1: % use preprint,12pt for submission
2: %\documentclass[preprint,12pt]{aastex}%
3: %\documentclass[preprint]{aastex}
4: \documentclass{emulateapj}
5: % NOTE: emulatepj requires the use of deluxetable* for a two-column
6: % table; which won't work in preprint format, so you need to
7: % change it below
8: %\usepackage{natbib}
9: %
10: % \etal is no longer necessary, a) use \citet etc b) no font changes anymore
11: %\newcommand{\etal}{\mbox{et al.}}
12: \newcommand{\uv}{\mbox{$u$-$v$}}
13: \newcommand{\ex}[1]{\mbox{$\times 10^{#1}$}}
14: \newcommand{\cm}[1]{\mbox{cm$^{#1}$}}
15: \newcommand{\sn}[1]{\mbox{s$^{#1}$}}
16: \newcommand{\Msol}{\mbox{$M_\sun$}}
17: \newcommand{\Msolyr}{\mbox{$M_\sun$~yr$^{-1}$}}
18: \newcommand{\Rsol}{\mbox{$R_\sun$}}
19: \newcommand{\kms}{\mbox{km s$^{-1}$}}
20: \newcommand{\muas}{\mbox{$\mu$as}}
21: \newcommand{\muasyr}{\mbox{$\mu$as~yr$^{-1}$}}
22: \newcommand{\masyr}{\mbox{mas~yr$^{-1}$}}
23: \newcommand{\Jb}{\mbox{Jy~beam$^{-1}$}}
24:
25: \newcommand{\Ra}[4]{\mbox{${#1}^{\rm h} \; {#2}^{\rm m} \; {#3}\fs{#4} $}}
26: \newcommand{\dec}[4]{\mbox{${#1}\arcdeg \; {#2}\arcmin \; {#3}\farcs{#4} $}}
27: \newcommand{\rhoCSM}{\mbox{$\rho_{\rm CSM}$}}
28: \newcommand{\thu}{\mbox{$\theta_{\rm us}$}}
29: \newcommand{\dotthu}{\mbox{$\dot \theta_{\rm us}$}}
30: \newcommand{\thd}{\mbox{$\theta_{\rm disk}$}}
31: \newcommand{\ths}{\mbox{$\theta_{\rm shell}$}}
32: \newcommand{\thout}{\mbox{$\theta_{\rm o}$}}
33: \newcommand{\thin}{\mbox{$\theta_{\rm i}$}}
34:
35:
36: %MFB change from deprecated lefthead etc
37: \shortauthors{Bartel and Bietenholz}
38: \shorttitle{SN1979C VLBI}
39: \begin{document}
40:
41:
42: \title{SHELL REVEALED IN SN~1979C}
43:
44: \author{N. Bartel\altaffilmark{1} and M. F. Bietenholz\altaffilmark{1,2}}
45: \slugcomment{Received 2008 January 31; accepted 2008 April 14}
46: \journalinfo{{\em The Astrophysical Journal, 682, 2008 August 1}}
47: \altaffiltext{1}{Department of Physics and Astronomy, York University, Toronto,
48: M3J~1P3, Ontario, Canada}
49: \altaffiltext{2}{Hartebeesthoek Radio Observatory, PO Box 443, Krugersdorp,
50: 1740, South Africa}
51:
52: % puts in version number
53: %\slugcomment{Version 1c \today}
54:
55: \begin{abstract}
56: VLBI observations at 5 GHz have revealed that supernova 1979C, in the
57: galaxy M100 in the Virgo cluster, has shell structure. The shell is
58: approximately circular with the 50\% contour deviating from a circle
59: by an rms of no more than 8\% of the radius. The brightness
60: distribution along the ridge may vary. The position of the center of the shell is at
61: R.A.=\Ra{12}{22}{58}{66758} and decl.=\dec{15}{47}{51}{9695} (J2000), with a standard
62: error of 0.8 mas in each of the coordinates. No isolated central
63: component is visible above a flux density limit of $\sim$150 $\mu$Jy which
64: corresponds to an upper limit of $\sim$15 times the corresponding spectral
65: luminosity of the Crab Nebula. The radio lightcurve is clearly falling again
66: and the radio spectrum is now flatter than
67: at earlier times. SN~1979C is only the fourth, and oldest (optically
68: identified) supernova of which a detailed image could be obtained.
