A&A 455, 1053-1058 (2006)
W. W. Tian1,2 - D. A. Leahy2
1 - National Astronomical Observatories, CAS, Beijing 100012, PR China
2 - Department of Physics & Astronomy, University of Calgary, Calgary, Alberta T2N 1N4, Canada
Received 3 March 2006 / Accepted 18 May 2006
New images of the radio Supernova Remnant (SNR) G65.1+0.6 are presented, based on the 408 MHz and 1420 MHz continuum emission and the HI-line emission data of the Canadian Galactic Plane Survey (CGPS). A large shell-like structure seen in the 2695 MHz Effelsberg map appears to have nonthermal spectral index. HI observations show structures associated with the SNR G65.1+0.6 in the radial velocity range of -20 to -26 km s-1 and suggest a distance of 9.2 kpc for the SNR. The estimated Sedov age for G65.1+0.6 is yr. The pulsar (PSR) J1957+2831 is possibly associated with G65.1+0.6, with consistent distance and kinematic age estimate, but different characteristic age than the SNR. The EGRET source 3EG J1958+2909 and -ray source 2CG 065+00 are also near the eastern edge of the SNR but do not agree in position with the pulsar and are likely not associated with the SNR. The SNR's flux densities at 408 MHz ( Jy), 1420 MHz ( Jy) and 2695 MHz ( Jy) have been corrected for flux densities from compact sources within the SNR. The integrated flux density based spectral index ( ) between 1420 MHz and 408 MHz is and agrees with the T-T plot spectral index of . The nearby SNR DA495 has a T-T plot spectral index of .
Key words: ISM: general - radio line: ISM - radio continuum: galaxies
The continuum and HI emission data sets come from the CGPS, which is described in detail by Taylor et al. (2003). The data sets are mainly based on observations from the Synthesis Telescope (ST) of the Dominion Radio Astrophysical Observatory (DRAO). The spatial resolution of the continuum images of G65.1+0.6 is at 1420 MHz and 2.8 at 408 MHz. The synthesized beam for the HI line images is and the radial velocity resolution is 1.32 km s-1. DRAO ST observations are not sensitive to structures larger than an angular size scale of about 3.3 at 408 MHz and 56' at 1420 MHz. Thus the CGPS includes data from the 408 MHz all-sky survey of Haslam et al. (1982), sensitive to structure greater than 51', and the Effelsberg 1.4 GHz Galactic plane survey of Reich et al. (1990, 1997), sensitive to structure with resolution 9.4' for large scale emission (the single-dish data are freely available by http://www.mpifr-bonn.mpg.de/survey.html). The low-order spacing HI data is from the single-antenna survey of the CGPS area (Higgs Tapping 2000) with resolution of 36'. See Taylor et al. (2003) for detail of the method of combining the synthesis telescopes and single dish observations.
We analyze the continuum and HI images of G65.1+0.6 and determine its flux densities and distance. For G65.1+0.6, integrated flux density's errors are found by comparing results for several different choices of background region. For compact sources, the flux density's errors are taken as the formal Gaussian fit errors. The influence of compact sources within the SNR is much reduced by employing similar methods to Tian & Leahy (2005).
|Figure 1: The first row of images shows the CGPS maps at 408 MHz ( left) and 1420 MHz ( right). The left of the second row shows the 1420 MHz map convolved to the same resolution as the 408 MHz map. The right of the second row is the 2695 MHz Effelsberg map. The 6 boxes labeled with letters A-F and used for T-T plots are shown in the left. The right of the first row shows positions and error circles of EGRET source 3EG J1958+2909 (95 confidence level, marked by number 1) and -ray 2CG 065+00 (90 confidence level, marked by number 2). Pulsar J1957+2831 is marked by letter X. The direction of North (N) and South (S) is marked on the lower right image.|
|Open with DEXTER|
The CGPS continuum images at 408 MHz and 1420 MHz are shown in the upper left and right panels of Fig. 1. The 6 boxes labeled with letters A-F and used for T-T plots are shown in the left. The upper right of Fig. 1 shows positions and error circles of EGRET source 3EG J1958+2909 (marked by number 1) and -ray 2CG 065+00 (marked by number 2). Pulsar J1957+2831 is marked by letter X. The lower left shows the 1420 MHz map convolved to the same resolution as the 408 MHz map. The 2695 MHz Effelsberg map is reproduced in the lower right for reference (Fürst et al. 1990). The Effelsberg map has a resolution of 4.3 and a sensitivity of 50 mK.
