(click on image for full resolution)
4/22/98 color composite image from
Håkon Dahle's supernova page. Used by permission.
Exploding Star Fans:
The analysis (my own) of SN 1998S, along with light curve(s) are presented. Needless to say this was an interesting event to catalog, and displayed some peculiar behavior that I have not witnessed for any event I have studied. The intial light curve will be attached to this message (.gif format), with a second curve (blueprint) following under separate cover. The latter will connect all mean value points to display a "cyclic" behavior that appears to have existed in the descending branch of the light curve. My heartfelt thanks are offered to all who had monitored this event, and provided the data to make this analysis possible. Thanks also to the individuals who offered their input and expertise. This material may be used freely by anyone who deems the material worthy of presentation.
Steve H. Lucas
The Bright Supernova 1998S In NGC 3877; An Identity Crisis?
ABSTRACT:
A visual light curve and "eyeball analysis" of the bright supernova in NGC 3877 (SN 1998S) discovered by Chinese professional astronomer Zhou Wan of the Beijing Observatory (BAO) is presented. The visual magnitude estimates were gleaned from the supernovae chat and alert forums of the VSNet (variable star network/Japan), the ISN (International Supernovae Network/Italy), French astronomers (Site Aude des Supernovae [tabs98s.html@www.visuanet.com]) courtesy of Jean-Marie Llapasset, and The Norweigan Astronomical Society, Variable Star Section, courtesy of Bjorn Granslo ( http://www.astro.uio.no/~bgranslo/sn1998s.html). Worldwide cooperation assisted in making this endeavor a meaningful one in that the extensive data on the rise and decay behaviour of this event could be cataloged. The bright supernova (SN) in NGC 3877 has displayed some interesting behaviour which is presented in the accompanying light curve.
The light curve is an assemblage of submissions in the Visual band. Observational estimates were compared to magnitude sequencing provided by various organizations. Initial information involving comparison magnitude sequencing was provided by B. Granslo in vsnet chat message #1324 which displayed information from a variety of modern reference sources. There was sparce data in other photometric bands, some will be touched on lightly or will not be addressed at all.
This event presented considerable full rise (~3.3 magnitudes) and decay information (~3.2 magnitudes). SN1998's signature, via a spectrograph (IAUC# 6830), indicated a type II-n event, however in viewing the light curve it appears to mimic a type Ia event (verses a visual mean trend), and is somewhat fainter than a mean SN type II-L (Linear) event. A subtle cyclic appearance, with a amplitude of ~0.5mv exists in the overall shape of the descending branch of the light curve and can only be speculated upon as to its true nature. (first implied on 4/19/98 by Dr. T. Kato in a VSNET message)
Of notable interest is a pre-maximum "glitch" (on the the rising branch of the light curve) which was observed in the visual, on 3/14/98 from various observers (six days before estimated maximum light [3/20/98]), and also in the V-band estimates 9.4 days before maximum light (3/10.6/98). Professional analysis will hopefully determine if this anomoly is real.
An analysis and discussion of various mean average trends are discussed, plus notable fluctuations during the events evolution. These are apparent on the light curve, and will be brought to light with some conjecture applied (by this author)(see Appendix A). It is the hope of this excercise to define the shape of the light curve and hence the event, rather than any detailed analysis of the data points....
key words: supernova, amateur observations, light curve, mean average templetes, analysis.
INTRODUCTION:
In VSNET alert message #1519 the following notice was posted, courtesy of Dr. Yamaoka: "According to IAU Circular 6829, Supernova 1998S was detected in NGC 3877 on Mar. 3 UT by BAO Supernova Survey. It was about 15.2 mag (unfiltered) on the discovery frame, and then rose to 13.5 mag (also unfiltered) on Mar. 4.3 UT. The location of SN 1998S is: R.A. =3D 11h46m06s, Decl. =3D +47o29'.0 (2000.0), which is 16" west and 46" south of the nucleus of the host galaxy. The spectroscopy shows a featureless continuum but probably with a broad H-alpha emission, which indicates that this SN is of type-II in a very early phase. The location of SN 1998S is embedded on the host galaxy. So the photometry is rather difficult; but it is very challenging and the results should be very useful....."
NGC 3877 is a ScII.2 type galaxy shining at 11.02 (Apparent blue magnitude in the Bt system corrected for galactic and internal absorption) and has an absolute magnitude of -20.35 (absolute magnitude in the Bt system). A supernovae at the estimated distance of NGC 3877(1)[~18.79 Mpc, or ~61M light years](Distance Modulus - 31.37 [Ho - 50+/km/sec/Mpsec]) indicate an estimated peak brightness for a type I event of ~11.57, and a type II event at ~13.57 (utilizing a -19.8, and -17.8 absolute value [author's estimate]). This event's estimated maximum of 11.9mv on 3/20/98 is then similiar, by observable inferrance, (observations) to a type Ia event?
The weighted mean observed recessional velocity for NGC 3877 is ~887(H/50+) which places this entity a bit closer to us than Cluster B of the Virgo Cluster (M49) which has a velocity of ~963+/-81. (1) and (2). Being that this event displayed a preliminary type II signature there is another consideration, by implication of the light curve, that 1998S MIGHT? be a bright linear type II event ~1.67 magnitudes brighter than expected. Reference (3), displays some similarities that exist with other bright type II events which have been studied and analyzed.
