Sausage radio relic: radio observations and modelling Andra Stroe [email protected]

Stroe et al. 2013, A&A, 555, A110 Stroe et al. 2014b, MNRAS Letters, 441, L41 Stroe et al. 2014c, to be submitted to MNRAS this week??

Leiden clusters meeting, 10 June 2014 Leiden clusters meeting, 10 June 2014 2

Overview • The field of cluster radio relics • Information about the Sausage cluster • Radio observations

• Modelling WSRT

WSRT

AMI GMRT Leiden clusters meeting, 10 June 2014 3

Galaxy clusters across wavelengths ●Soup of: ● Dark matter ● gas: ● Thermal bremsstrahlung: X-ray emission ● Non-thermal synchrotron emission: radio emission ● (optical, infra-red, radio)

Abell 3667

X-ray intensity in colour, radio emission in white contours (Rottgering et al. 1997) Leiden clusters meeting, 10 June 2014 4

Structure formation leads to shocks!

● Structure formation is a very violent process which leads to energy releases of up to 1064 erg (e.g. Hoeft et al. 2004) ● Some of the energy is released in the form of shocks ● Cosmological simulations predict M=1–10 shocks to be common in clusters and the filaments that connect them (e.g. Pfrommer et al. 2006) ● Also inject turbulence into the intracluster medium

Spatial Mach number distribution in a cosmological structure formation simulation (Pfrommer et al. 2006) Leiden clusters meeting, 10 June 2014 5

Diffuse radio emission in clusters

Radio relics Abell 2256 ● Diffuse radio synchrotron emission ● Mpc-sized, extended ● Located at cluster outskirts ● No optical counterpart ● Strongly polarised ● Related to cluster shocks

Radio haloes ● Diffuse, located at cluster centres ● Unpolarised ● Follows the ICM X-ray distribution ● Formed via turbulent reacceleration of ICM 1 Mpc

X-ray intensity in colour, radio emission in white contours (Clarke & Ensslin 2006) Leiden clusters meeting, 10 June 2014 6

Radio relics ● Extended (Mpc-wide), diffuse patches of radio synchrotron emission ● Located at the outskirts of merging clusters ● No obvious optical counterpart ● Associated with the electrons in the ICM ● Highly magnetised

A754 A1664 A548b

A2744 A115 RXCJ1314.4-2515

Feretti et al. 2012 Leiden clusters meeting, 10 June 2014 7

Shocks+synchrotron

● Formation mechanism: ● Two/more clusters collide ● Shock waves travel through the ICM ● Accelerate thermal electrons → emit synchrotron ● Similar to supernova remnants → but very different scales ● Spectrum: ● Initially a linear function SN1006 ● Spectral index steepening ● Spectral curving

Radio (red) and X-ray (blue) emission on top of an optical image (ESO) Leiden clusters meeting, 10 June 2014 8

Why are relics important?

● The largest particle accelerators in the world! ● Complementary way to discover clusters ● Probe magnetic fields + turbulence ● Shock are ubiquitous → shock efficiency? injection spectra? ● Study effect of cluster merger on galaxies ● Basic physics applicable to other astronomical fields: ● Injection mechanism ● Acceleration mechanism The LHC is not impressed with radio relics! Maybe it's just jealous! LHC Leiden clusters meeting, 10 June 2014 9

Injection ● Continous injection (CI, Pacholczyk 1970) ● Kardashev-Pacholczyk (KP, Kardashev 1962, Pacholczyk 1970) ● Jaffe-Perola (JP, Jaffe & Perola 1973) ● Tribble (Tribble 1993)

Injection mechanisms Leiden clusters meeting, 10 June 2014 10

Acceleration

● Adiabatic compression (Ensslin & Gopal-Krishna 2001)→ radio phoenixes ● Diffusive shock acceleration (Ensslin et al. 1999)→ elongated radio relics ● Phoenixes vs relics: curved vs straight integrated radio spectra

