CAVITIES AND SHOCKS IN THE GALAXY GROUP HCG 62 AS REVEALED BY CHANDRA, XMM-NEWTON, AND GIANT METREWAVE RADIO TELESCOPE DATA

HCG 62 was observed with Chandra for ∼48.5 ks on 2000 January 25 (ObsID 921) with ACIS-S in imaging mode operating at the focal plane temperature of −110°C. The data were reprocessed with CIAO 4.1 using CALDB 4.1.0 and corrected for known time-dependent gain problems following techniques similar to those described in the Chandra analysis threads.⁵ No charge transfer inefficiency (CTI) correction is available for data taken at −110°C. Screening of the event files was applied to filter out strong background flares. Blank-sky background files, filtered in the same manner as in the HCG 62 image and normalized to the count rate of the source image in the 10.0–12.0 keV band, were used for background subtraction. The final exposure time is 46.5 ks. We only use data from the S3 CCD since it covers the central part of the group emission where the cavity system and the radio source are located. We identified and removed the point sources on S3 using the CIAO task WAVDETECT, with the detection threshold set to the default value of 10⁻⁶.

Updated gain correction at −110°C only become available with CIAO 4.0 CALDB version, so previous work has been limited by calibration uncertainties. In particular, spectra were extracted using the SPECEXTRACT task and, as recommended for −110°C data,⁶ new RMF files were computed using the task MKACISRMF with the latest gain file, acisD2000-01-29p2_respN0005.fits.

HCG 62 was observed by XMM-Newton in 2007 June during revolutions 1382 (ObsID 0504780501, hereafter 501) and 1383 (ObsID 0504780601, hereafter 601) for nominal exposure times of 122.5 ks and 32 ks, respectively. In this paper, only data from the European Photon Imaging Camera (EPIC) are presented. The MOS detectors were operated in Full Frame Mode and the PN detector in Extended Full Frame Mode, both with the MEDIUM filter. We used the SASv8.0.0 processing tasks emchain and epchain to generate calibrated event files from raw data. Throughout this analysis single and double pixel events for the PN data (PATTERN <=4) were selected, while for the MOS data sets the PATTERNs 0–12 were used. The removal of bright pixels and hot columns was done by applying the expression (FLAG==0). To reject soft proton flares, we generated a light curve in the 10.0–12.0 keV band where the emission is dominated by the particle-induced background, and excluded all the intervals of exposure time having a count rate higher than a certain threshold value (the chosen threshold values are 0.20 (0.18) cps for MOS and 0.35 (0.32) cps for PN in obs. 501 (601)). The total remaining exposure times after cleaning are 112.0 ks for MOS1, 97.6 ks for MOS2, and 60.2 ks for PN. Starting from the output of the SAS detection source task, we made a visual selection on a wide energy band MOS and PN image of point sources in the field of view. Events from these regions were excluded directly from each event list. The source and background events were corrected for vignetting using the weighted method described in Arnaud et al. (2001), the weight coefficients being tabulated in the event list with the SAS task evigweight. This allows us to use the on-axis response matrices and effective areas, computed in the central 30'' with the tasks rmfgen and arfgen.

2.2.1. Background Treatment

HCG 62 was observed by the GMRT in 2008 February at frequencies of 610 MHz and 235 MHz (project 13SGa01) for a total of 125 and 250 minutes, respectively. The data were acquired in spectral line mode with 128 channels/band and a bandwidth of 16 MHz at 610 MHz and 8 MHz at 235 MHz. Data reduction was done using the NRAO Astronomical Image Processing System (AIPS) package. Accurate editing of the UV data was applied to identify and remove bad data. After bandpass calibration, the central channels were averaged to six channels of ∼2 MHz at 610 MHz and ∼1 MHz at 235 MHz. After further careful editing in the averaged data sets, images were produced using the standard procedure (calibration, Fourier transform, clean, and restore) and the wide-field imaging technique. Phase-only self-calibration was applied to remove residual phase variations. The rms noise level (1σ) in the final images is 50 μJy beam⁻¹ at 610 MHz and 170 μJy beam⁻¹ at 235 MHz. Following Chandra et al. (2004), we estimated that the amplitude calibration error at 610 MHz is ∼5%, while a calibration error of 8% was assumed at 235 MHz. The calibration uncertainties were included in the flux density and spectral index errors determined in Section 3.2. The GMRT observations of HCG 62 will be presented and discussed in more detail in S. Giacintucci et al. (2010, in preparation).

