3.5 cm with the Australia Telescope Compact Array have an angular resolution of $200. We compare our radio images with the optical images from the Cerro ...
The Astrophysical Journal, 596:287–298, 2003 October 10 # 2003. The American Astronomical Society. All rights reserved. Printed in U.S.A.
SUPERNOVA REMNANT CANDIDATES IN THE 30 DORADUS NEBULA J. S. Lazendic Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138
J. R. Dickel Astronomy Department, University of Illinois at Urbana-Champaign, 1002 West Green Street, Urbana, IL 61801
and P. A. Jones CSIRO Australia Telescope National Facility, Box 76, Epping, NSW 1710, Australia Received 2002 December 2; accepted 2003 June 19
ABSTRACT We have searched the 30 Doradus nebula, a giant extragalactic H ii region, for the presence of embedded supernova remnants using high-resolution radio, optical, and X-ray images. Radio images obtained at 6 and 3.5 cm with the Australia Telescope Compact Array have an angular resolution of 200 . We compare our radio images with the optical images from the Cerro Tololo Inter-American Observatory to distinguish thermal and possible nonthermal components in the nebula. We have also mapped, for the first time, the extinction across the region at high angular resolution. We identify two sources as supernova remnant candidates in the nebula. Derived physical parameters for the adiabatic phase imply relatively young remnants (103 yr) evolving in a very dense environment (e102 cm3). We compare our results with high-resolution data from the Chandra X-Ray Observatory, but with only a short exposure these preliminary X-ray data lack the sensitivity to resolve our supernova remnant candidates from intervening hot gas of the diffuse thermal X-ray component of the nebula. Subject headings: dust, extinction — H ii regions — ISM: individual (30 Doradus) — Magellanic Clouds — supernova remnants
signatures may arise only after many SNe have gone off and the remnants begin to overlap to clean out the surrounding gas. Fewer than a half-dozen SNRs have been confirmed inside giant extragalactic (Yang, Skillman, & Sramek 1994) or Magellanic Cloud H ii regions (Chu et al. 1993, 1999). It is therefore important to locate more SNRs within young H ii region complexes to evaluate their statistics and energy inputs. As the nearest giant H ii region, 30 Dor is an obvious candidate for a more detailed search for new SNRs. The small detection numbers may be partly a result of observational selection. Standard techniques for SNR detection such as the presence of nonthermal radio emission, X-ray emission, and an enhanced [S ii]/H ratio show less obvious results when applied to giant H ii regions like 30 Dor because of confusion with the dominant thermal emission. The radio spectral index (where the flux density S / ) of shell supernova remnants ranges between 0.3 and 0.8, those for pulsar wind nebulae are flatter with values between 0.0 and0.3; and the optically thin thermal continuum radio emission from H ii regions has values near 0.1. In addition, large remnants are difficult to detect because they become less bright as they expand and age, which reduces their contrast with the surrounding H ii region. Small, bright remnants need to be observed with high resolution so that their nonthermal emission is not averaged over the larger area of the H ii region. An alternative kinematic technique, based on the profiles and widths of optical spectral lines, has revealed the presence of high-velocity material that could well have originated from SNRs in 30 Dor (Meaburn 1984; Chu & Kennicutt 1988). These results need to be followed up with complementary high-resolution radio imaging to look for more individual SNRs that might not be near the rapidly
1. INTRODUCTION
The 30 Doradus nebula (Bode 1801) in the Large Magellanic Cloud (LMC) is the nearest giant extragalactic H ii region and one of the largest star-forming regions known in the Local Group of galaxies. The nebula is 200 pc in diameter (using a distance of 50 kpc to the LMC) and is part of the 30 Dor region, which is 1 kpc in size (Walborn 1991). The nebula consists of expanding ionized shells, molecular gas, and warm dust concentrated in dense nebular filaments. It contains the most luminous and massive stars known (e.g., Rubio et al. 1998; Brandner et al. 2001). The massive young stars believed to be responsible for the ionization of the nebula are concentrated within a diameter of 40 pc in the 30 Dor cluster (NGC 2070) with a very compact core of massive stars, R136, in its center. The region around 30 Dor contains a number of supernova remnants (SNRs) (e.g., Mills & Turtle 1984; Chu et al. 1995), including the young supernova remnant 1987A. However, N157B (SNR 05386911) remains the only identified SNR in 30 Dor and is well separated, at a distance of 70 (105 pc), from the main body of the ionized gas in the nebula (see Fig. 6). It was distinct enough even in low angular resolution optical observations to be cataloged as a separate feature by Henize (1956). SNRs are often found near OB associations and H ii regions (Montmerle 1979; van den Bergh 1988; Chu 1997). Next to stellar winds, SNRs are expected to be a major influence on the global kinematics and morphology of the giant H ii regions in which they are located (Kennicutt 1984). A few individual SNRs may have trouble causing much largescale effect, however, because they will have small sizes and irregular shapes due to the confining high density of the surrounding material. It may be that significant kinematic 287
