The Physics of the ISM, ISM-SPP 1573, Köln, 14 Feb. 2017

Magnetic Fields Throughout Filament Evolution

Thushara PillaiMax-Planck-Institut für Radioastronomie, Bonn

H. Wiesemeyer, D. Seifried, S. Reissl, J. Kauffmann, S. Wolf, F. Alves, G. Franco, J.C. Tan, R. Banerjee, S. Walch, K. Sugitani, F. Nakamura, K.M. Menten, P.F. Goldsmith, S.J. Carey

Supported by DFG ISM-SPP 1573

Recipe for Star Formation

Gravity Magnetic FieldsTurbulence

==> collapse possible

Canonical Knowledge about Magnetic FieldsAA50CH02-Crutcher ARI 27 July 2012 9:32

101100

101

102

103

102 103 104

nH (cm–3)105 106 107

|BLO

S| (µG

)

Figure 6The set of diffuse cloud and molecular cloud Zeeman measurements of the magnitude of the line-of-sight component BLOS of themagnetic vector B and their 1σ uncertainties, plotted against nH = n(HI) or 2n(H 2) for HI and molecular clouds, respectively(Crutcher et al. 2010). Although Zeeman measurements give the direction of the line-of-sight component as well as the magnitude,only the magnitudes are plotted. The solid blue line shows the most probable maximum values for BTOT (nH ) determined from theplotted values of BLOS by the Bayesian analysis of Crutcher et al. (2010). Also shown (plotted as light blue shading) are the ranges givenby acceptable alternative model parameters to indicate the uncertainty in the model.

There are additional ways to test whether ambipolar diffusion starting from magnetically sub-critical clouds is the driver of star formation. Figure 7 shows BLOS versus NH from the five majorZeeman surveys of HI, OH, and CN (Bourke et al. 2001, Heiles & Troland 2004, Falgarone et al.2008, Troland & Crutcher 2008; K.L. Thompson, T.H. Troland, unpublished observations) andthe compilation by Crutcher (1999); the straight line is for a critical M/". At first glance, thisfigure may seem to show exactly what the ambipolar diffusion model predicts. On the left side,with NH ! 1021 cm−2, mass-to-flux ratios M/" are subcritical; these clouds are almost exclusivelylower density HI clouds. On the right side, with NH " 1021 cm−2, the M/" are overwhelminglysupercritical; these clouds are mainly higher density molecular clouds and cores. Hence, the dataappear consistent with the strong magnetic field model with neutrals gravitationally contracting,leaving the magnetic flux behind and, hence, increasing M/" in the higher density molecular gas.However, there are several problems with this picture. First, the cold HI clouds in the Heiles& Troland (2004) survey are in approximate pressure equilibrium with the warm ISM and arenot self-gravitating, so they could not gravitationally collapse through the magnetic field. Their

www.annualreviews.org • Magnetic Fields in Molecular Clouds 47

Ann

u. R

ev. A

stro

. Ast

roph

ys. 2

012.

50:2

9-63

. Dow

nloa

ded

from

ww

w.a

nnua

lrevi

ews.o

rgby

Cal

iforn

ia In

stitu

te o

f Tec

hnol

ogy

on 0

4/30

/13.

For

per

sona

l use

onl

y.

Crutcher (2010, 2012):

B 100 µG(n/104 cm�3)0.65different regions, different methods

=> hard to probe density dependence

Magnetic Fields in Pristine IRDCs

Infrared Dark Clouds (IRDCs)=> probe magnetic field in early stages of SF

Outline

❖ Setting the Stage: Low Virial Parameters

❖ Observing Magnetic Fields in IRDCs

❖ Alternative Approach: High–Mass Starless Cores

❖ Summary

Setting the Stage: Low Virial Parameters

Starting Point: Low Turbulence in High–Mass Cores

mass & kinematics in G35.20w=> 1.6 mG required

Pillai, Kauffmann, Wyrowski et al. (2011)

virial parameter α ~ 0.3

Evidence for Strong Fields II: Virial Analysis

virial analysis:Bertoldi & McKee (1992)Kauffmann, Pillai & Goldsmith (2013)

Pillai et al. 2011, Kauffmann et al. 2013: "Starless" massive IRDC cores are "overbound"

α ≪ 2 => strong magnetic fields? many cores are in the unstable domain=> short lifetime?

Using Polarized Dust to Observe Magnetic Fields

Recent Surveys

Planck intermediate results. XXXV (Soler)

Taurus MC

Planck resolution of 5-10 arcmin=> beam of 6-12 pc @ 4 kpc!

No. 2, 2009 ANCHORING MAGNETIC FIELD IN TURBULENT MOLECULAR CLOUDS 893

(A)(B)

(C)

(E) (F)

(G)

(H)

(D)

Figure 1. Magnetic fields in the Orion molecular cloud region. The background image shows the IRAS (Neugebauer et al. 1984) 100 µm map in logarithmic scale. Wesuperpose on this map the magnetic field directions inferred from optical data (blue vectors), and the mean of all the optical data is shown as the thick gray vector. TheHertz polarimeter (Dotson et al. 2009) at the Caltech Submillimeter Observatory mapped eight clouds (see labels A through H on the IRAS map) in this region at 350 µmwith 20′′ resolution, and these CSO results are shown as insets, using red vectors on individual false-color intensity maps. The mean direction of all the 350 µmpolarization detections from a given core is shown as a white vector superposed on each core’s map, and these white vectors are also plotted on the IRAS 100 µm map.All the false-color Hertz intensity maps are plotted to the same scale: 140 arcseconds across (approximately 0.3 pc). Note that the spatial scales and mass densities arevery different between the regions probed by the two wavelengths, but the field orientations are very similar.Li et al. (2009, 2015)

maps of individual cores=> no comprehensive overviews

See Poster S2-5; G. Li

Magnetic Fields in Pristine IRDCs

μ ≤ 0.6 => sub-critical

MA ≤ 0.4 => magnetic fields dynamically as important as turbulence

Galactic Center cloud=> measurement in extreme environment

Pillai et al. (2015)

Hertz archival data from archival database Dotson+2010

Magnetic Fields in Pristine IRDCs

Pillai et al. (2015)

archival data: Dotson+ (2010) and Matthews+ (2009)

Pillai et al. (2016)

CN Zeeman

• clouds are sub–Alfvenic• clouds are about critical

Also Busquet's talk (Poster S3-2; Chen)

See Poster S4-1: Brauer on limitations of CN Zeeman

POLSTAR: Magnetic Fields in IRDCs

POLSTAR Survey

(a) simulated dust intensity map

λ = 1.3 mm

0.4 pc

(b) simulated field orientation map

λnear-IR

(diffuse gas)sub-mm

(dense gas)far-IR

(envelope and cores)

simulations(enabled by ISM–SPP)

Magnetic fields remain the major unknown!

(Largely)Quiescent ! High Mass Clusters !Low Mass Clusters!

POLSTAR: POLarization Survey of STAR Forming Filaments wavelengths: Near-IR (Pico Dos Dias), far-IR (SOFIA), sub-mm (APEX) PI: T. Pillai

Gutermuth 2008, Kirk+2013Fernandez-Lopez 2014

Leurini+2011Pillai+06,Kainulainen+2011

Summary

virial parameter=> strong field suggested

dust and CN–Zeeman in IRDCs=> strong field observed

POLSTAR=> multi–scale field studies ongoing

field orientation about constant, and ordered

CoCoAsurvey of all known massive 70 μm dark cores

=> massive starless cores are rare=> magnetic fields do not delay SF?