The raw 0.5–2.0 keV ACIS-S image in Figure 1 (top left panel) shows that the hot gas in the central region of HCG 62 is not smoothly distributed but is instead perturbed with cavities and edges. In particular, as best revealed by the unsharp masked image shown in Figure 1 (top right panel), we confirm the presence of two clear cavities, first discovered by Vrtilek et al. (2002), and we also notice the presence of surface brightness discontinuities or "fronts" which suggest the presence of large-scale gas motions or shocks in the group halo. In particular, by means of the surface brightness profiles presented in Section 5.2, we identify a front at 36 kpc (∼2') from the center toward the southwest (SW) at a distance of about 10 kpc (∼35'') from the southern radio lobe, and one at 20 kpc (∼70'') from the center toward the northeast (NE) close to the N cavity. These features are significant at the 5σ and 15σ levels, respectively. In the images, there is also a hint of an edge about the same distance SW of the group center as the front to the NE. The spectral analysis presented below indicates that the outer SW edge is a shock front (Section 5.2). As already estimated by previous studies (Bîrzan et al. 2004; Rafferty et al. 2006; Morita et al. 2006) the two cavities have similar sizes, with diameters of ∼10 kpc, and are located at a projected distance of ∼8.5 kpc from the group center. We extracted the surface brightness profiles along and orthogonal to the cavities, and at the radius of ∼8.5 kpc we estimate a brightness decrement of ∼30% along the cavities relative to the orthogonal direction.

Zoom In Zoom Out Reset image size

Because of the poorer XMM-Newton spatial resolution, the cavities are not evident in the raw 0.5–2.0 keV image shown in Figure 1 (bottom left panel). However, they do appear clearly in the unsharp masked image (Figure 1, bottom right panel) at positions coincident with those determined by the Chandra data.

The existing higher frequency (1.4 GHz) observations of HCG 62 show the central compact radio source with some hint of emission extending toward the southern cavity (Vrtilek et al. 2002). With the new GMRT observations at 235 MHz and 610 MHz, we detect clearly extended emission emanating from the core in the typical form of a bipolar flow. The two radio lobes point toward the N and S directions, filling the cavities. The overlay of the 610 MHz radio contours on the unsharp masked Chandra image is shown in Figure 2 (left), whereas the overlay of the 235 MHz radio contours on the smoothed Chandra image is shown in Figure 2 (right). It is evident that far more extensive structures become visible at lower frequencies: the radio emission at 235 MHz is more extended with two faint regions located outside the cavities and apparently bent toward the east (E). In particular, we identify inner lobes clearly filling the well-defined X-ray cavities (best visible in the left panel of Figure 2), and outer lobes having no associated X-ray cavities (best visible in the right panel of Figure 2). The connection of the radio source with the X-ray features will be discussed in Section 5.

Zoom In Zoom Out Reset image size

At 610 MHz, the source has a total flux density of 13.5 ± 0.7 mJy. At 235 MHz, the total flux density is 42.4 ± 3.4 mJy. Neglecting the K-correction term (for 0.3 < α < 1.7, the K-correction is <1%; Petrosian & Dickey 1973), the monochromatic radio power at each frequency is calculated as

$$\begin{equation} P_{\nu }=4 \pi \, D_{\rm L}^2 \, S_{\nu }, \end{equation} \tag{ 1 }$$

where D_L is the luminosity distance, which gives P_610 MHz = (5.7 ± 0.3) × 10²¹ W Hz⁻¹ and P_235 MHz = (1.8 ± 0.1) × 10²² W Hz⁻¹. This radio source is thus classified as a weak FR-I, and its spectral index α^610 MHz_235 MHz = 1.2 ± 0.1 is steep compared to that of typical radio galaxies.

Since the outer radio lobes are only detected at one frequency in our GMRT observations, we are unable to perform a detailed radio spectral analysis. New, deeper GMRT observations at 235, 327, and 610 MHz, scheduled during Cycle 17, will allow us to study with unprecedented detail the spectral properties of the source. Based on the analysis of the integrated radio spectrum and spectral index distribution, an estimate of the radiative ages of the electron population can be obtained (e.g., Giacintucci et al. 2007, 2008, and references therein) for both pairs of radio lobes. This will help to shed light on the relationship of the outer lobes to the inner ones.