69:
70: \end{abstract}
71:
72: \keywords{supernovae: individual (SN~1979C)
73: --- galaxies: individual (M100) --- radio continuum: supernovae}
74:
75: \section{INTRODUCTION}
76:
77:
78: \objectname{SN~1979C} (type II-L) was discovered in the galaxy
79: \objectname{M100} \citep[\objectname{NGC~4321}, distance: 15.2 Mpc as
80: estimated by][]{Freedman+2001} in the Virgo cluster on 1979 April 19
81: by G. E. Johnson \citep{Mattei+1979} and subsequently became one of
82: the most optically and radio luminous Type II supernovae ever
83: found. On the basis of its radio lightcurve, its explosion date is
84: estimated to be 1979 April 4 \citep[see][]{Weiler+1986}, hereafter
85: $t=0$~yr. VLBI observations started as early as 1982. They resulted in
86: a determination of the angular expansion rate \citep{Bartel1985,
87: Bartel+1985a} and deceleration. This determination allowed an estimate
88: of the mass-loss to wind-velocity ratio of the progenitor, which was
89: found to be an order of magnitude lower \citep{SN79C} than that
90: reported from radio lightcurve fitting.
91:
92: However, despite a quarter century of VLBI observations, the
93: relatively far distance to Virgo has prevented determining the
94: detailed structure of \objectname{SN~1979C} with VLBI\@. The
95: determination of the expansion rate and the deceleration had to be
96: based on size estimates made by using a model fit of the supernova
97: brightness distribution. Clearly, a determination of the supernova's
98: morphology is important for a more accurate determination of the
99: expansion velocity and deceleration. Such more accurate determination
100: would also allow for a direct, geometric, estimate of the distance to
101: \objectname{SN~1979C} and Virgo using the expanding shock front method
102: \citep[ESM; see, e.g.,][]{SN93J-4} and resolve the ambiguity in an
103: earlier estimate \citep{SN79C}. It is also important for revealing the
104: morphological type of the supernova: purely shell-like or center
105: filled, perhaps with a bright compact component. Previously, only
106: three other optically-identified supernovae could be imaged with an
107: angular resolution high enough to unambiguously determine the
108: morphological type. All three, \objectname{SN 1986J}
109: \citep{Bartel+1991, SN86J-Sci}, \objectname{SN 1987A}
110: \citep{Gaensler+1997, Gaensler+2007}, and \objectname{SN 1993J}
111: \citep{Marcaide+1995a, BartelBR1995, SN93J-Sci, SN93J-3} have shell
112: structure, with SN~1986J having a central component in addition to the
113: shell. SN~1979C is the oldest optically-identified supernova that can
114: still be monitored with VLBI\@. As such, it provides a bridge between
115: our understanding of supernovae (interacting with the circumstellar
116: medium left over from the progenitor) and supernova remnants
117: (interacting with the interstellar medium).
118:
119: Only a few supernova remnants which are less than a century old are
120: known, all extragalactic, e.g., in M82 \citep{McDonald+2001} and in
121: Arp220 \citep{Parra+2007}. The youngest Galactic remnants are several
122: centuries old, e.g., the Crab Nebula, which shows a pulsar wind
123: nebula, Cas A, which shows a shell, and G21.5$-$0.9, which shows both
124: a pulsar wind nebula \citep[e.g.,][]{G21.5} and likely also a shell
125: \citep{MathesonS2005, Bocchino+2005}. Since all radio supernovae
126: originate from the explosions of massive stars, they are expected to
127: leave a neutron star or a black hole. Either a young neutron star or
128: a black hole may be accompanied by a bright, compact radio source near
129: the center of the expanding shell. Young neutron stars are likely to
130: be associated with radio-bright pulsar wind nebulae, and black holes
131: may have radio bright jets due to an accretion-disk system.
132:
133: Indeed, pulsar wind nebulae are found around a number of young
134: pulsars, although the youngest known pulsars have ages near 1000~yr.
135: A central radio source possibly associated with a neutron star or black hole was
136: recently found in SN~1986J \citep{SN86J-Sci}. Such a central radio
137: source might also be seen near the centers of other older supernovae
138: when clearly resolved images are obtained. Here we report on such an
139: image of SN~1979C, discuss its morphology, radio lightcurve, and
140: spectral evolution, and search for signs of a central compact remnant
141: of the explosion. The expansion rate, deceleration, and distance
142: determination will be reported elsewhere.