The 408 MHz image for G65.1+0.6 reveals more structures in comparison with Landecker's (1990) 408 MHz image, especially, side-lobe effects which are obvious in Landecker's maps are much reduced on our image. The 1420 MHz map shows much better detail of the fine structure in G65.1+0.6 than any previous image. More compact sources within the SNR are resolved in the 1420 MHz map. More filament structures appear in the northern, northwestern and southwestern parts of the SNR. Figure 2 shows that a ring-shape structure inside box C is clearly detected near the western diffuse emission regions of the SNR. The outlines of G65.1+0.6 at 408 MHz and 1420 MHz are very similar.
|Figure 2: Image of box C at 1420 MHz.|
|Open with DEXTER|
The 3 2695 MHz Effelsberg map appears to show a circular-shell (Radius about 1) SNR-like emission region including the two known SNRs G65.1+0.6 and G65.7+1.2 (DA495). But the CGPS maps at both 408 MHz and 1420 MHz don't show this circular shell. We mark the large emission region (A) and its two inner areas (D and E) related with G65.1+0.6 in order to study the emission further.
Table 1: List of the 16 brightest compact sources (the first 11 sources are within G65.1+0.6) and their integrated flux densities within box A.
Bright compact sources affect the measured integrated flux densities for G65.1+0.6 and its measured spectral index. Thus we correct for the effects of compact sources. Table 1 lists properties of the 16 brightest compact sources which are detected in box A, and the first 11 of which are within G65.1+0.6 at both 408 MHz and 1420 MHz.
First we discuss spectral indices between 408 MHz and 1420 MHz based on the T-T plot method (Turtle et al. 1962). The principle of the T-T plot method is that spectral indices ( ) are calculated from a fit of a linear relation to the T1-T2 values of all pixels within a given map region. T1 is the brightness temperature of a map pixel at one frequency and T2 is for the second frequency. The higher resolution image has been smoothed to the lower resolution for the T-T plot comparison. The brightness temperature spectral index is derived from the slope of the line. The error in spectral index is derived from the uncertainly in slope of the line. The flux density spectral index ( ) is related to by . Spectral index refers to flux density spectral index in this paper unless specifically noted otherwise.
For the T-T plot spectral index analysis, we select a single large region (A) including the two SNRs G65.1+0.6 and DA495, region B including G65.1+0.6, region F including DA495 alone and other 3 areas (C, D, E) around G65.1+0.6, as shown in Fig. 1. Region A yields the T-T plots shown in the upper half of Fig. 3. Region B yields the T-T plots shown in the lower half of Fig. 3.
In the T-T plot analysis, two cases are considered: using all pixels including compact sources; and excluding compact sources listed in Table 1 from the images. The compact sources are bright compared to the SNR emission. Since the compact sources have a steeper spectrum than the SNR, they are seen in the T-T plot (Fig. 3 left panel) as the steeper lines of points extending to higher . We completely remove regions of pixels including the compact sources from the analysis. Each region is taken to be a few beamwidths across, so that any contribution from the compact source is below 1 of the diffuse SNR emission.
|Figure 3: 408-1420 MHz T-T plots for region A ( the upper half) and region B ( the lower half). The left plots for maps including compact sources; The right plots for compact sources removed from analysis.|
|Open with DEXTER|
Table 2 lists the results for two cases of analysis: including compact sources and removing compact sources. Visual inspection of the T-T plots confirms that the second method produces the most reliable results. The compact sources' influence on the spectral index calculation is obvious in the T-T plots of Fig. 3, and also seen in Table 2 for areas A, B, D and E. From now on we discuss spectral indices derived with compact sources removed, unless specified otherwise.
Region A includes several emission regions. The spectral index for a box F including DA495 alone is given in Table 2 ( ). From the right panels of Fig. 3, it is obvious that the T-T plot of G65.1+0.6 is only left lower part of region A's T-T plot. The upper right part of region A's plot is from the SNR DA495.
Table 2: 408-1420 MHz T-T plot spectral indices with and without Compact Sources (CS).
We have derived integrated flux densities of G65.1+0.6 from the 408 MHz and 1420 MHz maps. Values given have diffuse background subtracted. The resulting 408 MHz to 1420 MHz spectral index, using flux densities without compact sources, is . Table 3 lists the flux densities and spectral indices of G65.1+0.6 and the compact sources within G65.1+0.6. Compact sources contribute about 33 at both 408 MHz and 25 at 1420 MHz to the SNR's flux densities, and have a significant effect on the spectral index. It is noted that the SNR spectral index derived from integrated flux densities is consistent with the SNR spectral index ( ) derived by the T-T plot method.
We have calculated total compact source flux densities for 2695 MHz frequencies, using the 408-1420 MHz spectral index upper and lower limits and flux densities from Table 1. The compact sources contribute 21 at 2695 MHz to the SNR's flux densities. We have recalculated the flux density values of G65.1+0.6 at 2695 MHz by subtracting the compact source flux density. We note that Fürst et al. (1984) gave a flux density based on the 2965 MHz Effelsberg image with resolution arcmin, but we obtain a new value from the Effelsberg 2695 MHz image with a little higher resolution arcmin: Jy and use this instead. We fit the resulting flux density values with a power-law to obtain spectral index. Figure 4 shows the corrected flux densities and the best-fit power-law. The best fit spectral index is 0.49 with 1 uncertainty range of 0.40 to 0.57. The fit to G65.1+0.6 spectrum including compact sources flux density gives . The steeper is expected since compact source has the steeper spectrum than SNR (see Table 3).