288
LAZENDIC, DICKEL, & JONES
moving gas and may have already broken out from confinement diluting their synchrotron radio intensity. Another important aspect of H ii regions is the mixture of gas and dust in them. The environs of 30 Dor, particularly a large ridge to the south, contain a lot of molecular material (Cohen et al. 1988; Johansson et al. 1994, 1998; Mizuno et al. 2001), although the region right around 30 Dor appears relatively clear. To see whether a significant amount of dust remains in clumps in 30 Dor, the visual extinction should be mapped by comparing H and H images. This procedure improves on using a radio continuum/H ratio, which can suffer from contamination by radio synchrotron emission coming from SNRs, for which we are also searching. The independent H image can separate the two effects. The first high-resolution radio and comparative optical study of the 30 Dor by Dickel et al. (1994) involved a detailed comparison of a 13 cm radio image (FWHM of 7>5) from the Australia Telescope Compact Array (ATCA),1 the only aperture synthesis telescope capable of seeing the low declination of the LMC, and an optical H image from the 1.5 m telescope at the Cerro Tololo InterAmerican Observatory (CTIO).2 They found an average extinction across the nebula of 1.1 mag, consistent with other investigations (Kennicutt 1984; Parker 1993). However, two regions appeared to show a large deviation from the mean extinction value, prompting further investigation. Thus, we have performed new observations at 6 and 3.5 cm, the two shorter wavelengths available on the ATCA, to study this complex region with better sensitivity and spatial resolution. The study also included SNR 157B, for which results have already been published (Lazendic et al. 2000). Both H and H images taken with the Curtis-Schmidt and 1.5 m telescopes at Cerro Tololo, respectively, were used to evaluate the extinction. The observations are described in x 2. In x 3, we derive the extinction across the region and compare the results of the H/H ratio measurements with the radio continuum results. The SNR identification results are presented in x 4, where we also compare them to previous studies and high-resolution X-ray data obtained with the Chandra X-Ray Observatory.
1 The ATCA is part of the Australia Telescope funded by the Commonwealth of Australia for operation as a National Facility, managed by CSIRO. 2 The CTIO is operated by the Association of Universities for Research in Astronomy (AURA), Inc., under a cooperative agreement with the National Science Foundation as part of the National Optical Astronomy Observatory (NOAO).
Vol. 596 2. DATA
2.1. Radio Data The observations at 6 cm were obtained with the ATCA between 1993 January and July and at 3.5 cm between 1998 May and 1999 June. The array consists of six 22 m antennas that can be arranged in configurations giving baselines from 31 m to 6 km (for more details see Frater, Brooks, & Whiteoak 1992). Observing parameters are summarized in Table 1. Seven array configurations were used for the 6 cm observations, and six were used for the 3.5 cm observations to provide good u-v sampling of this complex region. For all observations PKS B1934638 was used for amplitude and bandpass calibration, and PKS B0454810 was used for phase calibration. For the observations at 6 cm with the 750 m arrays we used a single pointing, while for the observations with the longer baseline arrays we used two different pointing centers. We obtained the observations at 3.5 cm in mosaic mode with 20 pointing centers to cover the whole region of 30 Dor, since the ATCA primary beam is limited to 50 at this wavelength. Accordingly, all the data were gridded and deconvolved using the mosaicked maximum entropy procedure (Sault, Staveley-Smith, & Brouw 1996). A joint deconvolution of the individual pointing fields was performed, as the individual deconvolution did not give satisfactory image quality because of a sidelobe response from sources just outside the antenna field of view in this complex region. We note that in the joint approach the primary-beam attenuation is corrected only partially, since large noise variations will be introduced across the mosaicked image by the noise in the individual pointings. The result is limited in dynamic range to 100:1 in our final image. The resulting images were restored with Gaussian beams of FWHM 1>8 1>7 (P:A: ¼ 65 ) at 6 cm and 1>5 1>2 (P:A: ¼ 84 ) at 3.5 cm. The full size of the images is 200 , but in Figures 1a and 1b we show only the region with the nebula of 60 in size. The average rms noise level is 0.1 mJy beam1 at both 6 and 3.5 cm. The two images are very similar and show the familiar filamentary structure, with bright knots spread across the nebula. To investigate possible variations of the spectral index distribution across 30 Dor, the 3.5 cm image was smoothed to the resolution of the 6 cm image. Both images are then corrected for variations in the background level determined from the average of several areas around the nebula. The nebula has a spectral index of about 0.1, which is consistent with previous investigations (Mills, Turtle, & Watkinson 1978; Klein et al.
