Chemical Enrichment RGS cluster sample (CHEERS): Constraints on turbulence

1.2 MB·English

by  Ciro Pinto

Astronomy&Astrophysicsmanuscriptno.paper˙cheers˙v2 (cid:13)c ESO2015
January7,2015
Chemical Enrichment RGS cluster sample (CHEERS):
Constraints on turbulence
CiroPinto1,JeremyS.Sanders2,NorbertWerner3,4,JelledePlaa5,AndrewC.Fabian1,Yu-YingZhang6,
JelleS.Kaastra5,AlexisFinoguenov7,andJussiAhoranta7
1 InstituteofAstronomy,MadingleyRoad,CB30HACambridge,UnitedKingdom,e-mail:[email protected].
2 Max-Planck-InstitutfurextraterrestrischePhysik,Giessenbachstrasse1,D-85748Garching,Germany
3 KavliInstituteforParticleAstrophysicsandCosmology,StanfordUniversity,452LomitaMall,Stanford,CA94305-4085,USA
5 4 DepartmentofPhysics,StanfordUniversity,382ViaPuebloMall,Stanford,CA94305-4060,USA
1 5 SRONNetherlandsInstituteforSpaceResearch,Sorbonnelaan2,3584CAUtrecht,TheNetherlands.
0 6 Argelander-Institutfu¨rAstronomie,Universita¨tBonn,AufdemHu¨gel71,53121Bonn,Germany.
2 7 DepartmentofPhysics,UniversityofHelsinki,FI-00014Helsinki,Finland

n Received5November2014/Accepted24December2014
a
J ABSTRACT

6
Context.Feedbackfromactivegalacticnuclei,galacticmergers,andsloshingarethoughttogiverisetoturbulence,whichmayprevent

] coolinginclusters.
E Aims.We aim to measure the turbulence in clusters of galaxies and compare the measurements to some of their structural and
H evolutionaryproperties.
Methods.ItispossibletomeasuretheturbulenceofthehotgasinclustersbyestimatingthevelocitywidthsoftheirX-rayemission
.
h lines.TheReflectionGratingSpectrometersaboardXMM-Newtonarecurrentlytheonlyinstrumentsprovidedwithsufficienteffective
p areaandspectralresolutioninthisenergydomain.Webenefitedfromexcellent1.6MsnewdataprovidedbytheChemicalEnrichment
- RGSclustersample(CHEERS)project.
o
Results.Thenewobservationsimprovethequalityofthearchivaldataandallowustoplaceconstraintsforsomeclusters,whichwere
r notaccessibleinpreviouswork.One-halfofthesampleshowsupperlimitsonturbulencelessthan500kms−1.Forseveralsources,
t
s our data areconsistent withrelatively strong turbulence withupper limitson the velocity widthsthat are larger than 1000kms−1.
a
TheNGC507groupofgalaxiesshowstransonicvelocities,whicharemostlikelyassociatedwiththemergingphenomenaandbulk
[
motionsoccurringinthisobject.Wherebothlow-andhigh-ionizationemissionlineshavegoodenoughstatistics,wefindlargerupper
1 limitsforthehotgas,whichispartlyduetothedifferentspatialextentsofthehotandcoolgasphases.Ourupperlimitsarelarger
v thantheMachnumbersrequiredtobalancecooling,suggesting thatdissipationofturbulencemayprevent cooling, althoughother
9 heatingprocessescouldbedominant.Thesystematicsassociatedwiththespatialprofileofthesourcecontinuummakethistechnique
6 verychallenging,thoughstillpowerful,forcurrentinstruments.Inaforthcomingpaperwewillusetheresonant-scatteringtechnique
0 toplacelower-limitsonthevelocitybroadeningandprovidefurtherinsightsonturbulence.TheASTRO-HandAthenamissionswill
1 revolutionizethevelocityestimatesanddiscriminatebetweendifferentspatialregionsandtemperaturephases.
0
Keywords.X-rays:galaxies:clusters–intergalacticmedium
.
1
0
5
1. Introduction backmaypreventcoolingthroughtheproductionofturbulence
1
(see, e.g., Ruszkowskietal. 2004, Zhuravlevaetal. 2014, and
: Clusters of galaxies are the most massive, individual, bound
v Gasparietal.2014).Otherworksuggeststhatturbulentmixing
i objects in the Universe. In their gravitational potential well, mayalsobeanimportantmechanismthroughwhichAGNheat
X the gas, called the intra-cluster medium (ICM), is heated to
clustercores(see,e.g.,Banerjee&Sharma2014).
r temperatures of 107−8K and, therefore, strongly emits at X-
a
ray energies. The ICM is commonly thought to be in hydro- Itispossibletomeasurevelocitybroadeningontheorderof
static equilibrium, but there are several factors that may af- few hundreds kms−1 directly in the X-ray emission lines pro-
fect the dynamical state of the gas. The feedback from active duced by the hot ICM. The Reflection Grating Spectrometers
galactic nuclei (AGN) creates bubbles that may drive turbu- (RGS, denHerderetal. 2001) aboard XMM-Newton are cur-
lenceuptoabout500kms−1 (see,e.g.,Bru¨ggenetal.2005and rently the only X-ray instruments, which have enough col-
Fabianetal.2005).Sloshingofgaswithinthegravitationalpo- lecting area and spectral resolution to enable this measure-
tential may produce similar velocities, while galactic mergers ment. However, the spatial extent of clusters complicates the
cangiverisetoevenhighervelocitiesofabout1000kms−1(see, process due to the slitless nature of the RGS. Sandersetal.
e.g.,Lauetal.2009,Ascasibar&Markevitch2006). (2010) made the first measurement of cluster velocity broad-
The AGN feedback is thought to offset radiative losses ening using the luminous cluster A1835 at redshift 0.25. Due
and to suppress cooling in isolated giant elliptical galax- to thelimitedspatialextentofits brightcore,anupperlimitof
ies and in larger systems up to the richest galaxy clus- 274kms−1 wasobtained.Sandersetal.(2011)thenconstrained
ters (see, e.g., McNamara&Nulsen 2007 and Fabian 2012). turbulent velocities for a large sample of 62 sources observed
Simulations and observations have confirmed that AGN feed- with XMM-Newton/RGS,which includedclusters, groups,and
2 CiroPintoetal.:ChemicalEnrichmentRGSclustersample(CHEERS):Constraintsonturbulence
ellipticalgalaxies.Halfofthemshowvelocitybroadeningbelow
700kms−1.Recently,Sanders&Fabian(2013)usedcontinuum-
subtracted emission line surface brightness profiles to account
forthespatialbroadening.Thistechniqueisaffectedbysystem-
aticerrorsofupto150kms−1.
