The Perth Automated Supernova Search Andrew Williams, BSc. (Hons) This thesis is presented for the degree of Doctor of Philosophy of the University of Western Australia Physics Department, 1997 (Accepted 1998)

i

Abstract: An automated search for supernovae in late spiral galaxies has been established at Perth Observatory, Western Australia. This automated search uses three low-cost PC-clone computers, a liquid nitrogen cooled CCD camera built locally, and a 61-cm telescope automated for the search. The images are all analysed automatically in real-time by routines in Perth Vista, the image processing system ported to the PC architecture for the search system. The telescope control software written for the project, Teljoy, maintains open-loop position accuracy better than 30" of arc after hundreds of jumps over an entire night. Total capital cost to establish and run this supernova search over the seven years of development and operation was around US$30,000. To date, the system has discovered a total of 6 confirmed supernovae, made an independent detection of a seventh, and detected one unconfirmed event assumed to be a supernova. The various software and hardware components of the search system are described in detail, the analysis of the first three years of data is discussed, and results presented.

We find a Type Ib/c rate of 0.43 +/- 0.43 SNu, and a Type IIP rate of 0.86 +/- 0.49 SNu, where SNu are ‘supernova units’, expressed in supernovae per 1010 LBŒ galaxy per century. These values are for a Hubble constant of 75 km.s-1 per Mpc, and scale as (H0/75)2. The small number of discoveries has left large statistical uncertainties, but our strategy of frequent observations has reduced systematic errors – altering detection threshold or peak supernova luminosity by +/- 0.5 mag changes estimated rates by only around 20%. Similarly, adoption of different light curve templates for Type Ia and Type IIP supernovae has a minimal effect on the final statistics (2% and 4% change, respectively).

ii

Contents: 1. INTRODUCTION ..................................................................................................................................1 1.1 SUPERNOVA CLASSIFICATION AND MECHANISMS .....................................................................................4 1.2 HISTORICAL SUPERNOVAE.......................................................................................................................7 1.2.1 SN185..............................................................................................................................................7 1.2.2 SN1006............................................................................................................................................8 1.2.3 SN1054 (Crab Supernova)..............................................................................................................8 1.2.4 SN1181............................................................................................................................................8 1.2.5 SN1572 (Tycho’s Supernova) .........................................................................................................9 1.2.6 SN1604 (Kepler’s Supernova) ........................................................................................................9 1.2.7 Cas A...............................................................................................................................................9 1.3 OTHER SUPERNOVA SEARCH PROGRAMS ...............................................................................................10 1.3.1 Zwicky ...........................................................................................................................................10 1.3.2 Asiago and Sternberg ...................................................................................................................11 1.3.3 New Mexico – Digitized Astronomy Supernova Search................................................................11 1.3.4 Evans visual search ......................................................................................................................11 1.3.5 Berkeley ........................................................................................................................................12 1.3.6 Supernova Cosmology Project......................................................................................................12 1.3.7 MSSSO search for supernovae in Abell clusters...........................................................................13 1.4 REASONS AND STRATEGIES FOR THE SUPERNOVA SEARCH .....................................................................13 1.5 HISTORY OF WORK ON THE SEARCH .......................................................................................................17 2. SYSTEM HARDWARE.......................................................................................................................19 2.1 TELESCOPE AND DOME .........................................................................................................................22 2.2 OPTICAL COUPLER .................................................................................................................................28 2.3 CCD CAMERA .......................................................................................................................................29 2.3.1 CCD Basics...................................................................................................................................30 2.3.2 Perth CCD camera .......................................................................................................................33 3. SYSTEM SOFTWARE ........................................................................................................................37 3.1 TELESCOPE CONTROL ............................................................................................................................38 3.1.1 Teljoy structure .............................................................................................................................38 3.1.2 The user interface layer ................................................................................................................40 3.1.3 The ‘DetermineEvent’ layer..........................................................................................................41 3.1.4 The real-time layer........................................................................................................................42 3.1.5 Coordinates, Flexure, and Tracking.............................................................................................44 3.1.6 Performance .................................................................................................................................45 3.1.7 Conclusions...................................................................................................................................47 3.2 CCD CAMERA AND OPTICAL COUPLER CONTROL SOFTWARE ................................................................47 3.3 INTERPROCESS COMMUNICATION ..........................................................................................................49