TABLE 1 ATCA Observational Parameters Parameter
4.4 GHz
4.8 GHz
8.6 GHz
Date......................................................
1993 May 21, Jun 27 1993 July 16
1993 Jan 16, 19, 29 1993 Jul 23
Configurations...................................... Field center (J2000.0) ............................
1.5B, 6 (A, C) 05h 38m 45s , 69 060 0000 05h 38m 05s , 69 080 4500 128 32 100
750 (B, C, A, D) 05h 38m 45s , 69 060 0000
1998 May 1, 31, Aug 27 1998 Sep 26, Oct 22 1999 Jun 24 750 (A, E), 6 (C, A), 1.5D, 375 20 pointings
128 32 100
128 32 50
Total bandwidth (MHz)........................ Number of frequency channels.............. FWHM of primary beam ......................
No. 1, 2003
SNR CANDIDATES IN THE 30 DOR NEBULA
289
Fig. 1a Fig. 1.—Radio images of 30 Dor at (a) 6 cm and (b) 3.5 cm. The beams shown by the tiny dot in the bottom left corners have a HPBW of 1>8 1>7 at 6 cm and 1>5 1>3 at 3.5 cm. The gray scale is in units of Jy beam1.
1989). We found no significant variations of the spectral index across the region; i.e., we found no evidence for the presence of embedded nonthermal objects with steeper radio spectra in the nebula from the spectral index map. This is not surprising because of both the very bright thermal emission and the fact that some SNRs, such as N157B with a spectral index of 0.18 (Lazendic et al. 2000), would be barely distinguishable from an H ii region on the basis of the spectral index alone. 2.2. Optical Data The H image of the region, shown in Figure 2a, was provided by R. C. Smith as a preliminary image from the Magellanic Cloud Emission Line Survey (MCELS; Smith et al. 1999). The image was obtained in 1993 December with the CTIO 0.6 m Curtis-Schmidt telescope in conjunction with a Thomson CCD camera with an exposure time of ˚ with a FWHM 350 s. The H filter was centred on 6564 A ˚ . The image has a pixel size of 1>8, and the pointof 20 A spread function (PSF) was 3>5. The full image was 1024 pixels or 314, and the seeing is 200 , determined from fitting the Gaussian profile to stars in the field. The full image was 2048 pixels or 131.4 The absolute calibration of both line images was performed by measuring the flux in an area of 48 and a seeing of 3>5. The H image has a pixel size of 0>4 and a seeing of 200 . The stars are more prominent in the H image due to the difference in wavelength of the frame used for continuum subtraction (see text). The gray scale is in units of ergs s1 PSF1.
comparisons of the H and H images, the H image was convolved using the IRAF task GAUSS to match the poorer resolution of the H image. To correct for wavelength dependence of the extinction curve, we produced a color excess image across 30 Dor using the relation ð1Þ CHH ¼ 2:5 log FH =FH =2:92 ; where F is the integrated flux in units of ergs s1 cm2. The resulting image is shown in Figure 3a. There are no large variations in extinction apparent across the nebula, except toward the core, i.e., the R136 region, where it is difficult to subtract the continuum emission completely. The visual extinction is then (e.g., Seaton 1979) AV ðmagÞ ¼ 2:17CHH :
ð2Þ
The observed distribution of extinction in Figure 3a is quite uniform with a mean value near the value of 1.1 mag found previously (Dickel et al. 1994), which confirms the conclusion that there is little absorbing material within 30 Dor. A few dense clumps within the extended gas may have up to 2–3 mag more absorption that the mean. These may be spots to check for further star-forming activity.