Werneretal. (2009) anddePlaaetal. (2012) measuredtur-
bulent velocities through the ratio of the Fexvii emission lines
at15and17Å.Whenthevelocitybroadeningislow,thegasis
opticallythickinthe15Ålineduetoresonantscattering,while
the17Ålinesremainopticallythin.Thecomparisonofobserved
withsimulatedlineratiosfordifferentMachnumbersconstrains
the level of turbulence. This method is very efficient for cool
core clusters rich in Fexvii emission lines, but it is partly lim-
itedbythesystematicuncertainty(∼20%)inthelineratioforan Fig.1. RGS extraction regions and MOS1 stacked image of
opticallythinplasma. M87.
In this work, we measure the velocity broadening for the
44 sources of the CHEmical Enrichment RGS cluster Sample 3.1.RGSandMOS1datareduction
(CHEERS), which is connected to a Very Large Program ac-
ThedatawerereducedwiththeXMM-NewtonScienceAnalysis
cepted for XMM-Newton AO-12. We model the line spatial
System(SAS)v13.5.0.WeprocessedtheRGSdatawiththeSAS
broadeningusingCCDimages.Thismethodhassystematicsdue
taskrgsprocandtheMOS1datawithemproctoproduceevent
tothespatialprofileofthecontinuum,whichmayoverestimate
files,spectra,andresponsematricesforRGSandMOSdata.
the line spatial broadening, but it is still a useful technique to
Tocorrectforcontaminationfromsoft-protonflares,weused
measurethelevelofvelocitybroadeningwhendeep,high-spatial
theSAStaskevselecttoextractlightcurvesforMOS1inthe10–
resolutionmapsarelacking.Wealsotestanalternativemethod,
12 keV energy band, while we used the data from CCD num-
whichusesavariablespatial-broadening.Thepaperisorganized
ber 9 for RGS where hardly any emission from each source
asfollows.InSect.2,wegiveabriefdescriptionoftheCHEERS
is expected. We binned the light curves in 100s intervals. A
project.InSect.3,wepresentthedatareduction.Ourmethodis
Poissonian distribution was fitted to the count-rate histogram,
described in Sect.4. We discuss the results in Sect.5 and give
and all time bins outside the 2σ level were rejected. We built
ourconclusionsinSect.6.Furtherusefulmaterialisreportedin
thegoodtimeintervals(GTI)fileswiththeacceptedtimeevents
AppendixAtospeedupthepaperreading.
fortheMOSandRGSfilesthroughtheSAStasktabgtigenand
2. TheCHEERSproject reprocessedthedataagainwithrgsprocandemproc.TheRGS1
totalcleanexposuretimesarequotedinTable1.
The current catalog includes 44 nearby, bright clusters, groups
ofgalaxies,andellipticalgalaxieswithavalueofa&5σdetec-
tionfortheOviii1s–2plineat19Åandwithawell-represented 3.2.RGSspectraextraction
varietyofstrong,weak,andnoncool-coreobjects.Thiscatalog
We extractedtheRGSsourcespectraintwoalternativeregions
alsocontains19newobservationsof1.6Msin total,whichare
centeredontheemissionpeak:abroader3.4’region,whichin-
takenduringAO-12,PI:J.dePlaa(seeTable1).Moredetailon
cludes most of the RGS field of view and a narrower 0.8’ re-
thesamplechoiceisprovidedbyanotherpaper(dePlaaetal.,in
gionthatprovidestheclustercoresbutwithhighstatistics.This
preparation).Amongthe severalgoals of this large project, we
was done by launching rgsproc twice by setting the xpsfincl
mentionthefollowingones:
mask to include 99% and 90% of point-source events inside
– To understand the ICM metal enrichment by different SN
the spatial source extraction mask, respectively. We have used
types, (see,e.g.,Mernieretal.accepted)
the model background spectrum created by the standard RGS
– tostudysubstructures,asymmetriesandmultiphaseness,
rgsprocpipeline,whichisatemplatebackgroundfile,basedon
– tostudyheatingandcoolinginclustercores,
the count rate in CCD9. The RGS spectral extraction regions
– tomeasureturbulence(thispaper),
andtheMOS1imageofM87areshowninFig.1.Thespectra
– toimprovethecross-calibrationbetweenX-raysatellites. were convertedto SPEX1 formatthroughthe SPEX task trafo.
During the spectral conversion,we chose the option of sectors
3. Data
in the task trafo to create as many sectors as the different ex-
ThedatausedinthispaperarelistedinTable1.Inboldface,we posuresofeachsource.Thispermitsustosimultaneouslyfitthe
showthenewobservationstakenduringAO-12.Afewarchival multipleRGSspectraofeachsourcebychoosingwhichparame-
exposureshavenotbeenused,sincetheyweretooshort. terstoeithercoupleorunbindinthespectralmodelsofdifferent
The XMM-Newton satellite is equipped with two types observations.
of X-ray detectors: The CCD-type European Photon Imaging 3.3.MOS1spatialbroadeningprofiles
Cameras (EPIC) and the Reflection Grating Spectrometers
The RGS spectrometers are slitless, and, therefore, the spectra
(RGS). The European photon imaging cameras are MOS1,
arebroadenedbecauseofthespatialextentofthesourceinthe
MOS2,andpn(Stru¨deretal.2001andTurneretal.2001).The
dispersiondirection.Theeffectofthisspatialbroadeningisde-
RGScameraconsistsoftwosimilardetectors,whichhaveboth
high effective area and spectral resolution between 6 and 38Å scribedbythefollowingwavelengthshift
(denHerderetal.2001).TheMOScamerasarealignedwiththe 0.138
RGS detectors and have higher spatial resolution than the pn ∆λ= ∆θÅ, (1)
m
camera. We have used MOS1 for imaging and RGS for spec-
tralanalysis. 1 www.sron.nl/spex
CiroPintoetal.:ChemicalEnrichmentRGSclustersample(CHEERS):Constraintsonturbulence 3
Table1.XMM-Newton/RGSobservationsusedinthispaper.
