(5) From G. de Vaucouleurs (gav@astro.as.utexas.edu): Foreground (Galactic) color excess in the M81 field. To assist in the interpretation of the photometry/spectrophotometry of SN1993J, I have estimated - by comparison with the standard main sequence of Johnson and Morgan (ApJ 117, 313, 1953) - the color excess of 10 field stars for which UBV magnitudes are available. Sources: J. Neff, McDonald (ApJS 24, 421, 1972), 3 stars (N); A. Sandage, Palomar (AJ 89, 621, 1984), 4 stars (S); H. Corwin, McDonald (IAUC 5742), 3 stars (C). The range of individual values of E(B-V) is from 0.00 to +0.26, with means 0.065 (S), 0.077 (C), and 0.167 (N). For the total sample the unweighted mean is = 0.099 +/- 0.025 (n = 10, s.d. = 0.080) or, after rejection of two largest values, 0.069 +/- 0.017. With the standard Av/E = 3.3, this implies Av = 0.30 or 0.23, and A(B) = 0.43 or 0.30. For comparison, published values for the reddening in the M81 field range from A(B) = 0.43 (Jacoby et al. ApJ 344, 704, 1988) to 0.30 (RC2, 1976), 0.17 (Tonry, ApJ 373, L1, 1991), 0.16 (RC3, 1991 after Burstein and Heiles), and 0.07 (Sandage, loc.cit), for an unweighted average of = 0.23 +/- 0.06. These figures may be compared with the much higher total extinction quoted by R. Humphreys for the suspected precursor star. ************************************************************************* (3) From Roberta Humphreys (roberta@aps1.spa.umn.edu) Notes on the Progenitor: Reddening, Variability, and Luminosity Roberta Humphreys, Greg Aldering, and Kris Davidson University of Minnesota 2. Reddening: As Richmond reported in IAU Circ 5739 there is high reddening across the face of M81, as indicated from the UV images- Av 1.5 mag ( Hill et al 1992); the radio and Halpha measurements by Kaufman et al (1987) show an average reddening of 1.1 mag +/- 0.4 with no gradient from the nucleus. In a study of the red supergiants Humphreys et al (1986) found Av from 1 -1.5 mag. Thus there is significant reddening throughout M81. SN1993J may be in a hole, but given its location in a spiral arm near a prominent dust lane this seems unlikely. Reddening should be considered when discussing the luminosity of the supernova and the possible progenitor. ******************************************************************************** Hi Alex I have some remarks on the early IUE spectra of SN 1993J that you might like to distribute in your electronic journal which I've been calling SN~1993J Updates. I would like to emphasize that many others are examining the IUE spectra, and they will have seen more of them than I, and in most cases will know far more about analyzing interstellar and circumstellar matter from IUE spectra. So any valid remarks I make will probably be old news to many. I have examined low resolution short wavelength (\sim 1150--1950 A) IUE spectra from Mar.~30 to April 4 and low resolution long wavelength (\sim 2000--3300 A) IUE spectra from Mar.~30--31. These spectra were emailed to me by George Sonneborn who collaborates with Robert Kirshner (my boss) on IUE observations of supernovae. The March~30 spectra seem to be the earliest taken; March 30 was about the epoch of maximum light. The first spectrum rises blueward right across the IUE band from about 3300 A to about 1240 A. (I consider the spectra in the f_lambda representation.) From about 1200 to 1220 A, there is the solar coronal Lyalpha emission feature which is of no interest here. Below 1200 A the IUE spectra are quite noisy, but more examination than I have done might yield some information. By March 31, the UV spectrum had become rather flat, but declining somewhat redward. After March 31, the short wavelength IUE spectra decline redward and there appears to be a broad (i.e., about 80 A) depression centered at about 1630 at least before April~4. Apart from the Lyalpha emission, the 1630 depression, and another broad depression center at about 2200 A (see below), there are no other obvious broad features in the spectra. An interpretation of the early SN~1993 UV spectra, that is already widely discussed (e.g., Wheeler 1993), is that we are seeing a circumstellar wind at least for March 30--31. The circumstellar wind has been heated by the UV flash of the supernova outburst and is cooling rapidly on March 30--31. The wind is moving slowly. High resolution spectra from March 30 show strong narrow emission features at the wavelengths of the N V doublet, 1238.81 and 1242.80 A; I have only seen these spectra on a FAX. The 1238.81 emission is weaker and narrower (only about 100 km/s); the 1242.80 feature is about 20 % higher, has a breadth of about 200 km/s, and exhibits a narrow blueshifted absorption of about 50 km/s in width which appears to be a P~Cygni absorption. It is odd that the 1242.80 feature is more prominent since the N V 1242.80 line has only half the gf of the the N V 1238.81 line. If these features are due to N V in the circumstellar wind, then they suggest wind velocities up to 200 km/s and a wind temperature (assuming LTE) of order 20000--30000 K for March 30. It may be that