3.1 Radio morphology

MC SNR J0519–6902 has a ring-like morphology with three bright regions towards the north, east, and south (Bozzetto et al., Reference Bozzetto, Filipović, Urosevic and Crawford2012). In Figure 1, the new ATCA images at 5 500 and 9 000 MHz are compared with the Chandra and HST images. Root Mean Squared (rms) noise for these high resolution ATCA images at 5 500 and 9 000 MHz are $\sim$ 20 and $\sim$ 17 $\mu$ Jy beam $^{-1}$ , respectively; one order of magnitude better than that obtained by Bozzetto et al. (Reference Bozzetto, Filipović, Urosevic and Crawford2012).

The figure clearly shows a radio emission coincident with X-ray emission suggesting relativistic electrons are associated with the shock-heated gas. Interestingly, the optical emission is outside both the radio continuum and X-ray emission (Figure 1), indicating the optical emission is located at the forward shock. The radio morphology of this SNR is asymmetric in terms of variation in brightness. As noted by Bozzetto et al. (Reference Bozzetto, Filipović, Urosevic and Crawford2012), it features three bright regions located in the north, east, and south (Figure 1), resembling the structure of LMC SNR N 103B (Alsaberi et al., Reference Alsaberi, Barnes and Filipović2019) and Kepler in the Milky Way (MW) (DeLaney et al., Reference DeLaney, Koralesky, Rudnick and Dickel2002). Moreover, our new images show a faint structure on the north-east side of MC SNR J0519–6902 which was not present in previous images (see Figure 1). Although not statistically significant ( ${\lt3\sigma}$ ), this feature suggests that the SNR extends further in the north-east direction compared to the other sides (see text for details).

We used the Minkowski tensor analysis tool BANANAFootnote ^f (Collischon et al., Reference Collischon, Sasaki, Mecke, Points and Klatt2021) to determine the centre of expansion. This tool searches for filaments and calculates normal lines and line density maps. For a perfect shell, all lines should meet at the centre position inside the SNR, where the expansion must have started (see Collischon et al., Reference Collischon, Sasaki, Mecke, Points and Klatt2021, for more details). Since real sources deviate from this ideal picture, we circumvented this problem by smoothing the line density map using a circle with a diameter of 40 pixels. We then took the centre position of the pixel where the line density was the highest. Using the same method, we performed the tensor analysis separately for the 5 500 MHz, 9 000 MHz, and Chandra broad-band images. From this, we calculated the mean centre position and the $1\sigma$ uncertainty of the mean.

The resulting calculated centre is RA (J2000) = 05 $^{h}$ 19 $^{m}$ 34.85 $^{s}$ , Dec (J2000) = $-$ 69 $^\circ$ 02’08.22” with an uncertainty of $\sim$ 0.1 arcsec. This is $\sim$ 0.32^′′ ( $\sim$ 0.07 pc at the distance of 50 kpc) south-west from the previous estimation by Bozzetto et al. (Reference Bozzetto, Filipović, Urosevic and Crawford2012) (RA(J2000) = 05 $^{h}$ 19 $^{m}$ 34.9 $^{s}$ , Dec (J2000) = –69 $^\circ$ 02’07.9”). To calculate the radio continuum radius of this SNR, we used the miriad task cgslice to plot 16 equispaced radial profiles in 22.5 $^\circ$ segments around the remnant at 5 500 MHz. Each profile is 25” in length (Figure 2). We divide the remnant into four equal regions: south-west (profiles 1–5), south-east (profiles 5–9), north-east (profiles 9–13), and north-west (profiles 13–1) (Figure 2a). We identify the cutoff as the point where each profile intersects the outer contour line ( $3\sigma$ ATCA image contour or 60 $\mu$ Jy beam $^{-1}$ at 5 500 MHz). These cutoffs are represented as dashed vertical lines (see Figure 2b, c, d, and e). The thick black vertical lines represent the average of the cutoffs in each region. The resulting radii vary from $16.46\pm1.19''$ ( $3.99\pm0.28$ pc) towards the south-west (Figure 1b) to $16.08\pm1.14''$ towards the south-east ( $3.89\pm0.27$ pc; Figure 2c) and $18.44\pm0.75''$ ( $4.47\pm0.18$ pc) towards the north-east (Figure 2d) to $16.34\pm0.93''$ ( $3.96\pm0.22$ pc) towards the north-west (Figure 2e) with an average of $16.83\pm1.08''$ ( $4.08\pm0.26$ pc) for the entire SNR. This is consistent with a previous estimation by Bozzetto et al. (Reference Bozzetto, Filipović, Urosevic and Crawford2012). The size of this SNR is close to SNRs of similar age, J0509–673 (Bozzetto et al., Reference Bozzetto, Filipović, Urošević, Kothes and Crawford2014) in the LMC and Kepler (Patnaude et al., Reference Patnaude, Badenes, Park and Laming2012) in the MW.