143:
144: \section{OBSERVATIONS AND DATA ANALYSIS}
145:
146: The VLBI observations were carried out on 2005 February 25 with a
147: global array of 21 antennas. The array consisted of the Very Long
148: Baseline Array (VLBA; each of 25~m diameter), phased Very Large Array
149: (VLA; 130~m equivalent diameter), and the Robert C. Byrd telescope
150: (GBT; $\sim$105~m diameter), all of the National Radio Astronomy
151: Observatory, NRAO, in addition to the Arecibo Radio Telescope (305~m
152: diameter) and eight telescopes of the European VLBI Network, namely
153: the Effelsberg Radio Telescope, Germany
154: (100~m diameter), the Westerbork Array, The Netherlands (92~m
155: equivalent diameter), and the telescopes at Medicina and Noto, Italy
156: (32~m diameter), at Torun, Poland (32~m diameter), Jodrell Bank,
157: Great Britain (26~m diameter), and Hartebeesthoek, South Africa (26~m
158: diameter). We observed for a total of $\sim$15~h at a frequency of
159: 5~GHz with a bandwidth of 32~MHz and recorded both senses of circular
160: polarization with a total recording bit rate of
161: 256~Mbits~s$^{-1}$. The data were correlated with the NRAO VLBA
162: processor in Socorro and further analyzed with NRAO's Astronomical
163: Image Processing System (AIPS)\@. The initial flux density
164: calibration was done through measurements of the system temperature at
165: each telescope, and then improved through self-calibration of the
166: reference sources.
167:
168: We phase-referenced our VLBI observations to \objectname{J1215+1654},
169: which is 2.2\arcdeg\ away on the sky, and used, through most of the
170: run, a repeating cycle of 250 s, pointing for 170~s to SN~1979C and
171: for 80~s to the reference source. Finally, we also used the VLA,
172: which was in the B configuration, to determine the total flux density
173: of SN~1979C at 1.43, 4.99, and 8.43~GHz. In order to obtain the best
174: possible image, we phase self-calibrated the data after
175: phase-referencing.
176:
177:
178: \section{RESULTS}
179:
180: \subsection{The Brightness Distribution}
181:
182: In Figure~\ref{snf} we show the VLBI image of SN~1979C juxtaposed with
183: an earlier VLA image of the host galaxy M100. The supernova is located
184: in a southern spiral arm of the galaxy. The VLBI image shows that the
185: supernova has shell structure
186: with a brightness minimum located only slightly to the west of the
187: center. The outer
188: contour at half of the maximum brightness is approximately circular
189: with rms deviations from circularity of 8\% of the radius. We
190: think that this estimate is somewhat conservative since it is based on
191: an image convolved with an elliptical beam, which tends to elongate the
192: structure along the long axis of the beam, which is oriented
193: north-south. The 50\% contour is indeed slightly elongated in the
194: north-south direction. An ellipse fit to the 50\% contour has a ratio
195: between the major and minor axes of 1.10 and a p.a.\ of 173\arcdeg.
196:
197: The maximum brightness is only 187 $\mu$\Jb\ and the rms noise
198: brightness in the background is 11~$\mu$\Jb. How much could possible
199: deconvolution errors contribute to the uncertainty in the brightness
200: distribution of the supernova? We investigated the significance of the
201: appearance of the shell structure and of features in the brightness
202: distribution by using simulations. We used a disk model as a test
203: source, converted it to visibility data for our \uv~coverage, added
204: noise, and then imaged and deconvolved the simulated data in the same
205: way as the observed data. In this way we produced 36 images of our
206: model source with different noise realizations but with the rms
207: background noise brightness being always close to our observed background rms of
208: 11~$\mu$\Jb. The average of the 36 images resembled the original disk
209: model, indicating that there are no systematic distortions or
210: biases. The individual images, however, varied in their appearance
211: with the rms of the brightness variations on the image being twice as large
212: as the rms of the background noise brightness. None of these images
213: had a closed ridge with a clear minimum in or near the center. We
214: therefore exclude the possibility that the true brightness
215: distribution is uniform across the source or has a filled
216: center. Instead, with high probability, it is indeed a shell.
