According to Phillips et al. (1999), the decline rate parameter for SN 1991bg is Delta m_15(B) = 1.93 (0.10). This makes it one of the fastest declining Type Ia supernovae known. It is the prototype of the subclass of fast declining Type Ia SNe.
At the January, 1993, meeting of American Astronomical Society in Phoenix/Tempe Alain Porter and his colleagues presented a poster on SN 1991bg. Here is their abstract (Porter et al. 1992):
JHK Photometry and the Bolometric Light Curve of SN 1991bg by A. C. Porter, M. Dickinson, S. A. Stanford, E. A. Lada, G. A. Fuller, and P. C. Myers Bulletin of the Amer. Astron. Society, vol. 24, p. 1244 (1992) (meeting 181, abstract number 76.07) We have obtained infrared images of the unusually subluminous Type Ia supernova 1991bg in NGC 4374 = M84 on 5 nights in December 1991 and January 1992. The telescopes and instruments used were the 1.3m KPNO reflector with the SQIID detector, and the 2.1m KPNO reflector with the InSb IRIM detector. The December photometry confirms the low luminosity at maximum, and the January observations confirm the rapid fading of this object. Normal SNe Ia have a second infrared maximum that keeps them near their maximum J magnitude for a month. In contrast, SN 1991bg faded by 3 J magnitudes in the first month after the explosion. We have constructed a bolometric light curve of SN 1991bg by combining our infrared photometry with optical photometry and spectrophotometry (Filippenko et al. 1992, AJ 104 1543, and Leibundgut et al. 1993, AJ in press). We are comparing this light curve to the analytic models of Woosley, Taam, and Weaver (1986, ApJ 301 601) to find interesting constraints on the (56) Ni mass, or the ejected mass, produced in the explosion. We also discuss the similarity of SN 1986G in Cen A and SN 1971I in NGC 5055 to SN 1991bg, and the disturbing implications of such objects for attempts to use SNe Ia as standard candles.Sadly, Alain Porter died October 10, 1993, of thyroid cancer, shortly before his 35th birthday. The full, refereed paper related to the abstract reproduced above was never published.
The December 1991 data were taken by Adam Stanford, Mark Dickinson, and Peter Eisenhardt as part of a separate program on infrared observations of galaxies. The paper based on that observing run is Stanford et al. (1995). They used the 1.3-m telescope at Kitt Peak, and an instrument called SQIID, which provided simultaneous imaging in the near infrared J, H, K, and L' bands. They did not save the L' frames. SQIID had four PtSi (platinum silicide) 256 X 256 arrays. While the quantum efficiency was low (6.6 percent in J, 5.5 percent in H, and 3.4 percent in K), these arrays were four times larger than the then-available 62 X 58 InSb arrays. With the 1.3-m telescope SQIID gave a field of view larger than 5.5 arcmin on a side. The plate scale was 1.36 arcsec per pixel. (Because of the large plate scale, the images were considerably under-sampled. This eliminates the option of PSF photometry.)
In the (northern) spring of 2004 Mark Phillips discovered that Adam Stanford had on Exabyte tape the raw data files from December 1991, including the SN 1991bg imagery. I have recently heard from Elizabeth Lada that she did indeed observed SN 1991bg for Alain Porter. Perhaps she will be able to find the images on tape.
Because of the importance of Type Ia supernovae to modern extragalactic astronomy, and because of the rarity of the fast decliners, I thought I would make available what I have derived from the raw imagery obtained by Stanford et al. For other IR data on fast decliners, see observations of SN 1986G by Frogel et al. (1987), and the ApJ-paper-in-press on SN 1999by by Garnavich et al. (2004); this paper without the IR data was originally made available as astro-ph/0105490. Last year we obtained many nights of optical and IR imagery of the fast decliner SN 2003gs using the ex-2MASS 1.3-m telescope at Cerro Tololo, but we have not yet reduced the IR data.
For optical photometry of SN 1991bg see Leibundgut et al. (1993). In particular, note their chart of the field stars near the SN. They find that "star A" has V = 14.59 (0.04) , B-V = 1.42 (0.03), where values in parentheses are one-sigma error bars. For "star J" they find V = 12.31, B-V = 0.38, but with uncertainties of 0.3 and 0.4 mag, respectively.
The first thing we needed to do with the SQIID data was to characterize the performance of the instrument. When Stanford et al. took their data they knew of the non-linearities in the arrays, so they intentionally defocussed the camera for observations of the bright Elias et al. (1982) IR standards. Some of their observations of the standards were also taken in focus. So that gives us a chance to double check the non-linearity corrections. When the SN and out of focus standards were observed, typically less than 2700 counts were obtained, so no non-linearity corrections had to be applied. For "star A" and "star J" near SN 1991bg we needed to apply the corrections, as they typically gave 10,000 and 20,000 counts, respectively.
