On the Distance and Age of the Pulsar Wind Nebula 3C58
aa r X i v : . [ a s t r o - ph . GA ] O c t **Volume Title**ASP Conference Series, Vol. **Volume Number****Author** c (cid:13) **Copyright Year** Astronomical Society of the Pacific On the Distance and Age of the Pulsar Wind Nebula 3C 58
Roland Kothes National Research Council Canada, Herzberg Institute of Astrophysics,Dominion Radio Astrophysical Observatory, P.O. Box 248, Penticton, BCV2A 6J9, Canada
Abstract.
There is a growing community of astronomers presenting evidence thatthe pulsar wind nebula 3C 58 is much older than the connection with the historicalsupernova of A.D. 1181 would indicate. Most of the strong evidence against a youngage for 3C 58 relies heavily on the assumed distance of 3.2 kpc determined with H i absorption measurements. I have revisited this distance determination based on new H i data from the Canadian Galactic Plane Survey and added newly determined distances toobjects in the neighbourhood, which are based on direct measurements by trigonometricparallax. This leads to a new more reliable distance estimate of 2 kpc for 3C 58 andmakes the connection between the pulsar wind nebula and the historical event fromA.D. 1181 once again much more compelling.
1. Introduction
Of all the supernova remnants (SNRs) which are linked to historically observed su-pernova events the connection between the pulsar wind nebula (PWN) 3C 58 and thesupernova explosion observed A.D. 1181 by Chinese and Japanese astronomers is prob-ably the most disputed one. Stephenson (1971) claimed that there is a high probabilityfor a connection between the Guest Star from A.D. 1181 and the supernova explo-sion that created 3C 58. This has been revisited and confirmed by Stephenson & Green(2002). The main arguments are the length of the visibility of the 1181 event, indicat-ing a supernova explosion, and the lack of any other supernova remnant candidates. Forsuch a recent supernova there should be a visible SNR. Although it is very di ffi cult toargue with this argument, strong evidence against such a young age for 3C 58 has beenmounting (for a list of the main arguments see e.g. Fesen et al. 2008, Table 3).Most arguments against a young age for 3C 58 rely heavily on the assumed dis-tance of 3.2 kpc (Roberts et al. 1993). This distance was determined kinematically fromH i absorption measurements by comparing the resulting radial velocity with a flat ro-tation curve for the Outer Galaxy. However, in particular for Perseus arm objects, thisleads to a significant overestimate of the distance. A spiral shock in the Perseus armis “pushing” objects towards us, giving them a higher negative radial velocity, whichmakes them appear to be farther away (Roberts 1972). Examples for this e ff ect on dis-tances for SNRs and PWNe can be found in Kothes et al. (2002, 2003) and Foster et al.(2004). SNRs in particular rely on kinematic distance estimates, since their distancescannot be easily determined directly or by related stars as in the case of H ii regions.In this short article I present a new distance estimate for 3C 58 and based on thisnew distance I will discuss and re-evaluate evidence presented in the literature against1 Kothesthe historical connection of 3C 58. I will show that the new distance changes the char-acteristics of this PWN quite dramatically, which leads to a higher probability for itshistorical connection.
2. Observations
The H i data I used for the determination of the new absorption profile towards 3C 58were obtained with the Synthesis Telescope at the Dominion Radio Astrophysical Ob-servatory (DRAO, Landecker et al. 2000) as part of the Canadian Galactic Plane Survey(CGPS, Taylor et al. 2003). Single antenna data are incorporated into the interferom-eter maps to ensure accurate representation from the largest structures down to theresolution limit of about 1 ′ . The low spatial frequency data was drawn from the Low-Resolution DRAO Survey of the CGPS region, which was observed with the 26-m radiotelescope at DRAO (Higgs & Tapping 2000). The resolution in the final maps variesslightly across the final images close to 1 ′ × ′ cosec(DEC). At the centre of 3C 58 theresolution of the H i data is 58 ′′ × ′′ at an angle of 75 ◦ for the major axis (counter-clockwise from the Galactic longitude axis). The RMS noise is about 3 K T B in eachvelocity channel of width 0.82446 km s − at a velocity resolution of 1.3 km s − .