217:
218: Our simulations also showed that the northern and southern
219: low-brightness extensions are likely due to corrugations in the image
220: plane caused by our incomplete \uv~coverage, and are therefore
221: probably not significant. The brightness varies along the ridge
222: between 187~$\mu$\Jb\ and 112~$\mu$\Jb, representing a
223: range of only $\pm 1.7\times$ the estimated on-source brightness
224: uncertainty of $\sim22$~$\mu$\Jb. It is not clear whether the variation is significant.
225:
226: The brightness at the local minimum slightly to the west of
227: the center of the shell is 80~$\mu$\Jb. The brightness increases
228: toward the east and is $\sim130~\mu$\Jb at the center, although no isolated
229: central component is visible. We determine an upper limit of the flux
230: density of any compact component in the central area of the shell of
231: $\sim150~\mu$\Jb.
232:
233: \subsection{The Position}
234:
235: Phase-referencing allowed us to determine the position of SN~1979C
236: relative to that of our calibrator, J1215+1654. The calibrator's
237: position was taken from the VLBA Calibrator Survey
238: \citep[VCS1,][]{Beasley+2002} as R.A.=\Ra{12}{15}{03}{979123} and
239: decl.=\dec{16}{54} {37}{95733} (J2000). The standard
240: error\footnote{\citet{Beasley+2002} call this error ``inflated formal
241: error.''} is 0.5 and 0.6~mas in RA and decl., respectively. In our
242: VLBI image of SN~1979C in Figure \ref{snf}, we place the origin of
243: the coordinate system at the estimated center of the outer 50\%
244: contour. This coordinate origin is at the position of
245: R.A.=\Ra{12}{22}{58}{66758} and decl.=\dec{15}{47}{51}{9695} (J2000). We
246: estimate that our standard error of the position of SN~1979C relative
247: to J1215+1654 is 0.5~mas in each coordinate \citep[compare
248: with][]{PradelCL2006}.
249: Adding in quadrature the standard position error of J1215+1654's to
250: that of SN~1979C, we estimate a standard error of the position of the
251: center of the image in Figure~\ref{snf} to be 0.8~mas in each
252: coordinate.
253:
254: As a check, we also determined the position of SN~1979C relative to
255: J1215+1654 by using our earlier 1.7~GHz data from 2001. At this
256: frequency, SN~1979C was barely resolved, and contributions to the
257: position error due to unmodeled effects of the ionosphere expected to
258: be much larger. We estimate a standard error of 2~mas in each
259: coordinate. Our estimated relative position of SN~1979C at 1.7~GHz in
260: 2001 was within 1~mas of the more accurate position given above, well
261: within the uncertainties.
262:
263:
264:
265:
266: \subsection{VLA Total Flux Densities}
267: We determined the total flux density, $S_{\nu}$, of SN~1979C from the
268: interferometric observations with the VLA made during the time of our
269: VLBI observations. We list these flux densities for frequencies,
270: $\nu$, of 8.43, 4.99, and 1.43~GHz in Table~\ref{fluxt}. In addition, we
271: give the flux densities for $\nu = 8.46$ and 1.67~GHz measured in 2001, for
272: comparison. Then we convert the flux densities to $\nu = 4.89$ and 1.47~GHz and plot
273: them in Figure~2, together with the much larger set of
274: flux density determinations at these frequencies of \citet{Weiler+1986, Weiler+1991},
275: \citet{Montes+2000}, and \citet{SN79C}.
276:
277:
278: % need deluxtable* for emulateapj and plain deluxetable for preprint
279: % make sure to change the end{deluxetable} accordingly
280: \begin{deluxetable*}{l c c c c c@{\hspace{0.5in}}c}
281: %
282: %\begin{deluxetable}{l c c c c c@{\hspace{0.5in}}c}
283: %
284: \tabletypesize{\small}
285: %\rotate
286: \tablecaption{Total flux densities and spectral indices of SN~1979C}
287: %\tablewidth{0pt}
288: \tablehead{
289: \colhead{Date} & \colhead{Age (yr)\tablenotemark{a}} & \multicolumn{4}{c}{Flux
290: densities (mJy)\tablenotemark{b}}
291: & \colhead{$\alpha$\tablenotemark{c}\phn\phn}\\
292: % dunno why the above doesn't center nicely, add the \phn's to make it nice
293: & & \colhead{1.43 GHz} &
294: \colhead{1.67 GHz} & \colhead{4.99 GHz} & {8.43 GHz}
295: % don't use colhead for 8.43 GHz for nicer spacing
296: }
297: %
298: \startdata
299: 2001 Feb 23& 21.89 & & 4.28$\pm$0.23 & & \phn2.3$\pm$0.4\tablenotemark{d}\tablenotemark{e} & $-0.38\pm0.15$ \\
300: 2005 Feb 25& 25.90 & 3.19$\pm$0.22 & & 1.68$\pm$0.09 & 1.37$\pm$0.10 & $-0.49\pm0.09$ \\
301: \enddata
302: %
303: \tablenotetext{a}{Time since assumed explosion date of 1979 April 4 (1979.26).}
304: \tablenotetext{b}{Total flux density, $S_\nu$, measured with the VLA\@.