after March 31, we are seeing the supernova ejecta and that accounts for the depression that appears at about 1630 A. Broad absorptions are what one would expect from rapidly expanding ejecta and are what I had expected to see in the earliest IUE spectra. However, optical P~Cygni lines were only seen on April~6.3 (Wheeler \&~Clocchiatti 1993). Thus, 1630~A depression may just be some sort of modulation of the circumstellar flux. The low resolution spectra are quite noisy but show several narrow absorption lines. The low resolution mode of the IUE has a resolution of about 1 in 300 or 1000 km/s (5 A at 1500 A and 10 A at 3000 A). It is probable that none of the narrow features is resolved. One absorption at about 1545 A is probably the C IV 1549.1 ground level multiplet and is due to the circumstellar material. This absorption disappears after April 1; presumably, the circumstellar material temperature has fallen too low by then for C IV. The other narrow absorptions are probably interstellar lines. I have identified the Mg II 2798 resonance doublet with the two lines barely, but definitely resolved. The Fe II 2343, 2382, and 2599 ground level lines also appear in the long wavelength IUE specta; these Fe II lines are all strongest lines of widely spread multiplets; I can't identify any of the weaker multiplet lines. In the short wavelength IUE spectra, I identify the Si II 1263, 1308, and 1531 ground level multiplets and the Al II 1671 and C II 1335 ground level multiplets. These lines are typical UV lines due to the interstellar medium (e.g., De Boer, Jura, \&~Shull 1987). Another narrow absorption is at about 1795. It seems too far to the blue to be due to the comparatively weak Si II 1814 multiplet, but there is a difficulty with the wavelength calibration that may allow this identification (see below). The absorption may, of course, be circumstellar. Except for the Mg II resonance lines, none of the absorptions are clearly resolved into component absorptions either due to the various multiplet lines or different interstellar media at different velocities in the Milky Way and M81. This is not surprising given the low resolution of the low resolution mode of the IUE and the fact that M81 has a small heliocentric velocity (-43 km/s; Tully 1988). The centroids of absorption features in the IUE spectra all seem to be shifted by varying amounts of order a few angstroms both to the red and the blue. This probably because a definitive wavelength calibration has not been done. High resolution IUE and HST spectra will probably allow a much better study of the interstellar lines. HST spectra were scheduled to be acquired on April~15 as part of the SINS program (Kirshner~\etal 1988). Since all the people connected with that observation have been away from CfA since then, I have had no information on what was acquired; presumably the HST observations will soon be subjected to study. The broad depression in the IUE spectra centered at about 2200 A is almost certainly due to the 2175 A bump in the interstellar extinction (e.g., Cardelli, Clayton \&~Mathis 1989; Mathis 1990). This bump is believed to be due to carbon grains possibly in the form of graphite. The bump is a regular feature of UV extinction and is present for all values of R_V. It has a FWHM varying between about 400 and 800 A. The depression in the IUE spectra seems narrower than this and has a breadth at the base line I draw of only about 400 A. The IUE spectra are particularly noisy in the 2000--2400 A region, and so the existence of the depression could be challenged. However, given the presence of the interstellar lines and the existence of foreground extinction (Burstein \&~Heiles [1984] report a foreground E(B-V) of 0.0375), an extinction depression is expected. I have tried to eliminate the depression by correcting the spectra using R_V=3.1 (the standard value for the diffuse interstellar medium) and extinction law of Cardelli~\etal (1989). The most satisfactory E(B-V) value I find is 0.10 and I estimate an uncertainty of 0.03. An E(B-V) value as high as 0.2 must I think be completely ruled out since correcting the spectra with E(B-V)=0.2 creates a prominent bump centered at about 2150 A. I hope people will quickly emend an errors I have made. I'd like to thank Jason Pun for some information on IUE observations. References \refpaper Cardelli, J. A., Clayton, G. C., \&~Mathis, J. S. 1989, ApJ, 345, 245. % extinction law. \refpaper Burstein, D., \&~Heiles, C. 1984, ApJS, 54, 33. % Galactic foreground extinctions. \refedited De Boer, K. S., Jura, M. A., \&~Shull, J.~M. 1987, in Exploring the Universe with the {\it IUE} Satellite, ed.~Y.~Kondo, W.~Wamsteker, A.~Boggess, M.~Grewing, C.~De~Jager, A.~L.~Lane, J.~L.~Linsky, \&~R.~Wilson (Dordrecht: Reidel), 485. % diffuse and dark clouds in the interstellar medium \refindent Kirshner, R.~P., Blades, J.~C., Branch, D., Chevalier, R. A., Fransson, C., Panagia, N., Wagoner, R.~V., \&~Wheeler, J.