iii

3.4 SCHEDULING AND DATABASES ..............................................................................................................51 3.5 IMAGE ANALYSIS...................................................................................................................................52 3.5.1 Pre-processing ..............................................................................................................................53 3.5.2 Spatial matching ...........................................................................................................................54 3.5.3 Scaling ..........................................................................................................................................54 3.5.4 Supernova search..........................................................................................................................55 3.5.5 Automated photometry..................................................................................................................56 4. DATA REDUCTION............................................................................................................................58 4.1 THE DATA .............................................................................................................................................59 4.2 LIGHT CURVES ......................................................................................................................................63 4.3 PHOTOMETRY ........................................................................................................................................66 4.4 ANALYSIS ..............................................................................................................................................68 4.5 RESULTS ................................................................................................................................................72 4.6 ANALYSIS OF SEARCH TECHNIQUE .........................................................................................................76 4.7 CONCLUSION .........................................................................................................................................83 5. CONCLUSION .....................................................................................................................................86 6. REFERENCES .....................................................................................................................................89 APPENDIX A - TARGET GALAXIES ..................................................................................................92 APPENDIX B - CCD IMAGES.............................................................................................................103 APPENDIX C - DISCOVERY IAU CIRCULARS..............................................................................110

iv

Figures: FIGURE 1-1 BEFORE AND AFTER IMAGES OF SN1994AI IN NGC 908 ............................................................3 FIGURE 1-2 SUPERNOVA TYPES .....................................................................................................................5 FIGURE 2-1 PERTH-LOWELL AUTOMATED TELESCOPE ................................................................................20 FIGURE 2-2 SYSTEM SETUP .........................................................................................................................21 FIGURE 2-3 ELEVATED DOME FOR THE PERTH-LOWELL TELESCOPE ............................................................24 FIGURE 2-4 CCD CAMERA OPTICAL COUPLER .............................................................................................26 FIGURE 2-5 OPTICAL COUPLER AND CCD CAMERA ON THE PERTH-LOWELL TELESCOPE ............................27 FIGURE 2-6 CCD CROSS SECTION ................................................................................................................31 FIGURE 2-7 TWO DIMENSIONAL CCD READOUT STRUCTURE ......................................................................32 FIGURE 2-8 CCD RESPONSE CURVES ...........................................................................................................34 FIGURE 2-9 CCD READOUT ELECTRONICS ..................................................................................................36 FIGURE 3-1 TELJOY INTERNAL STRUCTURE AND PROGRAM FLOW ................................................................40 FIGURE 3-2 FLOW DIAGRAMS FOR SOFTWARE AUTOMATION .......................................................................51 FIGURE 4-1 RADIAL VELOCITY DISTRIBUTION IN THE ESO/UPPSALA CATALOGUE .......................................60 FIGURE 4-2 RA DISTRIBUTION OF IMAGES ...................................................................................................61 FIGURE 4-3 IMAGES TAKEN PER MONTH ......................................................................................................62 FIGURE 4-4 DAYS BETWEEN IMAGES ...........................................................................................................63 FIGURE 4-5 SUPERNOVA LIGHT-CURVE TEMPLATES ....................................................................................66 FIGURE 4-6 SN1994AI PHOTOMETRY IN R AND I .........................................................................................67 FIGURE 4-7 DAYS BETWEEN SEARCH IMAGES ..............................................................................................71 FIGURE 4-8 INCLINATION DISTRIBUTIONS ....................................................................................................75 FIGURE 4-9 OVERSAMPLING RATIO, BY IMAGE ............................................................................................79 FIGURE 4-10 OVERSAMPLING RATIO, BY GALAXY .......................................................................................80 FIGURE 4-11 SUPERNOVA SEARCH COVERAGE MAP.....................................................................................83

v

Acknowledgements Many people in the Physics Department at the University of Western Australia have provided invaluable help during the project. I would never have succeeded without the aid of many of the folk down in ‘the dungeon’, from whom I borrowed countless tools and asked countless questions. The skilled staff in the mechanical and electronic workshops also deserve congratulation for the fine work they have done on the hardware for the project – especially Derek Newman and Graeme Warburton, who built the CCD dewar and optical coupler. In addition, I must mention the University Computer Club – a great place to spend lunchtimes, and to absorb a wealth of information. I picked up most of the programming and electronics knowledge I needed to complete this project from assorted friends amidst the noise, dirt, and confusion of the UCC over the last 10 years.