305: The uncertainties are approximate standard errors and were computed by adding
306: in quadrature the statistical standard error and a systematic calibration
307: error of $\ge 5$\%.}
308: \tablenotetext{c}{The spectral index, $\alpha$, determined from
309: the flux density measurements, $S_{\nu}$, at frequency, $\nu$:
310: $S_{\nu}\propto\nu^{\alpha}$. For our last epoch a weighted fit was used. The errors
311: are one standard deviation, determined by varying the flux density values by their standard errors.}
312: \tablenotetext{d}{Data are from \citet{SN79C}.}
313: \tablenotetext{e}{The observing frequency was 8.46 GHz.}
314:
315: \label{fluxt}
316: %
317: \end{deluxetable*}
318: %\end{deluxetable}
319:
320: The new measurements of the flux density show that the supernova
321: is again fading, reaching at 1.5~GHz its lowest level ever. The bumpy
322: pattern seen previously is extended downward.
323:
324: \subsection{Radio Spectra}
325:
326: \citet{Montes+2000} found that the optically-thin spectrum had a
327: spectral index, $\alpha$, of $-0.75$ with a deviation range of $-0.76$
328: to $-0.63$ for $t<12$~yr (i.e., before 1991; $\alpha$ being obtained
329: from the fit of an empirical model to the multi-frequency radio
330: lightcurves), and $\alpha \sim -0.70$ for $t\geq12$~yr. At their
331: latest times, $t=17.7$ and 18.8~yr, they obtained $\alpha =
332: -0.63\pm0.03$. Our latest measurements give still flatter spectra
333: (see Table~\ref{fluxt}). The difference of our spectral indices to
334: that of $-0.63$ may be significant. In Figure~\ref{spectraf} we plot
335: our flux density measurements and the corresponding spectra for our
336: two latest epochs. For comparison we show the flux density
337: measurements at $t=17.7$ and 18.8~yr from \citet{Montes+2000} and the
338: corresponding average spectrum. Clearly, there is evidence that the
339: spectrum of SN~1979C has flattened further since $t\sim$17~yr.
340:
341: \section{DISCUSSION}
342:
343: SN~1979C is the fourth (optically identified) extragalactic supernova
344: for which an image with sufficient angular resolution could be made to
345: determine the nature of the brightness distribution. The other three,
346: as mentioned in the introduction, are SN 1986J, SN 1987A,
347: and SN 1993J\@. Our VLBI image shows that SN~1979C has a morphology
348: similar to the other three, namely a fairly to highly round shell,
349: with the brightness possibly being modulated along the ridge. The
350: possible brightness modulation likely arises from density
351: irregularities either in the circumstellar medium or in the ejecta.
352:
353: The flux density of SN~1979C is now clearly again on the downward
354: trend. Earlier, it was suggested that the radio lightcurve exhibits a
355: quasi-periodic pattern \citep{Weiler+1992} which was interpreted as
356: possibly being due to the dynamical modulation of the wind from the
357: red supergiant progenitor by a binary companion
358: \citep{SchwarzP1996}. With more data up to $t\sim19$~yr, it was
359: noticed that any periodic variation of the radio lightcurve was not
360: extended beyond $t\sim10$~yr. Instead, it was suggested that, with
361: the flattening of the radio lightcurve, the supernova had apparently
362: entered a new stage of evolution \citep{Montes+2000}.
