Pre-Supernova Ring Around PSR0540-69
ABSTRACT SNR0540-69 is a supernova remnant in the LMC, harbouring a young (~1600 yrs) radio/optical/X-ray pulsar (P=50 ms). Ground based Halpha imaging of the region has shown a unique spiral-like structure centered around the pulsar. In narrow band HST imaging, the feature seems resolved in a ring-like structure, probably ejected by the progenitor star in a pre-supernova phase (> 10^4 yrs ago).
arXiv:astro-ph/9810016v1 1 Oct 1998
PRE-SUPERNOVA RING AROUND PSR0540-69.
PATRIZIA A. CARAVEO1,2, ROBERTO MIGNANI3AND GIOVANNI F. BIGNAMI4,5
1Istituto di Fisica Cosmica del CNR , Milano, Italy
2Istituto Astronomico, Roma, Italy
3STECF-ESO, Garching, Germany
4Agenzia Spaziale Italiana, Roma, Italy
5Universit´ a di Pavia, Italy
ABSTRACT. SNR0540-69 is a supernova remnant in the LMC, harbouring a young (τ ∼ 1600 yrs)
radio/optical/X-ray pulsar (P=50 ms). Ground based Hα imaging of the region has shown a unique
spiral-like structure centered around the pulsar. In narrow band HST imaging, the feature seems resolved
in a ring-like structure, probably ejected by the progenitor star in a pre-supernova phase (≥ 104yrs
SNR0540-69 is the remnant of the last supernova explosion occurred in the Large Mag-
ellanic Cloud before SN1987A (Kirshner et al, 1989). High resolution Hα images of the
remnant unveiled a spiral-like feature, centered on the pulsar optical counterpart, and
definitely not present in other wide (B and V) or narrow-band filter (OIII and SII)
images (Caraveo et al, 1992). The spiral-like feature was later confirmed by a deeper
Hα exposure, taken with sub-arcsec seeing conditions. The resulting image is shown in
To investigate the nature of the Hα structure, we have pursued both spectroscopy and
high resolution, narrow band imaging.
2. The data.
Spectroscopy of SNR0540-69 was performed on 1995 January, at ESO (La Silla) using
EMMI (ESO Multi Mode Instrument) at the 3.5m NTT. Two 50-minute spectra were
obtained with a 1.5” slit centered at the pulsar position and the long axis oriented
North-South and East-West, as shown in Fig.1.
Globally, our results are in excellent agreement with the comprenhensive work of
Kirshner et al (1989) but for the region around Hα. The identification of the feature
observed at ∼ 6580˚ A is not straightforward. It was first attributed to [NII] (6584˚ A)
by Mathewson et al (1980), then to CaI (6572.8˚ A) by Dopita & Tuohy (1984) and,
finally, to Hα (6562.8˚ A) by Kirshner et al (1989). However, the width of the feature
Fig. 1. 30 min Hα exposure of SNR0540-69 taken in Jan 1994 with the SUperb Seeing Imager
(SUSI) of the ESO/NTT and a ∼ 0.6” seeing. North to the top, East to the left. The spiral-like
feature is ∼ 4” in size. The two stripes show the slit orientations.
(∼ 3500±50 km/s as opposed to the values of ∼ 1800−2000 km/s measured for single
unblended lines) is far too large to allow the identification with a single line. In order
to better assess the structure of the of the 6580˚ A region we can exploit the excellent
quality of our 2-D spectra. Figs.2a and b show a blow-up of the NS and EW 2-D spectra
in the region 6500-7000˚ A including both the SII doublet and the 6580˚ A region. While
the SII lines are blended and significantly redshifted (to wit their origin in a young
SNR which is both expanding at v ∼ 1200 km/s and receding at v ∼ 600 km/s), the
6580˚ A complex is composed of three lines which appear both narrower and less (if at
all) redshifted. As in the Crab Nebula (e.g. Nasuti et al. 1996), we identify the lines
as an Hα, unrecoverably polluted by LMC emission, in the middle of a better defined
NII doublet (6548,6584˚ A). The three lines appear unblended and the troughs, although
partially filled by the synchrotron continuum of the plerion, are deeper than expected
if the lines were affected by the same velocity dispersion as the SII. Moreover, looking
at the brighter 6584˚ A component of the NII doublet, we see that, while its appearence
is markedly different in the two spectra, its centroid is not significantly redshifted w.r.t.
the LMC lines. However, the structure of the 6584˚ A line points towards a non-spherical
emitting region which is not expanding at the rate measured for SNR0540-69 neither is
it receding at the same speed.
Fig. 2. Blow-up of the 2-D spectra, showing the Hα complex and the SII doublet, for the
N-S (a) and E-W (b) orientations of the slit. x-axis: wavelengths, increasing from left to right
(1 pixel = 1.2˚ A); y-axis: spatial coordinate along the slit (1 pixel=0.27”) (a) North at top,
South at bottom (b) East at top, West at bottom. The passbands of the HST filters 656N
(λ = 6562˚ A;∆λ = 22˚ A) and 658N (λ = 6590˚ A;∆λ = 28.5˚ A) are also shown together with the
ESO/NTT Hα (λ = 6552˚ A;∆λ = 60˚ A).
New imaging of SNR0540-69 was performed in October 1995 with the HST/WFPC2.
To image the different components of the λ = 6580˚ A feature, clearly blended in the
wider ESO Hα filter (see the passband schematically shown in Fig. 2a), we have used
the WFPC2 ”Hα” (656N) and “N[II]” (658N) filters. The resulting images are shown
as contour plots in Fig.3 a,b where all field stars, falling within or close the remnant,
were removed. The difference in the remnant brightness between the two filters confirms
our spectroscopy, where the NII [6584˚ A] line appeared most prominent than the Hα
one. The most notable feature in Fig. 3a,b is a bright emission knot ∼ 1′′south of
the pulsar position, clearly visible also in the Hα NTT image (Fig.1). Since the knot
does not appear in our reference HST V-band image, we attribute it to the remnant
structure rather than to a fore/background object. Although the S/N is too low to claim
(λ = 6562˚ A;∆λ = 22˚ A), t=3400s; (b) 658N ”N[II] filter” (λ = 6590˚ A;∆λ = 28.5˚ A),t=4000s.
The frames are aligned in RA and DEC. Axis units are PC pixels (0.045 ”) for a corresponding
field of view of about 5.5” ×5.5”. Each isophote corresponds to an increment of 0.2 counts/pixel
against a background level of 1. Point sources falling within or near the SNR have been removed
via PSF subtraction.
any definite association, a physical link between this knot and PSR0540-69 seems to be
present in the 656N image (Fig.3a). The interaction of PSR0540-69 with the surrounding
medium is clearly shown in the 658N image (Fig.3b) where a conical structure is seen
to originate from the pulsar and to connect it with an arc-shaped feature ∼ 1” to the
NW. A similar, albeit fainter, arc is visible on the East side of the pulsar and the
combination of the two yields a ring-like structure (∼ 3” in diameter and ∼ 0.2” thick)
centered roughly on PSR0540-69. An additional, faint North-South, bar-like, emission
is seen to go through, or originate from, the pulsar. Although definitely brighter in the
”[NII] filter”, a ring-like structure of similar dimensions can be inferred in the noisy
656N image (Fig. 3a).
Since the passbands of the 656N and 658N filters encompass respectively the Hα and
[NII] line, we can directly compare the spectra, shown in Fig.2 a,b, with the HST images,
shown in Fig. 3 a,b, once the different pixel sizes of the two intrument are properly ac-
counted for. Since the NTT/EMMI spatial resolution (y-axis on Fig 2) is 0.27” per pixel
and the PC pixel size is 0.045”, the 1.5” slit covers ∼ 33 PC pixels. In the N-S spectrum,
shown in Fig.2a, the [NII] line has a spatial extent identical to the ring diameter and
its brightest part corresponds to the southern knot. Going from North to South, the
average velocity of the line shifts redward continously for a total displacement of 3-4
EMMI pixels, corresponding to ≤ 5˚ A. In the E-W spectrum, the [NII] line is dominated
by the emission of the knot, now appearing in the upper part of Fig.2b, owing to the
rotation of the CCD. The less intense (and slightly bluer) part of the line is due to
the SW part of the arc-shaped feature, which we have tentatively interpreted as a ring
around the pulsar.
All in all, the velocity of the [NII] feature differs very little from the LMC one and the
overall shear along the N-S diameter can be quantified to be ≤ 100 km/s. However, a
ring expanding ≤ 50 km/s could not have been formed at the time of the SN explosion.
To account for the ring dimensions, an expansion velocity of ∼ 200 km/s would be
needed and our data do not support a velocity shear of ∼ 20˚ A, corresponding to 16
EMMI pixels. Moreover, the [NII] line emission originated from the ring is not affected
by the overall SNR0540-69 recession speed. Therefore this ”ring” must have formed be-
fore the event responsible for such a recession, i.e. before the SN event itself.
Our slowly expanding ring would thus trace a pre-supernova stage of the progenitor star
like, probably, for the rings around SN1987A (e.g. Plait et al,1995, Panagia et al, 1996)
and for the hourglass nebula around the blue supergiant (BSG) Sher 25 (Brandner et
al, 1997a and b). SNR0540-69 would thus be the second example, after SN1987A, of a
supernova remnant retaining memory of a pre-supernova mass ejection similar to that
undergone by the BSG Sher25 ∼ 6,600 year ago (Brandner et al, 1997b). Indeed, both
the dimensions and the slow expansion rate of the SNR0540-69 ring are reminiscent of
the inner ring of SN1987A as well as of that around Sher 25.
If we now concentrate on the comparison between SNR0540-69 and SN1987A, we find
that their many similarities in dimensions, composition and velocity point towards sim-
ilar progenitors for these two supernovae which happen to be in the same star forming
region in the LMC (Kirshner et al, 1989). Assuming that the progenitor of the ∼ 1600yrs
old SNR0540-69 experienced mass ejection ≥ 104yrs before the supernova explosion and
using the distance value of 50.9 kpc, as inferred by Panagia et al (1997) for SN1987A,
the measured angular size of the ring (∼ 3”) implies a velocity ∼ 30−40 km/s. This is
larger to the expansion velocity measured for the inner ring of SN1987A (∼ 11 km/s,
Panagia et al, 1996) but similar to the value of ∼ 30km/s found for the ∼ 6,600 y old
ring around Sher25 (Brandner et al, 1997b).
Coupling high resolution imaging with spectroscopy, we have shown that the ring-like
structure around PSR0540-69 was formed during the pre-supernova phase and survived
to the present day. We can exclude that the ring be simply due to SN ejecta since this
would imply an expansion velocity too high to be consistent with our spectral data
but far too low to be consistent with the average expansion rate measured for the SNR.
Moreover, the absence of any significant redshift for the NII line argues strongly in favor
of a pre-supernova origin. In the assumption that the receding speed of SNR0540-69 was
acquired during the supernova explosion, the ”ring” must have originated before such
explosion, during the pre supernova phase of the progenitor as it is believed to be the
case for the rings around SN1987A.
However, considering also the case of the ring detected around the BSG Sher 25, we
can study such structures well before the SN explosion, and follow them when they are
lighted first by the SN flash and than by the interaction with the SN ejecta and, possibly,