The best way to derive the non-linearity corrections is to use their frames taken on December 17, 1991 (UT), where were integrations taken of the inside of the dome in the middle of the night. They took exposures of 5, 10, 20, 40, 80, 160 sec for JHK, and also 320 and 640 sec for some J exposures, and 320 sec for the H-band. Using the shortest exposures we can determine the rate at which the counts increased (plotting total counts vs. exposure time) and use those slopes to predict what we would have detected in the longer exposures if there were were no non-linearities of response. The bottom line is that at 10,000 counts, we need to correct the instrumental magnitudes by about 0.02 mag (making them brighter), while at 20,000 counts the correction is about 0.05 mag. Take a look at this graph. The blue dots are J-band data, the green squares are H-band data, and the red triangles are from K-band exposures.
As a check of the accuracy of our non-linearity corrections, let us consider J-band measures of HD 1160 made on December 15, 1991. Four out of focus images give a mean instrumental magnitude of 15.310 (0.011). The in focus images give a mean instrumental magnitude of 15.338 (0.013), but with non-linearity corrections obtained from the maximum counts in the raw frames, we get a corrected mean instrumental magnitude of 15.308 (0.013) for the in focus images, in excellent agreement with the value from the out of focus images.
December 15, 1991, was apparently a photometric night at Kitt Peak. Stanford et al. observed standards of Elias et al. (1982) on 10 occasions. This gives us the option of deriving the extinction values and judging the constancy of the zeropoints over the course of the night. From the observations of the first 8 standards I find a J-band extinction of 0.121 (0.028) mag/airmass; the K-band extinction similarly derived was 0.090 (0.011) mag/airmass. From the observations of Gl 105.5 at three airmasses I derived an H-band extinction of 0.055 mag/airmass. These are three very sensible values of the near-IR extinction.
I confirmed that there are no color terms that need to be taken into account to derive good J- and K-band magnitudes with SQIID, but there is a substantial color term for H. If we consider a transformation of the following form:
H (standardized) = h (instrumental) - extinction * airmass + (color term)* (H-K) ,Stanford (private communication) found a color term of -0.23. (Their published value was -0.21, however.) I derived a color term of -0.261 (0.053) from observations of the first 8 standards observed on December 15 UT.
There appears to have been a slight zeropoint change (0.02 to 0.03 mag) prior to the 9th standard observed that night. Then SN 1991bg was observed, then the 10th and last standard. The best possible zeropoint for the purpose of the 91bg observations is obtained using the 9th and 10th standards. I note that the 9th and 10th standards and the 91bg field were observed at almost the same airmass (1.1 to 1.2), so any errors due to extinction corrections would be negigibly small.
I find the following values for "star A" and "star J", applying my derived values of extinction and my color term for the H-band magnitudes. For comparison I also give the 2MASS values:
Object J H K J(2MASS) H(2MASS) K(2MASS)
star A 11.803 11.164 11.002 11.782 11.136 10.901
(0.010) (0.008) (0.028) (0.035) (0.037) (0.034)
star J 10.822 10.585 10.574 10.867 10.598 10.490
(0.005) (0.005) (0.011) (0.036) (0.036) (0.035)
You will note that there is excellent agreement for J and H, but
that 2MASS values of K are systematically brighter. I shall
stand by my K-band magnitudes. Consider that Gl 105.5 had a B-V
color of 1.19 according to Gliese's (1969) catalogue. Its H-K
color is 0.110 according to Elias et al. (1982). We find H-K
= 0.162 for "star A" near 91bg, rather than the 0.235 given
by 2MASS. Recall that the B-V color of "star A" is 1.42, a bit redder
than that of Gliese 105.5, but not very much redder.
Also, if we can believe that B-V ~ 0.4 for "star J"
(at least we know it is not a red star) according to Leibundgut
et al. (1993), then it should have H-K ~ 0 (as we observed), not 0.11 as given
by 2MASS. Thus, the 2MASS values of the K-band magnitudes might
be systematically too bright.On the night of December 15, 1991 (UT), we can tie the brightness of SN 1991bg directly to the 9th and 10th standards observed. For the record, those were HD 105601 and HD 106965. For the nights of December 16th and 17th UT (which were not photometric) we can calibrate the SN using "star A" as the principal comparison star, with "star J" as a check star. We obtain the following results:
Infrared Photometry of SN 1991bg
1991
UT date Julian Date J H K sky conditions calibrator(s)
Dec 15 2448606.05 13.488 13.445 13.490 photometric HD 105601/106965
(0.024) (0.038) (0.052)
Dec 16 8606.06 13.509 13.440 13.407 non-photometric "star A"
(0.024) (0.022) (0.052)
Dec 17 8607.05 13.595 13.71 barely there non-photometric "star A"
(0.042) (0.15)
I would not really trust the H-band measurement of December 17th. Note
its very large uncertainty. I also note that this photometry is based on aperture photometry in a 3 pixel radius for J and H, and in a 2 pixel radius for K, including r = 3 --> 6 or r = 2 --> 6 pixel aperture corrections based on "star A" and "star J". The sky annulus for the SN 1991bg photometry had an inner radius of 7 px and a "width" of 4 px. Thus, it is possible we have included a small amount of light of the host galaxy in the r = 2 or 3 px apertures containing SN 1991bg.