363:
364: With our new data up to $t\sim26$~yr, which show that the lightcurve
365: is again falling after $t \sim 16$~yr, a new look at the
366: characteristics of the radio lightcurve is warranted. In particular,
367: we think that the evidence for a periodicity in the radio lightcurve
368: or a new stage of evolution now seems weaker. If at all, then the
369: radio lightcurve appears to be periodic on a logarithmic time scale.
370: However, a closer inspection leads us to think that, rather than being
371: periodic on a logarithmic time scale, the radio lightcurve is
372: modulated on a whole range of time scales. For instance, there appear
373: to be variations on time scales of less than a year. There are also
374: variations on time scales of several years with e.g., a 5-yr bump in
375: the radio lightcurve at $t\sim7$~yr and a bump lasting at least 15 yr
376: at the end of the observed radio lightcurve. These variations are
377: likely caused by variations in the density of the ejecta \citep[for
378: discussions in the case of SN~1993J, see][]{MioduszewskiDB2001} and/or
379: of the circumstellar medium. In case of the variations in the latter,
380: the radio lightcurve can be thought of as a ``time machine''
381: which records in reverse the mass-loss to wind-velocity ratio of the
382: progenitor for thousands of years before the star died.
383:
384: The spectrum appears to have flattened somewhat with the spectral
385: index being above the deviation range of $-0.76$ to $-0.63$ given by
386: \citet{Montes+2000}. A similar behavior is also seen for SN 1993J
387: where the spectrum flattens while the source dimming
388: accelerates\footnote{Please note that although \cite{Weiler+2007}
389: claim that the spectral index of SN~1993J remains constant over time,
390: an increase of the spectral index consistent with the measurements of
391: \citet{SN93J-2} and \citet{Perez-TorresAM2002} can be clearly seen in
392: their data with the smallest errors and/or scatter.} \citep{SN93J-2,
393: Perez-TorresAM2002}.
394:
395: No clearly isolated component was found in the central region of the
396: projected shell of SN~1979C\@. However, the brightest portion of the
397: shell appears to extend toward the center in projection,
398: and could overlap with any compact component that may be present in
399: the interior. An isolated central component was in fact found for
400: SN~1986J, with a spectrum peaking at $\sim$20~GHz while the emission
401: from the remaining parts of the supernova has an optically thin
402: spectrum from at least 1.7 to 22~GHz \citep{SN86J-CosparII,
403: SN86J-Sci}. This isolated central component may be emission
404: associated with the compact remnant of the explosion.
405:
406: It is intriguing that the spectrum of SN~1979C shows evidence of
407: flattening with time. Such flattening would be expected for a
408: supernova with emission from a compact remnant that is starting to
409: appear in the center of a shell which is expanding and becoming
410: increasingly transparent at progressively higher frequencies.
411: However, no central component can clearly be seen in SN~1979C, so the
412: observed spectral flattening cannot be ascribed to such a component.
413:
414: It is probable, therefore, that the spectral flattening is intrinsic
415: to the shell's synchrotron emission. The flattening must then be due
416: to a change in the population of accelerated particles which produce
417: the synchrotron emission. Calculations have shown that physical
418: conditions in supernova shocks can effect the spectral index of radio
419: emission \citep[e.g.,][]{Ellison+2007, EllisonBB2000}.
420:
421:
422: Our new upper limit on the flux density of any central compact source
423: in SN~1979C of 150~$\mu$Jy puts strong constraints on the luminosity
424: of such a source. \citet{BandieraPS1984} argued that young
425: pulsar wind nebulae would have spectral luminosities of 10 to 1000
426: times that of the Crab Nebula.
427: The spectral luminosity of the central component of SN 1986J is
428: $\sim$200 times that of the Crab Nebula between 14~GHz and 43~GHz
429: \citep{SN86J-CosparII}. In contrast, the upper limit on the 5~GHz
430: spectral luminosity of any central component in SN~1979C is only 15
431: times the corresponding spectral luminosity of the Crab Nebula, more
432: than an order of magnitude fainter than both that in SN~1986J and our
433: previous limit for SN~1979C \citep{NB-COSPAR}, and at the low end of
434: \citet{BandieraPS1984}'s estimate. Any pulsar wind nebula that may
435: exist inside the shell can probably not yet be detected because the
436: shell is likely still opaque at radio frequencies. Since the
437: supernova shell becomes less dense and more transparent as it expands,
438: any central component will become more easily detectable in the
439: future. Searching for emission from the compact central remnant of the
440: explosion therefore remains an intriguing prospect.