We produced projected radial temperature and metallicity profiles by extracting spectra in six circular annuli centered on the peak of the X-ray emission. The annular regions are described in Table 1. The data from the three XMM-Newton cameras were fitted simultaneously to an absorbed vapec thermal plasma model, where O, Si, S, and Fe are treated as free parameters with the Ni abundance linked to the Fe abundance and the Mg and Ne abundances linked to the O abundance. The best-fitting parameter values and 90% confidence levels derived from the fits to the 360° annular spectra are summarized in Table 1.

The azimuthally averaged projected temperature profile (Figure 3, left) shows a positive temperature gradient at small radii, with temperatures increasing up to ∼1.5 keV at a radius of ∼50 kpc. This behavior is similar to that observed in other groups (e.g., Gastaldello et al. 2007; Finoguenov et al. 2007; Rasmussen & Ponman 2007; Sun et al. 2009), although we do not observe the typical decline in temperature at larger radii. Previous ROSAT observations of HCG 62 indicate a temperature decline beyond ∼120 kpc (Ponman & Bertram 1993), so our result is consistent with the earlier literature.

Zoom In Zoom Out Reset image size

The azimuthally averaged projected abundance profiles (Figure 3, right) do not show strong gradients, but are instead relatively flat around values of ∼0.5 solar for both Si and S, and ∼0.3 solar for Fe. We were able to constrain O only in the central region, with a value of ∼0.4 solar.

To correct for the effect of projection along the line of sight, we also performed a deprojection analysis on the same annular spectra used in Section 4.1 by adopting the XSPEC projct model. Under the assumption of ellipsoidal (in our specific case, spherical) shells of emission, this model calculates the geometric weighting factor, according to which the emission is redistributed amongst the projected annuli. As expected, the deprojected central temperature is lower than the projected one, since in the projected fits the spectrum of the central annulus is contaminated by hotter emission along the line of sight. We also note that the low value that we measure probably represents the very coolest gas in the core, owing to emission weighting in the large bin adopted to account for the XMM-Newton point-spread function. Due to the limited photon statistics relative to the high number of parameters in the projct model we could only poorly constrain the deprojected abundance profiles, which therefore are not shown.

We also estimate various quantities derived from the deprojected spectral fits. The electron density n_e is obtained from the Emission Integral EI = ∫n_en_pdV given by the vapec normalization: 10⁻¹⁴EI/(4π[D_A(1 + z)]²). We assume n_p = 0.82n_e in the ionized intra-cluster plasma. By starting from the deprojected density and temperature values, we can then calculate the gas pressure as p = nkT, where we assume n = 2n_e, and the gas entropy from the commonly adopted definition S = kT n^−2/3_e. The results are reported in Table 2 and the corresponding deprojected temperature, density, pressure, and entropy profiles are shown in Figure 4.

Zoom In Zoom Out Reset image size

The interaction of the radio source with the thermal gas can produce significant effects on the X-ray properties of the group halo that are thus expected to show azimuthal variations. To investigate this, we performed a spectral analysis in annular sectors each having angular aperture of 45°. Starting from the W direction with position angle (P.A.) = 0° and counting counterclockwise, the sectors are labeled as WN (westnorth), NW (northwest), NE (northeast), EN (eastnorth), ES (eastsouth), SE (southeast), SW (southwest), and WS (westsouth). Each sector was divided in the same six annular regions used in Section 4.1, obtaining the regions shown in Figure 5 (left panel). We produced projected radial temperature and metallicity profiles by extracting spectra in these regions. In particular, we compared the radial properties along and orthogonal to the radio lobes and cavity system. The northern and southern cavities fall entirely within the first annulus in the NE and SW sectors, respectively. The profiles derived along the different sectors are shown in Figure 5 (top right), while the temperature and metallicity wedge maps are shown in Figure 5 (bottom left and right, respectively). The gas properties do not appear to show variations obviously associated with the cavities and radio lobes. This indicates that the spectral analysis in large sectors is probably not the most sensitive way to study the two-dimensional distribution of temperature and metallicity in such a disturbed system as HCG 62. Higher-resolution spectral maps are presented and discussed in Section 5.3.

