THE FIRST 400 DAYS OF OBSERVATIONS OF SN 1993J IN M81 AT NF/ OBSERVATORY
A. William Neely EMAIL neely@astro.wnmu.edu ABSTRACT 1993J was first thought to be a Type II supernova. Subsequently, helium lines appeared in its early evolution, and 1993J was re- classified as a Type IB supernova. Our observatory runs completely automatically, using ATIS (automatic telescope instruction set) software, and fortuitously imaged the parent galaxy (at UD March 28.30, 1993) only a few hours after shock outbreak and again 20 hours later. After the supernova was discovered (March 28.9), and the news dispersed via internet, frequent observations in V, R, and I were taken over the next 400 days. Early data showed a steep climb in brightness to a first peak at 2.3 days after outbreak, a dip to a early minima at 8 days, then a slow climb to the second peak at 23 days. The (V-R) data is more complex, showing cooling phases after the first two peaks, another cooling period from day 85 to 118, and then a final cooling starting at day 165 and continuing to the end of the observational period. The data suggest a small ejection mass with several phases of photosphere regression to hotter layers of gas. 1. Introduction NF/ observatory is a automatic robotic telescope (.45 meter F4.5 newtonian) with a radio link to Silver City, NM. Its operation has been described before (Neely 1989). The observatory is presently producing approximately 10,000 images per year. The nightly observation list is determined in advance and compiled in the ATIS (automatic telescope instruction set) format (Epand 1992 and Boyd 1993). The 'ATIS' file can be updated by radio, as needed, to modify the target list. At the time of shock outbreak from SN 1993J, the telescope was engaged in imaging all the Messier objects. Late March is the usual time for 'Messier Marathons', which are generally done on-site without computer aid. SN 1993J was discovered by F. Garcia in Lugo, Spain, on March 29.9, 1993. He was a member of a supernova search group, and reported his observation to J. Ripero who measured the magnitude at Vmag = 10.8 (Ripero 1993). Once the report of the supernova was relayed to us by Alain Porter at Kitt Peak, a new instruction set was sent to the observatory, and SN 1993J was assigned a high priority for nightly observations. The first two observations were un-filtered. All subsequent nights also used V, R, and I filters. 2. Instruments and Lightcurves The camera uses a 1024x1024 Craf-Cassini CCD, with a total noise of 25 electrons. The filters are the Kitt Peak 'Harris set', which is well matched to Johnson V and Cousins R. The I filter consists of 3mm of RG9 Schott glass. Our initial reported value for the supernova on UD March 28.3 was Vmag = 13.7 (Neely 1993a). This value was approximated from 'Richmond Star B and C' (GSC 0928 and 0574), assuming a good color match for the supernova. After better estimates of initial color of the SN were received from G. Vaucouleurs (personal communication), this value was refined (Neely 1993b). Exposures of the cluster PG 1633+099 (Lanhold 1992) were made at a zenith angle of 23 degrees. Several stars of different colors in the cluster were used to solve the equation of the form, Vmag = Mag(unfiltered) + a + c(B-V). This gave a color term (c) of .51 +/- .15. Assuming that the (B-V) of the SN was .15 at the time of the observation, the final corrected value on March 28.30 is 13.43 V +/- .15. The next two nights of data were somewhat over-saturated because of the rapid rise of the supernova. These nights had to be modeled, using over-exposures taken later in the course of observations. All other measurements to day 200 were done as compared to Richmond 'B'. After day 200, the magnitudes were compared to Richmond 'C', which remained constant during the observational period. This star still has somewhat uncertain color indicies (De Vaucouleures 1994). Daily measurements and uncertainties are listed in Table 1. Over seven hundred images of the supernova were taken to compile the database. All images were taken automatically using ATIS instructions. The images were then analyzed with PCVISTA (Treffers 1989), on a weekly basis. Uncertainties in the photometric