I would also like to thank all of the staff of Perth Observatory, for their friendship and advice, and Ralph Martin in particular, whose work, especially on the image processing software, made the supernova search project possible. As well, I would like to thank the Perth Astronomy Research Group for sponsoring this project to begin with and promoting astronomical research in Western Australia – in particular, David Blair, Rolf Koch, and Ron Burman.

This research has made use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, CalTech, under contract with the National Aeronautics and Space Administration. I also thank the Astronomical Data Center (ADC) and the National Space Science Data Center through the World Data Center A for Rockets and Satellites, for providing catalogue data on the ADC ‘Selected Astronomical Catalogs’ CD-ROM. This project has also benefited from Australian federal funding, including three years of ARC small-grant support and a DITAC travel grant.

Most of all, I would like to thank my family for all their support and encouragement over the years, without which this work would never have been completed.

Dedicated to Freda Jacobs vi

1. Introduction If we were able to watch a galaxy like our Milky Way over a time span long enough to see its rotation, we could hardly miss seeing the importance of supernovae. Even the most conservative current estimates for supernova rates have over a million supernova explosions occurring in the Milky Way during each slow galactic revolution. Every one comes close to outshining the combined light from the billions of other stars in the Milky Way, at least for a brief instant. Even by more ephemeral, human time scales, supernovae are still amazingly frequent – up to 1000 supernovae per second are estimated, out to a redshift of z=2 (Blair, 1997). Despite this, details about supernova mechanisms, progenitors, rates, and distribution are still hotly debated, and only around 1000 supernovae in total have ever been discovered.

The reasons for this lack of knowledge are varied. One problem is that supernovae are transient events, visible for only a few months or days before fading below detectability. In any given galaxy, they occur very rarely – current estimates range from one to ten supernovae per century for a typical galaxy. In addition, it is difficult to use supernovae discovered ‘accidentally’ during other observations in any estimate of supernova frequency – how do you estimate how many galaxies were observed, and for how long, for each accidental discovery?

An extra complication is that observations of supernovae over the last hundred years have revealed an increasing number of sub-classes, all with varying light curve shapes, and with peak luminosities ranging over 5 magnitudes or more. In general, supernovae discovered before a subdivision (like the distinction between Ia and Ib/c events) cannot be adequately classified. This means any rate estimate using supernovae discovered before the date the sub-classification was recognised will be a weighted average of two, often very different, rates. To make things worse, the weight of each sub-rate used to find the average will have been a complex function of the selection effects biasing the discovery probabilities of the two sub-classes. For example, bright Type Ia supernovae occur much less often than the dimmer Type Ib and Ic supernovae, but because of their higher luminosity, are much easier to detect. Because of these selection effects, the early estimates of the merged ‘Type I’ rate were far too low. The same problem occurs with 1

other averaged properties, like light curve shapes and peak luminosities, in that the more easily detected objects dominate the averaging process. More common, but less easily detected sub-classes are dismissed as rare, anomalous events.

This thesis describes the establishment of an automated search for supernovae in the southern hemisphere. It has been designed to allow a low-cost search system to be built, allowing detection of supernovae without a large investment of human resources. Frequent observations are used to avoid systematic bias in the final result due to unknown parameters, and to increase the chances of early supernova detections for further study. Late spiral galaxies (Sbc-Sd) have been targeted as they have the highest supernova rates. Limiting the search to these galaxies results in more supernova discoveries, and also removes the need to divide the supernovae into separate galaxy type ‘bins’ and increase statistical error.

This system has been built on a very low budget – total hardware costs over the seven years of development and observations are only around US$30,000, and include automation of an existing 61-cm telescope and construction of a liquid nitrogen cooled CCD camera for imaging. The system takes and analyses images completely automatically in real time, but still requires human intervention at the beginning and end of the night (dewar filling, flatfields, etc.). The image processing and control systems can also be used for automatic photometric imaging in parallel with the supernova search. So far, the system has discovered six confirmed supernovae, and detected two additional events – one without external confirmation, but assumed to be a supernova, and SN1994AD independently detected here after its discovery. Before and after images of one of our discoveries, supernova SN1994AI in NGC 908, are shown in Figure 1-1, and the rest of the supernova discovery images are in Appendix B.