3. Results3.1. H i Absorption and Emission Associated with 3C 58
The latest distance determination for 3C 58 by H i absorption measurements resulted ina systemic velocity of ∼ −
38 km s − and a Perseus arm location (Roberts et al. 1993).This was translated with a flat rotation model for our Galaxy and the IAU supportedvalues for the Sun’s Galacto-centric distance of R ⊙ = . v ⊙ =
220 km s − to a distance of 3.2 kpc.I used the CGPS data to determine a new H i absorption profile towards 3C 58 tostudy not only the integrated absorption spectrum, but also the changes over the PWN.The resulting profiles are shown in Fig. 1. H i channel maps taken at the peaks of theabsorption spectrum and channels averaged over the interarm areas are displayed inFig. 2. The integrated absorption spectrum is literally identical to the one published byRoberts et al. (1993). However, I found that the weak peak, seen at around −
41 km s − is actually a real absorption feature. It is not seen over the entire PWN, but only onthe right hand side at position 3 (see Figs. 1 and 2). This is likely the reason why thisfeature was not obvious in previously published integrated absorption spectra.Wallace et al. (1994) discussed the possible association of 3C 58 with a large in-terstellar bubble. An inspection of the CGPS H i data set reveals that this double shellstructure is actually a very smooth object (see Fig. 3). The bubble is visible from about −
33 to −
38 km s − and 3C 58 is located in the top left area of it. If 3C 58 was locatedwithin this bubble, the last absorption feature at about −
41 km s − is likely produced bythe shell of the bubble expanding towards us. Roberts et al. (1993) found a Perseus arm location at a systemic velocity of about v LS R = −
38 km s − , which they translated to a distance of 3.2 kpc. This systemic veloc-ity was confirmed with the discovery of the large bubble by Wallace et al. (1994). Theistance and Ageof3C58 3 -0.8-0.6-0.4-0.20.0 -80-60-40-20 0 20 R e l a t i v e A b s o r p t i on Radial Velocity [km s -1 ]Pos. 3-0.8-0.6-0.4-0.20.0 -80-60-40-20 0 20 R e l a t i v e A b s o r p t i on Radial Velocity [km s -1 ]Pos. 2-0.8-0.6-0.4-0.20.0 -80-60-40-20 0 20 R e l a t i v e A b s o r p t i on Radial Velocity [km s -1 ]Pos. 1 -0.5-0.4-0.3-0.2-0.1 0 -80-60-40-20 0 20 R e l a t i v e A b s o r p t i on Radial Velocity [km s -1 ]806040200 -120-100-80-60-40-20 0 20 B r i gh t ne ss T e m pe r a t u r e T b Radial Velocity [km s -1 ]Pos. 3806040200 -120-100-80-60-40-20 0 20 B r i gh t ne ss T e m pe r a t u r e T b Radial Velocity [km s -1 ]Pos. 2806040200 -120-100-80-60-40-20 0 20 B r i gh t ne ss T e m pe r a t u r e T b Radial Velocity [km s -1 ]Pos. 1 Figure 1.
Top : H i absorption profiles of 3C 58. The relative absorption of theradio continuum signal at 1420 MHz is displayed as a function of radial velocity v LS R . Left: Absorption profiles of 3 di ff erent positions are displayed as indicatedin the bottom right panel. Right: Profile calculated using all pixels on 3C 58 with T b ≥