3.2. The 6 cm/H Comparison If all of the radiation from a certain region is thermal emission from an ionized gas, then standard formulas allow us to determine the ratio of the radio continuum to H, and any apparent deficit in the H emission can be attributed to interstellar reddening. Extra nonthermal radio emission can produce the same observed effect, however, so we need to compare the apparent radio/H extinction with that calculated from the H/H ratio in order to evaluate the nonthermal radio contribution. Therefore, we evaluate the apparent visual extinction from 30 Dor as if the emission were all thermal and select sources that appear to have excess extinction over that seen from the H/H determination as possible SNR candidates. The optical and radio (thermal) emission from the ionized hydrogen can be related through the emission measure (EM) and the electron temperature (e.g., Dickel, Harten, & Gull 1983), EMH ðcm6 pcÞ ¼ 920 104 FH ðergs cm2 s1 Þ ; EM6 cm ðcm6 pcÞ ¼ 8350Tb ðKÞ ;
ð3Þ ð4Þ
where Tb is the observed surface brightness temperature
292
LAZENDIC, DICKEL, & JONES
Vol. 596
Fig. 2b
from Tb 2k S / 2 ;
ð5Þ
where S is observed radio flux in Jy, is the wavelength in cm, and k is Boltzmann’s constant. The equations are approximations for the Te 8000 K case. The visual extinction can then be found, AV ðmagÞ ¼ 3:2 logðEM6 cm =EMH Þ :
ð6Þ
The image obtained in this way is shown in Figure 3b. In Figure 4 we show the image of extinction across the 30 Dor nebula, overlaid with contours from the 6 cm radio image and the optical H image. 3.3. The Sources Selected sources all show a distinct excess in radio emission as indicated by the contours in Figure 4. They are not just areas of high extinction. In order to eliminate the possibility of bad pixels and noise in the images or data analysis, we have limited candidate SNRs to have a diameter of more than 400 and to have an excess of more than 1 mag in the apparent radio/H extinction as compared with the H/H extinction. The four sources with higher extinction, i.e., higher radio/optical ratio, are
listed in Table 2 in order of increasing right ascension and are also described below. The first column in Table 2 lists the name of each source and the type of object; the second and third columns list the coordinates; the third and fifth columns list the peak radio emission at 6 and 3.5 cm, respectively; the fourth and sixth columns list the integrated radio emission at 6 and 3.5 cm, respectively; and the last column lists the angular size of each source. While the coordinate system should be accurate to better than 100 , the listed positions are the centroids found by eye from Figure 3b and will have somewhat less precision. The source designations in Table 2 are given by the acronym MCRX (Magellanic Cloud radio–X-ray source), followed by the truncated J2000.0 coordinates (e.g., Dickel et al. 2001). While none of the identified sources shows an obvious shell structure, we do have only a few resolution elements across them, and there is still a lot of thermal emission present along the line of sight, so we cannot use that characteristic in any classification. Nonround outlines may be the result of confinement by the surrounding material. 3.3.1. MCRX J053831.8690620
This source appears only in radio images and is one of the two regions of apparent high extinction from Dickel et al.
No. 1, 2003
SNR CANDIDATES IN THE 30 DOR NEBULA
293
MCRX J053848.6−690412
MCRX J053848.3−690442
MCRX J053831.8−690620
MCRX J053838.8−690730
Fig. 3a Fig. 3.—Images of visual extinction AV across 30 Dor, assuming the emission is all thermal, from (a) H/H comparison and (b) 6 cm/H comparison. The sources with apparently higher radio/optical extinction discussed in the text are marked with circles and labeled as in Table 2. The gray scale is in units of magnitude.