2

Figure 1-1 Before and after images of SN1994AI in NGC 908 These images were taken during the Perth Automated Supernova Search. Both are subregions of 180second unfiltered CCD images using a 576x384 pixel CCD with 22µm pixels

3

1.1 Supernova classification and mechanisms Supernovae are split into two main classes, based on their spectra near maximum light. These are the ‘Type I’ events, that show no evidence of hydrogen in their spectra, and the ‘Type II’ events that do show hydrogen lines. SNe I are found in all galaxy types – including those without evidence for recent star formation – and therefore must have old, highly evolved progenitors. SNe II occur only in star-forming galaxies, and presumably have young, massive progenitors. Other types (III, IV, V) were once used to classify supernovae using various spectral features – some only had one or two known members – but are no longer used. Instead, the presence or absence of hydrogen now defines the class (Type I or Type II) and other features define subclasses.

This split into subclasses, and the current classification system for SNe, is now rather complex. The Type I class has been split into Ia, Ib, and Ic events, based on the presence of Si II and He I lines in the spectra, and the Type II class has been split into IIP (‘Plateau’) and IIL (‘Linear’) events based on the light curve shape. Upper case letters are used to indicate classification by photometric, not spectroscopic, features, although recently distinctions between IIP and IIL spectra near maximum light have been indicated. Table 1-1 gives a summary of supernova classes, and Figure 1-2 shows the history of type-splitting for the classifications currently in use.

Table 1-1 Supernova types, in order of increasing envelope thickness Type Spectral Features Probable Mechanism Ia Si II Ignition of white dwarf after accretion or No H or He coalescence. Ic No H, He, or Si Core collapse in massive star with no H and thin or missing He envelope. Ib He I Core collapse in massive star with no H No H or Si envelope. IIL Strong H Core collapse in massive star with thin H envelope. IIP Strong H, with P-Cygni profile Core collapse in massive star with thick H envelope.

4

Figure 1-2 Supernova types This figure illustrates the history of the current divisions between supernova types, and indicates the progenitors believed to be involved. Only supernova types currently in use are shown.

The Type Ia events are believed to be thermonuclear explosions of C-O white dwarfs, caused either by a coalescence of a binary pair of white dwarfs, or accretion onto a white dwarf from a giant companion. Type Ia’s are believed to make good ‘standard candles’ and have very uniform properties, with a few easily recognised exceptions (Branch et al., 1995).

5

The rest of the SN types (Ib, Ic, IIP and IIL) are thought to be core-collapse events in massive stars. The core exceeds the Chandrasekhar limit and collapses to a neutron star or black hole. The rest of the stellar material initially falls in, but ‘bounces’, forming a shockwave that, unassisted, would run out of energy before reaching the stellar surface. The shockwave is prevented from stalling by heating from neutrinos emitted by electron capture in the core, and the re-heated shockwave tears the star apart.

For a core-collapse event, after shock breakout and initial rapid cooling, the light curve and spectrum is dominated by the outer envelope – a thick hydrogen envelope results in an extended plateau in the light curve (a Type IIP) as the hydrogen shell expands outwards (increasing the black-body emission radius) and cools (becoming transparent and moving the photosphere inwards through the ejected envelope). If these two effects balance, there is a long period of essentially constant luminosity. When the photosphere reaches the inner regions where

56

Ni and

56

Co generated in the explosion are mixed,

there is an extension of the plateau, or even a secondary maximum, as the radioactive decay heats the material. After this, the light curve shape is an exponential tail from the 56

Ni -

56

Co -

56

Fe decay, modified by changes in the transparency as the capture

efficiency for the gamma photons from the decay is reduced (Eastman et al. 1994, Blanton et al. 1995).

The other SN types (IIL, Ib, and Ic) are thought to result from core-collapse in stars that have lost all or most of their envelopes. The IIL events are SNe where the H shell is fairly thin, not leading to a light curve plateau. The Ib and Ic events are from stars where all or most of the H and H/He envelopes respectively, are missing, either from normal mass-loss in massive stars or from Roche-lobe overflow in a binary system (Woosley et al. 1993, 1995).

There is a good chance that some or all of the core-collapse SN types form a continuum rather than distinct classes, as the envelope thickness of the progenitor is varied. Several supernovae have been observed to change their spectral type as they evolved. SN1987K, for example, changed from a Type II to a Type Ic spectrum, probably indicating a very thin envelope. Some intermediary SN have been observed, and given a new class, IIb (e.g. SN1993J) – they seem to have a combination of the II and Ib properties. There is also another class of objects (IIn) that seem to be SNe embedded in a massive stellar 6

outflow, presumably occurring in stars that are losing mass rapidly. These events are distinguished by narrow-line P-Cygni profiles (hence the ‘n’ designation) and some of these events are strong radio sources. The narrow lines are probably due to ‘clumps’ in the outflow. Some of the Type IIn light curves decline very slowly.