Figure 2. Estimate of the radio continuum radius of MC SNR J0519–6902 at 5 500 MHz. a: Radial profiles around the remnant from the centre (see Section 3.1) overlaid on the $3\sigma$ ATCA image contour (60 $\mu$ Jy beam $^{-1}$ ) at 5 500 MHz. The central position of the remnant is RA (J2000) = 05 $^{h}$ 19 $^{m}$ 34.85 $^{s}$ , Dec (J2000) = $-$ 69 $^\circ$ 02’08.22”. b, c, d, and e: Radial profiles. The dashed vertical lines represent profile cut-offs and the thick black vertical lines represent the average of the cutoffs (in arcsec) for different parts of the shell; south-west, south-east, and north-east, and north-west, respectively.

3.2 Polarisation

The fractional polarisation for MC SNR J0519–6902 was calculated using the Equation:

(1) \begin{equation}P=\frac{\sqrt{S^2_Q+S^2_U}}{S_I},\end{equation}

where P is the average fractional polarisation, and $S_{Q}$ , $S_{U}$ , and $S_{I}$ are integrated intensities for the Q, U, and I stokes parameters, respectively.

MC SNR J0519–6902 fractional polarisation vectors appear prominent in 5 500 and 9 000 MHz maps (see Figure 3), we also present polarisation intensity maps in the same figure. The average fractional polarisation values are $5\pm1$ % and $6\pm1$ % for 5 500 and 9 000 MHz, respectively. These are fractionally lower than the values reported by Bozzetto et al. (Reference Bozzetto, Filipović, Urosevic and Crawford2012) but consistent within the errors, and higher than the values of low frequencies (1 472 and 2 368 MHz) reported by Dickel & Milne (Reference Dickel and Milne1995). They are similar to the values of N 103B of $\sim$ 8% at 5 500 MHz (Alsaberi et al., Reference Alsaberi, Barnes and Filipović2019) in the LMC, Kepler of $\sim6$ % at 4 835 MHz (DeLaney et al., Reference DeLaney, Koralesky, Rudnick and Dickel2002), and G1.9+0.3 of 6% at 5 500 MHz (De Horta et al., Reference De Horta, Filipovic and Crawford2014) in the MW.

Figure 3. Fractional Polarisation vectors of MC SNR J0519–6902 overlaid on the intensity ATCA images at 5 500 MHz (upper left) and at 9 000 MHz (upper right). The blue circle in the lower left corner represent a synthesised beam of $5''\times5''$ and the blue line below the circle represents 100% polarisation. The bar on the right side represents the greyscale of the intensity images in Jy beam $^{-1}$ . Polarisation intensity maps of MC SNR J0519–6902 at 5 500 MHz (bottom left) and at 9 000 MHz (bottom right) are shown at bottom with intensity image contour lines overlaid. Respective contour levels are 0.2, 0.6, 1, 1.4, 1.8, and 2.2 mJy beam $^{-1}$ for both images. The colour bar represents gradients of polarisation intensity.