(1994). Rosa & Mathis (1987) found that this region has a much lower electron temperature and higher helium content, which implies enrichment of the heavy elements in the ISM of this region. They suggested that this material could originate from the ejecta of very massive stars through their presupernova evolution, but no star has been detected. We note that there is a trough of optical emission running apparently northeast-southwest through that area, with this radio source being the only apparent object. We therefore suggest that there is a little material in the trough and that
this feature is indeed a SNR rather than a region with very high extinction. We compare our images with near-infrared (NIR) images from Hubble Space Telescope observations from Rubio et al. (1998). There is no Br emission toward this source, but there is a knot of H2 that might coincide with the source. The absence of Br emission implies that there is no strong ionization toward this region and that the object could be nonthermal. The H2 emission in that case could be shock excited from the SNR blast wave, as found in galactic SNRs
TABLE 2 Properties of the Higher Extinction Regions Listed in x 3.3
Source (1)
Type (2)
R.A. (J2000.0) (3)
Decl. (J2000.0) (4)
6 cm Ipeak (Jy beam1) (5)
6 cm Iint (Jy) (6)
3 cm Ipeak (Jy beam1) (7)
3 cm Iint (Jy) (8)
Size (arcsec) (9)
MCRX J053831.8690620 .............. MCRX J053838.8690730 .............. MCRX J053848.3690442 .............. MCRX J053848.6690412 ..............
SNR SNR Knot 2 H ii P4 H ii
05 38 31.8 05 38 38.8 05 38 48.3 05 38 48.6
69 06 20.0 69 07 30.5 69 04 42.2 69 04 12.7
0.02 0.01 0.04 0.01
0.08 0.10 0.20 0.07
0.03 0.01 0.04 0.02
0.09 0.11 0.23 0.09
10 8 11 9
Note.—Units of right ascension are hours, minutes, and seconds, and units of declination are degrees, arcminutes, and arcseconds.
294
LAZENDIC, DICKEL, & JONES
Vol. 596
Fig. 3b
(see, e.g., Lazendic et al. 2002a). We note that Ye, Turtle, & Kennicutt (1991b) report a possible SNR centered about 1500 southwest of this source by using a ‘‘ scaled subtraction ’’ method. In this method the H image is scaled by a radio/H ratio factor determined from the intensity in the H ii region unaffected by the emission from an SNR and then subtracted from the radio image (Ye et al. 1991a). Because the Molonglo Observatory Synthesis Telescope used by Ye, Turtle, & Kennicutt (1991b) had an angular resolution of 4500 and the source lies on a steep intensity gradient at the western edge of 30 Dor, we conclude that they are the same source at the refined coordinates given here. 3.3.2. MCRX J053838.8690730
This source is well defined in the radio images and has also been reported by Dickel et al. (1994) as a region of higher extinction. In H, however, it seems to be situated in a void. It has a high radio/optical and low H/H ratio, both of which point to a possible nonthermal emission, i.e., a SNR. To investigate this further, we compared our 6 cm radio image with the 2MASS J-band (1.25 lm) image in Figure 5. There is some emission toward the region around the source, but at the radio continuum peak there is a void
of NIR continuum emission, supporting the inference that radio emission from this source is nonthermal.
3.3.3. MCRX J053848.3690442 (Knot 2)
This region is a well-defined source in 6 and 3.5 cm images, but it is not very prominent in the optical images. A second SNR candidate was reported at this location by Ye et al. (1991b). However, the feature seems to have somewhat high H/H extinction in Figure 3a, which suggests the source is an H ii region with above-average extinction. Indeed, the region coincides with the location of the protostellar object ‘‘ Knot 2 ’’ of Walborn & Blades (1987) and the location of an H2O maser (Whiteoak et al. 1983; Whiteoak & Gardner 1986; Lazendic et al. 2002b).
3.3.4. MCRX J053848.6690413 (P4)
As in the previous case, this region is a well-defined source in the radio images only. It has a high radio/optical and slightly high H/H ratio. The region coincides with the protostellar object P4 (Hyland et al. 1992; Rubio, Roth, & Garcia 1992; Rubio et al. 1998), and we conclude that it is a young H ii region with moderately high extinction.
No. 1, 2003
SNR CANDIDATES IN THE 30 DOR NEBULA
295
Fig. 5.—Shows J-band (1.25 lm) image from 2MASS overlaid with the 6 cm image convolved to a resolution of 3>5. Contour levels are 4, 6, 8, 10, 12, 14, 16, 18, and 20 mJy beam1. Two SNR candidates are marked as crosses. The radio emission at the location of SNR candidate MCRX J053838.8690730 is clearly anticorrelated with the NIR emission from 2MASS, supporting the nonthermal nature of radio emission from this source. There is also no NIR emission at the location of SNR candidate MCRX J053831.8690620.
