Acta Polytechnica doi:10.14311/AP.2013.53.0617 Acta Polytechnica 53(Supplement):617–620, 2013 © Czech Technical University in Prague, 2013 available online at http://ojs.cvut.cz/ojs/index.php/ap RX-J0852−4622: THE NEAREST HISTORICAL SUPERNOVA REMNANT – AGAIN Bernd Aschenbach∗ PR Vaterstetten, Mozartstraße 8, 85591 Vaterstetten, Germany ∗ corresponding author: bernd.aschenbach@t-online.de Abstract. RX-J0852−4622, a supernova remnant, is demonstrated to be closer than 500 pc, based on the measurements of the angular radius, the angular expansion rate and the TeV γ-ray flux. This is a new method of limiting the distance to any supernova remnant with hadronic induced TeV γ-ray flux. The progenitor star of RX-J0852−4622 probably exploded in its blue supergiant wind, like SN 1987A, preceeded by a red supergiant phase. A cool dense shell, expected around the outskirts of the red wind, may have been identified. The distance (200 pc) and age (680 yr) of the supernova remnant, originally proposed, are supported. Keywords: supernova remnants, stellar evolution, multi-wavelength, X-rays, TeV γ-rays, cosmic rays, individual: RX-J0852−4622, Vela Jr., SN 1987A, SN 1006. 1. Introduction RX-J0852−4622, nicknamed Vela Junior (Vela Jr.), is a supernova remnant (SNR) discovered in the ROSAT all-sky survey [3]. It is located along the direction towards the south-eastern corner of the Vela SNR, and it is completely covered, even beyond its boundaries, by the emission from the Vela SNR, such that Vela Jr. does not shine up against the very bright soft X-ray emission from the Vela SNR. Only at X-ray energies > 1.3 keV does Vela Jr. become X-ray visible. It is a patchy narrow shell-type source with a diameter of 2° of almost perfect circular shape. Since the dis- covery multi-wavelength obsersations have confirmed the SNR status, including non-thermal radio emis- sion; hard X-ray emission with soft X-ray emission apparently missing; MeV γ-ray line emission from ra- dioactive 44Ti, 26Al, 44Ca, of low significance, though; GeV γ-ray continuum emission; and TeV γ-rays. A summary of the observations and interpretations can be found e.g. in [2, 10, 12]. Aharonian et al. [2] pro- vide evidence that the TeV γ-rays are likely not to be of leptonic origin (inverse Compton scattering of electrons and photons) but of hadronic origin (nuclear collisions between SNR produced cosmic ray protons and nuclei of the ambient matter). This is supported by the MeV γ-ray measurements of Tanaka et al. [12]. An open question still is the distance and the age of Vela Jr.; the numbers range between 200 pc and 1 kpc, and 680 yr and 5000 yr. I show that Vela Jr. is very close, and likely to be as young as originally suggested [4]. 2. Implications of observations Three unambiguous key measurements of Vela Jr. have been made: the angular radius Ra = 1° [3], the angular velocity of the shock wave vs,a = 0.84′′/yr ± 0.23′′/yr [8] and the total energy of cosmic ray protons Wp to explain the TeV γ-ray flux, Wp = 1049 ergd22/n0 [2]. Units are cm−3 for the particle density n0 of the ambient matter; 200 pc for the distance to the source, d2. From X-ray measurements additional information is ob- tained for an upper limit of the ambient density of n0,x < 0.029 cm−3(d1 f)−1/2 [11], where d1 is distance in kpc, and f is a filling factor < 1. In contrast to other information, this is not definite, because it depends on the interpretation of the X-ray spectral measurements. 2.1. Distance The kinetic energy E0 available from an SN explo- sion is shared mainly between the acceleration of particles to cosmic rays (Wp) and the expansion of the explosion wave into the ambient medium driven by some energy Ek, such that E0 ≥ Wp + Ek with Ek = 4.1 × 1049 ergn0d52. Ek represents the sum of the kinetic energy and the associated thermal energy, the latter of which is due to the shock-wave heating of the ambient matter. As Wp scales with (d22/n0) and Ek scales with (d52n0) a single absolute maximum of d2 exists for any given E0, which also fixes n0. Any filling factor, additionally introduced, does not change the maximum distance. The normalization factors