Source ID(a) Totalcleantime(ks)(b) kT (keV)(c) z(c) N (1024m−2)(d)
H
2A0335+096 0109870101/02010147800201 120.5 3.0 0.0349 30.7
A85 0723802101/2201 195.8 6.1 0.0556 3.10
A133 01443101010723801301/2001 168.1 3.8 0.0569 1.67
A189 0109860101 34.7 1.3 0.0320 3.38
A262 0109980101/06010504780101/0201 172.6 2.2 0.0161 7.15
A496 0135120201/08010506260301/0401 141.2 4.1 0.0328 6.00
A1795 0097820101 37.8 6.0 0.0616 1.24
A1991 0145020101 41.6 2.7 0.0586 2.72
A2029 01112702010551780201/0301/0401/0501 155.0 8.7 0.0767 3.70
A2052 01099201010401520301/0501/0601/0801 104.3 3.0 0.0348 3.03
0401520901/1101/1201/1301/1601/1701
A2199 0008030201/0301/06010723801101/1201 129.7 4.1 0.0302 0.909
A2597 010846020101473301010723801601/1701 163.9 3.6 0.0852 2.75
A2626 00831502010148310101 56.4 3.1 0.0573 4.59
A3112 01056601010603050101/0201 173.2 4.7 0.0750 1.38
A3526 00463401010406200101 152.8 3.7 0.0103 12.2
A3581 02059901010504780301/0401 123.8 1.8 0.0214 5.32
A4038 02044601010723800801 82.7 3.2 0.0283 1.62
A4059 0109950101/02010723800901/1001 208.2 4.1 0.0460 1.26
AS1101 01478001010123900101 131.2 3.0 0.0580 1.17
AWM7 01359503010605540101 158.7 3.3 0.0172 11.9
EXO0422 0300210401 41.1 3.0 0.0390 12.4
Fornax 00128301010400620101 123.9 1.2 0.0046 1.56
HCG62 011227070105047805010504780601 164.6 1.1 0.0140 3.76
Hydra-A 01099803010504260101 110.4 3.8 0.0538 5.53
M49 0200130101 81.4 1.0 0.0044 1.63
M86 0108260201 63.5 0.7 -0.0009 2.97
M87(Virgo) 01141201010200920101 129.0 1.7 0.0042 2.11
M89 0141570101 29.1 0.6 0.0009 2.96
MKW3s 01099301010723801501 145.6 3.5 0.0450 3.00
MKW4 00930601010723800601/0701 110.3 1.7 0.0200 1.88
NGC507 0723800301 94.5 1.3 0.0165 6.38
NGC1316 03027801010502070201 165.9 0.6 0.0059 2.56
NGC1404 0304940101 29.2 0.6 0.0065 1.57
NGC1550 01521501010723800401/0501 173.4 1.4 0.0123 16.2
NGC3411 0146510301 27.1 0.8 0.0152 4.55
NGC4261 00563401010502120101 134.9 0.7 0.0073 1.86
NGC4325 0108860101 21.5 1.0 0.0259 2.54
NGC4374 0673310101 91.5 0.6 0.0034 3.38
NGC4636 0111190101/0201/0501/0701 102.5 0.8 0.0037 2.07
NGC4649 00215402010502160101 129.8 0.8 0.0037 2.23
NGC5044 00379501010584680101 127.1 1.1 0.0090 6.24
NGC5813 03024601010554680201/0301/0401 146.8 0.5 0.0064 6.24
NGC5846 0021540101/05010723800101/0201 194.9 0.8 0.0061 5.12
Perseus 0085110101/02010305780101 162.8 6.8 0.0183 20.7
(a)ExposureIDnumber.(b)RGSnetexposuretime.(c)RedshiftsandtemperaturesareadaptedfromChenetal.(2007)andSnowdenetal.(2008).
(d)Hydrogencolumndensity(seehttp://www.swift.ac.uk/analysis/nhtot/).Newobservationsfromourproposalareshowninboldface.
where m is the spectral order and θ is the offset angle of the 4. Spectralmodeling
sourceinarcmin(seetheXMM-NewtonUsersHandbook).
TheMOS1DETYdirectionisparalleltotheRGS1disper- Ouranalysisfocusesonthe7−28Å(0.44−1.77keV)firstand
siondirectionandcanbeusedtocorrectforthespatialbroaden- second order RGS spectra. We model the spectra with SPEX
ing. With evselect, we extracted MOS1 images for each expo- version 2.03.03. We scale elemental abundances to the proto-
Solar valuesof Lodders&Palme (2009), which are the default
surein the 0.5–1.8keV(7–25Å) energybandandtheirsurface
inSPEX.WeadoptC-statisticsand1σerrorsthroughoutthepa-
brightness profiles in the dispersion direction. We account for
per,unlessotherwisestated,andtheupdatedionizationbalance
spatialbroadeningusingthelpromultiplicativemodelinSPEX,
calculationsofBryansetal.(2009).
whichconvolvestheRGSresponsewithourmodelofthespatial
extentofthesource.Weshowsomecumulativeprofilesofspa- Clusters of galaxies are not isothermal, and most of them
tial broadeningin Fig. 2. We have also producedstacked Fe-L have both hot and cool gas phases (see e.g. Franketal. 2013).
band(10–14Å)imagesforeachsource.Thecentral10’region Therefore, we have used a two-temperature thermal plasma
containsmostoftheclusteremission(seeFig.A.1). model of collisional ionization emission (CIE). This model is
4 CiroPintoetal.:ChemicalEnrichmentRGSclustersample(CHEERS):Constraintsonturbulence
profile as given by the other available observations, but the δλ
parameterin thelpro componentisleftfree.Thisfactorallows
us to shift the modellines by the same amount(in Å) for each
specificspectrumandstronglydecreasesthesystematiceffects.
Theδλparameterisalwaysfreeinourfitstoaccountforanyred-
shiftvariation,which wouldotherwise affectthe line modeling
(seee.g.Sandersetal.2011).
The simultaneous modeling of multiple observations has
been done through the use of the sectors option in SPEX (see
alsoSect.3.2).TheRGS1and2spectraofthesameobservation
haveexactlythesamemodelandprovideasinglesector,while
RGSspectraofotherobservationscontributeadditionalsectors
andhavethecienormalizationsuncoupled.
4.1.Resultsusingafixedspatialbroadening
Fig.2.MOS1average7–25Åcumulativespatialprofiles.
We have successfully applied this multi-temperature model to
boththe3.4’and0.8’RGSspectra.Weshowthespectralmod-
elingforthe3.4’regionofthe44sourcesinFigs.A.5,A.6,and
able to fit all the spectra in our database. The cie model in
A.7inAppendixA.Wedisplaythefirst-orderstackedspectrato
SPEXcalculatesthespectrumofaplasmaincollisionalioniza-
underlinethehighqualityoftheseobservationsandtoshowthe
tionequilibrium.Thebasisforthismodelisgivenbythemekal
goodnessofthemodeling.
model, but several updates have been included (see the SPEX
Forsome sourceslike Fornax,M49,M86,NGC4636,and
manual). Free parameters in the fits are the emission measure
NGC5813,the 15 and 17Å Fexvii emission lines are notwell
Y =n n dV,thetemperatureT,theabundances(N,O,Ne,Mg,
e H fitted. Precisely, the model underestimates the line peaks and
andFe),andtheturbulentbroadeningvofthetwociemodels.
overestimatesthe broadening.This may be due to the different
Webindtheparametervandtheabundancesoftwociecom-
spatial distribution of the gas responsible for the cool Fexvii
ponents with each other and assume that the gas phases have
emission lines and for the one producing most of the high-
thesameturbulenceandabundances.Thisdecreasesthedegree
ionization Fe-L and Oviii lines. The cool gas is indeed to be
of degeneracy.This assumption is certainly not true, but some
foundpredominantlyinthecenteroftheclustersshowingapro-
clusters just need one CIE component and the spectra of sev-
file morepeakedthan thatoneof the hotter gas.The estimated
eral clusters do not have good enough statistics in both high-
spatialprofilesdependonthe emission ofthe hottergasdueto
andlow-ionizationemissionlines,whichprohibitsconstraining
its higher emission measure, and, therefore, they overestimate
thevelocitiesandtheabundancesforbothhotandcoolphases.
