Electra Panagoulia with Andy Fabian and Jeremy Sanders Wednesday - - PowerPoint PPT Presentation

Electra Panagoulia with Andy Fabian and Jeremy Sanders

Wednesday seminar 13th November 2013

1

• Introduction

– Why study clusters? – Cooling flows and heating – The Chandra telescope – The Centaurus cluster in X-rays – The Centaurus cluster in optical and IR

• Data analysis and results

– Imaging – Spectra

• Where is the missing iron?
• Conclusions
• Future work

2

• Largest gravitationally bound
• bjects in the Universe
• Trace primordial density

fluctuations

• Ideal for studying large structure

formation and evolution → powerful cosmological probes

• But: details of processes are not

clearly understood

• If clusters are to be used as a

reliable testbed for structure formation, internal processes need to be better understood

Abell 1689 (space.mit.edu)

3

• X-rays carry the

thermal imprint of processes such as AGN feedback and cluster mergers

• They provide a

window through which we can study the effects of some of the most energetic processes in the Universe on large scale structure formation

Hydra A (chandra.harvard.edu)

4

• Based on Sx profiles, two types
• f clusters: cool-core (CC) and

non-cool core clusters

• For CC clusters, high ne and low

T in cores → tcool << cluster age → high star formation rates, lots

• f cool and cooling gas
• But: observed properties do

not match theoretical predictions (cooling flow problem)

NGC5813 (z=0.0065)

5

• AGN feedback is thought to be

main heating mechanism (Birzan+ 2004, Dunn & Fabian 2006, McNamara & Nulsen 2012…)

• Outstanding problems:

transport mechanisms poorly understood, some star formation, optical filaments… → gas in clusters is mulAphase: A SIMPLE COOLING FLOW MODEL IS INSUFFICIENT

FFT-processed and smoothed 0.3-7 keV image of the Perseus cluster (Sanders & Fabian 2007)

6

• Cooling flows and AGN heating are primarily manifested in group

and cluster cores

• Need high spatial resolution to disentangle details of these

processes in innermost regions

7

• 2nd closest cluster after Virgo

(z=0.0114)

• Studied extensively at all

wavelengths because of brightness and proximity

• X-ray:

– weak cooling flow (Allen+ 2001) – central cooling time <0.5 Gyr – 2-temperature structure in core + central iron abundance drop (Sanders & Fabian 2006, Sanders+ 2008)

Multi-band X-ray image of NGC 4696 (Fabian+2005)

8

5 kpc

Hα image of NGC 4696 (Crawford+ 2005)

9

Herschel PACS multiband image (Mittal+ 2011)

0.5 – 7.0 keV composite images of NGC 4696

200 ks of clean exposure time

10

• Interested in radial profiles
• Created 22 concentric annuli,

each containing ~115000 counts in total (S/N≈340) in the 0.5-7.0 keV band

• Projection effects: spectra

deprojected using DSDEPROJ (Sanders & Fabian 2007, Russell+ 2008) – like peeling an onion!

• Joint fits performed in the 0.5-

7.0 keV band using an absorbed

• ptically thin thermal model

(altmd.com)

11

12

Panagoulia+ 2013

• Steep drop in Fe

abundance in central 7 kpc, particularly innermost 5 kpc

• Si and S show similar

abundance profiles, drops possibly have same origins as Fe

13

Panagoulia+ 2013

• Bimodal temperature in

central 5 kpc (e.g Sanders+ 2008)

• Joint fit of two innermost

annuli using 2-temperature model

• “Missing” iron mass in

inner 5 kpc is 1.4×106 Msolar

14

• Drop in Fe, Si and S abundances is co-spatial with the

brightest dust emission seen in the IR with Herschel

Smoothed multi-band PACS image (left) and SPIRE image (right) (Mittal+ 2011)

15

• 1. SNe and

stellar winds inject Fe in cluster core

• 2. Much of the iron

can remain in grains, and join the filaments, making them dusty

• 3. Filaments are

dragged outwards by buoyant bubbles

• 4. Filaments are

somehow destroyed at ~20 kpc, enhancing the local ICM

• 5. Approximate steady

state achieved if 10% of filaments are dragged

• ut per AGN cycle

16

Need to remove Fe from cluster core, but without removing H too

• Analysed ~200ks combined Chandra observation of

NGC4696

• Obtained temperature, Fe, Si and S abundance profiles
• Steep drop in Fe abundance in central 5kpc, equivalent to

1.4 x 10 Msolar

• Similar drops in Si and S abundances
• Part of this missing Fe could be locked up in cold dust, as

seen by Spitzer and Herschel

• Fe injected by stellar mass loss is in grains, becomes

incorporated in the filaments in the same region, which are then dragged outwards and destroyed

• Si and S abundance drops could have the same origin

17

• A 500 ks Chandra
• bservation was

approved in August 2013

• Will test our theory

regarding the depletion

• f Fe onto dust grains by

measuring radial abundances of Ne and Ar

• We hope to further study

the abundance structure

• f the Centaurus cluster

with ASTRO-H (launch in late 2015)

18

Astro-H (Illustration: Akihiro Ikeshita / JAXA)

19 O VIII Ne X Ar XVII Ar XVIII Ca XIX Si XIII Si XIV S XV S XVI

Left: Projected 0.5-7 keV spectrum for the innermost 2.7 kpc (200 ks Chandra)

Fe-L Si S Ar Ca

Right: Simulated 0.5-7 keV spectrum for the inner 10.8 kpc, for a 100 ks ASTRO-H

• bservation

Thank you! Any questions?

20