for Wp and Ek are known from the measure- ments. For Vela Jr. they are Wp,1 = 1049 erg and Ek,1 = 4.1 × 1049 erg, respectively, (see above) for d2 = 1 and n0 = 1. For E0 = 1051 erg the distance d is < 500 pc. For E0 = 3 × 1050 erg, d < 355 pc, and for E0 = 1050 erg, d < 260 pc. The maximum of d is realized if E0 is shared evenly between Wp and Ek. For instance, for Wp = 0.1E0 and E0 = 1051 erg, d < 435 pc, and 225 pc for E0 = 1050 erg, respectively. The re- verse case, i.e. Wp = 0.9E0 does not change the upper limit of d, but it changes the associated n0, slightly. Finally, the uncertainties of the measured 617 http://dx.doi.org/10.14311/AP.2013.53.0617 http://ojs.cvut.cz/ojs/index.php/ap Bernd Aschenbach Acta Polytechnica values of vs,a, ±0.23′′/yr, change the distances quoted by at most 10 %. This means that there exists a firm upper limit of d < 550 pc. Only if the event were a hypernova, the distance could be larger. The method outlined above can, of course, be ap- plied to any other SNR, which emits TeV γ-rays of hadronic origin. The upper limit of the distance, i.e., d2,max is generally given by the relation d2,max = ( 0.25 ( E20/(Wp,1Ek,1) ) v−2s,a R −3 a )1/7 ; E0 = 1051 erg is the maximum energy generally at- tributed to a SN explosion; Wp,1 is the cosmic ray proton energy, derived from the measured TeV γ-ray spectrum normalized to d2 = 1 and n0 = 1 cm−3 (for the procedure see [2]); Ek,1 = 5.8 × 1049 erg is the energy computed for d2 = 1 and n0 = 1 cm−3, and an angular expansion (shock velocity) of 1′′/yr and an angular radius of 1°. This method appears to be a powerful tool for setting a distance upper limit of TeV γ-ray emitting SNRs, because this method de- pends just on angular measurements. For instance, the TeV γ-ray flux, if of purely hadronic origin, and the measurements of vs,a and Ra quoted [1], predict an upper limit of d = 1.85 kpc to SN 1006, using this new method. 2.2. Density Thermal X-ray emission is expected from the shock wave heating of the ambient medium. The upper limit of n0,x of Slane et al. [11] sets further constraints to the distance when used in the expressions for Wp and Ek. For E0 = 1051 erg and f = 1 the upper limit of d is reduced to < 420 pc and n0,max = 0.05 cm−3. For E0 = 1050 erg, d < 250 pc and n0,max = 0.06 cm−3. Values of f < 1 lower the acceptable maximum dis- tance further, but do not change the acceptable n0 significantly. Slane et al. [11] note that the upper limit of n0,x is valid only for temperatures of kT > 1 keV, which corresponds to a shock velocity of vs = 860 km/s; with vs,a = 0.84′′/yr, the upper limit of n0,x is therefore not applicable for d < 235 pc. This is about the distance to the Vela SNR, and the temperatures of the Vela SNR across the area of Vela Jr. range between 0.5 and 0.9 keV [9]. This means that the thermal X-ray emission from Vela Jr. and the Vela SNR can be easily mixed up, and they cannot even be spectroscopically discriminated. The majority of the soft X-ray emis- sion received from the Vela Jr. covered area could be attributed to Vela Jr. rather than to the Vela SNR. Taking the ROSAT soft X-ray measurements n0 could be a factor of 10 or more higher than the Slane et al. limit, i.e., > 0.6 cm−3 [9]. 2.3. Age The evolutionary state of an SNR is often described by the expansion parameter β, defined by R ∼ tβ, with R the radius and t the age. This relation leads to β = vs/v, with vs the current shock velocity and the mean expansion velocity v = R/t. Given the angular value of vs, i.e. vs,a, and the angular radius Ra, the age would be calculated to t = βRa/vs,a. Formally, 0 ≤ β ≤ 1. The value of β close to 1 means that the SNR is almost freely expanding; β = 0.4 describes the adiabatic state with the SNR expansion slowed down by the ambient matter. For this case to apply the mass overrun by the shock wave (swept- up mass) should be much greater than the mass of the SN ejecta. The radiative phase starts around vs ∼ 250 km/s, equivalent to kT ∼ 0.1 keV. Given vs,a, the SNR would be located at a distance of 65 pc. This would have made Vela Jr. a very bright EUV source, which, however, is not listed in