Garnavich et al. (2004) give absolute magnitudes at maximum for SN 1999by of M_H = -17.87 (0.24) and M_K = -17.79 (0.25). Tonry et al. (2001) give a distance modulus for NGC 4374, the host of SN 1991bg, of m-M = 31.32 (0.11). On December 15, 1991 (UT), SN 1991bg had absolute magnitudes of M_J = -17.87, M_H = -17.90, M_K = -17.84. I have used the Schlegel et al. (1998) value of E(B-V) = 0.041 along the line of sight to NGC 4374, the host of SN 1991bg. The corresponding near-IR extinctions are A_J = 0.036, A_H = 0.024, and A_K = 0.014 mag, using the reddening law of Cardelli, Clayton, and Mathis (1989). As NGC 4374 is an elliptical galaxy, I assume that the host galaxy reddening is zero.
According to Leibundgut et al. (1993), the date of V-band maximum was December 14.7, or JD 2448605.2. Thus, the Stanford et al. observations correspond to 0.85, 1.86, and 2.85 days after the time of V-band maximum. If SN 1991bg were like SN 1999by, its B-band maximum would have occurred 2 days prior to T(Vmax). The SQIID observations were made roughly 3, 4, and 5 days after the time of B-band maximum. Garnavich et al. (2004) suggest that the IR maxima of SN 1999by occurred several days after T(Bmax), rather than 3 days before T(Bmax) like Type Ia supernovae of mid-range and slow decline rates. We give in the following table a comparison of the absolute magnitudes at maximum, or near maximum light, of Type Ia SNe of different decline rates.
Absolute magnitudes at maximum of Type Ia supernovae
Object(s) Decline rate M_J M_H M_K Notes
Delta m_15(B)
21 SNe 0.8-1.7 -18.60 (0.03) -18.27 (0.03) -18.43 (0.03) 1
SN 1999by 1.90 (0.05) < -17.70 (0.26) -17.87 (0.24) -17.79 (0.25) 2
SN 1991bg 1.93 (0.10) < -17.87 (0.11) < -17.90 (0.12) < -17.84 (0.12) 3
1. Based on the 16 SNe used by Krisciunas et al. (2004a) and five more from
Krisciunas et al. (2005).
2. We use the maximum observed brightness in the J-band, two days before the time
of B-band maximum (Hoeflich et al. 2002). It is difficult to extrapolate with
confidence to the J-band maximum, given the whole data set and the large gap in
the IR light curve. See Fig. 3 of Garnavich et al. (2004).
3. Based on the data 0.85 days after the time of V-band maximum.
We adopt the Tonry et al. (2001) SBF distance to the host.
We do not know exactly when the IR maxima of SN 1991bg occurred,
but suffice it to say that its absolute magnitudes near maximum
were very nearly equaly to those of SN 1999by, more than half a magnitude
fainter than Type Ia supernovae of mid-range and slow decline rates.If we consider the dereddened V-J, V-H, and V-K colors of SNe 1986G, 1991bg, and 1999by, along with the loci for mid-range decliners and slow decliners (Krisciunas et al. 2004b), we find that SN 1991bg not only one of the least luminous known Type Ia supernova in the infrared, but its intrinsic V minus near-IR colors are the reddest. Consider the following graphs. For SN 1986G we assume that A_V = 1.56 mag, as adopted by Krisciunas, Phillips, and Suntzeff (2004a). (This extinction value for SN 1986G has a large uncertainty, possibly as large as 0.4 mag, so the 86G points in the upcoming graphs might have systematic errors as large as 0.3 mag.) Scaling factors to V minus IR color excesses are given by Krisciunas et al. (2004b). For SNe 1991bg and 1999by we made small corrections for the reddening by dust in our Galaxy.
Another way to demonstrate the extreme redness of the fast decliners is to plot the pseudo-colors at maximum vs. the decline rate parameter. In the next plot we have taken the extinction corrected colors of the better sampled objects shown in Fig. 12 of Krisciunas et al. (2004b) and have added SNe 1999by (left pointing triangles) and 1991bg (right pointing triangles). SN 1986G has quite uncertain extinction corrections, as mentioned above, and we have left it off this plot. Note that the time of the near-IR maxima is typically 5 days prior to T(Vmax), so the blue dots in these plots are colors that do not correspond to an actual moment in time.