4. DISCUSSION
4.1. Identification of SNRs
Fig. 4.—Image of visual extinction across 30 Dor from the radio and H comparison (same as Fig. 3b) overlaid with (top) 6 cm radio contour levels at 2, 4, 8, 12, and 20 mJy beam1, and (bottom) H contour levels at 39, 77, 118, 231, and 385 1014 ergs s1 PSF1 level. The gray scale is in units of magnitude.
3.3.5. Other Possible Sources
Except for the first and third sources above, noted previously by Ye et al. (1991b), no other individual radio features within 30 Dor have been previously identified. There are a few spots to the east of MCRX J053848.6690413 that are prominent in the 6 cm/H comparison but that also show high extinction with approximately the same AV in the H/H comparison. Thus, they are probably H ii regions with a high extinction. We reach the same conclusion for another feature at about R:A: ¼ 05h 38m 30s and decl: ¼ 69 060 4500 that also has about the same value in both extinction maps. Finally, the features just north of MCRX J053838.8690730 are very uncertain because of the very low optical emission in that area.
Using new radio and optical data with improved angular resolution we have confirmed that the two regions with high radio-to-optical ratios from Dickel et al. (1994) are SNR candidates located in the main body of the 30 Dor nebula. A similar method for searching for embedded SNRs employed by Ye et al. (1991b) used the ‘‘ scaled subtraction ’’ method and suggest a presence of two SNRs in 30 Dor, but our more accurate determination of the extinction has excluded one of their candidates. We compared our results with the stellar content in 30 Dor and observations at other wavelengths. None of our SNR candidates are coincident with IR or optical stellar sources (Rubio et al. 1998; Brandner et al. 2001) or X-ray point sources (Portegies Zwart, Pooley, & Lewin 2002). SNRs are often expected to evolve in cavities of expanding shells formed by powerful stellar winds from the progenitor and previous explosions from other members of the stellar association, especially in a region crowded with OB stars such as 30 Dor. Thus, detailed spectroscopic observations have been used to investigate the kinematics of 30 Dor (Meaburn 1984, 1988; Chu & Kennicutt 1994). The nebula has a very complex structure, with isolated and connected expanding shells of various sizes (a few to 100 pc) and velocities (20–200 km s1). The shells with velocities less than 100 km s1 were suggested to originate from stellar winds (typical velocity 50 km s1), but velocities of more than 100 km s1 might be produced by SNRs alone (typical velocity 150 km s1) or in combination with stellar winds. However, a comparison between the LMC SNRs located in H ii regions and those in superbubbles showed that the former type of objects can have velocities larger than 100 km
296
LAZENDIC, DICKEL, & JONES
Vol. 596
s1, but the latter type of object may not, which is why Chu (1997) concluded that the kinematic method is not an unambiguous way to separate embedded SNRs. Chu et al. (1994) also used UV absorption lines to search for hidden SNRs in LMC superbubbles. The UV emission in a superbubble is produced by photoionization at the interface between the hot interior and cool shell, and all the observed lines should have the same velocity. In 30 Dor, Chu et al. (1994) found large velocity shifts between the highly ionized species, such as [C iv] and [Si iv], and the low ionized species, such as [S ii], [Si ii], and [C ii]. This was attributed to the presence of SNRs in superbubbles, because an embedded SNR will drive a fast shock into the walls of the superbubble and produce excess emission from highly ionized species. A similar signature is expected to occur in X-rays. An off-center SNR inside a superbubble whose shock wave has already encountered the walls of the superbubble will produce excess X-ray emission (Chu & Mac Low 1990; Wang & Helfand 1991a). However, it is not clear whether this type of remnant would produce radio emission because of the extremely low density of the ambient gas into which it is expanding. 4.2. Comparison with Chandra Data X-ray observations of 30 Dor have shown that the emission from the nebula is predominantly diffuse and thermal, with two discrete sources of stellar origin prominent in the central region of the nebula (Wang & Helfand 1991b; Wang ¨ gelman 1995). We compared our 1995, 1999; Norci & O radio data with preliminary Chandra data5 provided by L. Townsley (Townsley et al. 2002). These observations enabled X-ray studies to be performed on a subarcsecond (0>5) scale for the first time. Data were obtained with an exposure of 25 ks on 1999 September 19. The nebula was positioned on the ACIS-I array, which consists of four CCDs arranged in a square encompassing a 16