~C. 1988, {{\it HST} Proposal SINS: The Supernova INtensive Study} \refpaper Mathis, J. S. 1990, Ann. Rev. Aston. Astrophys., 28, 37. % on interstellar dust and extinction. \refbook Tully, R. B. 1988, Nearby Galaxies Catalog (Cambridge: Cambridge Univ.~Press). % catalog of nearby galaxies (the Tully catalog) with velocities % heliocentric and systematic and distances and more. \refindent Wheeler, J. C. 1993, SN~1993J Updates, Wed Apr 7 13:24:26 1993 \refindent Wheeler, J. C., \&~Clocchiatti, A. 1993, IAU Circ., No.~? % see also SN~1993J Updates, Wed Apr 7 23:22:27 1993 Notes on Blended Image and Possible Composite Spectrum Roberta Humphreys and Greg Aldering The image of the probable progenitor is defintely blended and the observed energy distribution must be a composite of two or more stars. The measured magnitudes of the progenitor object now cover the wavelength range from U to I and allow for the possibility of modeling the energy distribution. The recently reported I magnitude by Blakeslee and Tonry ( circ #5758) has been very important for tieing down the contribution from a red component. ( They reported V=20.8 and I=19.4) We have also measured UBV and R photographic magnitudes for several stars within about 20 arcsec of the SN in an attempt to get a realistic estimate of the reddening in that region. We have sufficient multi- color data on 3 of these objects to estimate their reddening. These 3 are 10 to 15 arcsec from the SN. We find Av values of .8, 1.0 and 1.5 mag. and adopt Av = 1 mag for fitting the observed energy distribution. Adopted magnitudes for the energy distribution: I= 19.4 , R= 20.0, V= 20.8, B= 21.8 , U= 21.7 mag. We decided to use V= 20.8 because of the many measurements of the progenitor at this brightness. We calculated a grid of different fits to the observed energy distribution for different combinations of supergiant spectral types using published colors and used a chi-squared test to select the best fits. A few acceptable fits included an O star or early B-type supergiant but because of the lack of Halpha emission we decided to rule them out ( see Richmond circ. # 5739 ). These fits are for two stars only. Of course more might be present. 1. For Av = 1.0 mag the range of acceptable combinations are: a. G8 I + B5 I 33% at V ( percent contribution of hot component at V) + B8 I 37% " + A2 I 48% " b. K2 I + B5 I 48% " + B8 I 53% " + A2 I 68% " c. K3 I + B5 I 57% " + B8 I 63% " + A2 I 80% " 2. As an example adopt the K2 star, we then derive for a true distance modulus of 27.6 mag: Mv Mbol Mv Mbol B5 -7.0 -7.8 + K2 -7.1 -7.7 B8 -7.1 -7.6 + K2 -7.0 -7.6 A2 -7.4 -7.6 + K2 -6.6 -7.2 Thus the red component ( which is probably most people's choice for the SN ) is a 4500K star with a luminosity appropriate for a star with an initial mass near 15 Msun. 3. If one wants a fit with no reddening, the results for the cool component aren't much different with Mbol about -7.2 but the blue component is much less luminous at Mbol of -5.5 mag. Needed: a good, deep CCD V image obtained under good seeing prior to outburst. We have good deep J and F plates under good seeing that show an association of stars 1 - 2 or more magnitudes fainter than the progenitor image within 10 arcsec. There are even a few faint images on our U plate. However our V plate ( IIaD ) doesn't go as deep. If we could get some V magnitudes for the same stars we would have the potential of measuring a C-M diagram for the association which would be very interesting. Anyone interested in a collaboration please get in touch with us. [Alex's editorial note: I think the evidence for a rather low reddening to SN 1993J is quite strong (weakness of 2200 A feature, weakness of interstellar lines, very blue color at maximum, shape of spectrum after maximum, etc.). Thus, I personally think that Roberta's low-reddening option 3 (see below) is more realistic than the high-reddening option, regardless of the derived reddening to other stars within 20" of SN 1993J.] >From Roberta Humphreys (roberta@aps1.spa.umn.edu) Hi Alex, Your points about the reddening problem in M81 are valid except that the 2200A feature cannot be used to set a limit or to define the extinction in the visual. It and the UV extinction relation are highly variable in our galaxy and the LMC and SMC have very different relations from the what we call the galactic UV extinction relation. Several years ago while working with UV spectra of some hot stars in M31 and M33 I found that the galactic UV extinction relation way over corrected the 2200A feature producing a bump in the energy distribution. The LMC relation worked much better. In subsequent work by Phil Massey both of these galaxies he has shown that a UV extinction laew like the LMC is the correct one for both M31 and M33! We don't have any idea what is appropriate for M81. While the visual/near-IR extinction extinction "law" appears to be very uniform in our galaxy and apparently in others not so for the UV. There are several stars even in our galaxy with high visual extinction and NO 2200 A feature. Not only is this strange but also can be very misleading without additional information. Regards, Roberta