For SN 1999by we plot the extinction corrected colors 2 days prior to T(Bmax) for V-J and V-H, and 3 days prior to T(Bmax) for V-K. For SN 1991bg we plot the extinction corrected colors at 0.85 days after T(Vmax).
What should be done next? We shall see if we can re-reduce the January, 1992, data taken by Lada et al. According the BAAS abstract, the SN was 3 magnitudes fainter in J a month after maximum light. We need to reduce the IR data for SN 2003gs. According to Jose Luis Prieto, the decline rate parameter for this object is Delta m_15 = 1.84. Given how rare the fast decliners are, we need to squeeze all we can out of the data available.
Acknowledgments: I thank Mike Merrill for sending me the SQIID manual. I thank Mike Merrill, Adam Stanford, and Mark Dickinson for useful discussions about SQIID and the December, 1991, run. Thanks to Gajus Miknaitis, we were able to get the data off tape.
References:
Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, "The Relationship between Infrared, Optical, and Ultraviolet Extinction," Astrophysical Journal, 345 , 245
Elias, J. H., Frogel, J. A., Matthews, K., & Neugebauer, G. 1982, "Infrared Standard Stars," Astronomical Journal, 87 , 1029
Frogel, J. A., Gregory, B., Kawara, K., Laney, D., Phillips, M. M., Terndrup, D., Vrba, F., & Whitford, A. E. 1987, "Infrared Photometry and Spectroscopy of Supernova 1986G in NGC 5128 - Centaurus A," Astrophysical Journal, 315 , L129
Garnavich, P., Bonanos, A. Z., Krisciunas, K., et al. 2004, "The Luminosity of SN 1999by in NGC 2841 and the Nature of 'Peculiar' Type Ia Supernovae, Astrophysical Journal, in press (The original version of this article can be found on the astrophysics preprint server as astro-ph/0105490. It has been accepted for publication. The revised version contains infrared data from the Fred L. Whipple Observatory 1.2-m telescope.)
Gliese, W. 1969, Catalouge of Nearby Stars , Veroeffentlichungen des Astronomischen Rechen-Instituts Heidelberg, Number 22
Hoeflich, P., Gerardy, C. L., Fesen, R. A., & Sakai, S. 2002, "Infrared Spectra of the Subluminous Type Ia Supernova 1991by," Astrophysical Journal, 568 , 791
Krisciunas, K., Suntzeff, N. B., & Phillips, M. M. 2004a, "Hubble Diagrams of Type Ia Supernovae in the Near-Infrared," Astrophysical Journal, 602 , L81
Krisciunas, K., Suntzeff, N. B., Phillips, M. M., et al. 2004b, "Optical and Infrared Photometry of the Nearby Type Ia Supernovae 1999ee, 2000bh, 2000ca and 2001ba," Astronomical Journal, 127 , 1664
Krisciunas, K., Suntzeff, N. B., Phillips, M. M., et al. 2005, "Optical and Infrared Photometry of the Nearby Type Ia Supernovae 1999ek, 2001bt, 2001cn, 2001cz, and 2002bo," in preparation
Leibundgut, B., Kirshner, R. P., Phillips, M. M., et al. 1993, "SN 1991bg: a Type Ia Supernova with a Difference," Astronomical Journal, 105 , 301
Phillips, M. M., Lira, P., Suntzeff, N. B., Schommer, R. A., Hamuy, M., & Maza, J. 1999, "The Reddening-Free Decline Rate Versus Luminosity Relationship for Type Ia Supernovae," Astronomical Journal, 118 , 1766
Porter, A. C., Dickinson, M., Stanford, S. A., Lada, E. A., Fuller, G. A., & Myers, P. C. 1992, "JHK Photometry and the Bolometric Light Curve of SN 1991bg," Bulletin of the Amer. Astron. Society, 24 , 1244
Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, "Maps of Dust Infrared Emission for Use in Estimation of Reddening and Cosmic Microwave Background Radiation Foregrounds," Astrophysical Journal, 500 , 525
Stanford, S. A., Eisenhardt, P. R. M., & Dickinson, M. 1995, "Evolution of Infrared-Selected Galaxies in z ~ 0.4 Clusters," Astrophysical Journal, 450 , 512
Tonry, J. L., Dressler, A., Blakeslee, J. P., et al. 2001, "The SBF Survey of Galaxy Distances. IV. SBF Magnitudes, Colors, and Distances," Astrophysical Journal, 546 , 681
Go back to Kevin Krisciunas home page by clicking here.