thespatialbroadeningofthe15–17Ålines.Itishardtoextracta
Weattempttoconstraintheturbulenceinthedifferentphasesin
spatialprofilefortheselinesbecauseMOS1hasalimitedspec-
Sect.5.1.
tral resolution, and the images extracted in such a short band
The cie models are multiplied by a redshift model and a
lack the necessary statistics (see e.g. Sanders&Fabian 2013).
model for Galactic absorption, which is provided by the hot
In Sect.5.1, we attempt to constrain the turbulencefor lines of
modelinSPEXwithT =0.5eVandNTOT,asestimatedthrough
H different ionization states. The 15Å/17Å line ratio is also af-
thetoolofWillingaleetal.(2013).Thistoolincludesthecontri-
fected by resonant scattering, which would require a different
butiontoabsorptionfrombothatomicandmolecularhydrogen.
approach.Werefertoaforthcomingpaperontheanalysisofthe
The redshifts and column densities that have been adopted are
resonantscatteringintheCHEERSsources.
shown in Table 1. To correct for spatial broadening, we have
multiplied the spectral model by the lpro component that re- Weskipthediscussionoftheabundancesandthesupernova
ceives as input the surface brightness profile extracted in the yieldsbecausethesewillbetreatedbyotherpapersofthisseries
MOS1images(seeSect.3.3andFig.2). (dePlaaetal.inpreparationandMernieretal.submitted).
WedonotexplicitlymodelthecosmicX-raybackgroundin In Fig.3 (left panel), we show the upper limits on the ve-
theRGSspectrabecauseanydiffuseemissionfeaturewouldbe locitybroadeningobtainedwiththesimultaneousfitsofthe0.8’
smearedoutintoabroadcontinuum-likecomponent. 7–28Å RGS spectra. We obtain upper limits for most clusters,
For a few sources, including the Perseus and Virgo (M87) while NGC507 shows high kinematics. More detail on our re-
clusters, we have added a further power-law emission compo- sults for the 3.4’ and 0.8’ regionsand their comparisonare re-
nent of slope ∼ 2 to take the emission from the central AGN portedinTableA.1andFig.A.2(leftpanel).The3.4’limitsare
into account.Thiscomponentis notconvolvedwith the spatial moreaffectedbythesourcecontinuum,asclearlyseenforM87,
profilebecauseitisproducedbyacentralpoint-likesource. AWM7,andA4038,whichmakesthemlessreliable.
To avoid the systematic effects due to the stacking of mul-
tiple observations with different pointing, we have simultane-
4.2.Resultsusingthebest-fitspatialbroadening
ouslyfittedtheindividualspectraofeachsourceextractedinthe
tworegionsdefinedinSect.3.2andshowninFig.1.Theplasma Itisknownthatthespatialprofileofthesourcecontinuummay
modeliscoupledbetweentheobservations.Theonlyuncoupled bebroaderthanthespatialdistributionofthelines.TheMOS1
parameters are the emission measures of the two collisional- imagesarestronglyaffectedbytheprofileofthesourcecontin-
ionizedgascomponents.Foreachobservationweadoptthespa- uum and, therefore,may overestimatethe spatial line broaden-
tialprofileextractedintheMOS1imagetakenduringthatexpo- ingandunderestimatetheresidualvelocitybroadening.Forin-
sure.Forthoseexposures,duringwhichtheMOS1detectorhad stance,NGC1316andNGC5846show1σlimitsof20kms−1,
a closed filter, we have adopted an exposure-weightedaverage whicharenotrealistic(seeTableA.1).
CiroPintoetal.:ChemicalEnrichmentRGSclustersample(CHEERS):Constraintsonturbulence 5
Fig.3.Leftpanel:Velocity68%(red)and90%(green)limitsforthe0.8’regionwiththespatialbroadeningdeterminedwithMOS1
images(seeSect.4.1).Rightpanel:Velocitylimitsobtainedusingthebest-fitspatialbroadening(seeSect4.2).
Toobtainmoreconservativelimits,wehavesimultaneously 4.3.Furthertests
modeledthespatialandthevelocitybroadening.Thiswasdone
by fitting the RGS 0.8’ spectra with a free s parameter in the
lprocomponent.Thisfactorsimplyscalesthewidthofthespa- Toestimatethecontributionofthespatialbroadeningtotheline
tial broadening by a factor free to vary (see the SPEX man- widths, we have temporarily removed the convolution of the
ual).Thefreesparameterincreasesthedegeneracyinthemodel spectral model for the spatial profile and re-fitted the data. In
but provides conservative upper limits on the residual velocity thesefitsthevparameteroftheciecomponentaccountsforany
broadening,which is measured with the v parameterof the cie contributiontothelinebroadening.Thetotal(spatial+velocity)
component.ThenewlimitsonthevelocitiesareplottedinFig.3 widthsarealsoquotedinTableA.1.
(right panel) and quoted in the last two columns of TableA.1.
In Fig.A.2 (rightpanel), we comparethe velocity upperlimits We have also tested the continuum-subtracted line surface
estimatedwiththestandardmethod(MOS1spatialprofilewith brightnessprofilesintroducedbySanders&Fabian(2013).This
s ≡ 1 in the lpro component) with this new approach using a new method consists of subtracting the surface brightnesspro-
freesparameter.Theygenerallyagree,butthenewupperlimits files of two regions that are clearly line-dominated (core) and
onthehotterAbellclustersaresystematicallylargerbyanaver- continuum-dominated(outskirts).Itcanbeappliedonlytothose
agefactor∼ 2.Thisconfirmsthatsomeofthepreviousvelocity objects with a narrow core where it is possible to distinguish
limits were underestimated due to the broader spatial profiles. between line-rich and line-poor regions. We have locally fit-
Therefore,webelievethenewupperlimitstobethemostconser- ted the Oviii 19.0Å emission line of A2597, A3112, Hydra-
vative. Interestingly, the conservative velocity limits of the hot A, Fornax (NGC1399), and NGC4636 and we have found a
clusters are generally higher than the cool galaxy groups with generalagreementwiththeresultsofSanders&Fabian(2013).
the exception of NGC507, which is expected since the sound However, our MOS1 images have much lower spatial resolu-
speedscalesasapowerofthetemperature(seeSect.5.2). tionthanthe Chandramapsusedbythem,whichincreasesthe
uncertaintiesthatarepresentinthismethod.Athorough,exten-
sive,analysiswouldrequiredeepChandramapsthatarenotyet
available.
6 CiroPintoetal.:ChemicalEnrichmentRGSclustersample(CHEERS):Constraintsonturbulence
5. Discussion
Inthisworkwehaveanalyzedthedataof44clusters,groupsof
galaxies,andellipticalgalaxiesincludedintheCHEERSproject,
aVeryLargeProgramthatwasacceptedforXMM-NewtonAO-
12(seeSect.2)togetherwithcomplementaryarchivaldata.
