H-alpha Filaments and Feedback in NGC4696 at the centre of the Centaurus cluster

One third of Brightest Cluster Galaxies (BCGs) show optical line emission (Crawford et al 1999; Donahue et al 2000). The host clusters of these BCGs are those which require feedback to prevent catastrophic radiative cooling of the inner intracluster medium (ICM). Feedback eventually stops the stellar growth of BCGs, and is crucial for understanding the cutoff of the high end of the galaxy mass function (see Fabian, 2012 for a review). The line emission is seen to originate in a filamentary nebula, sometimes stretching well beyond the effective stellar radius of the galaxy. In many cases, the dominant mass component of the nebula is found to be molecular (Edge 2001; Salom�Le & Combes 2003) which dominates the total gas mass of the innermost few kpc. The ICM surrounding the nebulae is typically at a temperature of a few keV. The best observed nebula surrounds NGC1275 at the centre of the Perseus cluster with spectacular H filaments stretching over 100 kpc. They have been imaged by Lynds (1970), Conselice et al (2001) and by us with HST/ACS (Fabian et al 2008). Our HST image in the red F625Wfilter, which includes H and [NII] is shown in Fig. 1b. Individual outer filaments are just resolved with HST (see lowest two rows of panels, Fig. 1), with a radius of about 35 pc yet are many kpc long. The magnetic field required to support the filaments is inversely proportional to the radius, and deduced to be B 100ƒÊG, which is close to equipartion with the hot gas. The infrared molecular hydrogen and CO emission closely parallels that from H (Lim et 2011; Salom�Le et al 2010). The realisation that the outer filaments around NGC 1275 are essentiallymagnetically-dominated molecular clouds, driven by HST imaging, has forced a drastic rethink of the excitation processes responsible for the emission at all wavelengths. It cannot be photoionization by young stars, for no young stars are seen there. Instead the crucial ingredient is a population of secondary particles (Ferland et al 2008, 2009) produced by the interpenetration of the hot gas (Fabian et al 2011). The results of computations involving heating by energetic particles agree very well with the observed line ratios and predicts that the surface brightness in H of an individual filament is constant and proportional to the incident energy flux from the hot gas, which for NGC1275 is about a factor ten higher than the observed flux. This provides a test of the model since HST can resolve overlapping filaments. Donahue et al (2011) conclude from Spitzer spectra that particle heating is the most plausible model. The total line and FIR luminosity for BCGs can be as high as 1043 to 1044 erg/s, dominating the total radiative losses of the inner 10 kpc. Emission-line nebulae are also common around radio galaxies and many distant luminous galaxies (e.g. Reuland et al 2003), so are a marker of AGN feedback. To use them as diagnostics we need to understand what the filaments are, how they form and their long lifetimes. The object of this proposal is to deepen our understanding of the filaments by resolving and studying the nearest example in which molecular gas has been detected. One model for the filaments assumes that the observed central bubbles (which are inflated in the ICM by the central radio source) buoyantly rise outward dragging the cooler/cooled gas with them (Churazov et al 2001). The filaments thus trace the path of the bubbles before slowly falling back (Fabian et al 2003b). The velocities of the filaments match this picture in NGC1275 (Hatch et al 2006). The dust content of many filaments means they are unlikely to form in situ by cooling. We propose here to observe NGC 4696, which is at almost half the distance of NGC 1275 and has a large filamentary nebula (Fig. 1), a molecular component and strong dust lanes related to the filaments. The filaments are, however, not yet resolved enough to allow the surface brightness of the emission to be measured, a key test of particle heating. The dust component, the bright soft X-ray emission associated with the filaments and the high ratio of [NII] to H, contrast with NGC 1275. (Dust lanes seen near the centre of NGC1275 are due to an unrelated intervening galaxy). Our aims are i) to resolve the width of the filaments and thereby ii) to determine the surface brightness of the emission in order iii) to test the particle heating model, iv) to trace the full extent of the filamentary system and v) to compare with a new deep Chandra image (500ks: PI Sanders) now being taken (Spring 2014). The unique advantage of HST imaging is the ability to resolve the filaments, enabling measurements of radius and surface brightness which are crucial for testing magnetic support and particle heating models.