Table 3: Integrated flux densities and 408-1420 MHz spectral indices of G65.1+0.6 and compact sources within the SNR.
|Figure 4: Radio spectrum of G65.1+0.6 with and without Compact Sources (CS). The power law model has a best-fit spectral index of 0.49 with 1 uncertainty range of 0.40 to 0.57.|
|Open with DEXTER|
We have searched the CGPS radial velocity range for features in the HI which might relate to the morphology of G65.1+0.6. There is emission which is coincident with the boundary of G65.1+0.6 in the velocity range -20 to -26 km s-1, and only in this velocity range. The spatial relation with the edge of G65.1+0.6 indicates that the HI is associated with the SNR. These HI observations show a good association of HI features with G65.1+0.6, similar to other accepted HI associations with SNRs and with similar HI velocity ranges (10 km s-1), e.g. HI associated with G126.2+1.6 (Tian Leahy 2006), with G127.1+0.6 (Leahy Tian 2006) and with DA530 (Landecker et al. 1999). Figure 5 shows maps of HI emission in twelve channels. Each map has superimposed on it contours of continuum emission at 1420 MHz chosen to show G65.1+0.6.
|Figure 5: HI emission in the field centered on G65.1+0.6 from -20 to -26 km s-1 and -41 to -46 km s-1. The radial velocity of each map is indicated at its top left corner. The outline of G65.1+0.6 in 1420 MHz continuum emission is indicated by the contour at 7.2 K.|
|Open with DEXTER|
The SNR is of low surface brightness and large radius, so probably exploded in a low density part of the ISM. A Sedov model (Cox 1972) gives yr if n0=0.01 cm-3, or yr if n0=0.001 cm-3. The radius at which cooling affects the dynamics of the SNR is 68 pc if n0=0.01 cm-3, or 127 pc if n0=0.001 cm-3. So it is possible that the SNR is still in the Sedov phase or has already entered the cooling phase depending on the density n0. The shock temperature is K if n0=0.01 and K if n0=0.001. Since the density is low it would be faint in X-rays, and since it is quite distant it would be absorbed in X-rays by the intervening ISM. The non-detection in the Rosat All-sky Survey of G65.1+0.6 is consistent with a large distance to G65.1+0.6.
Lorimer et al. (1998) discovered PSR J1957+2831 which is located at the eastern edge of G65.1+0.6, and commented that the pulsar is not likely to be associated with G65.1+0.6 due to the large dispersion measure (DM) distance ( kpc) and old characteristic age (1.6 Myr) of the pulsar, and their statistical grounds of pulsar-SNR pairs in the case of absence of G65.1+0.6's distance and age parameters. But Lorimer et al. (1998) also admitted a possibility that the PSR J1957+2831 may indeed be younger than its characteristic age because the pulsar has a spectral index typical for other young pulsars. If we assume that the pulsar birth-place is at geometrical centre of G65.1+0.6, and moves away toward east of the SNR at typical velocity of 500 km s-1 (Lyne et al. 1994), the kinematic age of the pulsar approximately is yr and consistent with the estimated SNR's above age. Although previous studies (Tian Leahy 2004; Mereghetti et al. 2002; Kaspi et al. 2001) have suggested that pulsar characteristic ages can be poor age estimators for young pulsars, the pulsar J1957+2831 is not young. The pulsar's DM distance ( kpc) and kinematic age agree with G65.1+0.6, and its characteristic disagrees with the age of the SNR, so it is possible that the pulsar is associated with the SNR.
The 2695 MHz Effelsberg map appears to show a circular-shell SNR-like emission region most of which is included in region A shown in Fig. 1. The 408-1420 MHz T-T plot spectral index of areas C, D, E in Table 3 shows they are probably all nonthermal emission regions (each is about 2 different from thermal spectral index). This supports the possibility that there is a large diffuse SNR including the two known SNRs. But the HI images don't show any association of HI features with the large circular shell. Thus current evidence is inconclusive.
The ring-like structure in box C detected in the 1420 MHz image (Fig. 2) may be part of the western edge of G65.1+0.6. However, the ring has a good association with HI features in the velocity range -41 to -46 km s-1 as shown in Fig. 5. The HI velocity corresponds to a distance of 11kpc (-43 km s-1). Now we name the emission structure as a possible new SNR G64.5+0.9 which certainly needs to be confirmed by future high resolution and sensitivity observations.
We acknowledge support from the Natural Sciences and Engineering Research Council of Canada. We thank Dr. R. Kothes at the DRAO provided technical help on the CGPS data processing. The DRAO is operated as a national facility by the National Research Council of Canada. The Canadian Galactic Plane Survey is a Canadian project with international partners.