Because classification of supernovae has changed so much, existing databases of SNe, and estimates of SN rates, must be treated cautiously. Older supernovae classified as ‘Type I’ could have been Ia, Ib, or Ic events. Later SNe classified as ‘Type Ib’ could have been either Ib or Ic events. These events are often described as ‘Type Ibc’ or ‘Type Ib/c’ because of the similarity between the two. Also, differences between IIP and IIL spectra have only recently been recognised, and are often not definitive, so most are simply described as ‘Type II’ unless long-term light curve photometry has been taken.

1.2 Historical Supernovae Supernovae have been observed and written about by humans as far back as the second century AD. Chinese records include descriptions of ‘guest stars’, that appear, are visible for some time, then fade away. Most of these are probably common novae, but a few of the ‘guest stars’ described are almost certainly supernovae. The historical events considered as supernovae in our galaxy are described below. Details have been taken from Schaefer (1995, 1996), van den Bergh (1993), Bethe (1990), and Stephenson (1971).

1.2.1 SN185 The earliest event considered as a supernova was reported in 185 AD, on December 7th. It was described as being near the stars α and β Cen, and as disappearing “... in the sixth month of the Hounian year.” This phrase was originally translated as “year after next”, giving a long, 20-month duration that implies a supernova. A supernova this bright (apparent magnitude brighter than -11) should have been recorded world-wide, not just by the Chinese. More recent analysis (Schaefer 1996) suggests a mistranslation occurred, possibly combined with an error in the records, giving a duration of only 6 months. With this shorter duration, a nova or comet is a possible explanation. Indeed, 7

the planetary nebula He 2-111 could well have been formed by a nova in 185 AD, as it is approximately 2000 years old, and is exactly in between α and β Centauri. Comet P/Swift-Tuttle, which disappeared in 188 AD is another explanation, possibly in combination with the nova. Schaefer now considers that the event in 185 AD was fairly unlikely to have been a supernova.

1.2.2 SN1006 This event occurred on May Day in the year 1006 AD, and was seen around the world, comparable with the moon in brightness. It remained visible for up to several years after peak light, and is believed to have occurred in the constellation of Lupus, associated with supernova remnant G327.6+14.6, at a distance of 1.59 +/- 0.13 kpc. This supernova is believed to have been a Type Ia supernova for reasons that include: • Iron absorption lines from the remnant are seen, consistent with the mass of iron expected in a Type Ia remnant. • No evidence of pre-supernova mass loss is seen, as the remnant is expanding into a thin interstellar medium. • The remnant is away from the galactic plane, so the progenitor was unlikely to have been a young, massive star.

1.2.3 SN1054 (Crab Supernova) The Chinese recorded a supernova in 1054 AD that has been identified with the Crab Nebula, a bright, easily visible remnant at a distance of about 2 kpc. No record survives of detection elsewhere, but Europe was not very politically stable at that time, and any records could have been lost. There is also a pulsar at the centre of the remnant – a neutron star, the core of the progenitor. This indicates that the supernova must have been a Type II or Ib/c event, to leave a neutron star. Hydrogen emission in the ejecta requires a Type II event. The pulsar is providing energy to the whole nebula, causing it to remain visible for a long period.

1.2.4 SN1181 This event was first seen around August 6th in the year 1181 AD by many observers in China and Japan, appearing in what we know as the constellation Cassiopeia. The 8

remnant of this supernova is believed to be the radio source 3C58, very similar to the Crab nebula. This makes it likely that SN1181 was also a Type II event.

1.2.5 SN1572 (Tycho’s Supernova) First seen in November of 1572, in the constellation of Cassiopeia, this supernova is reported as being as bright as Venus at peak light, and visible during the daytime. The event was recorded around the world, but Tycho Brahe made and collected detailed observations. The supernova remnant has been identified using Tycho’s position data. The supernova type is unknown – light curve, remnant position and structure are all ambiguous. Some evidence that it was not a Type II event exists, but it could have been either a Type Ia or Ib/c. Earlier arguments pointing strongly to a Type Ia event are now discounted. The distance to the remnant has been calculated at 2.35 +/- 0.20 kpc.