We have measured upper limits of velocity broadening for
theseobjectswithamethodsimilartothepreviousoneusedby
Bulbuletal.(2012)andSanders&Fabian(2013).Thisconsists
of fitting high-quality grating spectra by removing the spatial
broadeningthroughsurfacebrightnessprofilesofthesourcesas
provided by CCD imaging detectors. These profiles are unfor-
tunatelyaffectedbythe sourcecontinuumandtendtooveresti-
matethelinespatialbroadeningwithaconsequentshrinkingof
the residual velocity broadening. Sanders&Fabian (2013) ad-
dressed this point in their Sect.2.2 on A3112 where they de-
creased these systematic effects by using Chandra continuum-
subtracted line spatial profiles. We have tested this method by
using the MOS1 observations that were taken simultaneously Fig.4.90%upperlimitsonvelocitybroadeningobtainedinthe
withtheRGSspectra(seeFig.A.1).XMM-NewtonCCDshavea 0.8’ region for the Oviii lines compared with those measured
spatialresolutionlowerthanChandraCCDs,whichincreasethe for high-ionization Fexx (open red triangles) and for the low-
systematiceffectsinthecreationofcontinuum-subtractedmaps. ionization Fexvii (filled black triangles) line systems. For six
DeepChandraobservations,enablinganaccuratesubtractionof sourceswecouldalsomeasurethelimitsforthehot(opengreen
different energy bands, are missing for most sources. We have circles)andthecool(openmagentaboxes)CIEcomponents(see
thereforetriedtousetheMOS1integralmapsandtofitthecon- Sect.5.1).
tributionofthespatialbroadeningasanalternativemethod.
5.1.Temperaturedependenceoftheupperlimits 5.2.Turbulence
Sofarweadoptedthesamevelocitybroadeningforalltheemis- In Fig.3 we show the velocity broadening of the RGS spectra
sionlines.Formostsourcesitispossibletomeasurethevelocity extractedinthe0.8’coreregion.Wefindupperlimitstotheve-
broadeningoftheOviiiandFexx-to-xxivemissionlines,which locitybroadeningwiththepossibleexceptionofNGC507.They
aremainlyproducedbyhotgas.Onlyafewsourceshavehigh- generallyrangebetween200and600kms−1.Forseveralobjects
statisticsFexviilinesproducedbycool(T <1keV)gas.Sixob- like A85, A133, M49, and most NGC elliptical we found ve-
jectsexhibitbothstronglow-andhigh-ionizationlinesandallow locitylevelsbelow500kms−1,whichwouldsuggestlowturbu-
to fit the velocitybroadeningof the two cie components,sepa- lence.Thebroader3.4’regionismoreaffectedbyspatialbroad-
rately,inthefull-bandspectralfits.17sourcesallowtomeasure ening as shown by the higher upper limits, but non-detection,
90%upperlimitsonturbulencefortheOviii,Fexvii,andFexx ofA4038,AWM7,andM87(see TableA.1andFig.A.2).For
lines,byfittingthe18.0−23.0Å,14.0−18.0Å,and10.0−14.3Å these sourcesit is difficultto constrainthe velocitybroadening
becausetheirlargeextentsmearsouttheemissionlines.
rest-frame wavelength ranges, respectively. For each local fit
we adopt an isothermal model and correct for spatial broad- Tounderstandhowmuchenergycanbestoredinturbulence,
eningby usingadditionalsurface-brightnessprofilescalculated we compare our upper limits with the sound speeds and the
throughMOS1 images extracted in the same rest-frame wave- temperatures of dominant cie componentin these objects. The
lengthrangesusingthesamemethodshowninSect.3.3.These soundspeedis givenbycS = γkT/µmp, whereγ isthe adi-
profiles are still affected by the continuum but provide a bet- abatic index, which is 5/3 for pideal monoatomic gas, T is the
ter description of the spatial broadening in each line. In Fig.4 RGS temperature, µ = 0.6 is the mean particle mass, and mp
we compare the Oviii velocity limits with those measured for isprotonmass.Theratiobetweenturbulentandthermalenergy
theFexviiandFexxlinesystems.Thehigh-ionizationFe lines isεturb/εtherm = γ/2M2,where M = vturb/cS istheMachnum-
clearlyshowhigherupperlimits,whichisconfirmedbythere- ber (see also Werneretal. 2009). In Fig.A.2 we compare our
sultsofthefull-bandfits:thehotter(T1)CIEcomponentallows 2σupperlimitsonthevelocitiesinthecentral0.8’regionwith
for highervaluesof velocitybroadening.The hotter gasis dis- thesoundspeedandsomefractionsofturbulentenergy.Wealso
tributedoveralargerextentthanthatofthecold(T2)gasandhas showthemoreconservativevelocityupperlimitsthatweremea-
largerspatialbroadening,whichaffectstheT1-T2resultsshown sured with a variable spatial broadening. At least for half the
inthisplot.TheOviii,Fexvii,andFexxlineswerefittedbysub- sample,our90%upperlimitsarebelowthesoundspeedinthe
tractingthespatialbroadeningextractedexactlyintheirenergy system.Inabouttenobjectstheturbulencecontainslessthanthe
band,whichshouldpartlycorrectthissystematiceffect,butitis 40%ofthethermalenergy.Thisissimilartothepreviousresults
difficultto estimate the systematic uncertaintiesdue to the low ofSanders&Fabian(2013).Apparently,thehotterobjectsallow
spatial (andspectral) resolutionof the CCD data. On some ex- forhighervelocities.
tent,thehotterphasemaystillhavelargerturbulence.Forclarity, Wenotethatthespectralextractionregionhadafixedangle,
we also tabulate these line-bandfits in TableA.2. We also note and, therefore, the actual physical scale – where we estimated
thatthethevelocitylimitsoflow-andhigh-ionizationironlines the velocity broadening – depends on the source distance. In
fallatoppositesidesoftheFe–Oviii1:1line,whichmeansthat Fig.A.4 (left panel), we show the Mach numbers for the 90%
the metallicity distribution in the sources should not affect our conservativeupper limits as a functionof the temperature,and
broad-band,multi-ion,limitsshowninFig.3. wecomparetheaverageupperlimitsonMachnumbercalculated
CiroPintoetal.:ChemicalEnrichmentRGSclustersample(CHEERS):Constraintsonturbulence 7
withindifferentrangesofphysicalscales.Thereisnosignificant
trend with the temperature, but the average upper limit on the
Mach number is lower for narrowerphysicalscales. Assuming
a Kolmogorovspectrumfortheturbulenceintheseobjects,the
root-mean-squarevelocityscale dependsonthe1/3rdpowerof
the physical length. Therefore, we scaled the upper limits by
(sc/sc )1/3,wheresc istheminimumphysicalscaleperarc-
min min
sec∼0.07kpc/1”ofNGC4636,thenearestobjectinoursample.