any of the EUV catalogues. For β = 0.4, Vela Jr. being in the adiabatic phase, t ∼ 1700 yr (1300 ÷ 2400 yr). The maximum explosion energy of E0 = 1051 erg and the maximum density of n0,x = 0.05 cm−3 allow a maximum distance of 420 pc. vs,a then corresponds to 1700 km/s equivalent to kT ∼ 4 keV. The expected thermal hard X-ray flux is fairly low at the level of n0,x and is probably not detectable given the sensitivity of present instruments, but see [7]. But it is noted that the energy in cosmic ray protons would be 8.8 × 1050 erg. A cosmic ray acceleration efficency of 88 % would be a surprise, and an overrun mass of just 2.4 solar masses casts doubts on the assumption of the adiabatic state. Finally, for Vela Jr. being close to free expansion, the ambient density, which needs to be low for free expansion to apply, is nevertheless high, such that it is inconsistent with d > 500 pc and Wp ≤ 1051 erg. Summarizing, the analysis shows that the concept of an explosion of the Vela Jr. progenitor star into a medium of constant isotropic matter density dis- tribution is not applicable, and estimates of age and density on this basis are irrelevant. 2.4. Circumstellar environment of the SN progenitor star Recalling SN 1987A, it is proposed that the Vela Jr. progenitor was a star with a very low density wind, like a blue supergiant, shortly before its explosion. After passing the low density region the SN shock wave may have hit the base of the preceeding red wind which is of much higher density, preserving pressure equilibrium, i.e., (nv2) = const. The relevant density before the velocity slow-down is the mean density delivered by the blue wind up to the density jump and the density of the red wind at the radius of the density jump afterwards. The relevant velocities are the initial quasi-free expansion velocity vfr of the SN taken over the full distance up to the location of the density jump, on the one hand, and the current shock velocity vs on the other hand. The jump condition 618 vol. 53 supplement/2013 RX-J0852−4622: the Nearest Historical Supernova Remnant – Again translates into Q = dMb/dt vb dMr/dt vr = 16 27 ( vs,a vfr,a )2 with vs,a the current angular shock velocity and vfr,a the initial angular velocity during the quasi-free ex- pansion phase. Chevalier & Fransson [6] published typical mean values for the wind parameters, i.e., mass loss rate over wind velocity, of red supergiants and of blue supergiants, respectively, which are around dMb/dt vb = 10−6 solar masses/yr 550 km for a blue wind and dMr/dt vr = 10−5 solar masses/yr 20 km for a red wind. The blue wind parameters are con- sistent with the observations of the SN 1987A ob- servations, the progenitor star of which had been identified as a blue supergiant. With this set of wind parameters and d2 = 1, one gets vs = 800 km/s and vfr = 10 200 km/s for the quasi-free expansion phase velocity. For t = 680 yr, the encounter of the quasi- free expanding shock wave with the red wind base would have occurred at 0.91R after 333 yr. A dense shell, made up mainly of the red wind matter, with a width of 0.09R, is traversed by the shock wave at a speed of 800 km/s. The outer zone of a stellar wind is expected to be surrounded by a thin, dense and cool shell, made up of the ambient swept-up interstellar matter. In the north-westerly direction of Vela Jr., just outside the outer boundary of the Vela SNR, the SNR Puppis-A is located, which is a very bright X-ray source with a diameter of 45′. A spatially resolved spectral study of the ROSAT measurements across the surface of Puppis-A shows an excess of low energy absorption above the mean interstellar absorption. The excess absorption is spatially confined to a region looking like a slighty curved lane, or filament, which stretches right through the middle of Puppis-A from its north- eastern boundary to its southwestern boundary. This indicates the presence of some cool matter leading to extra, spatially confined absorption on top of the inter- stellar absorption [5, 9]. This absorption lane could be considered as a fraction of a complete shell produced by the red wind. The majority of