VISUAL OBSERVATIONS
The amateur/professional community began visual and photometric monitoring of this supernova (SN) almost immediately after its discovery. The initial estimates were provided by French observers (Site Aude des Supernovae [tabs98s.html@www.visuanet.com]), the Spanish Supernova Group M-1 (http://personales.jet.es/drodrig/sn98) (from private communication), and individuals associated with The International Supernovae Network/Italy and VSNET/Japan. The latter provided extensive magnitude estimates for this event (over 400). The light curve is a reflection of the abovementioned teams and their visual observational estimates and are displayed as a mean value per day.
On the light curve a separate set of information is rendered for the French observers, who (as far as I can determine) performed their own magnitude sequencing from their own photometric calibrations.
Detailed magnitude sequencing (VSNET message #1324) (with noted catalog magnitude uncertainty were taken from: "The Hipparcos Catalogue, Tycho Catalogue, Tycho Input Catalogue, and Guide Star Catalogues. The Colour indices (b-r) were based on the PPM USNO-A1.0 catalogue", were presented by Bjorn Granslo (Norwegian Astronomical Society - Variable Star Section). Visual estimates from various observers were also presented by Mr. Granslo and were made accessible on VSNET message #1344, ISN chat forums, sci.astro amateur, sci.astro research usenet groups and later on VSNET message #1376 (part 2).
The French observers presented magnitude estimates in both visual and photometric (Site Aude des Supernovae [tabs98s.html@www.visuanet.com]) modes. The visual (visuel) estimates were gleaned from a variety of observers. The photometric data was also presented with methods described under the section "Photometrie differentielle avec la Moyenne ponderee" in their web page, courtesy of Jean-Marie Llapasset (private correspondance) Only the visual estimates are presented on the light curve and are differentiated from other data by employing a separate notation.
Dr. Dalibor Hanzl representing observers from the Czech Republic posted comparison magnitude sequencing around the site of SN 1998S (vsnet-alert #1710), this data was obtained by CCD and V-band data reductions (using V filters) and are available for inspection at: (http://astro.sci.muni.cz/hotnew.html).
The bulk of the magnitude estimates were freely accessed from the VSNET and
ISN chat forums. Dr. Tachi Kato (VSNET) presented almost a daily rendering
of estimates that he and his collegues had enmassed. These estimates
thusly, are used primarily to construct the light curve of the bright SN in
NGC 3877 that will accompany this article. In addition daily averages were
periodically displayed, but where not used by this author for this analysis
due to several CCD based magnitude estimates, this light curve will reflect
observations in the visual mode only.
----------------------------------------
(Author's Note: It might be prudent at this time to mention that with the
consciousness of the participants in this endeavor, several comparison
magnitude sequences appear to have been provided. It becomes somewhat
difficult to pinpoint the exacting or most productive sequence for our
purposes. It is evidenced that scatter from magnitude estimations when
drafting up the light curve...are in evidence. So rather than evaluate this
data or question the methods used, I will concentrate on displaying and
commenting on the observed shape OF the light curve, and later (in this
article) discuss what it could all possibly mean? [SHL]).
----------------------------------------
THE EVENT (Analysis of)
Visual observations of this event began on 3/5/98 and ended on 5/26/98. Thusly, extensive monitorings on the rise and decay behaviour of this SN where accomplished.
The rapid rise of this event appears to have been halted about 3.2 days (-14 days before maximum light) after its discovery and shortly after the visual monitorings began, and displayed a more "gentle" slope upwards. This trend continued until -5.8 days (3/14) where a SUDDEN rise (0.36mv) from the trend was observed and cataloged. It might be noted that this "glitch" was not only observed by the VSNET/ISN team but also by French observers at about the same time frame. Also, V-band (VSNET/ISN data) photometry for this event indicated a similiar "glitch" occuring at -9.4 days (3/10) with a 0.22V increase from the normal trend. In both cases the sudden rise (glitch) lasted for ~2days, until the normal trend(?) once again continued upward toward maximum light. Was this glitch real? Three separate sources seem to indicate it was, however no explanation can be presented at this time....perhaps theorists/professional astronomers can provide information about this anomoly. Was it a beginning of a pre-maximum plateau? The visual observations then took a turn downward (fainter) by 0.16mv (0.06m in the V-band) for ~2 days (-2.8 days before maximum). At which point the visual trend continued to rise in an unprecedented manner toward visual maximum on 3/20/98 increasing 0.34mv in that period of time (2.8d).
Both the vsnet/ISN team and the French obsevers placed maximum light on ~3/20/98. The V-band estimates did not indicate a maximum, and only by observational inferrance (looking at the light curve) can we suspect a later phase for this to have occured (possibly 1-2 days after visual maximum).