1.2.6 SN1604 (Kepler’s Supernova) Detected on October 9th, 1604, Kepler’s supernova was also detected around the world in the constellation of Ophiuchus. Like Tycho’s supernova, this event had accurate position information recorded, and the remnant has been identified at a distance of 3.4 +/- 0.3 kpc. The type of supernova is unknown – arguments about its nature include: • The remnant’s high space velocity could imply a young ‘runaway’ progenitor, hence a Type Ib/c or II event. However, the space velocity seems to be too high to be accounted for in a young star, implying an old high velocity halo star progenitor and a Type Ia mechanism. • There is evidence of pre-supernova mass loss forming a circumstellar shell that the remnant is currently expanding into, indicating a Type II or Ib/c event. • There seems to be insufficient quantities of iron in the ejecta for a Type Ia explosion.

1.2.7 Cas A The supernova responsible for the remnant in Cassiopeia known as Cas A, a strong radio source, does not appear in the historical record as having been detected by contemporary observers. It is sometimes described as SN1670, with the year derived from the approximate remnant age. Analysis of the remnant points to a core collapse

9

explosion (Type II or Ib/c), and the scarcity of hydrogen lines in the remnant’s spectra indicates a progenitor with a thin or missing hydrogen shell, either a Ib/c or a IIL event.

In addition to these events, various other nebulae in the Milky Way have been identified as supernova remnants, and dated with varying degrees of precision. It is likely that some of the other ‘guest stars’ listed in Chinese records were supernovae, but precise identification is difficult. There were also a handful of extragalactic supernovae detected between 1885 and 1930. The most famous recent supernova was, of course, SN1987A in the LMC, which was visible to the naked eye. Analysis contributed greatly to the understanding of supernova mechanisms, and the remnant will be studied intensely for decades to come.

1.3 Other supernova search programs Several consistent, organised searches for supernovae have been undertaken in the past 60 or so years, and these, combined with accidental discoveries, have led to a total of over 1000 known supernovae (1035 to October 1996). Some of these supernova searches are described below:

1.3.1 Zwicky The first organised search for supernovae began almost exactly 60 years ago, using an 18” Schmidt telescope located on Mount Palomar. It had only been a few years since the term ‘supernova’ had been coined, and no distinctions between supernova types had been recognised. Results of this first search were reported by Zwicky (1938), for the period from September 5th, 1936 to January 31st, 1938, during which time three confirmed supernovae were discovered. The search technique was to take 30-minute fields of galaxies and galaxy clusters (each field was 9.5º square), and visually compare each new plate with a previous reference image.

10

1.3.2 Asiago and Sternberg Two independent supernova search groups in Europe are at the Asiago Astrophysical Observatory (Italy) and the Crimean Station of the Sternberg Institute of Moscow (Russia). Both started around the same time (1959 for the Asiago search, 1961 for the Sternberg search), and because of their similarities and the improved statistics from a larger search sample, these groups co-operated in 1993 to release combined supernova rates from the two searches (Cappellaro et al. 1993a, 1993b). They had found totals of 32 and 21 supernovae for the Asiago and Sternberg searches respectively to that date. Both searches use photographic plates or films, with visual comparison to reference images.

1.3.3 New Mexico – Digitized Astronomy Supernova Search Feasibility studies for this project began in 1965, and the initial version of the search system was complete by 1973. This was the first completely automated supernova search, using surplus military equipment to build a computer controlled telescope in the Magdalena Mountains in New Mexico. Imaging systems used were Orthicon, Isocon, and Vidicon image tubes, which had fairly low readout noise for the time, but have since been superseded by CCDs. An IBM 360/44 computer was used for telescope control and image processing, which was done in real time. In 1973, soon after the software was working intermittently, NSF funding was terminated and the IBM 360 was retired as obsolete. New funding in 1977 allowed the purchase of a PRIME 300 computer and the project was re-opened, requiring a complete software re-write. By February 1987 it began regular observations for a time (Pearce 1987), but it is no longer in operation.

1.3.4 Evans visual search The Rev. Robert Evans, in Coonabarabran, NSW, Australia, has been conducting a visual search for supernovae for many years. Using a 25-cm telescope up to 1986, and a 41-cm telescope after that, he has spent 15-25 hours per month visually observing 300 to 800 galaxies looking for supernovae. Evans’ detection limit ranges from around mag 14 to mag 15, depending on the instrument used and observing conditions. Evans has discovered a total of 28 supernovae to date – 12 with the 25-cm telescope, and 16 with

11

the 41-cm telescope after 1986. Rates from the Evans search were reported by Evans et al. (1989), and van den Bergh (1993).