100 K. Each pixel was weighted by its intensity.
Bottom : Left: H i emissionprofiles used to determine the absorption profiles in the top left. These were averagedover the positions marked in the right panel. Right: Continuum image of 3C 58 takenfrom the CGPS. The ”on” and ”o ff ” positions used to calculate the H i absorptionprofiles are indicated by solid and dashed circles, respectively. CGPS H i data (Figs. 1 and 2) show an additional absorption feature at about −
41 km s − and a systemic velocity of about −
36 km s − for the bubble. There is no absorption atvelocities beyond −
41 km s − . In Fig.2 the map averaged around −
25 km s − , repre-sentative of the interarm between the Local arm and the Perseus arm shows significantabsorption, however, the map averaged over velocities beyond −
41 km s − is free ofabsorption even though this velocity range shows brighter emission than the interarmaround −
25 km s − . This confirms the Perseus arm location for 3C 58. Presumably3C 58 is located inside the bubble shown in Fig. 3 and the last absorption feature, whichis only partly visible in the absorption of 3C 58 is caused by a part of that bubble mov-ing towards us. That would give this bubble an expansion velocity of about 5 km s − .Using the equations from McClure-Gri ffi ths et al. (2002) we calculate a dynamic age ofabout 3 × yr assuming it is the result of a single supernova explosion and 6 × yrfor a stellar wind bubble. Kothes Figure 2. H 1 channel maps of 3C 58 taken at the peaks of the absorption spectraat + − (bottom centre).The right panels show 10 channels averaged over the interarm velocities betweenthe Local and Perseus arm (top) and 10 channels averaged over the velocities justoutside the last absorption feature. The continuum emission of 3C 58 is indicated bythe white contours. The black dashed lines represent the H i contours at the levelsindicated by the labels of the colour bars.
4. Discussion4.1. The Distance to 3C 58
The method by Foster & MacWilliams (2006) was used to determine a distance-velocityrelation in the direction of 3C 58 (Fig. 4). For a systemic velocity of −
36 km s − thereare two possible distances: in the Perseus arm shock at about 2 kpc and beyond Perseusarm at about 2.5 - 2.8 kpc. Interstellar material is compressed in the spiral shock, formsmolecules, and then stars. It takes a long times – several 10 years – to migrate beyondthe Perseus arm. Therefore, considering the dynamical age of the bubble and the highprobability that the progenitor star of 3C 58 formed in the spiral shock, the likelihoodof 3C 58 being at the farther distance is very low.There is a second independent method to determine a distance to 3C 58, to relatethe PWN to a nearby object with a more reliable distance estimate. The W 3 / / ii region complex and the related SNR HB 3 are just a few degrees away from 3C 58.Both, HB 3 and W 3 / /
5, have very similar systemic velocities (e.g. Routledge et al.1991). They are all located in the Perseus arm and for W 3 the distance was determinedwith trigonometric parallax to related masers to be 1 . ± .
04 kpc and 2 . ± .
07 kpcby Xu et al. (2006) and Hachisuka et al. (2006), respectively. This nicely agrees withthe 2 kpc I estimated for 3C 58 from H i absorption.istance and Ageof3C58 5 Figure 3. H 1 channel maps of the area around 3C 58. Each channel is0.82446 km s − wide and the centre velocity is indicated. The location and emis-sion structure of 3C 58 are marked by the white contours. There are a number of arguments against a relation between the PWN 3C 58 and thehistorical supernova explosion from A.D. 1181 as outlined in Fesen et al. (2008, Table3). I will discuss the two major groups of these arguments below.
Figure 4. Plot of the distance-velocity relation in the direction of 3C 58 deter-mined with the method byFoster & MacWilliams (2006).