Zoom In Zoom Out Reset image size

5.1.1. Energetics

5.1.2. Pressure Balance

Since the radio source is filling the cavities, we can compare directly the radio pressure of the relativistic plasma internal to the lobes with the X-ray pressure of the surrounding thermal gas. The pressure of the hot gas is measured from the density and temperature derived from the X-ray data as p = 2n_ekT (Section 4.2). In particular, for the X-ray pressure in the northern and southern cavities we assume the values estimated in the first bin by performing a deprojection analysis along the NE and SW sectors, respectively (see the top left panel of Figure 5). We find p_X,N = 2.5 × 10⁻¹¹ erg cm⁻³ and p_X,S = 3.2 × 10⁻¹¹ erg cm⁻³. These values are consistent with the gas pressure at the radius of the cavities estimated by Morita et al. (2006).

The total pressure in a radio lobe is the sum of the magnetic pressure, p_B, and the total particle pressure, p_part, and can be written as

$$\begin{equation} p_{\rm radio} = p_{B} + p_{\rm part} = \frac{E_{B}}{\phi V} + \frac{1}{3} \frac{E_{\rm part}}{\phi V} = \frac{B^2}{8 \pi } + \frac{1}{3} \frac{(1+k) E_{\rm e}}{\phi V}, \end{equation} \tag{ 2 }$$

where k is the ratio of the energy in protons to that in electrons (E_e), V is the volume of the radio lobe, and ϕ is the volume filling factor. Using the expression for E_e given in Pacholczyk (1970), Equation (2) determines the lobe pressure in terms of the magnetic field strength and the factor k/ϕ, once the volume V of the radio lobe is known. This calculation is usually performed under the widely adopted minimum energy conditions, in which the relativistic plasma is in equipartition with the magnetic field (B_eq). For historical reasons the frequency band adopted to calculate the standard equipartition field is ν₁ = 10 MHz to ν₂ = 100 GHz, i.e., roughly the frequency range observable with radio telescopes. From a physical point of view, the adoption of this frequency band in the calculation of the minimum energy is equivalent to the assumption that only electrons emitting between 10 MHz and 100 GHz, i.e., with energy between γ_min ∝ (ν₁/B_eq)^1/2 and γ_max ∝ (ν₂/B_eq)^1/2, are present in the radio source. This approach neglects the contribution of the electrons emitting below 10 MHz and, as a more serious bias, in radio sources with different B_eq selects different energy bands of the electron population because the energy of the electrons which emit synchrotron radiation at a given frequency depends on the magnetic field intensity (Brunetti 2002). We therefore adopt a different approach to calculate the minimum energy conditions, in which B_eq does not depend on the emitted frequency band but directly on the low-energy cutoff of the electron spectrum. These so-called "revised" equipartition conditions select also the contribution to the energetics due to the low-energy electrons (Brunetti et al. 1997). In our calculations, we assume γ_min = 100 and the observed spectral index (see Table 4).

Table 4. Results of Equipartition Calculations

Quantity	Cavity N		Cavity S
	(revised)	(standard)	(revised)	(standard)
S235(mJy)	6.0	6.0	5.4	5.4
S610(mJy)	1.8	1.8	1.7	1.7
α610235	1.3	1.3	1.2	1.2
P235(W Hz−1)	2.5 × 1021	2.5 × 1021	2.3 × 1021	2.3 × 1021
γmin	100	900	100	850
Beq(μG)	6.7	3.1	6.6	3.3
pradio(erg cm−3)	2.3 × 10−12	4.8 × 10−13	2.3 × 10−12	5.5 × 10−13
pX/pradio	10.9	52.1	13.9	58.2
kbal	318	1830	500	2450

Notes. We assume ϕ = 1, k = 1, and V = V_X estimated in Section 5.1.1. A low-energy cutoff γ_min = 100 of the electron spectrum is assumed to calculate the "revised" equipartition field, which implies a spectral frequency range of ∼280 kHz–100 GHz. We also show for comparison the results of the standard equipartition calculations, which adopt instead a fixed emitted frequency band of 10 MHz–100 GHz.