Diffusive shock acceleration

Diffusion

Reacceleration

t n ro f ck o h S Leiden clusters meeting, 10 June 2014 11

Why is it hard? ● Faint & extended → difficult to detect ● Lack of telescopes ● Lack of suitable systems: ● Equal mass systems ● Merging in the plane of the sky ● Low impact parameter ● At the right moment Leiden clusters meeting, 10 June 2014 12

The Sausage cluster • z~0.2 • X-ray luminous • Disturbed morphology • Merger in the plane of 1.4 Mpc the sky → twin, outward travelling shock waves (Mach ~4.5)

X-ray intensity, radio overlays (Akamatsu & Kawahara 2013, Ogrean et al. 2013) Leiden clusters meeting, 10 June 2014 13

Sausage cluster - Pretty pictures • Radio maps with contours drawn at [4, 6, 8, 16, 32]xRMS

150 MHz 300 MHz

Stroe et al. 2013 Leiden clusters meeting, 10 June 2014 14

Integrated spectrum - Sausage

α = -1.06±0.05 M = 4.58 ± 1.09 From the integrated spectrum you can get the Mach number of the shock!

Stroe et al. 2013 Leiden clusters meeting, 10 June 2014 15

7 frequency spectral index & curvature • Fit a spectrum to every pixel in our 7 maps • Consistent with DSA – indicative of ageing behind a shock • Fit second order function to every pixel • Predicted by DSA, but never observed

Spectral index Spectral curvature

Stroe et al. 2013, van Weeren et al. 2010 Leiden clusters meeting, 10 June 2014 16

Colour-colour plot • Bin pixels based on their spectral index values • Trace the relic in a shell-like pattern

Stroe et al. 2013 Leiden clusters meeting, 10 June 2014 17

Colour-colour plot • KGJP model fit the data best • Multiple populations of Sausage electrons of different ages • All populations follow a JP injection model

Stroe et al. 2013 Leiden clusters meeting, 10 June 2014 18

But, some disagree! • You could see spectral index trends just be caused by projection effects

Edge-on Face-on view view

Spectral index of simulated radio relic emission at different viewing angles (Skillman et al. 2013) Leiden clusters meeting, 10 June 2014 19

In the case of the sausage, NO! • We fit the Tribble ageing model pixel by pixel • Clear trends of increasing spectral electron age from the shock front into the downstream area → shock is clearly moving northwards • Age varies very little along the relic → maximum ICM inhomogeneities of 10% in density/temperature at 1.5 Mpc cluster-centric distance • Self-consistently constrain the B field to 1 nT • Shock moves at ~2500 km/s speed → cluster core passage ~800Myr ago

Left: Age (time since last acceleration) of the electrons; Right: Shock speed averaged along the length of the Sausage (Stroe et al. submitted) Leiden clusters meeting, 10 June 2014 20

We now have a consistent picture! Some violent mergers lead to travelling shock waves. The shock waves accelerate thermal particles from the intra-cluster medium through the diffusive shock acceleration mechanism. The particles radiate synchrotron emission within an ordered , with isotropisation of the pitch angle between the electrons and the B field. Or maybe we don't? Leiden clusters meeting, 10 June 2014 21

16 GHz detection! • Radio maps with contours drawn at [4, 8, 16, 32]xRMS • Recover northern relic at high S/N

16 GHz 16 GHz 3' resolution 40'' resolution

(Stroe et al. 2014b) Leiden clusters meeting, 10 June 2014 22

Integrated spectra – 16 GHz AMI • All the radio, X-ray data and simulations consistent with simple DSA • But, steepening towards 16 GHz? ● Inhomogeneous medium - density, temperature gradient across the relic ● Inhomogeneous magnetic field – turbulence? Intermittent magnetic field?

AMI More modelling required!

12σ away from DSA prediction!

(Stroe et al. 2014b) Leiden clusters meeting, 10 June 2014 23

“Extreme” frequency data are required! • Up until now, we have been focusing on a very narrow range in spectrum • We need to explore very low and very high frequencies to answer: ● Injection spectrum ● Shock efficiency ● Electron ageing model ● Relic formation model

We looked LOFAR only MWA here! LWA

ATCA, JVLA, AMI, PdBI, ALMA (Giacintucci et al. 2008)