Inotherwords,wedividedourupperlimitsbytherelativephysi-
calscaleperarcsecrelativetoNGC4636,whichisequivalentto
normalizingbytheratiobetweenthesizeofthespectralextrac-
tionregionofeachclusterandthatoneofNGC4636.Thescaled
upperlimitsontheMachnumbersaretabulatedinTableA.2and
plottedinFig.A.4(rightpanel).Theyarerandomlydistributed
around Ma ∼ 0.8and donotdependanymoreon the physical
scale.Wecodedthepoint-sizeandthecolorswiththevaluesof
r andK takenfromtheliterature.Ther istheradiuswithin
500 0 500
whichthemeanover-densityoftheclusteris500timesthecriti-
caldensityattheclusterredshift,andK isthevalueofthecen- Fig.5.Ratiosbetweenthe90%conservativeupperlimitsonthe
0
tralentropyinthesamecluster.Alltheadoptedvaluesandtheir Mach number (velocity / sound speed) that are scaled by the
referencesarereportedinTableA.2.Wedonotfindanysignifi- 1/3rd power of the spatial scalel assuming Kolmogorov turbu-
cantrelationbetweentheupperlimitsontheMachnumberand lence (see Sect.5.2), and the Mach number, which is required
thesephysicalproperties,possiblyduetothelimitedsample. to make a heating–cooling balance (see Eq.2). The point size
To understand whether dissipation of turbulence may pre- providesther500,andthecoloriscodedaccordingtothecentral
ventcoolinginoursample,wecomputedtheMachnumberthat entropy,K0,inunitsofkeVcm2.
isrequiredtobalancetheheatingandcooling,accordingtothe
followingequation:
by other authors, who use the measurements of resonant scat-
n 1/3 c −1 l 1/3 tering (Werneretal. 2009 and dePlaaetal. 2012). In particu-
Ma ≈0.15 e s (2)
REQ (cid:18)10−2cm−3(cid:19) 103kms−1! 10kpc! lar,ourlimitsforM49(alsoknownasNGC4472),NGC4636,
and NGC5813 agree with the 100kms−1 upper limit ob-
wheren isthedensityatthecavitylocation,c thesoundspeed tained by Werneretal. (2009). We also found upper limits of
e s
thatwe haveestimatedthroughtheRGStemperature,andlthe a few 100skms−1 for A3112, which is similar to the results
characteristiceddysize,whichwetakeastheaveragecavitysize of Bulbuletal. (2012). However, we measured higher limits
(seeZhuravlevaetal.2014).TheMachnumbersrequiredtobal- with a variable spatial broadening that agree with continuum-
ancecoolingaretabulatedinTableA.2.Mostcavitysizeswere subtractedprofilesmethodofSanders&Fabian(2013).
takenfromPanagouliaetal.(2014b).Forclusterswithmultiple Recently, Zhuravlevaetal. (2014) used the surface bright-
cavities,weusedanaveragesize.Forthe19sourcesoutsideof nessfluctuationsintheChandraimagesofthePerseusandVirgo
theirsample,weusedtheirr−T relationtodeterminethecavity clusterstoderiveturbulentvelocitiesintherange70–210kms−1
size.MostdensitiesweretakenfromtheACCEPTcatalog. forPerseusand43–140kms−1forVirgo,wherethesmallerval-
In Fig.5, we compare the ratios between the conservative uesrefertothecentral1.5’region.Ourupperlimitsinthecores
upperlimitsofthescaledMachnumbersassumingKolmogorov oftheclustersareconsistentwiththeirvalues,especiallywhen
turbulence,and those that are requiredto balance cooling with normalizedbythephysicalscalefactor1.5′/0.4′.Theyshowthat
the RGS temperatures. For most sources, our upper limits are theseturbulentmotionsshoulddissipateenoughenergytooffset
largerthan the balancedMachnumbers,which meansthatdis- thecoolingofthecentralICMintheseclusters.Fortenobjects,
sipation of turbulence can provide enough heat to prevent the thescaledMachnumbercanbetransonic,andamajorfraction
coolingofthegasinthecores. ofenergycanbestoredinturbulence,whichcouldsignificantly
Itisdifficulttoknowwhichisthemainmechanismthatpro- heat the gas through dissipation (see, e.g., Ruszkowskietal.
duces turbulence in these objects. Our scaled upper limits are 2004).Recently,Gasparietal.(2014)notedthatevenifthetur-
mostly below 500kms−1, which can be produced by bubbles bulence in the hot gas is subsonic, it may be transonic in the
inflated by past AGN activity (see, e.g., Bru¨ggenetal. 2005). cooler gas phases. Zhuravlevaetal. (2014) reported that dissi-
For some objects, our upper limits are consistent with veloci- pation of turbulence may balance cooling even under subsonic
ties upto 1000kms−1, which wouldcorrespondto Mach num- regime. Our upper limits on Mach number are larger than the
berslargerthanone.ForNGC507,wedetecttransonicmotions valuesnecessarytobalancecoolingandareconsistentwiththis
presumably due to merging (see, e.g., Ascasibar&Markevitch scenario. However,it is possible that other processes are dom-
2006).Inaforthcomingpaper,wewillanalyzetheresonantscat- inant, such as turbulent mixing (see, e.g., Banerjee&Sharma
teringoftheFexviilinesexhibitedbyhalfofoursampletoplace 2014).
lowerlimitsonturbulentbroadeningandprovidemoreinsights The NGC507 group exhibits velocities larger than
onitsoriginanditsroleinpreventingcooling. 1000kms−1 inboththe0.8’and3.4’regions,correspondingto
a scaled Mach number Ma = 4.2±1.7 (1σ). The 15Å Fexvii
line is stronger than the one at 17Å, which would suggestlow
5.3.Comparisonwithpreviousresults
resonantscattering (see Fig.A.7) and, therefore,high kinemat-
Our velocity limits broadly agree with the previousresults ob- ics in the galaxy group. This object is known to have a dis-
tained bySanders&Fabian (2013) using a similar methodand turbedshape and to hostradio lobespresumablyin a transonic
8 CiroPintoetal.:ChemicalEnrichmentRGSclustersample(CHEERS):Constraintsonturbulence
NGC5846 (10kms−1), NGC4636 (100kms−1), and NGC507
(1000kms−1, see Fig.7 right panel). We have used the model
fittedforthefull(-1.7’,+1.7’)RGSspectraasatemplate,which
areshownin Fig.A.7, becausethisextractionregioniscompa-
rableto the3.05’×3.05’field-of-viewofthemicrocalorimeter.
Thespatialbroadeningwasexcludedfromthemodel.Thesim-
ulated SXS spectra are characterized by a richness of resolved
emission lines, which providesvelocity measurements with an
accuracyof50kms−1 orbetter.Thelinewidthsclearlyincrease
throughoutNGC5846,NGC4636,andNGC507.Thehottergas
presentinthePerseusclusterproducesstronghigher-ionization
lines above 1keV, which constrain the turbulence in different
(Fe-LandFe-K)gasphases.