1.3.5 Berkeley The automated supernova search project was begun in 1980. By 1986 the system had discovered its first supernova, operating in a semi-automatic mode. This early version of the search system used a 76-cm telescope at Leuschner Observatory controlled by computer, with a 512x320 pixel CCD camera writing images to video tape. These tapes were examined later manually by visual comparison with reference images. Four supernovae were found by this semi-automatic search between over 1986 and 1987, and by 1988 the search system was fully automated, using new control and analysis computers. This automatic detection software scaled, aligned and convolved each test image, then subtracted a reference image of the field. The search was known as the ‘Berkeley Automatic Supernova Search’ (BASS). Any supernova appeared as a residual star in the subtracted image. The first supernova detection by a fully automatic search was made in 1988. The statistics for the first three years of fully automatic operation were presented in 1992 (Muller et al.).

After some period without funding, the telescope began being used again for supernova searching using new control software written for the ‘Berkeley Automatic Imaging Telescope’ (BAIT), a 0.5-m telescope at the same site used for general-purpose remote and automatic observing. This new search was known as ‘Leuschner Observatory Supernova Search’ (LOSS). Automated use of both telescopes was later abandoned in favour of development of a new 30-inch telescope known as KAIT (‘Katzman Automatic Imaging Telescope’), recently completed and soon to be used for supernova searching and other observing projects.

1.3.6 Supernova Cosmology Project This project’s goal is to use distant Type Ia supernovae (redshift z ≥ 0.35) to measure the cosmological parameters Ω and Λ. Run by the Lawrence Berkeley National Laboratory in collaboration with many other groups around the world, the project uses the Cerro Tololo 4-meter telescope in Chile to detect supernovae in distant galaxies, identifies them with spectra from the Keck 10-meter telescope in Hawaii, and takes 12

photometric data from Kitt Peak (Arizona), Apache Point (New Mexico), the Canary Islands, and the Anglo-Australian Telescope. Over 28 supernovae have been discovered so far, 11 of them in the first 48-hour 1995 detection run. The supernova are found by taking deep images of thousands of galaxies during new moon, then re-observing them at the next new moon and looking for new objects. (Perlmutter et al., 1997)

1.3.7 MSSSO search for supernovae in Abell clusters This search uses the MACHO (Massive Compact Halo Object) gravitational microlensing search system at Stromlo (see, for example, Alcock et al., 1996) to look for supernovae in Abell clusters. Designed to detect medium distance supernovae (redshift 0.02 < z < 0.08), this search fills in the gap between nearby supernova searches and the high-redshift programs. So far, 16 supernovae have been discovered. The goal is to use the Type Ia supernovae as distance estimators to check independent distance estimates to the clusters, and to study the spectral and photometric evolution of all the suitable supernovae.

1.4 Reasons and strategies for the supernova search There are a host of reasons why supernova searches are valuable, both because of the intrinsic value of supernovae for study (supernova mechanisms are still poorly understood), and because accurate knowledge of the rates is important in a variety of other astrophysical problems. Because supernova discoveries are both intrinsically valuable and useful for determining SN rates, any search for supernovae must strike a balance between optimisation for early and nearby supernova detection, and increasing the number of discoveries. Early (pre-peak) discovery of supernovae is important for accurate light-curve fitting, even for Type Ia events used only as distance estimators.

Differing goals leads to many different types of supernova search, ranging from a wide, shallow search (amateur astronomers world-wide would detect a moderately bright supernova in our galaxy within hours, even minutes) to a narrow, deep search like the survey of high-redshift galaxy clusters. Our system, by design and through use of available resources, covers nearby galaxies. It is unlikely to detect supernovae in our 13

galaxy, or the extremely distant Type Ia events useful for Hubble constant determination. Instead, we aimed for a moderate number of discoveries close enough (and discovered early enough) to be useful for further study, and enough discoveries in total to determine supernova rates.

Supernova mechanisms are still hotly debated in the scientific literature. Information about progenitor stars for the various events is poor, especially for Ia, Ib and Ic events. Analysis of supernova light curves, spectra, and radio emission helps to constrain explosion models and progenitors. The interaction of the shockwave with circumstellar material can, in principle, be used to view the previous mass-loss history of the progenitor in reverse, as the shockwave moves out through the slower moving material ejected before the supernova. This is especially important for the Type IIn supernovae, believed to be embedded in a very dense, clumpy circumstellar material created by very strong mass loss from the progenitor.