data were taken from the program output. All the figures use a zero time, signifying shock outbreak, of UD March 28.22, 1993. The initial time calculated was 28.0 (Wheeler 1993a), but this figure has been recently revised (Vaucouleurs personal communication). Figure 1, shows the complete 400 days of data in Unfiltered and V. Richmond 'B' star was used as a comparison, until the supernova became fairly faint after day 200. Then Richmond 'C' was used. Initial worries that Richmond 'C' might be variable, were not realized during our observational period. Figure 2 shows the first 400 days in R and I. Figure 3 shows the first 28 days in Unfiltered and V. Figure 4 shows the same interval, first 28 days, in R and I. We see a steep rise to an initial peak of 10.96 V, 2.3 days after the shock outbreak, then a rapid fall to 11.87 V, at 7.9 days. Then a gradual rise to the second and highest peak, 10.83 V, at 22.5 days. Following the second peak, the magnitude falls gradually, reaching 17.1 V, on day 383.9. Following the second peak. The fall of the light curve is exponential, with a half-life of 49 days. Figure 5 shows values for (V-R). There is shift to R (increase in V-R) after the first peak (Porter 1993), which then reverses as the rise to the second peak starts. After the second peak, the R shift starts again until day 50, when the supernova's emission starts to shift toward V again. There is a plateau of the V shift from day 88 to day 120, then the supernova continues to shift to V again until day 170, after which the cloud continues to shift to the red till the end of our observing period. 3. Discussion The first dramatic feature of SN 1993J consisted of a sharp intensity spike after shock outbreak. Early spectra during this phase were featureless and consistent with a well defined blackbody (Wheeler 1993). A hint of the same type of spike may have been seen in SN 1987A. It is likely that this early feature would have been missed in most supernova observations because of its transient nature. After this short-lived burst, the light-curve climbed to a second and more recognizable peak at 22.5 days. By day 8, Hydrogen Alpha lines pre-dominated, suggesting a type II supernova. Instead of reaching a plateau, SN 1993J started to fade rapidly after day 23. By day 28, data from La Palma showed strong He I lines. Clearly, the supernova was no longer consistent with its initial classification, and it was redefined as a type Ib supernova. Some researchers have proposed a new classification of type IIb, which would cover explosions of stars with some hydrogen present, which exhibit helium lines later in the evolution (Balonek 1993). SN 1987K might have been another such event. Other pieces of data point to a event with a low mass ejection. After the second maximum, the decline in brightness of 1993J matches a half-life of 49 days. In SN 1987A the decline was consistent with a half-life of 78 days, from the decay of Co56. Gamma rays at the correct energy levels were detected for the first time in the 1987A event. Because of the greater distance to M81, the models of the Ni-Co production do not predict a sufficient gamma ray flux to be detectable with CGRO(Hoffich 1993). The rapid fall of the lightcurve in 1993J is probably secondary to a fairly transparent ejecta in gamma (Fillepenko 1993 personal communication). The final interesting feature our dataset lies in the decrease in V-R seen from day 60 to day 165. This probably represents a progressive relative regression in the photosphere, uncovering hotter layers, or could represent the effect of a buried source. Possibilities include a pulsar with a period of 5 ms.