the shell cannot be observed because sufficiently bright background X-ray sources are missing. From this interpretation it follows that dMr/dt vr = Ml (Rabs,ad) with Ml the total mass loss of the progenitor star by its red wind, or numerically, dMr/dt vr = 9 × 10−6 solar masses/yr 20 km Ml,10 d2 , with Ml,10 in units of 10 solar masses. The den- sity nr of the red wind at its base hit by the SN shock wave is then nr = 0.1 cm−3Ml,10/d32. Using the measured flux of the TeV γ-rays [2] it follows that Wp = 1050 ergd52/Ml,10. This demonstrates that even under extreme conditions, i.e., large values of Wp and Ml,10, d can be only somewhat larger than 200 pc. In fact, the result suggests that d < 200 pc. If so, the requirements for the associated total SN energy and cosmic ray energy would be somewhat relaxed. The age t of the SNR is limited by the acceptable value of Q = 16/27(vs,a/Ra)2(t/A)2 with A = vfr/v > 1, v = R/t. dMr/dt vr is basically fixed by the observa- tion of the absorbing outer shell built up by the red wind. For vfr = 10 000 km/s the required density jump is consistent with the blue wind parameters suggested for SN 1987A, shown above. Observations have shown that vfr ∼ 40 000 km/s over the first 6 yr for SN 1987A with an abrupt deceleration following. Applying this extreme case of such a quasi-free exansion velocity to Vela Jr. for a duration of up to about 100 yr after the explosion one gets dMb/dt vb = 10−7 solar masses/yr 1000 km . But an upper limit of vfr can be derived from the width of the SNR shell. Figure 3 of Iyudin et al. [7] shows that the hard X-ray emission is con- fined to a shell of < 6′ to 7′ thickness measured by XMM-Newton. This would result in vfr less than 15 000 km/s. It appears that the Chandra images show a shell as thin as 5′ [10], implying vfr ∼ 10 000 km/s. In summary, the blue wind parameters of the Vela Jr. progenitor star were probably in a range of 10−6 solar masses/yr 550 km and 10−7 solar masses/yr 1000 km during its final evolutionary phase. The wind param- eters of the red wind and blue wind respectively, are surprising close to the typical mean values. SN 1987A appears to be the only example so far of a blue supergiant SN, and Vela Jr. may be a further example. Such SNe might be very hard to be detected in the radio and low energy X-ray regimes, because the luminosity depends heavily on the ambient matter density. For instance, SN 1987A escaped detection for 6 years and might have escaped so for a much longer period of time. But after these 6 years the shock wave started to encounter the ‘inner ring’ with its much higher density, which slowed down the propagation of the shock wave dramatically. When, and whether, it will have reached the base of the red wind, like in the case of Vela Jr., remains to be seen. In this sense the observation of an absorbing shell produced by the red wind in case of Vela Jr. appears to be unique. The absorbing shell attributed to the red wind is located at an angular distance of Rabs,a = 6.5° measured from the center of Vela Jr., which corre- sponds to Rabs = 22 pc for d2 = 1. It shows a core 619 Bernd Aschenbach Acta Polytechnica width of ∼ 5′ and a full width of ∼ 30′, the latter of which is limited by the sensitivity of the measure- ments. The absorbing column density in the core region is NH = 1.71021 cm−2 (0.8 ÷ 1.91021 cm−2) [5]. This allows computing the ambient density nam of the matter through which the red wind propagated, i.e., nam = 0.55 cm−3NH,21d−12 . The analysis of the Vela SNR at a distance of 250 pc reveals a mean ambient density of ∼ 0.35 cm−3, derived also independently from the interstellar absorption column density, with spatial variations of a factor of > 2. It appears that the identification of the absorption lane as fraction of an absorption shell produced by the red wind of the Vela Jr. progenenitor star and the distance of ∼ 200 pc are fairly plausible. 