Kothes
One important aspect of expansion studies that is not always fully appreciated is thefact that a supernova explosion is a one-time event. For any expansion study, observedfeatures related to the explosion could have been decelerated but not accelerated. Onthe other hand, this may not be true for synchrotron emitting filaments related to aPWN, because those have a continuous source of energy in the central pulsar. Hence,the interpretation of expansion studies of synchrotron filaments is not straightforward.A comparison of the two major expansion studies in optical (Fesen et al. 2008, ⇒ age t ≈ ⇒ t ≈ erg released in a supernova and the approximate energy released bythe pulsar since birth of ∼ erg (Chevalier 2004) negates that possibility.The optical filaments are created by material accelerated through the supernovaexplosion. Because those filaments could have been decelerated but not accelerated asimple averaging of the expansion velocity of individual filaments would not necessar-ily lead to a good age estimate unless the scatter is entirely produced by uncertaintiesin the observations or systematic errors, which is certainly not the case in the study byFesen et al. (2008). The fastest filament should be taken, since it presumably showsthe lowest deceleration. There is quite a wide spread in velocity in the optical study byFesen et al. (2008), which again indicates that a lot of the emitting material must havebeen decelerated. Therefore the assumption of insignificant deceleration for the opticalfilaments is invalid as well. This significantly weakens the case for a large age basedon the optical and radio expansion studies. Chevalier (2004) discussed theoretical calculations of the evolutionary path of a PWNand compared their results for swept up mass and internal energy with those from ob-servations. Both of these results rely heavily on the assumed distance. Bocchino et al.(2001) determined from their X-ray observations a swept-up mass of M sw = . d . . M ⊙ ,assuming a radius R of 2 . ′ , which results in 0.1 M ⊙ for a distance d of 3.2 kpc. The the-oretical value from Chevalier (2004) equates to: M sw = ˙ ER − t ( ˙ E : rotational energyloss rate of the pulsar) resulting in 0.005 M ⊙ . Even with the relatively high uncertaintyof this kind of calculation the large discrepancy suggests a much larger age for 3C 58.To increase the theoretical value to the observed one the age of the PWN has to beabout t ≈ erg from equipartition considerations. Thisvalue depends on the radio luminosity and thereby on d . The total energy releasedfrom the pulsar into the nebula, however, which can be approximated by ˙ Et is about0 . × erg, which is much smaller. This value is distance independent.For a distance of 2 kpc the minimum energy required to produce the synchrotronemission decreases to 0 . × erg, which is now lower than the total energy releasedby the pulsar, which remains unchanged. The values for the swept up masses equateto 0.030 M ⊙ and 0.013 M ⊙ for the X-ray observations and the theoretical calculations,respectively. The two results for the swept-up mass still di ff er but not by much. To getthose values to be equal requires either an age of about 1100 yr or a distance of 1.7 kpc.For a distance of 1.9 kpc the required age would be reduced to 1000 yr.istance and Ageof3C58 7
5. Conclusion
I derived a new more reliable distance of 2 kpc to the PWN 3C 58, by means of H i absorption in combination with a newly determined distance-velocity relation and byrelating this PWN to a nearby H ii region SNR complex of well known distance. Thisnew distance changes many characteristics of this PWN quite dramatically which onceagain makes the connection between 3C 58 and the historical supernova explosion from1181 A.D. much more likely. Acknowledgments.
I would like to thank Tyler Foster for providing me with adistance-velocity diagram in the direction of 3C 58 using his modeling of Galactic hy-drogen distribution. The Dominion Radio Astrophysical Observatory is a National Fa-cility operated by the National Research Council. The Canadian Galactic Plane Surveyis a Canadian project with international partners, and is supported by the Natural Sci-ences and Engineering Research Council (NSERC).
References
Bietenholz, M. F. 2006, ApJ, 645, 1180Bocchino, F., Warwick, R. S., Marty, P., Lumb, D., Becker, W., & Pigot, C. 2001, A&A, 369,1078Chevalier, R. A. 2004, Advances in Space Research, 33, 456Fesen, R., Rudie, G., Hurford, A., & Soto, A. 2008, ApJS, 174, 379Foster, T., & MacWilliams, J. 2006, ApJ, 644, 214Foster, T., Routledge, D., & Kothes, R. 2004, A&A, 417, 79Hachisuka, K., et al. 2006, ApJ, 645, 337Higgs, L. A., & Tapping, K. F. 2000, AJ, 120, 2471Kothes, R., Reich, W., Foster, T., & Byun, D.-Y. 2003, ApJ, 588, 852Kothes, R., Uyaniker, B., & Yar, A. 2002, ApJ, 576, 169Landecker, T. L., et al. 2000, A&AS, 145, 509McClure-Gri ffiffi