(Wheeler 1993b). The internet is an invaluable tool in the study of SN events. Several coordinated groups are now scanning galaxies for evidence of SN's. This coupled with the automated system at Berkeley, provides early notification wordwide. SN 1994D and SN 1994I are two bright SN's which were recently spotted before maximum light. Extensive coverage of these events in multiple wavelengths are essential to increase our understanding of the scope of stars involved in these events. Because weather is always an independent variable, a network of observing sites is desirable. The particular site must also be able to alter the night's observing list. NF/ Observatory has a radio telemetry link to Silver City. Normally the ATIS set runs completely "hands-off", but a new ATIS set can be sent to the telescope via radio or phone at a moment's notice. This study of SN 1993J has demonstrated the utility of ATIS software for the long term study of a supernova event. The ability to image a SN event nightly, from a single site, without interrupting other observing schedules, further supports the concept of automated telescope systems. References: Boyd, L.J., Epand, D., Bresina, J., Drummond, M., Swanson, K., Crawford, D.L., Genet, D.R., Genet, R.M., Henry, G.W., McCook, G.P., Neely, A.W., Schmidtke, P., Smith, D.P., Trueblood, M. 1993, IAPPP Communication No 52, pp 23-81, Summer. De Vaucouleurs, G., Corwin, H.G. Jr., Skiff, B.A. 1994, P.A.S.P. 106:156-160, February. Epand, D., Neely, A.W. 1991, IAPPP Communications No. 45 p92-103 Sep.- Nov. Hoflich P., Muller, E., Khokhlov A. 1993, A&A 268,570. Landholt, A.U. 1992, A. J., 104, pp.340-371 and pp.436-491. Neely, A.W., and Treasure, F.A. 1989, Remote Access Automatic Telescopes, Fairborn Press, Mesa, Ariz., p141-150. Neely, A.W., 1993a, IAU Circ. No. 5740. Neely, A.W. 1993b, IAU Circ. No. 5832. Porter, A.C., Neely, A.W. 1993, IAU Circ. No. 5742. Ripero, , J. 1993, IAU Circ. 5730. Treffers, R.R., and Richmond, M.W. 1989, Pub. A. S. P, 101, p725. Wells, L.A., Balonek, T.J., Tweedy, R., Hintz, E.G., Joner, M.D., Porter, A.C., Neely, A.W., Tremonti, C.A., Koch, J.L., Silvestri, N., Roming, W.A., Nelson, K.A., Kwitter, K.B. 1993,"The Early Light of SN 1993J", Presented at the IAU Colloquium No 145 at Xian, PRC. Wheeler, J. C. 1993a, "At the IAU Colloquium No 145 at Xian, PRC". Wheeler, J.C., Barker, E., Boisseau, J., Clocchiatti, B.A., de Vaucouleurs, G., Gaffney, N., Harkness, R.P., Khokhlov, A.M., Lester, D.F., Smith, B.J., Smith, V.V., Tomkin, J. 1993b, A.J., 417:L71-L74, November 10. TABLE 1 Comparisons are to Richmond Star 'B'. (Unfiltered 11.4, V 11.9, R 11.6, I 11.3) after Oct. 10 the Comparisons are to Richmond Star 'C' (Unfiltered 14.0, V 14.5, R 14.2, I 13.9) Uncertainties U Date Unfilt V R I Unfil V R I 1993 Mar 28.30 13.30 13.45 .03 .15 29.15 10.86 11.05 .2 .2 31.05 10.63 10.96 10.79 10.67 .2 .2 .2 .2 Apr 1.11 10.65 11.10 10.85 10.72 .03 .03 .04 .04 2.10 11.00 11.47 11.20 10.88 3.12 11.27 11.72 11.52 11.14 5.15 11.33 11.87 11.50 11.24 8.15 11.09 11.60 11.25 11.07 9.15 10.97 11.49 11.03 10.16 10.88 11.43 11.03 10.80 11.17 10.76 10.92 10.78 12.17 10.69 11.23 10.83 10.72 13.17 10.62 11.10 10.75 10.66 17.20 10.42 10.87 10.55 10.48 19.20 10.38 10.83 10.51 10.52 20.20 10.38 10.83 10.50 10.49 21.20 10.42 10.90 10.52 10.47 24.20 10.62 11.14 10.58 10.49 25.18 10.70 11.27 10.76 10.59 26.20 10.77 11.36 10.84 10.67 29.25 10.92 11.62 11.02 10.79 30.16 11.02 11.72 11.12 10.86 May 1.17 11.07 11.80 11.16 10.88 2.16 11.10 11.83 11.22 10.91 .04 .04 .05 .05 6.17 11.26 12.03 11.36 11.06 7.17 11.42 11.04 8.17 11.32 12.11 11.47 11.11 9.13 11.36 12.12 11.49 11.14 10.15 11.38 12.16 11.51 11.16 11.18 11.42 12.19 11.55 11.20 16.16 11.52 12.29 11.70 11.30 21.15 11.62 12.47 11.78 11.36 22.16 11.68 12.43 11.81 11.41 25.17 11.71 12.47 11.93 11.47 Jun 2.15 11.89 12.63 12.00 11.62 7.16 11.97 12.67 12.10 11.77 8.16 11.98 12.74 12.17 11.79 11.17 12.05 12.87 12.23 11.79 22.16 12.27 12.93 12.45 12.03 23.16 12.29 12.98 12.53 12.06 25.16 12.35 29.16 12.38 13.08 12.59 12.04 .07 .08 .06 .06 Jul 8.16 12.57 13.30 12.81 12.37 21.14 13.42 12.60 22.14 12.90 13.54 13.01 12.68 23.16 12.93 24.15 12.94 13.57 13.18 12.76 25.16 13.65 13.18 12.80 26.16 12.97 Aug 9.40 13.95 14.37 14.04 13.73 .10 .15 .12 .10 18.40 14.17 13.95 14.08 Sep 8.3 14.35 14.98 14.47 14.17 9.3 14.40 14.94 14.39 14.31 24.3 14.55 15.52 14.77 14.62 .20 .20 .20 .20 Oct 8.3 14.60 15.60 14.70 18.3 15.20 15.80 15.20 15.10 19.3 15.20 16.00 15.20 15.00 25.2 15.20 15.40 15.40 Dec 9.2 15.40 16.70 15.30 15.60 .25 .30 .25 .25 25.2 15.60 15.60 1994 Jan 10.2 15.60 16.90 15.40 11.2 15.80 15.70 16.00 16.8 15.80 15.70 Feb 1.3 15.90 9.2 16.17 16.04 15.88 19.2 16.56 16.31 20.2 16.39 16.26 Mar 6.3 16.75 16.87 16.87 16.34 .40 .50 .45 .45 Apr 17.4 17.10 17.30