2.5. Interstellar absorption The fits of models, available in standard X-ray spec- tral software packages, to the measured broad band X-ray spectra (0.5 ÷ 10 keV) of Vela Jr. appear to be consistent with a pure non-thermal spectrum with an absorption column density of NH > 2×1021 cm−2 [10], which is at least 20 times higher than the absorption in front of the Vela SNR at a distance of 250 pc. This value of NH, interpreted as interstellar absorption, and therefore some measure of the distance, has raised the claim that Vela Jr. lies behind the Vela SNR, and that it is even much more distant. But even at the largest distances suggested the mean interstellar den- sity needs to be close to or to exceed 1 cm−3, for which there is no observational evidence otherwise. Narrow band (0.2 ÷ 2 keV) analysis of the ROSAT spectra is consistent with a much flatter non-thermal spectrum, and NH is consistent if not lower than that measured for the Vela SNR [4]. In particular at the low energy end (< 2 keV), the statistical quality of the fits to the broad band measurements is not convincing, and one may wonder whether the source spectrum is not much more curved than the modified power-law models of the standard software packages offer. The flattening of the observed spectrum for E < 2 keV, which is the energy range extremely sensitive to interstellar absorption, is more likely not be caused by interstellar absorption but to be of intrinsic origin, i.e., the SNR produced cosmic ray electron spectrum for which a turn-over might occur at energies ∼ 0.1 keV (see, e.g., the models shown in Fig. 17 of Aharonian et al. [2]). One could turn the argument around and can perhaps learn something about the electron acceleration to cosmic ray energies and the shape of the spectrum close to the turn-over at the highest energies, through the measurements. 3. Conclusions Angular radius, angular shock velocity and TeV γ-ray flux limit the distance to Vela Jr. to < 500 pc for an explosion with an available energy of < 1051 erg. The upper limits on the density associated with a high temperature plasma (> 1 keV) reduce the distance further. The SN explosion into a stellar wind environ- ment can explain the observational results. The wind parameters, i.e., mass loss rate over wind velocity, are in agreement with those known for blue supergiant winds and red supergiant winds. It appears that the progenitor star of Vela Jr. exploded in its blue wind developing a few thousand years before the end of the progenitor’s life. The blue wind phase was preceeded by an expected red wind phase. The base of the red wind was probably hit by the SN explosion shock wave after 200 to 300 years. Because of the high density at the red wind base the shock wave velocity signifi- cantly dropped and the bulk of the expected thermal X-ray emission is expected at temperatures < 1 keV, and may be hidden among the Vela SNR emission as the temperatures are very similar. The wind model, in particular the explosion of a blue supergiant is reminiscent of the findings for SN 1987A. The appli- cation of a wind model also suggests that protons are preferentially accelerated to cosmic ray energies in an environment of continuous high shock velocity and low density, which keeps helping the losses by collisions with ambient ions low, and, in addition, little energy is being spent for expanding and heating. This could make blue supergiant explosions favourable TeV γ-ray emitters, explaining their rareness among the SNR populations. Acknowledgements References [1] Acero, F. et al.: 2010, A&A 516, 26 [2] Aharonian, F.A. et al.: 2007, ApJ 661, 236 [3] Aschenbach, B.: 1998, Nature 396, 141 [4] Aschenbach, B., Iyudin, A.F., Schönfelder, V.: 1999, A&A 350, 997 [5] Aschenbach, B., 1995, ASP Conf. Ser. 80, 432 [6] Chevalier, R.A., Fransson, C.: 1987, Nature 328, 44 [7] Iyudin et al. et al: 2005, A&A 429, 225 [8] Katsuda S. & Tsunemi, H.: 2012, MmSAI 83, 277 [9] Lu, F.J. & Aschenbach, B.: 2000, A&A 362, 1083 [10] Pannuti, T.G. et al.: 2010, ApJ 721, 1492 [11] Slane, P. et al.: 2001, ApJ 548, 814 [12] Tanaka, T. et al., 2011, ApJ 740 620 Acta Polytechnica 53(Supplement):617–620, 2013 1 Introduction 2 Implications of observations 2.1 Distance 2.2 Density 2.3 Age 2.4 Circumstellar environment of the SN progenitor star 2.5 Interstellar absorption 3 Conclusions Acknowledgements References