. day week month 0.1% Cody and Hillenbrand (2018)
In Infrared ed Va Variability Too (Spitzer data) ~50% of identified variables also vary in color ~33% are periodic ~25% “dip” and ~25% “burst” ~50% “trend” over a month ~20% “stochastic” [ Rebull et al 2014, 2015 ]
STARS ST S <10 MY MYR OL OLD, BY CON ONTRAST, ARE A MESS, EVEN EV EN WHEN EN LOCATED ED IN CLUSTER ERS • Young stars are active, including underlying spottedness plus any superposed accretion effects, both of which cause blue-ingat short wavelengths. Also dust/gas causing red excess at longer wavelengths. • Question of at what wavelengths we can measure mostly the stellar photosphere (vs disk effects) and hence how to best determine extinction correction to account for reddening • bolometric correction from measured flux to luminosity. • • Complication of variability: Median RMS values in the ONC: • use median magnitude? <0.19> mag at 0.8 um <0.14> mag at 1.2, 1.6, 2.2 um • use bright state for dippers/faders? <0.07> mag at 3.6, 4.5 um • use faint state for bursters? Variability tail extends to >2 mag! • Need to be thoughtful regarding techniques.
Amard et al. 2019 PR PRE-MA MAIN SE SEQUENCE EV EVOLUT UTIONARY TR TRACKS STILL CA ST CARRY SY SYST STEMATICS S BE BETWEEN MO MODEL SETS MODEL PHYSICS INPUTS “BIRTHLINE” EFFECTS
ST STELLAR CONTRACTION TH THEORY Despite improvements, pre-main sequence evolutionary tracks still do not reproduce cluster luminosity vs effective temperature sequences. This directly impacts IMF studies for all M < 1 Msun, especially M < 0.1 Msun.
STELLAR CONTRACTION THEORY ST Inferred mass is age-dependent. Results can differ if simply interpolate HRD vs adopt a single age or age +/- sigma(age).
HO HOW ARE WE TRYING TO TO DO DO THE E MEA EASURING? (T (THE EXTRAGALACTIC WAY) Meneses-Goytia et el., 2015 Integrated Light methods 2 Gyr 10 Gyr
HO HOW ARE WE TRYING TO TO DO DO THE E MEA EASURING? (T (THE EXTRAGALACTIC WAY) Attempt to distinguish metallicity, age, and the IMF from single spectra! Dries et el. 2016
HO HOW ARE WE TRYING TO TO DO DO THE MEASURING? (TH THE GALA LACTI TIC / NEARBY STAR FORMING REGIONS WA WAY) Luminosity Function method Maia et el., 2016 / DR17
HO HOW ARE WE TRYING TO TO DO DO THE MEASURING? (TH THE GALA LACTI TIC / NEARBY STAR FORMING REGIONS WA WAY) Infrared CMD Isochrone method Jose et el., 2017 / Stock 8
HO HOW ARE WE TRYING TO TO DO DO THE MEASURING? (TH THE GALA LACTI TIC / NEARBY STAR FORMING REGIONS WA WAY) Optical CMD Isochrone method Suarez et el., 2019 / 25 Ori Log (<M>/ Msun)
HO HOW ARE WE TRYING TO TO DO DO THE MEASURING? (TH THE GALA LACTI TIC / NEARBY STAR FORMING REGIONS WA WAY) Ratio of [stars : brown dwarfs] Muzic et el. 2019 / Rosette Nebula
HO HOW ARE WE TRYING TO TO DO DO THE MEASURING? (TH THE GALA LACTI TIC / NEARBY STAR FORMING REGIONS WA WAY) Spectroscopy method Hosek et al. 2019 / Arches
SP SPECTROSC SCOPY IS S RE REQUIRE RED Dahm & Hillenbrand 2015 Photometry is cheap. Spectroscopy is still the bottleneck!
WH WHAT ARE WE WE TRYI YING TO MEASURE? dN(M)/dM α M - α dN(logM)/dlogM α M Γ Kouwenhoven et al. Hopkins, 2018
WH WHAT IS THE BEST WAY OF COMPARING IMFS AMONG ST STAR FORMING REGIONS S AND ENVIRONMENTS? S? • Discrete IMF shapes è m peak • Gamma-plot (or alpha-plot) • Ratio of stars to brown dwarfs
COMPARING THE MEASURED STELLAR/ SU CO SUB-ST STELLAR IN INIT ITIA IAL MASS FU FUNCTION ON IN THE MILKY WAY dN(M)/dM α M - α dN(logM)/dlog M α M Γ Galactic clusters and OB assoc. Γ = (1 −α ) h/ χ Per; Pleiades; M35; Praesepe Stars in molecular clouds Orion Nebula Cluster NGC 3603, Wd1, Wd2, R136, etc. typical random error in slope after Scalo (1998), Hillenbrand (2004) Log (<M>/ Msun)
Some claims of “top heavy” IMFs in extreme star-forming environments, but evidence is not consistent. Would a ”logistic” type function better describe the form over all masses?
CO COMPARING THE MEASURED STELLAR INIT INITIA IAL MASS FUNC NCTIO ION N AROUND ND THE LOCAL GROUP Weisz (2015)
Recommend
More recommend