Wealsonotethatthe1’spatialresolutionofASTRO-Hpro-
vides,forthefirsttimethemeansforaspatially-resolvedhigh-
resolutionspectralanalysisandthemeasurementsofturbulence
in differentregionsof the clusters. The ATHENA X-ray obser-
vatorythatisto belaunchedbythe late 2020swill furtherrev-
olutionizeour measurementsdue to its combinedhigh spectral
(2.5eV)andspatial(<5”)resolution.
Fig.6.Line-of-sightvelocityversusprojecteddistancefromthe
central cD galaxy for the member galaxies of NGC507 group.
Optical spectroscopic redshifts are taken from Zhangetal.
6. Conclusion
(2011).
Wehavepresentedasetofupperlimitsandmeasurementsofthe
velocity widths for the soft X-ray emitting gas of a sample of
clusters, groups of galaxies, and elliptical galaxies included in
expansion/inflation (Kraftetal. 2004). However, our high val- theCHEERSproject.We havesubtractedtheinstrumentalspa-
uessuggestthepresenceofbulkmotions.InFig.6,weshowthe tialbroadeningthroughtheuseofsurfacebrightnessprofilesex-
velocities of the galaxies in the NGC507 group as taken from tractedintheMOS1images.
Zhangetal.(2011).Theyarenotnecessarylinkedtothatofthe Formostsources,weobtainupperlimitsrangingwithin200-
ICM,buttherearehighkinematicsandhintsofinfallingclumps, 600kms−1, where the turbulence may originate in AGN feed-
which indicate a substructure extended toward the observer. In back or sloshing of the ICM. However,for some sources, such
thisgroup,thegalaxyvelocitiesgenerallydoublethoseobserved asNGC507,wefindupperlimitsof1000kms−1 orlarger,sug-
inNGC4636,wherewemeasurelowervelocitybroadening(see gesting other origins, such as mergers and bulk motions. The
alsothedifferentlinewidthsinFig.A.7). measurementsdependontheangularscaleandthetemperature.
For a small sample producing strong high- and low-ionization
lines,wemeasuredsignificantlybroaderupperlimitsforthehot
5.4.TowardASTRO-H
gasphase,whichmaybepartlyduetoitslargerspatialextentas
TheRGS gratingsaboardXMM-Newtonarecurrentlythe only comparedtothecoolphase.WhenwenormalizetheMachnum-
instruments that can measure 100skms−1 velocities in X-ray bers for the physical scale, assuming Kolmogorov turbulence,
weconstrainupperlimitsrangingwithin0.3< Ma<1.5.These
spectra of extended sources like clusters of galaxies. However,
they are slitless spectrometers and, therefore, affected by spa- valuesare abovethe Mach numbersnecessaryto balancecool-
tial broadening.We have partly solved this issue by using line ing,which meansthatthe dissipation ofturbulencemaybe the
dominantmechanismtoheatthegasandquenchcoolingflows.
surface brightness profiles, but there are still systematic un-
certainties larger than 100kms−1. Our models provide an im- However,itispossiblethatadditionalprocessesareheatingthe
portant workbench once the new ASTRO-H X-ray satellite ICM.Inaforthcomingpaper,wewillusetheresonant-scattering
techniquetoplacesomelowerlimitsonthevelocitiesinhalfof
(Takahashietal.2010)islaunched.Thespectra,asprovidedby
itsmicrocalorimeter(SXS),donotsufferfromspatialbroaden- oursample, whichare characterizedbystrongFexviiemission
ingasfortheRGSandwillrevolutionizethemethod.Moreover, lines, and make an extensive use of higher-resolutionChandra
its constant spectral resolution in terms of energy increases maps.Thiswillprovidealternativemeasurementsandfurtherin-
sightsontheoriginandtheroleofturbulenceinclusters,groups
the sensitivity at high energies, which allows us to use higher-
ionizationlinesupto6-7keV(Fe-Klinecomplex)necessaryto ofgalaxies,andellipticalgalaxies.
constraintheturbulenceinhottergasphases.Thepositionofthe Thecurrenttechniquesarepartlylimitedbysystematicsas-
linesunveilevidenceofbulkmotions. sociatedwiththespatiallinebroadening,butthefutureASTRO-
In Fig.7 (left panel), we compare the effective area of the HandATHENAmissionswilldramaticallyimprovethemethod
due to their high broadband spectral resolution and to the ab-
ASTRO-H SXS with that of the first order RGS 1 and 2.
senceofissuesconcerningthespuriousspatialbroadening.They
ASTRO-HprovidesclearlybetterresultsthanthesumofRGS1
willalsoenablespatially-resolvedturbulencemeasurements.
and 2 below 14Å (above 1keV). The RGS has still a better
spectralresolutionthantheSXSinthewavelengthrangethatin-
cludestheFexviilinesofthecoolgas,buttheabsenceofspuri- Acknowledgements. Thisworkisbasedonobservations obtainedwithXMM-
Newton,anESAsciencemissionfundedbyESAMemberStatesandtheUSA
ousline-broadeningintheSXSmakesitagreatalternativetool.
(NASA). YYZ acknowledges the BMWi DLR grant 50 OR 1304.We thank
We have simulated a 100ks exposurewith the ASTRO-H SXS Electra Panagoulia forkindly providing the values ofcentral entropy andthe
forfourinterestingobjectsinourcatalog:Perseus(500kms−1), anonymousrefereeforveryusefulcommentsonthepaper.
CiroPintoetal.:ChemicalEnrichmentRGSclustersample(CHEERS):Constraintsonturbulence 9
3
m)2 0.010.02 RGS 2 Astro−H Counts s keV−1−1 110 Perseus 12 NGC 5846
Effective Area ( 5×10−3 RGS 1 0.1 1 Ene2rgy (keV) 5 00.6 0.8 Ene1rgy (keV)1.2 1.4
3 6
102×10−3−3 5 10 15Wavelength 2(Å0) 25 30 35 Counts s keV−1−1 12 NGC 507 24 NGC 4636
00.6 0.8 1 1.2 1.4 00.6 0.8 1 1.2 1.4
Energy (keV) Energy (keV)
Fig.7.Leftpanel:RGS1and2firstorderandASTRO-HSXSeffectivearea.Rightpanel:ASTRO-HSXS100kssimulationsfor
thebrightPerseusclusterandthreegroupsofgalaxieszoomedintheRGSenergyband(see Sect.5.4).We adoptedthe3.4’RGS
modelsasatemplate(seeFig.A.7).
References The MOS1 images (Fig.A.1) are obtained by stacking the
Fe-L (10–14Å) band images extracted in each exposure (see
Ascasibar,Y.&Markevitch,M.2006,ApJ,650,102
Banerjee,N.&Sharma,P.2014,MNRAS,443,687 Sect.3.3).Wezoomedonthecentral10’region.
