HOT TIMES FOR COOLING FLOWS Mateusz Ruszkowski

Cooling flow cluster Non-cooling flow cluster gas radiates X-rays & loses pressure support against gravity gas sinks towards the center to adjust to a new equilibrium COOLING FLOW PROBLEM

PROBLEMS “COOLING FLOWS” –No evidence for large mass dropout Stars, absorbing gas –Temperature “floor’’ Temp. drops by factor ~3 Sanders & Fabian 2002

CLUSTER HEATING appears to be: RELATIVELY GENTLE –No shock heating –Cluster gas convectively stable –Abundance gradients not washed out DISTRIBUTED WIDELY – not too centrally concentrated –Entropy “floor” manifest on large scales –Needed to avoid cooling “catastrophe”

HEATING CANDIDATES AGN heating (Tabor & Binney, Churazov et al.) Thermal conduction (Bertschinger & Meiksin, Zakamska & Narayan, Fabian et at., Loeb) Turbulent mixing (Kim & Narayan)

WE CALL THIS “EFFERVESCENT HEATING” Cluster gas heated by pockets of very buoyant (relativistic?) gas rising subsonically through ICM pressure gradient –Expanding bubbles do pdV work Dependent on two conditions: –Buoyant fluid does not mix (much) with cluster gas persistent X-ray “holes” –Acoustic & potential energy is converted to heat by damping and/or mixing

EFFERVESCENT HEATING: 1D MODEL “Bubbles” rise on ~ free-fall time Assume –Number flux of CR conserved –Energy flux decreases due to adiabatic losses –Dissipation converts motion to heat ~locally

Volume heating rate: Compare to cooling rate: HEATING MODEL TARGETS PRESSURE GRADIENT STABILIZES COOLING

Ruszkowski & Begelman D ZEUS SIMULATIONS Includes: Spitzer/4 Simple feedback in center

Ruszkowski & Begelman 2002 AGN, not conduction, dominates heating

ENTROPY PROBLEM IN THE ICM – entropy “floor” –Supernova heating may be inadequate Roychowdhury, Ruszkowski, Nath & Begelman 2003 Possible solutions Possible solutions: Cooling --- gas cools and forms galaxies, low entropy gas is removed; Voit et al. Turbulent mixing (Kim & Narayan) AGN heating --- gas is heated; entropy increases

relation ? Roychowdhury, Ruszkowski, Nath & Begelman 2003

Testing assumptions of the model ‘‘Pure’’ theory requires Lateral spreading of the buoyant gas must be significant Spreading must occur on the timescale comparable to or shorter than the cooling timescale BUT Heating must be consistent with observations No convection Preserved abundance gradients Cool rims around rising bubbles Radio emission less extended spatially than X-rays Sound waves

THE TOOL – the FLASH code Crucial to model mixing and weak shocks accurately –PPM code with Adaptive Mesh Refinement, e.g., FLASH, better than lower-order, diffusive code, e.g., ZEUS

Chandra image 3C 84 and Perseus Cluster Fabian et al Note multiple “fossil” bubbles, not aligned with current radio jets

RAPID ISOTROPIZATION RAPID ISOTROPIZATION – buoyant gas spreads laterally on dynamical timescale until it covers steradians Ruszkowski, Kaiser & Begelman 2003

Chandra image 3C 84 and Perseus Cluster Fabian et al Cold rims, not strong shocks

COOL RIMS COOL RIMS – entrainment of lower temperature gas Ruszkowski, Kaiser & Begelman 2003

THE DEEPEST VOICE FROM THE OUTER SPACE Fabian et al Unsharp masked Chandra image X-ray temperatures 131 kpc

MEDIA CRAZE

SOUND WAVES Ruszkowski, Kaiser & Begelman 2003

Chandra image +1.7 GHz radio 3C 338 and Abell 2199 Johnstone et al “fossil” bubbles

Ruszkowski, Kaiser & Begelman 2003 Conditions emulate Abell 2199, with cooling; Myr

Radio: Higher contrasts, detectable only close to jet axis X-rays: spread out laterally “Ghost cavities” do not trace previous jet axis 3C Abell 2199 (Johnstone et al. 2002)

CONCLUSIONS No need for large mass deposition rates Minimum temperatures around 1 keV Entropy floor Significant and fast lateral spreading Sound waves Cool rims Mismatch between X-ray and radio emission SEMI-ANALYTICAL MODELS SEMI-ANALYTICAL MODELS NUMERICAL SIMULATIONS NUMERICAL SIMULATIONS