Download table as:  ASCIITypeset image

The results of the equipartition calculations are reported in Table 4, where for comparison we also show the standard equipartition values. We estimate that the X-ray pressure is about 1 order of magnitude higher than the radio pressure, as typically found in cavity systems (e.g., Blanton et al. 2001; De Young 2006; Croston et al. 2008). We also find that with revised equipartition the cavities are closer to pressure balance than they are with standard equipartition. On the other hand, by assuming the lobes are in pressure equilibrium with the ambient gas we can obtain constraints on the particle content within the radio lobes (Dunn & Fabian 2004; Bîrzan et al. 2008). In particular, we determine the ratio k_bal of the energy in protons to that in electrons that is required to achieve pressure balance under revised equipartition conditions. We find k_bal∼ 300–500, whereas higher values (k_bal∼ few thousands) are expected with standard equipartition. These values are in the range found by Bîrzan et al. (2008), who studied the energetics and particle content of the lobes of 24 radio galaxies in cooling cores, obtaining values of k_bal up to approximately 4000 (with standard equipartition). Based on an analysis of nine FR-I radio galaxies in a sample of groups of galaxies, Croston et al. (2008) found that the radio lobes are underpressured by a factor up to 70 (with revised equipartition, assuming γ_min = 10 and α = 0.5), and that the pressure imbalance appears to be linked to the radio-source morphology, i.e., "plumed" sources typically have larger pressure deficits than sources where the jets are embedded in the lobes ("bridged" sources). The authors interpret this result as evidence that plumed sources have a higher entrainment rate due to the larger fraction of the jet surface which is in direct contact with the external medium, leading to an increase in k/ϕ. Although the classification into bridged and plumed morphologies may not directly apply to radio sources at the center of cooling core systems, typically having amorphous structures, this picture is consistent with the results of Dunn et al. (2006) who argue that the large pressure imbalance observed in radio bubbles as those of the Perseus cluster is more likely to be due to entrainment rather than a relativistic proton population. Unfortunately, we cannot derive good constraints on the energy in relativistic particles without knowing the break frequency in the radio spectrum. However, we note that HCG 62 shows a disturbed radio morphology with inner lobes clearly filling the well-defined X-ray cavities, but with outer lobes having no associated X-ray cavities (see Section 3.2). Assuming their detection is not limited by the sensitivity of the current Chandra image, this suggests the possibility of mixing between ambient gas and radio plasma in the lobes. Therefore, the k_bal > 0 values that we measure in the lobes might be the results of entrainment of thermal gas through the group atmosphere.

We wish to investigate in more detail the morphological features presented in Section 3.1 by measuring the surface brightness and temperature profiles.

The surface brightness profile extracted along the SW sector within an aperture of 260°–330° (counting counterclockwise from P.A. = 0° toward the west) is shown in the top left panel of Figure 6. By fitting a broken power-law density model, at 36 kpc from the center we identify a clear break in surface brightness which corresponds to a jump in density of 1.65. To unambiguously determine the nature of this front, we extracted the temperature profile along the same sector (Figure 6, bottom left) and find that the region immediately interior to the front is significantly hotter than the undisturbed region just outside of it, with a temperature jump across the front of ∼14%. As discussed below, this is consistent with the interpretation of the front as a shock.

Zoom In Zoom Out Reset image size

We also identify a front to the NE at 20 kpc from the center (see Figure 6, top right), that appears to have a symmetric feature to the SW (the surface brightness profile shows an edge at 16 kpc from the center, see the inner box in Figure 6, top left). We note that the detection of a temperature rise in the regions immediately inside of these fronts, which would be expected if they are shocks, is complicated by the underlying rising temperature profile of the global group atmosphere (see Figure 6, bottom). For this reason, despite the lack of evidence for a temperature jump across the inner fronts, we cannot rule out the presence of shocks. In fact, the symmetric placement of these fronts and their proximity to the cavities (3–5 kpc outside the cavity edge) support their interpretation as weak shocks. Alternatively, they could be interpreted as cold fronts, caused by the subsonic expansion of the cavities displacing cool, low entropy gas outward. However, since we cannot place conclusive constraints on these features with the present data, we will not discuss them further in the paper and in the following we will focus only on the outer shock detected to the SW (simply referred to as "the shock").

A hydrodynamic model for the shock was made by initiating an explosion at the center of a hydrostatic, isothermal atmosphere with a power-law density profile. The power-law index for the density for the unshocked gas was determined from the broken power-law fit to the surface brightness profile. Our data are consistent with the presence of a shock having a Mach number $\mathcal {M} = 1.45$ and an age ≳2.7 × 10⁷ yr. Compared to a model with continuous energy injection, the point explosion produces a stronger, i.e., faster, shock at early times so its age estimate is a lower limit. For such a Mach number our model predicts the emission measure weighted temperature to rise by ∼15% across the front, which is in agreement with our measurements. As it appears evident from the comparison with the temperature profile obtained with XMM-Newton along the same sector (overlaid in green in Figure 6, bottom left), the detection of the temperature jump is possible only with the superb spatial resolution of Chandra. We estimate an energy ∼3.4 × 10⁵⁸ erg, and a power ∼4.0 × 10⁴³ erg s⁻¹, which is about 1 order of magnitude higher than the cavity power. The unusually large difference between the shock power and the cavity power is essentially due to the fact that the shock is much further away than the cavities from the group center (R_shock = 36 kpc versus R_cav ∼ 8.5 kpc). Such a difference may be reduced by a factor of ∼2 if we adopt the cavity volume estimated from the radio data instead of that estimated from X-rays.