3.3 Spectral index

The radio spectra of MC SNR J0519–6902 can be described as a pure power-law of frequency.

We used the miriad (Sault et al., Reference Sault, Teuben, Wright, Shaw, Payne and Hayes1995) task imfit to extract a total integrated flux density from all available radio continuum observations of MC SNR J0519–6902 listed in Table 2. This includes observations from the Murchison Widefield Array (MWA) (For et al., Reference For, Staveley-Smith and Hurley-Walker2018), Molonglo Observatory Synthesis Telescope (MOST) (Clarke et al., Reference Clarke, Little and Mills1976; Mauch et al., Reference Mauch, Murphy and Buttery2003), Australian Square Kilometre Array Pathfinder (ASKAP) (Pennock et al., Reference Pennock, van Loon and Filipović2021), and ATCA. We measured the MWA flux density for each sub-band (76–227 MHz) and we also re-measured the ASKAP flux density at 888 MHz from Pennock et al. (Reference Pennock, van Loon and Filipović2021). All arrays are sensitive to angular scales much larger than the size of the SNR. Therefore, the sampling of the uv plane for different arrays and at different frequencies would not affect flux density measurements of extended structures. For cross-checking and consistency, we used aegean (Hancock et al., Reference Hancock, Trott and Hurley-Walker2018) and found no significant difference in integrated flux density estimates. Namely, we measured MC SNR J0519–6902 local background noise (1 $\sigma$ ) and carefully selected the exact area of the SNR which also excludes all obvious unrelated point sources. We then estimated the sum of all brightnesses above 5 $\sigma$ of each individual pixel within that area and converted it to SNR integrated flux density following (Findlay, Reference Findlay1966, Equation 24). We also estimate that the corresponding radio flux density errors are below 10% as examined in our previous work (Filipović et al., Reference Filipović, Payne and Alsaberi2022; Bozzetto et al., Reference Bozzetto, Filipović and Sano2023).

In Figure 4, we present the flux density vs. frequency graph for MC SNR J0519–6902. The relative errors are used for the error bars on a logarithmic plot. The best power-law weighted least-squares fit is shown (thick red line), with the spatially integrated spectral index $\langle\alpha\rangle= {-0.62 \pm 0.02}$ ; slightly steeper but within range when compared to Bozzetto et al. (Reference Bozzetto, Filipović, Urosevic and Crawford2012). This spectral index is consistent with values of similar aged SNRs such as Kepler (–0.64, Dickel et al., Reference Dickel, Sault, Arendt, Matsui and Korista1988) and SN 1006 (–0.6, Gardner & Milne, Reference Gardner and Milne1965) within the MW.

We also produced a spectral index map using 5 500 and 9 000 MHz images. To create this, all images were first re-gridded to the finest pixel size ( $0.26''\times0.26''$ ) using the miriad task regrid. The images were then smoothed to a common resolution ( $3''\times3''$ ) using convol. Finally, maths was applied to create the spectral index map and its corresponding error map as shown in Figure 5. The average spectral index value across the remnant was found to be ${-0.7\pm0.2}$ . We note that the uncertainty value is an order of magnitude higher than the estimate from Figure 4. This is because the error map of spectral index was generated using only two images (5 500 and 9 000 MHz).

3.4 Rotation measure and magnetic field

To calculate the RM of MC SNR J0519–6902, the 2 048 MHz bandwidth at 5 500 MHz was split into four 512 MHz sub-bands (4 723, 5 244, 5 756, and 6 268 MHz). Using position angle measurements associated with fractional polarisation maps for these frequencies, the miriad task imrm was applied. RM values with an error $\geq$ 120 rad m $^{-2}$ were flagged. The resulting RM map is shown in Figure 6. We note the RM values are mostly negative with some positive values towards the north and west of the remnant. Values of RM are concentrated in three areas of the remnant; north-east, north, and north-west (see Figure 6), with an average value of ${-124\pm83}$ rad m $^{-2}$ , different than the value reported by Bozzetto et al. (Reference Bozzetto, Filipović, Urosevic and Crawford2012). Once the RM had been determined, the vectors were rotated back to their intrinsic (zero wavelength) values. In Figure 7 an additional 90 $^\circ$ was added to the electric vectors so that the vectors represent the direction of magnetic-fields within the remnant. These vectors are oriented in a radial direction (see Figure 7), as expected from a young SNR.

To calculate the electron density ( $n_e$ ) for MC SNR J0519–6902, we used Equation 2, where EM is the emission measure in cm $^{-3}$ , D is the distance to the SNR; $n_e$ and $n_p$ are electron and proton densities, respectively.

(2) \begin{equation} EM = \int n_e * n_p\;dV/(4\pi D^2) ,\end{equation}

By assuming $n_e = 1.2\,n_p$ , a volume filling factor = 1, an isotropic and spherical distribution of ionised gas, and using an $EM\sim\! 22.5 \times 10^{58}$ cm $^{-3}$ (Maggi et al., Reference Maggi, Haberl and Kavanagh2016), we obtain $n_e = 6.8\,cm^{-3}$ .

The magnetic field strength can be estimated using the Equation for Faraday depthFootnote ^g:

(3) \begin{equation} RM=811.9\int_{0}^{L}n_{e}B_{\parallel}dl ,\end{equation}

where RM is the rotation measure in rad m $^{-2}$ , $n_e$ is the electron density in cm $^{-3}$ , $B_{\parallel}$ is the line of sight magnetic field strength in $\mu$ G, and L is the path length through the Faraday rotating medium in kpc (Clarke, Reference Clarke2004).

Using our average value of RM (–124 rad m $^{-2}$ ), an SNR thickness (L) of the compressed shell of $\sim$ 2 pc, and an $n_e$ value of 6.8 cm $^{-3}$ , we obtain an average magnetic field strength of 11.2 $\mu$ G.

Table 2. Flux density measurements of MC SNR J0519–6902. The asterisk ( $^{*}$ ) indicates that we re-measured this flux density.

Figure 4. Radio continuum spectrum of MC SNR J0519–6902. The blue dot indicates that we re-measured this flux density.

The equipartition modelFootnote ^h can also be used to estimate the magnetic field strength for MC SNR J0519–6902. This method uses modeling and simple parameters to estimate the intrinsic magnetic field strength and energy contained in the magnetic field and cosmic ray (CR) particles using radio synchrotron emission (Arbutina et al., Reference Arbutina, Urošević, Andjelić, Pavlović and Vukotić2012; Arbutina et al., Reference Arbutina, Urošević, Vučetić, Pavlović and Vukotić2013; Urošević et al., Reference Urošević, Pavlović and Arbutina2018).

This approach is purely analytical, described as a rough, only order of magnitude estimate because of assumptions used in analytical derivations and errors in the determination of distance, angular diameter, spectral index, a filling factor, and flux density, tailored especially for the magnetic field strength in SNRs. Arbutina et al. (Reference Arbutina, Urošević, Andjelić, Pavlović and Vukotić2012); Arbutina et al. (Reference Arbutina, Urošević, Vučetić, Pavlović and Vukotić2013); Urošević et al. (Reference Urošević, Pavlović and Arbutina2018) present two models; the difference is in the assumption whether there is equipartition, precisely constant partition with CRs or only CR electrons. Urošević et al. (Reference Urošević, Pavlović and Arbutina2018) showed the latter type of equipartition is a better assumption than the former.

Using the Urošević et al. (Reference Urošević, Pavlović and Arbutina2018) modelFootnote ⁱ, the mean equipartition field over the entire remnant is ${72\pm5}$ $\mu$ G with an estimated minimum energy of E $_\textrm{min}$ = 2.6 $\times10^{48}$ erg. However, using the Arbutina et al. (Reference Arbutina, Urošević, Andjelić, Pavlović and Vukotić2012) earlier model, which assumes CRs composed of electrons, protons, and ions, the estimate is ${156\pm5}$ $\mu$ G, with a minimum explosion energy of E $_\textrm{min}$ = 1.2 $\times10^{49}$ erg. This is consistent with previous values reported by Bozzetto et al. (Reference Bozzetto, Filipović, Urosevic and Crawford2012).

Of course, we cannot suppose we know the remnant’s magnetic field to the significant figures implied above as the models do not have any such accuracy. We do feel it is reasonable to say MC SNR J0519–6902’s magnetic field can be estimated between 10 and $10^2$ $\mu$ G. This is consistent with other relatively young Type Ia SNRs including Tycho, Kepler, and SN 1006 (Reynolds & Ellison, Reference Reynolds and Ellison1992).

3.5 $\Sigma$ –D relation

A previous $\Sigma$ –D diagram (Urošević Reference Urošević2022; Urošević, Reference Urošević2020; Pavlović et al., Reference Pavlović, Urošević and Arbutina2018, their Figure 3) was compared with our values, D $\sim$ 8 pc and $\Sigma_\textrm{1 GHz} =6\times10^{-20}$ W m $^{-2}$ Hz $^{-1}$ sr $^{-1}$ , for MC SNR J0519–6902. This comparison implies the MC SNR J0519–6902 is undergoing expansion within a surrounding environment characterised by a low-density of 0.005–0.02 cm $^{-3}$ with an initial energy of explosion of E $_\textrm{ 0}$ = $10^{51}$ erg (Figure 8). MC SNR J0519–6902 position at $\Sigma$ –D tracks can set it somewhere at the end of the free expansion phase, close to entering the early Sedov phase of evolution.

Figure 5. Left: Spectral index map of SNR MC SNR J0519–6902 created using 5 500 and 9 000 MHz images. Right: Error map of spectral index of SNR MC SNR J0519–6902. Contour lines are overlaid using our ATCA 5 500 MHz image (0.09, 0.2, 0.4, 0.6, and 0.8 mJy beam $^{-1}$ ).

Of course, multiple tracks can correspond to a single SNR on the $\Sigma$ –D plane. Assuming this is a young SNR, it is unlikely that the black and red lines would apply. If the evolutionary tracks are black or red in this position on the $\Sigma$ –D diagram, then this SNR would need to be old, which does not fit. Pavlović et al. (Reference Pavlović, Urošević and Arbutina2018) gives an explanation for 1-4 outliers. Alternatively, application of the method from Urošević (Reference Urošević2022, Reference Urošević2020) to determine the evolutionary status of newly detected SNRs, the position on the $\Sigma$ –D diagram (where a single SNR might align with multiple tracks) must relate to the equipartition magnetic field strength and spectral characteristics. According to the literature, MC SNR J0519–6902 is a young SNR, and its position on the $\Sigma$ –D diagram, equipartition strength, and steep spectrum (–0.62) support this classification. In this paper, the evolutionary phase using Urošević (Reference Urošević2022, Reference Urošević2020) is not estimated; instead, assuming it is a young SNR, environmental density is estimated based on the $\Sigma$ –D tracks.

3.6 Distribution of H i clouds

The distribution of H i in the LMC velocity range towards MC SNR J0519–6902 is shown in Figure 9a where bright H i clouds lie to the south-east. At this radial velocity from $\sim$ 220 to $\sim$ 270 km s $^{-1}$ alone, H i gas was predominantly present (see Appendix for details). However, we could not confirm an association because the angular resolution of the H i data is roughly twice the diameter of the SNR. The SNR shell H i integrated intensity is $\sim$ 600 K km s $^{-1}$ , corresponding to a column density of $\sim1 \times 10^{21}$ cm $^{-2}$ under an optically thin assumption (e.g., Dickey & Lockman, Reference Dickey and Lockman1990).

Figure 6. Rotation measure boxes of MC SNR J0519–6902 (ATCA; 4 723, 5 244, 5 756, and 6 268 MHz) overlaid on the ATCA image at 5 500 MHz (greyscale). Most boxes are open (negative values), ranging from a minimum of –314 rad m $^{-2}$ to a maximum of 33 rad m $^{-2}$ . The blue circle in the lower left corner represents a synthesised beam of $5''\times5''$ . The bar on the right side represents the greyscale of the 5 500 MHz ATCA image in Jy beam $^{-1}$ .

Figure 7. Magnetic field direction vectors of MC SNR J0519–6902 overlaid on the ATCA image at 5 500 MHz (greyscale). The blue circle in the lower left corner represents a synthesised beam of $5''\times5''$ . The bar on the right side represents the greyscale of the 5 500 MHz ATCA image in Jy beam $^{-1}$ .

Figure 8. Radio surface brightness–to–diameter diagram for SNRs at a frequency of 1 GHz (black triangles), obtained from numerical simulations (Pavlović et al., Reference Pavlović, Urošević and Arbutina2018, their Figure 3). MC SNR J0519–6902 is marked with an open red circle while the open triangle represents Cassiopeia A. The open black circle represents the youngest Galactic SNR, G1.9+0.3 (Luken et al., Reference Luken, Filipović and Maxted2020). Numbers represent the following SNRs: (1) CTB 37A, (2) Kes 97, (3) CTB 37B, and (4) G65.1+0.6.

Figure 9. (a) Integrated intensity map of H i towards MC SNR J0519–6902. The integrated velocity range is from 214.5 to 269.7 km s $^{-1}$ . The superposed contours are the same as shown in Figure 1a. (b) Position–velocity diagram of H i. The integrated Declination range is from $-$ 69.04 $^{\circ}$ to $-$ 69.03 $^{\circ}$ . The dashed curve in the position–velocity diagram indicates the boundary of the H i cavity (see Section 3.6).

Figure 9b shows a position–velocity diagram centred near the SNR shell on the edge of an incomplete cavity-like structure of H i that spans a diameter of $\sim$ 50 pc and a velocity of $\sim$ 30 km s $^{-1}$ , respectively. MC SNR J0519–6902 possibly exploded at the edge of what may be a wind bubble that might have originated from a fast outflow driven from an accreting white dwarf (WD); similar to that proposed for Tycho (Zhou et al., Reference Zhou, Chen and Zhang2016; Tanaka et al., Reference Tanaka, Okuno and Uchida2021) and N 103B (Sano et al., Reference Sano, Yamane and Tokuda2018). Using a model from Leahy (Reference Leahy2017) having an updated radius of 3.6 pc in a wind environment, we find new ages with the same model between 250 and 1,100 yrs for various values of the ejecta density profile index ( $r^{-n}$ with $n=6$ to 14), compared to $\sim 2\,700$ yrs for a uniform circumstellar medium.

At the same time, the tentative bubble may not originate from the progenitor star(s). Those winds eject of order one solar mass of material at low velocities with low energy. The H i bubble has a size of 30 pc at the distance of the LMC. We use Kwok (Reference Kwok2007) to calculate the radius and mass of the swept-up H i shell. Using his Equation (16.62), we obtain the shell radius of 1.9 pc for a wind speed of 20 km s $^{-1}$ , ISM density of 1 cm $^{-3}$ and wind duration of 1 Myrs and 55 pc for a duration of 100 Myrs. The mass of the H i shell would be 0.8 M $_\odot$ and 2 $\times 10^{4}$ M $_\odot$ for duration of 1 and 100 Myrs, respectively. The mass of the partial shell feature that is visible in H i in Figure 9b is estimated by integrating the H i over velocity and over the area of the shell (with radius $\sim$ 30 pc). The resulting mass in the feature is $\sim$ 2.5 M $_{\odot}$ , which is consistent with the theoretical estimate above if the duration of the wind is $\sim$ 3 million yrs. The radius of a shell expanding at 10 km s $^{-1}$ for 3 million yrs is 30 pc, consistent with a low-mass stellar wind origin of the shell. In any case, further H i studies with higher angular resolutions are needed to confirm this scenario for MC SNR J0519–6902.