The trend then began a sharp downward slope for 2.8 days (0.28mv), where a more linear or slowed posture began (3/23). This, more or less linear trend continued throughout the remainder of the evolution of the event, where at about day ~63+/-, magnitude estimates became sparce, and the event posture began "dropping off" (getting drastically fainter). Of minor notatation is the cyclic posture of the entire descending branch of the light curve. The initial (first) downward (fainter) "dip" began about day ~12+/- and lasted until ~day 27+/-. The extreme possibility that a very gentle plateau (first implied by Dr. T. Kato on a message posted to the VSNET chat forum on 4/19/98 from the observations) began to take shape at about day ~27 +/-, where the trend began an upward (brighter) swing is, at this time, pure speculation? The appearance of this cycle is so subtle that one would have to view the light curve "edge-on" to notice the minute variation. This upward departure lasted until about day ~53 +/-, where the trend once again began to "drop off" (get fainter). No explanation for this phenomenon is offered at this time, it would be fair to assess that this cyclic trend was independant of any personal equation, (scatter) in my humble opinion.... The full cycle deviation, or amplitude for the abovementioned event cycle is ~0.50mv. (Note: a "blueprint" light curve with full connections to the mean data points is provided in Appendix A, which also highlights the noticable cyclic data)
The visual type Ia mean average light curve of (4) was placed against the rise and decay slopes of this event. As one can observe, the shape of the event is in pretty good agreement with this slope. (Note: on the rising branch of the light curve, this author has arbitrarily adjusted the data point at the -1 magnitude level from -15.7d to -12.5d[see explantion in the "discussion section" analysis of SN 1998aq by this author]). The event then followed the decay trend of the mean curve, with some minor fluctuations.
It can be observed that the "slight plateau" (if thats what it was?) is evidenced when placed against the mean visual curve slope. At day ~53 the event trend "drops off" considerably (from the mean type Ia curve). till about day 67 +/-, where observational estimates ceased to be reported.
In the mean decay slope of (4) for a Linear (II-L) type II SN, another interesting development is apparent. The event slope follows this trend with some fluctuations throughout the entire monitoring due to the aforementioned cyclic appearance of the descending branch of the curve. The differance (in the trend, rather than individual data points) does not exceed 0.4mv at any one point to day ~59. After that point the separation increases as the estimates depart from the mean II-L slope into a more fainter posture. (Note: Is the sudden "drop off" in brightness estimates from ~day 53 onward real? In viewing SN's 1998aq, and 1998bu the same "drop off" occurs at or near the termination of submitted estimates. One educated guess, other than the host galaxy sinking near the horizon, would indicate that the event began more difficult to observe as it blended into the light of the host entity).
EPILOGUE
This event certainly was an interesting one to catalog. The variations in the shape of the light curve were challenging to say the least. Was this event a type II-n as was originally characterized? But why the possible decay behaviour similiar to a type Ia SN? And the near linear response similiar to a type II-L event? Was a subtle plateau beginning for this event? Or other manifestation? Truly the theorists will have a field day with this event based on the visual light curve. But I imagine detailed professional analysis will yield information on what I judge from the visual observations to be a very peculiar event.
For instance in (5) the researchers indicate in (6) that a subclass called SN IIB-L "have a progenitor and explosive mechanism more similiar to SNIa, rather than to II-P. Both IIL and Ia are thought to result from the deflagration of C-O degenerate core of relatively low mass stars, the only differance being that the II-L left a small hydrogen envelope at the time of the explosion". Did a residual envelope exist to influence the observations? With all things being equal (or not) the data begs the question: "SN 1998S!...what do you want to be, when you grow up?....
Many, many thanks to all observers who contributed in monitoring this event. And to the individuals who communicated their input and expertise on many issues.
Steve H. Lucas (77620.1721@ compuserve.com)
International Supernovae Network
Completed August 20, 1998
APPENDIX A
A second light curve is presented for the event 1998S. This blueprint version is a more analytical depiction of the mean average data mentioned in the above text analysis, and will contain connections to the mean average points (type Ia, and II-L). Also indicated is the "cyclic" trend as noticed by this author in the descending branch of the light curve. The type II-L connection decay curve will act as the high (brighter) side amplitude depiction, while an "eyeball" decay linearity curve will act as the low (fainter) side amplitude depiction of the observed ~0.50mv variation that occured during this cycle. The data points that were associated with the light curve and the event (rather than the trend of the event) where connected as closely as possible to display the cycle of what could be called an interesting phenomenon.
Also indicated is the 20-day magnitude decay value (v/20m) for this event-(1.06mv).
REFERENCES:
1.)A Revised Shapley-Ames Catalog of Bright Galaxies; Sandage and Tammann(1981).
2.)STUDIES OF THE VIRGO CLUSTER, by Binggeli, Sandage and Tammann; The Astronomical Journal, Vol. 90, No.9, Jan., 1984-1985,)
3.)"Model Light Curves of Linear Type II SNe" by Swartz and Wheeler, Preprint #129, Dept. of Astronomy and McDonald Observatory, Jan 10, 1991.
4.) Doggett and Branch, AJ. _90_, 11 Nov., 1985.
5.) E. Cappellaro et. al A&A,_293_723-732 (1995).
6.) T.R Young and D. Branch, ApJ_342_,L55 (1989).
Back to Bright Supernovae
David Bishop