From a physical point of view, the difference in location and energetics between the shock and cavity system can be explained by two possibilities: (1) the shock and cavities are created by the same, violent AGN outburst and (2) the shock and cavities originate from different, multiple episodes of gentle AGN outbursts. (1) In the first case, the jet is initially supersonic and inflates the cavity violently, with much of the energy driving a shock. Once the jet slows the cavity rises buoyantly and the shock detaches from its tip and continues to propagate in the group atmosphere at a velocity $\mathcal {M} c_s$. By assuming that the shock was never any weaker than it is now and that the group atmosphere has an average temperature of ∼1 keV, we estimate the time for the shock to travel from the AGN to its current location to be ≲4.8 × 10⁷ yr, which is consistent with the shock age estimated above. The shock age is also consistent with the refill timescale of the cavities, but slightly longer than the buoyancy timescale (see Section 5.1.1). The buoyancy timescale can be considered as an upper limit to the age of the outburst that produced the cavities, since it assumes the cavities were inflated at the group core and have risen slowly. If they formed close to their current location their age would be even shorter, and the conflict with the shock age more serious. (2) In the second case, it is possible that the shock originated during a previous AGN outburst which produced the outer radio lobes and perhaps an associated pair of cavities, though these are not detected in the existing data. In this scenario, the inner cavities are thought to be evacuated by the inner radio lobes during a more recent outburst, thus providing a natural explanation for the marginal difference between the buoyancy timescale of the cavities and the age of the shock. The expected older outer cavities might be detected with deeper exposures in a region close to the observed shock front.

However, with the current radio data we are not able to determine the relationship of the outer radio lobes to the inner ones. Therefore, considering also the uncertainties in the timescales estimated above, we cannot determine whether the observed radio/X-ray morphology is the result of one single AGN outburst, or multiple episodes. In either case, the position of the shock outside the S radio lobe makes plausible the interpretation of the shock as being directly driven by the lobe expansion triggered by an AGN outburst. Averaging over the past few 10⁷ yr, the cavity power alone thus provides a lower limit to the true total mechanical power of the AGN. Inclusion of the energy in the shock provides a more complete estimate. The shock plus cavity power of ∼4.4 × 10⁴³ erg s⁻¹ is much higher than the cooling luminosity (L_X = 1.5 × 10⁴² erg s⁻¹, estimated in Section 5.1.1). If the power level of 4.4 × 10⁴³ erg s⁻¹ is sustained, since it exceeds the power radiated by cooling core, either the core is currently being heated, or most of the power must be deposited in gas beyond the cooling core. The fact that the shock front is located at the outer edge of the cooling region is consistent with the latter.

On the other hand, the apparent absence of an outer shock front to the NE direction, which would be expected if the shock was created by the two outer radio lobes, requires some asymmetry. It is also possible that the shock strength varies with the position angle. In this case, the shock front may be in fact intrinsically symmetric and surrounding the whole central region of the group, but is detected only to the SW because of some difference in gas properties (e.g., higher density). We investigate further these possibilities also by means of a temperature map, which shows a lack of very cool (<0.9 keV) material on the SW side of the core (see Section 5.3.1) that might be explained by the shock.

Since no obvious feature associated with the cavities and radio lobes appear in the wedge maps (Section 4.3), we produced spectral maps with higher spatial resolution in order to investigate the two-dimensional temperature (Section 5.3.1) and iron abundance (Section 5.3.2) distribution independently of the choice of the sectors used to extract the spectra. Such spectral maps are suitable for identifying potentially interesting structures.

Chandra spectral maps were produced using a method similar to that described in O'Sullivan et al. (2005). The ACIS-S3 field of view was divided into a grid of square map pixels with side length 10 physical pixels (∼5''). For each of these map pixels a spectrum was extracted from a circular region centered on the pixel, with a radius chosen so as to obtain 800 net counts in the 0.3–5.0 keV band. Radii were allowed to vary between 10 and 100 physical pixels. Spectra and responses were created in an identical fashion to that used in our main spectral analysis and fitted with an absorbed apec model. Point sources were excluded from both spectra and the calculation of the spectral extraction region size and energies outside the 0.3–5.0 keV band were excluded. 90% uncertainties on parameters were estimated and any pixel with a temperature uncertainty greater than 15% was excluded. The colors of the remaining map pixels indicate the best-fitting temperature, or abundance.

5.3.1. Temperature

In Figure 7 (left), we show the temperature map obtained from Chandra spectra. We also show in Figure 7 (right) a temperature image of the group central region built from XMM-Newton colors. Specifically, we extracted mosaicked MOS images in four different energy bands (0.5–1.0 keV, 1.0–2.0 keV, 2.0–4.5 keV, and 4.5–8.0 keV), subtract the background, and divide the resulting images by the exposure maps. A temperature in each square map pixel, chosen to have side length of 5 physical pixels (=55), is then obtained by fitting a thermal plasma model to the count rate estimated in the above energy bands in a surrounding area with a fixed radius of 15 physical pixels (=165).

Zoom In Zoom Out Reset image size

The two temperature maps are consistent with each other and with previous results (Morita et al. 2006; Sanders et al. 2010), showing the presence of the cool core, but also the presence of cool gas in the region of the northern cavity. Owing to emission weighting in the large map extraction regions, this cold feature could be related to the presence of the cavity rims, which are indeed typically observed to be cold (McNamara & Nulsen 2007, and references therein). On the other hand, the lack of a comparable cold feature in the region of the southern cavity is notable. One possible explanation is that there was initially the presence of cold gas also to the south, but that this has then been heated by the passage of the shock. This argument could be taken as an indication of an asymmetric shock, or at least of an asymmetric shock strength, as discussed in Section 5.2.

5.3.2. Iron Abundance

In Figure 8, we show the iron abundance map obtained from Chandra spectra. The iron abundance distribution appears to be inhomogeneous, asymmetric, and lacking the usual central peak. We have reason to believe that the structures seen in the abundance map are real, as they are also observed in the maps of 90% upper and lower bound on abundance (not shown here). However, the uncertainties on the value of individual map pixels are large (30%–50%) and the variable sizing of spectral extraction regions produces an effect analogous to adaptive smoothing. It is therefore necessary to extract separate spectra to determine the true abundance of individual regions and the significance of any abundance differences. In particular, to confirm the presence of the arc-like region of enriched material visible at about 2' to the SW of the group center, first discovered by Gu et al. (2007), we extract spectra from two elliptical regions centered at 2' to the NE and SW of the group center along the direction of the inner radio lobes. The semimajor axis of both ellipses is orthogonal to the direction of the inner radio lobes and 1' in extent, with the semiminor axis being half of it. We measure abundances of 0.12^+0.06_−0.04 solar and 0.95^+0.26_−0.16 solar (1σ errors) in the NE and SW regions, respectively. This indicates that the SW arc is more abundant than the NE dip at 5σ significance.

Zoom In Zoom Out Reset image size

Gu et al. (2007) discuss several possibilities about the origin of the high-abundance gas, including entrainment during the cavity expansion triggered by the AGN activity, and ram pressure stripping due to tidal interaction of two central galaxies which have experienced a recent (∼10⁸ yr) minor merger, as indicated by optical studies (Spavone et al. 2006). Unfortunately, we do not have enough information in the current data to disentangle this problem and the origin of the high-abundance arc-like region is still an open question. However, we stress here that the enriched material has a location and shape consistent with that of the newly discovered shock. Recent papers by Kaastra et al. (2009) and Prokhorov (2009) claim that in some cases, for instance near interfaces of hot and cold gas and near shocks, the usual approximation of Maxwellian electron distributions adopted for calculating thermal X-ray spectra is no longer valid. In such situations, the presence of non-thermal electrons modifies the line emissivities hence the X-ray spectrum, thus affecting the measurements of metal abundances. In particular, the best-fit iron abundance for the isothermal model is about 30% higher than the actual abundance (Kaastra et al. 2009). This could explain the high abundance arc that we measure. A similar effect of apparent iron overabundance may also be produced by the presence of non-equilibrium ionization states below the shock (Prokhorov 2010). These interpretations of the arc-like feature observed in the abundance map of HCG 62 are consistent with the presence of the detected shock front.