2A 0335+096.-There is a nearby bright star in the image that results in large
areas of the galaxy being masked. Significant star formation was detected by
Romanishin & Hintzen (1988).
5.6 2A0335+096 - This is known to be a cold front cluster (Mazzotta, Edge &
Markevitch 2003), which also possesses a radio-lobe cavity (Birzan et al. 2004).
A recent deep XMM-Newton analysis of this system indicates that it is also
likely to be undergoing a merger with a subcluster (Werner et al. 2006).
The cluster 2A 0335+096 is poor but popular. Sarazin et al. (1995) presented VLA
images showing that this cluster contains a double-lobed source with a nucleus.
Each lobe lies about 12" on either side of the nucleus of the central galaxy.
That galaxy has a spectacular, filamentary emission-line system with a bar and
filaments in the central 20 kpc (Romanishin & Hintzen 1988). The X-ray images
show two prominent cavities along with filamentary structures that are not
correlated with the radio or optical line emission (Mazzotta et al. 2003). The
structure of the radio source suggests that it was produced by multiple
Chandra data on this cluster have previously been analyzed by two groups,
Kawano et al. (2003) and Mazzotta et al. (2003). Their results disagree somewhat
with each other. Our results agree more with the latter, who may have used a
later CALDB version than the former group.
Our comparison with the Kawano et al. (2003) analysis shows that our
projected 1-T fit agrees in the region 0 kpc < r < 20 kpc, with temperatures
between 1.65 and 1.76 keV. However, the temperature gradient we derive rises
more quickly over the 20 kpc < r < 120 kpc range. At r ~ 40 kpc we find a
best-fit temperature value T = 3.0 +/- 0.1 keV while Kawano et al. (2003) find T
= 1.9 +/- ^0.1^_0.05_ keV. At r ~ 54-70 kpc our best-fit temperature is
3.8+/-0.15 keV while Kawano et al. (2003) obtain 2.5+/-0.1 keV. Our overall
temperature fits are also more consistent with the ASCA temperature of
2.86+/-0.02 keV (Horner 2001). The difference between our best-fit temperatures
and those of Kawano et al. (2003) arises from the version of the CXC Calibration
Database (CALDB) used to calibrate the data and the energy range used for
spectral fitting. We used CALDB 3.0 in our work, while Kawano et al. (2003) used
CALDB 2.9. Also, we fitted the spectral range 0.7-7.0 keV, whereas Kawano et al.
(2003) fitted the 0.5-10 keV range. In our experience, the earliest results from
Chandra for the coolest clusters are the most affected by improvements in the
calibration. Also, including bins at the highest energies without many counts in
them can lead to unreliable results.
We also compared our projected and deprojected one-temperature MEKAL fit
results with those of Mazzotta et al. (2003), who used the Kawano et al. (2003)
data. They do not report the version of CALDB that they used. The temperature
gradient we measure has a similar slope to the one they obtained. Over the
region of 0"-240", Mazzotta et al. (2003) show that their best-fit temperature
rises from 1.8 to 4.2 keV with a flattening at 100" < R < 200". The metallicity
and N_H_ profiles are also similar in value and behavior. A second temperature
component in the central region marginally improved our deprojected fit (Table
4), but the lower limits on the normalization of this component and the modest
improvement of ?2 are not indicative of a secure detection, particularly since
the uncertainties of the calibration are probably highest at the energies of
RX J0338.7+0958 is the cluster 2A 0335+096.