Bru¨ggen,M.,Hoeft,M.,&Ruszkowski,M.2005,ApJ,628,153 In TableA.1, we quote all our velocity results. In Fig.A.2
Bryans,P.,Landi,E.,&Savin,D.W.2009,ApJ,691,1540 (left panel), we compare the 2σ upper limits on the velocity
Bulbul,G.E.,Smith,R.K.,Foster,A.,etal.2012,ApJ,747,32
broadening as measured for the 3.4’ and 0.8’ regions by sub-
Cavagnolo,K.W.,Donahue,M.,Voit,G.M.,&Sun,M.2009,ApJS,182,12
tractingtheMOS1spatialprofiles.InFig.A.2(rightpanel),we
Chen,Y.,Reiprich,T.H.,Bo¨hringer,H.,Ikebe,Y.,&Zhang,Y.-Y.2007,A&A,
466,805 compare the 0.8’ 2σ upper limits estimated by subtracting the
dePlaa,J.,Zhuravleva,I.,Werner,N.,etal.2012,A&A,539,A34 adoptedandthe best-fit, spatial-line-broadening(see Sect.4.2).
denHerder,J.W.,Brinkman,A.C.,Kahn,S.M.,etal.2001,A&A,365,L7 InTableA.1,wealsoshowthetotallinewidthsasresultofspa-
Fabian,A.C.2012,ARA&A,50,455
tialplusDopplerbroadeningwiththeir68%uncertainties.These
Fabian,A.C.,Sanders,J.S.,Taylor,G.B.,&Allen,S.W.2005,MNRAS,360,
L20 totalwidthsareclearlydominatedbythespatialbroadening.
Finoguenov, A.,Pietsch,W.,Aschenbach, B.,&Miniati, F.2004,A&A,415, In TableA.2, we report the values of r and K adopted
500 0
415 with with the physical scales, the RGS temperaturesestimated
Frank, K.A.,Peterson, J. R.,Andersson, K.,Fabian, A.C., &Sanders, J.S.
with an isothermal model, the conservativeupperlimits on the
2013,ApJ,764,46
velocities and the Mach numbers, the limits separately mea-
Gaspari,M.,Oh,S.P.,&Ruszkowski,M.2014,ArXive-prints
Kraft,R.P.,Forman,W.R.,Churazov,E.,etal.2004,ApJ,601,221 sured for the Oviii, Fexvii, and Fexx-to-xxiv emission lines,
Lau,E.T.,Kravtsov,A.V.,&Nagai,D.2009,ApJ,705,1129 and finally the velocity limits for the two CIE components
Lodders,K.&Palme,H.2009,MeteoriticsandPlanetaryScienceSupplement, whereitwaspossibletofitthemseparately.TheseparateFexvii
72,5154
and 2-T fits were accessible only for a very limited sample of
McNamara,B.R.&Nulsen,P.E.J.2007,ARA&A,45,117
Panagoulia,E.K.,Fabian,A.C.,&Sanders,J.S.2014a,MNRAS,438,2341 sourceswith bothstronghigh-andlow-ionizationFelines(see
Panagoulia,E.K.,Fabian,A.C.,Sanders,J.S.,&Hlavacek-Larrondo,J.2014b, Sect.5.1).InFig.A.3,wecomparethevelocityupperlimitses-
MNRAS,444,1236 timated with both the standard and the conservative methods
Ruszkowski,M.,Bru¨ggen,M.,&Begelman,M.C.2004,ApJ,615,675
with the average RGS temperature measured with an isother-
Sanders,J.S.&Fabian,A.C.2013,MNRAS,429,2727
mal model, the sound speed, and the fractions of thermal en-
Sanders,J.S.,Fabian,A.C.,&Smith,R.K.2011,MNRAS,410,1797
Sanders,J.S.,Fabian,A.C.,Smith,R.K.,&Peterson,J.R.2010,MNRAS,402, ergy stored in turbulence.In Fig.A.4 (left panel), we show the
L11 Mach number as a function of the temperature with the point-
Snowden,S.L.,Mushotzky,R.F.,Kuntz,K.D.,&Davis,D.S.2008,A&A, sizeand,itiscolorcodedaccordingtothephysicalscaleandthe
478,615
centralentropy,respectively.The lines show the averageMach
Stru¨der,L.,Briel,U.,Dennerl,K.,etal.2001,A&A,365,L18
Takahashi,T.,Mitsuda,K.,Kelley,R.,etal.2010,inSoc.ofPhoto-OpticalInst. numbercalculatedwithinparticularrangesofphysicalscales.In
Eng.Conf.Ser.,Vol.7732,Soc.ofPhoto-OpticalInst.Eng.Conf.Ser. Fig.A.4 (left panel), we show the Mach number scaled by the
Tashiro,M.,Ito,K.,Abe,K.,&Isobe,N.2006,inESASpecialPublication,Vol. 1/3rdpowerofthephysicalscale,assumingKolmogorovturbu-
604,TheX-rayUniverse2005,ed.A.Wilson,569
lence(seealsoSect.5.2).
Turner,M.J.L.,Abbey,A.,Arnaud,M.,etal.2001,A&A,365,L27
In Figs.A.5, A.6, and A.7, we show the RGS spectra ex-
Werner,N.,Allen,S.W.,&Simionescu,A.2012,MNRAS,425,2731
Werner,N.,Zhuravleva,I.,Churazov,E.,etal.2009,MNRAS,398,23 tracted in the 3.4’ cross-dispersion region (see also Sect.3.2).
Willingale,R.,Starling,R.L.C.,Beardmore,A.P.,Tanvir,N.R.,&O’Brien, The spectra were combined with the rgscombine task for plot-
P.T.2013,MNRAS,431,394 ting purposes. We have adapted the spectral model from the
Zhang,Y.-Y.,Andernach,H.,Caretta,C.A.,etal.2011,A&A,526,A105
bestfitobtainedfortheparallelmodelingoftheindividualexpo-
Zhuravleva,I.,Churazov,E.,Schekochihin,A.A.,etal.2014,Nature
sures.Mostspectrawerefittedwithamulti-temperature,two-cie
model(seeSect.4).Afewofthemrequirejustasingleciecom-
AppendixA: Maps,spectra,andvelocitylimits ponent.The spectralmodelinginvolvedthe 7−28Å band,but
we focuson the shorter 10−21Å bandcontainingthe Fe, Ne,
We have placed the MOS1 images, the RGS spectra, and the
andOlines.
velocitylimitsinthissectiontounburdenthepaperreading.
10 CiroPintoetal.:ChemicalEnrichmentRGSclustersample(CHEERS):Constraintsonturbulence
Fig.A.1.MOS1stackedFe-Lbandimages:Central10’×10’region.Thestarredclustersarepartofournewcampaign(seealso
Sect.3.3).
Fig.A.2.Leftpanel:Velocitybroadeningat2σupperlimitsforthe(-1.7’,+1.7’)andthe(-0.4’,+0.4’)regionsatcomparison.The
spatialbroadeningwasremovedthroughtheMOS1surfacebrightnessprofiles.Rightpanel:(-0.4’,+0.4’)velocity2σupperlimits
comparedwiththoseestimatedinthesameregionbutwiththevariablebest-fit,spatialbroadening(scaleparameter,s,isfreeinthe
lprocomponent,seeSect.4.2andTableA.1).