441:
442: \acknowledgements
443:
444: Research at York University was partly supported by NSERC\@. NRAO is
445: a facility of the National Science Foundation (NSF) operated under
446: cooperative agreement by Associated Universities, Inc. The Arecibo Observatory is
447: part of the National Astronomy and Ionosphere Center, which is
448: operated by Cornell University under a cooperative agreement with the
449: NSF\@. The European VLBI Network is a joint facility of European and
450: Chinese radio astronomy institutes funded by their national research
451: councils. We have made use of NASA's Astrophysics Data System
452: Abstract Service.
453:
454: \bibliographystyle{apj}
455: \bibliography{/home/bartel/paper/nb-bib}
456: %,/home/michael/tex/mybib1.bib,/home/michael/tex/proposals/g21.5.bib}
457:
458: %\end{thebibliography}
459:
460:
461: \clearpage
462:
463:
464:
465: \begin{figure}[ht]
466: \centering
467: \includegraphics[width=\textwidth, trim=0 0.in 0 0.in]{f1.eps}
468: \caption{Left side: A radio image of the spiral galaxy M100 (NGC~4321)
469: in the Virgo cluster, with SN~1979C in it, made with the
470: VLA in the C array at 1.7~GHz on 1996 March 15. The contour levels are at 1.3,
471: 2.5, 4, 8, 16, 32, and 64\% of the peak brightness of 15.8~m\Jb. The full-width at half-maximum
472: (FWHM) restoring beam is plotted in the lower left corner. The image
473: is taken from \citet{SN79C}. Right side: the VLBI
474: image of SN~1979C, made at 5.0~GHz on 2005 February 25. The contour
475: levels are at $-17$, 17, 30, 40, {\bf 50} (thick white contours), 60, 70, and 90\% of the peak
476: brightness of 187~$\mu$\Jb. The grey scale, in $\mu$\Jb,
477: is given at the top. The rms brightness of the background noise is 11
478: $\mu$\Jb. The FWHM of the restoring beam is 2.0~mas $\times$ 1.1~mas
479: oriented at a p.a.\ of $-10$\arcdeg. North is up and east to the
480: left. The origin of the coordinate system is at R.A.=\Ra{12}{22}{58}{66759}
481: and decl.=\dec{15}{47}{51}{9695} (J2000) with a standard error of 0.8 mas in each coordinate. It is
482: at the approximate center of the outer 50\% contour (see text).}
483: \label{snf}
484: \end{figure}
485:
486:
487: \begin{figure}[ht]
488: \includegraphics[height=3.5in]{f2.eps} \figcaption{The radio lightcurves
489: at 1.47 and 4.89~GHz. Open symbols represent values taken from
490: \citet{Weiler+1986, Weiler+1991} and \citet{Montes+2000}. Filled symbols
491: represent values from \citet{SN79C} and this paper.
492: The observed values were taken at a number of slightly different
493: frequencies. If necessary, they were converted to the above frequencies with
494: $\alpha=-0.75$ \citep{Montes+2000}, except for our values at $t=22$
495: and 26~yr, for which we used the spectral indices from
496: Table~\ref{fluxt}. The 4.89-GHz value at $t=22$~yr is interpolated
497: between the 1.67 and 8.43~GHz values from Table~\ref{fluxt}.}
498:
499: \bigskip
500: \bigskip
501: $^3$~\footnotesize{
502: Conversion with a slightly different spectral index of $-0.63$, as
503: used in \citet{SN79C} would result in values virtually
504: indistinguishable from those plotted.}
505: \label{fluxf}
506: \end{figure}
507:
508:
509: \begin{figure}[ht]
510: \centering
511: \includegraphics[height=3.5in]{f3.eps}
512: \caption{The flux densities from Table~\ref{fluxt} for the epochs
513: given in the lower left. For comparison and to show evidence for the flattening of
514: the spectrum with time, the flux densities given by
515: \citet{Montes+2000} for the time range from $t=17.68$ to 18.82 yr after
516: the explosion are also plotted. The straight lines indicate fits
517: to the corresponding data.}
518: \label{spectraf}
519: \end{figure}
520:
521:
522: \end{document}
523: