2 Observations

Figure 1: NGC 6951 showing SN 1999el (left panel) and SN 2000E (right panel).

2.1 Photometry

The CCD-photometric observations were obtained with four telescopes: the 28 cm Schmidt-Cassegrain at the campus site of University of Szeged (#1), the 60/90 cm Schmidt at Piszkésteto Station of Konkoly Observatory (#2), the 1 m Ritchey-Chrétien-Cassegrain at Piszkésteto (#3) and the 1.2 m Cassegrain at Calar Alto Observatory, Spain (#4). The small-aperture instrument was used only around the maximum of the SN. The CCD-frames were exposed through standard Johnson-Cousins filters, most often V and $R_{\rm C}$. $I_{\rm C}$ was used at early epochs, and a few B frames were also obtained at later phases. It would have been better to use the same telescope and setup to obtain a homogeneous dataset. However, this was strongly limited by the weather conditions and the availability of the instruments.

Transformation to the standard system has been performed by applying the following equations:

$$V =~ v + C_{\rm V} (V-R) + D_{\rm V}$$ (1)

$$R =~ r + C_{\rm R} (V-R) + D_{\rm R}$$ (2)

$$I =~ i + C_{\rm I} (V-I) + D_{\rm I}.$$ (3)

The occasional B data have been transformed in a similar way, except that (B-V) was used in the colour term. The instrumental coefficients (the C_i slopes and D_i zero points) have been determined by observing Landolt standard fields (Landolt 1992) or the M67 standard sequence (Montgomery et al. 1993) at each site. Actually, the use of the zero points have been eliminated by observing local comparison stars in the field of SN 2000E and computing differential magnitudes. Differential extinction correction between the SN and the local comparison stars was neglected, due to the small field of view. For each telescope the transformation slopes are collected in Table 1. The magnitudes of the standard stars could be recovered within $\pm$0.03 mag.

 

 

Table 1: Transformation slopes. See text for the telescope codes.

Tel.	$C_{\rm V}$	$C_{\rm R}$	$C_{\rm I}$
1	-0.11	-0.08	-0.04
2	+0.10	+0.08	-
3	-0.06	+0.06	+0.05
4	+0.06	+0.07	-

Figure 2 shows the field of NGC 6951 and SN 2000E with the local comparison stars labelled. The standard magnitudes of these stars were determined via Landolt standards observed with the Calar Alto telescope, where the photometric conditions were the best during our campaign. The results are listed in Table 2. Note that B1, B2 and B3 were used only for the frames taken with telescope #3.

 

 

Table 2: Standard magnitudes of local comparison stars. Errors are given in parentheses.

Star	V	(B-V)	(V-R)	(V-I)
F1	12.53 (0.01)	1.64 (0.02)	0.81 (0.01)	1.74 (0.01)
F2	14.86 (0.01)	0.87 (0.02)	0.52 (0.02)	1.01 (0.02)
F3	13.90 (0.03)	0.72 (0.04)	0.44 (0.01)	0.87 (0.02)
F4	15.55 (0.02)	0.74 (0.04)	0.43 (0.02)	0.92 (0.04)
F5	14.97 (0.02)	0.91 (0.03)	0.54 (0.02)	1.01 (0.02)
F6	15.71 (0.03)	0.88 (0.05)	0.49 (0.03)	1.07 (0.06)
F7	14.53 (0.02)	0.88 (0.02)	0.51 (0.02)	1.06 (0.03)
F8	saturated
B1	15.88 (0.03)	0.97 (0.05)	0.57 (0.03)	1.16 (0.04)
B2	16.42 (0.05)	1.01 (0.11)	0.54 (0.06)	1.21 (0.07)
B3	16.60 (0.05)	1.18 (0.11)	0.64 (0.06)	1.32 (0.07)

The magnitudes of SN 2000E were inferred with aperture photometry, and transformed to the standard system via the local comparison stars. A small aperture radius of 4 pixels has been used in order to minimize the effect of the host galaxy background. The background light was estimated within a 4 pixel-wide annulus having 5 pixels inner radius. This background was determined on each frame and subtracted from the SN flux. PSF-photometry was not applied, because the frames with telescopes #1 and #2 had undersampled and/or strongly varying PSF. The transformation equations (Eqs. (1-3)) were applied for all observed SN data including those that were obtained more than 1 month past maximum. The colour terms in Eqs. (1-3) resulted in magnitude corrections in the order of 0.05 mag in the nebular phase. It may caused some additional uncertainty, because at late epochs the spectral distribution of the SN light resembles more a nebula than a star. K-correction has been neglected, because at the redshift of NGC 6951 (z = 0.005) it does not exceed 0.01-0.02 mag in V (Hamuy et al. 1993), i.e. it is well below the photometric uncertainty. The effect of the host galaxy background on the SN flux was investigated on frames taken in October, 1999 showing SN 1999el only (see Fig. 1, left panel). An aperture with the same size was placed on the position of SN 2000E and the surrounding background was subtracted, as above. The remaining flux was negligible even in the R filter. Thus, the background subtraction with the small aperture-annulus combination gave acceptable magnitudes for SN 2000E, minimizing the effect of the host galaxy light.

Figure 2: Local secondary standard stars in the field of NGC 6951. North is up and east is to the left. See Table 2 for magnitudes.

The list of standard magnitudes of SN 2000E is given in Table 3. The errors were estimated from the rms deviations of the SN magnitudes from the comparison stars on each frame, and do not contain the possible systematic errors arise from the standard transformation. We believe that this latter error source is smaller than the uncertainites of the instrumental magnitudes measured with smaller telescopes.

The magnitudes in Table 3 were also compared with filtered CCD-magnitudes of Hornoch & Hanzl (2000) that were made between JD 51655 and 51666. Their R-magnitudes differ by about 0.4 mag from our values. This is a 4$\sigma$difference considering the given uncertainty of their data ($\pm$0.1 mag). Because the details of the standard transformation of the data of Hornoch & Hanzl were not published, the cause of this discrepancy cannot be studied in more detail here, except to underline that this may indicate a systematic error in either datasets at late epochs of SN 2000E. More published standardized measurements are needed to resolve this issue.

 

 

Table 3: Photometric data of SN 2000E. Errors are given in parentheses.

JD	B	V	$R_{\rm C}$	$I_{\rm C}$	Tel.
2451572.2	-	-	13.77 (0.03)	-	2
2451576.2	-	13.88 (0.07)	13.53 (0.04)	13.72 (0.06)	1
2451578.3	-	13.73 (0.04)	13.51 (0.02)	13.57 (0.06)	1
2451579.3	-	13.67 (0.04)	13.51 (0.08)	13.68 (0.05)	1
2451581.3	-	13.78 (0.10)	13.50 (0.16)	13.59 (0.15)	1
2451582.3	-	13.80 (0.11)	13.63 (0.08)	13.60 (0.05)	1
2451585.3	-	14.02 (0.07)	13.76 (0.04)	13.75 (0.05)	1
2451589.5	-	14.21 (0.02)	14.12 (0.02)	14.14 (0.03)	2
2451656.5	-	-	16.39 (0.02)	-	2
2451657.5	-	-	16.43 (0.02)	-	2
2451658.5	17.64 (0.02)	16.65 (0.03)	16.48 (0.02)	-	2
2451661.5	17.66 (0.02)	16.79 (0.02)	16.52 (0.02)	-	2
2451662.5	17.80 (0.03)	16.88 (0.02)	-	-	2
2451664.5	17.85 (0.02)	16.93 (0.02)	-	-	2
2451667.5	-	16.82 (0.10)	16.51 (0.15)	-	2
2451696.5	-	17.63 (0.02)	17.24 (0.03)	-	3
2451705.6	-	17.89 (0.04)	17.87 (0.08)	18.09 (0.08)	3
2451706.6	-	17.89 (0.06)	17.86 (0.10)	-	3
2451727.4	-	18.22 (0.02)	18.41 (0.05)	18.24 (0.05)	4

The light curves in V, R and I filters are plotted in Fig. 3. The continuous line is the optimal result of the Multi-Colour Light Curve Shape (MLCS) method (see next section). It can be revealed that the maximum brightness in V occurred around JD 51580, and the peak V-magnitude was about 13.7 mag, being in good agreement with the predicted peak brightness of SNe Ia at the distance of NGC 6951 (see Sect. 1).

Figure 3: Light curves of SN 2000E in B,V,R,I filters. The B,R,I data have been shifted vertically for better visibility. The continuous line is the template calculated with the MLCS method using $\Delta =-0.5$, E(B-V)=0.34 and $\mu _{\rm0}=32.70$ mag (see Table 4).

Figure 4: Comparison of $V(RI)_{\rm C}$ light curves of SN 2000E (filled circles) with those of SN 1998bu (open symbols). All SN 1998bu data have been shifted vertically by 1.88 mag in order to bring them into agreement with SN 2000E. The similarity of the light curves is apparent.

In order to test the reliability of our measurements and standard transformation, the light curves of SN 2000E were compared with those of SN 1998bu. For this bright SN, which appeared in the Leo I group galaxy M96, published light curves of very good quality are available, and it received considerable attention recently (Jha et al. 1999; Suntzeff et al. 1999). Figure 4 shows that the V, R, I magnitudes of SN 2000E (filled symbols) at the early part of the light curve agree satisfactorily with the data of SN 1998bu. This agreement was reached by adding 1.88 mag to the light curves of SN 1998bu in all filters. This suggests that the reddening of the two SNe were similar, about $E(B-V) \approx 0.3$ mag (Jha et al. 1999; see also Sect. 3). At present, the accuracy of inhomogeneous SNe light curves is usually not better than $\pm$0.1 (see e.g. Fig. 8 of Jha et al. 1999), so the small deviations in Fig. 4 (especially in the I band) are probably not significant. The light curves will be analysed further in Sect. 3.

2.2 Objective-prism spectroscopy

Spectroscopic observations were gathered with an objective prism attached to the 60/90 cm Schmidt telescope at Piszkésteto Station of Konkoly Observatory, between 26th and 28th February, 2000 (JD 2451601 - 03), when SN 2000E was about 1 month past maximum. The images were exposed onto an electronically cooled Thomson $1536 \times 1024$ CCD-chip (readout noise about 16 e^-). The dispersion axis was aligned in the north-south direction, along the shorter side of the CCD-chip.

An objective prism spectrograph is certainly not an ideal tool for SN spectroscopy. However, this was the only spectroscopic instrument available to us at that time.

A considerable number of image processing steps were necessary to extract the SN from the smeared spectrum of the host galaxy. The location of the SN spectrum was determined from an intensity plot along the line perpendicular to the dispersion axis. At first, this was possible only in the blue region where the host galaxy showed negligible contribution. The red side, however, was heavily contaminated by the smeared background spectrum of the host galaxy. Removal of this background was essential to obtain reliable SN spectra.

The easier way to correct for the galaxy background would have been the usage of an objective prism picture of the host galaxy without the SN. Since it was not possible for us, we had to choose another, more approximative approach. First, the central peak of the background galaxy spectrum was identified visually. Then, the galaxy image was cut into two pieces at this central ridge, and the western part (including the SN) was dropped. The eastern side was reflected and added back into the position of the western side, thus, generating a symmetric picture of the galaxy. This picture was then subtracted from the original one, resulting in a much cleaner SN spectrum. While the removal of the galaxy "spectrum'' was far from complete, its contribution at the position of the SN spectrum was considerably suppressed in this way.

Figure 5: Objective-prism spectra of SN 2000E. The epochs of the observations are indicated to the right of each spectrum. Some spectral features are marked (see text for reference).

The extraction of the cleaned SN spectrum was performed with the standard subroutines in IRAF/SPECRED. The intensities within a 3-4 pixel-wide aperture were summed and this aperture was slid along the dispersion axis, taking into account the tilt of the spectrum. The remaining background light was subtracted after a low-order polynomial fit. The wavelength calibration was performed using the lines in the spectrum of Vega taken with the same instrument and setup. The Vega spectrum was also used for the flux calibration. First, the telescope response function was determined by matching the measured continuum fluxes of the Vega spectrum with the tabulated ones (given in e.g. Gray 1992). Then, the SN spectrum was multiplied by the response function producing a flux-calibrated spectrum. The resulting spectra of SN 2000E are plotted together in Fig. 5, where an arbitrary vertical shift was applied for better visibility. Information for line identification was collected from Filippenko (1997).

It is apparent from Fig. 5 that SN 2000E shows the standard spectral features of a type Ia SN at one month after maximum light. A closer inspection of the features around 6000 Å with those of SN 1998aq taken at the same phase (Vinkó et al. 1999) showed good agreement, despite of the much lower resolution of the present spectra. This confirms the classification of type Ia, although a slightly confusing description of the presence of the $H\alpha$ line was also reported in IAUC 7353 by Polcaro et al. (2000). This was then revised in IAUC 7359.

There is continuously growing amount of evidence that SNe Ia show considerable diversity in peak brightness, decline rate, spectral features, etc. (see e.g. Filippenko 1997; Phillips et al. 1999; Nugent et al. 1995; Hatano et al. 2000 and references therein). The similarity between the spectra of SN 2000E and SN 1998aq may mean that SN 2000E is close to the "normal'' SNe Ia being neither SN 1991T-like, nor SN 1991bg-like event. It should be noted, however, that the spectroscopic diversity between these subclasses of SNe Ia is usually studied at earlier epochs, between $\pm$10 days around maximum. Therefore, the spectra presented here are too late for such a distinction. This issue will be studied in more detail using the shape of the light curve in the next section.