RGS observations of cool gas in cluster

cores

Jeremy SandersInstitute of Astronomy

University of Cambridge

A.C. Fabian, J. Peterson, S.W. Allen, R.G. Morris,J. Graham, R.M. Johnstone

Sanders et al, 2008, MNRAS, 385, 1186Sanders et al, in prep

Cooling in cluster cores Many cluster cores have steeply peaked X-ray

surface brightness profiles, i.e. short mean radiative cooling times.

Suppose there is a luminosity L emitted from within a cooling region rcool

To offset radiation lost through cooling, if the cluster is in steady state and there is no heating, there should be a mass deposition rate of

Measured values from SB profiles gave values of 10-1000s solar masses per year – a “cooling flow”

luminosity of region is: - radiation of thermal energy - PdV work done on gas entering rcool kT

mμLM

52

Lack of cool X-ray emitting gas

see also Peterson et al 01, 03, Kaastra et al 01, 03, Tamura et al 01,...

Spectra imply less than 10% of cooling rates expected from luminosity profiles

Only significant gas down to 1/2 to 1/3of outer temperature

AGN heating?

Fe XVII lines (indicatingtemperature ~ 0.7 keV)missing

de Plaa et al

2A 0335+096

The Centaurus clusterInner core of the

Centaurus cluster

200 ks Chandra observation

RGB image

Sanders & Fabian (2002, 2006)

Fabian, Sanders et al (2005)

10 kpc

Centaurus temperature map

Temperature (keV)

Crawford et al 2005

Chandra 200ks observation200 ks Chandra observation

Projected temperature in Centaurus varies from

3.7 keV in outskirts

to 0.7 keV in the cool plume

Centaurus cross dispersion image

wavelength

1D image

With the RGS spectrometer we get a spectrum of the cluster as a function of one spatial dimension.

150 ks total XMM exposure

Inner 160 arcsec width (30 kpc)Central inner 60 arcsec (13 kpc)60-160 arcsec (13-30 kpc)

Ionisation equilibrium

Values taken from Mazzotta et al 1998

By measuring the distribution of ionisation states of metals we canfind the gas temperature, or distribution of gas temperature

Inner 60 arcsec width (16 kpc)

strong

missing

Line ratios — constraints on kT

Implies average temperature of gas 3.5 to 5.2 × 106 K

Ratio of flux of emission lines and compared to CHIANTI model

Spectral fitting limits on gas kT

Multi temperature model(components fixed in temperature but

varying in emission measure)

Cooling flow model(mass deposition in absence of heating)

Factor of 10 in temperature > factor of 10 in mass deposition

Spectral fitting limits on gas kT

Multi temperature model(components fixed in temperature but

varying in emission measure)

Cooling flow model(mass deposition in absence of heating)

Factor of 10 in temperature > factor of 10 in mass deposition

tcool ~ 107 yr!

New RGS spectra

156ks

138ks

HCG 62

O VIII

Fe XVII

156 ksHCG 62 is a compact group of galaxies

Prominent radio bubbles with ~1057 erg energy content

Temperature varies in spatially resolved studies from 1.5 to 0.7 keV (Morita et al 2006)

High abundance arc (Gu et al 2007)

HCG 62: temperature distribution8 component VAPEC model

Shows clear break in temperature distribution at 0.6 keV

Emission measure drops by more than an order of magnitude

Lower temperature consistent with CCD results from Chandra/XMM (Morita et al 2006)

HCG 62: mass deposition rate8 component VMCFLOW cooling flow model in steps of temperature

In the absence of cooling, there is a very sharp drop in mass deposition rate at 0.6 keV, by at least an order of magnitude

tcool ~ 3×108 yr

Conclusions Varied behaviour of cool cores observed by RGS

Wide range in temperature in Centaurus (factor 10)

Narrow range in HCG 62 (factor 2)

Both show strong variation in emission measure with temperature

Implies there is always close feedback It is important to study a wide range of clusters

with deep observations with XMM We have a deep Abell 1835 observation soon

Star formation in this object ~ mass deposition rate (100 Solar masses/year)

Sound waves in Centaurus

Fourier high-pass-filtered image

Sanders & Fabian (submitted)

Centaurus SB profiles and residuals

λ ~ 9 kpc

Power calculation Use deprojection factor to convert from

ripple amplitude to intrinsic pressure ampl. Assuming spherical waves, this implies

9×1042 erg s-1 sound wave power at r=30 kpc, where tcool=3.3 Gyr

9 kpc wavelength implies 107 yr period (similar to Perseus)

Luminosity emitted within 30 kpc is 1.3×1043 erg s-1

Using period and energy of radio bubbles, implies power of 2×1043 erg s-1

HCG 62: extra absorption8 component VMCFLOW cooling flow model in steps of temperature

Extra column density of 5×1020 cm-2 added into model for three lowest temperature components

HCG 62: ignoring OVII8 component VMCFLOW cooling flow model in steps of temperature

Only fitting spectrum from 7 to 20 A

Very coolest gas

Very coolest X-ray gas at 0.35 keV has a mean radiative cooling time of only ~107 yr

Line profiles

Emission measure maps

Best fittingsmoothingsizes fromRGS linewidths

Centaurus temperature and metals25 kpc

Centaurus in different elements

Sanders & Fabian (2006)

Significant amount from Type II supernovae (~30%)

Therefore large amount of star formation in the Centaurus cluster

Either massive initial burst

Continuous ~5 M/yr Likely to have been

undisturbed for ~ 8 Gyr

Close heating/cooling balance

Sanders & Fabian (2006)Supernova fractions