One of the most pressing problems is explaining differences between Type Ia events – do the atypical (sub or super luminous) Type Ia events form a continuous spectrum, or discrete types, and what distinguishes them? In addition, the progenitors of the Type Ib and Ic events are still poorly defined – it is fairly certain mass loss has occurred, but it is not known whether that mass loss is the result of strong stellar outflow in extremely massive stars, or Roche-lobe overflow in binary systems.

Observations of supernovae are important not only for direct use in explosion and progenitor studies, but also to refine parameters necessary for other work. For example, light curves of various supernova types are important for determining supernova rates. It is important to know the average peak luminosities, average light curve shapes, and amount of random variation in both, for supernovae in a given class. As this information has improved over the last few decades, supernova rate estimates have changed by an order of magnitude or more.

Supernova rates themselves are important in many different fields. Supernovae are the only source of heavy elements, and act to enrich the interstellar medium with these elements for incorporation into new stars. The chemical evolution of galaxies depends strongly on the supernova rate, and knowledge of this rate is important, along with its 14

variation across different host-galaxy Hubble classes. Studies of pulsar evolution and behaviour are obviously interlinked with supernova rates, since pulsars are formed in core-collapse supernovae. Supernovae may also act to trigger star formation, and are certainly important in the dynamics of the star forming regions, where massive stars generate core collapse supernovae while star formation is still going on elsewhere. Supernova rates are also of interest to cosmic ray astronomers, since they are sources of cosmic ray acceleration. In addition, both neutron star coalescence and supernovae are important proposed sources for gravitational radiation, and accurate knowledge of the rate of both events in the Milky Way is essential.

Supernovae can also be valuable as distance indicators to help pin down the value of the Hubble constant. Type Ia supernova are understood to be homogenous, as a class, and can be used as ‘standard candles’ with a few recognisable exceptions. They are bright enough to be used as a distance measure at high redshifts. Nearby Type II supernova are also useful as distance estimators, though not through use as ‘standard candles’ – they vary too much as a class. Instead, the fact that they approximate black-body sources during the early stages of the light curve is used to determine their distance using a technique commonly known as the Expanding Photosphere method (see, for example, Schmidt et al. 1994). The angular radius of the object is derived using the luminosity and an assumption of black-body behaviour, while the actual radius is derived by extrapolating the radial velocity information from spectral lines superimposed on the mainly black-body spectrum.

The aim of the Perth search was not just to find as many SNe as possible, but to try to establish a system able to obtain supernova rates with less dependence on poorly known parameters than other existing rate determinations. The strategy has been to look at nearby galaxies and to repeat observations frequently, with observations generally separated by only a few days. To maximise the number of supernovae found, we chose to look only at late spiral galaxies (Sbc-Sd), because other SN searches have found that these galaxies produce more supernovae (Tammann et al. 1994). While this is an assumption based on existing SN statistics, it will not bias our final result, which is an independent estimate for SN rates in late spiral galaxies. Limiting our search to late spirals removes the need to split supernova discoveries into host galaxy class ‘bins’, and

15

thereby increase statistical errors. Also, rates can be applied to the Milky Way, a late spiral galaxy, albeit of debatable Hubble class.

The first classification of supernovae was the split into ‘Type I’ and ‘Type II’ events, and since then, each of those types has been further divided. Almost all of the SN type splitting started out with one or two ‘anomalous’ low luminosity events that later turned out to be fortuitously discovered samples of a new type, as common (or more common) than the original. Using existing SN statistics to calculate new data can lead to circular logic. For example, early supernova searches assumed that all type I SNe were equally bright and visible for long periods, so galaxy observations were widely spaced. This led to discovery of far more of the high luminosity SNe Ia than other (then indistinguishable) class I subtypes. The result was a final ‘Type I’ rate estimate that was approximately that of the bright, Type Ia events, instead of the sum of all of the Type I subclass rates.

These strong selection effects show that a supernova search must be structured so as to reduce dependence of the final statistics on initial assumptions. It is tempting to use existing supernova rate and luminosity estimates to maximise the number of supernovae discovered, by spacing observations as far as possible to avoid ‘overlap’. Doing this will bias the search towards confirming the initial assumptions, rather than discovering any new results.

Taking frequent observations also leads to early discoveries, an important goal in itself. The hardware available at Perth is obviously not suitable for finding the high-redshift supernovae most valuable as distance estimators. Discovering a smaller number of supernovae near maximum was thought to be better than finding a larger number of medium-distance supernovae weeks or months after maximum. It also gives a rate estimate with less dependence on light curves and other poorly known parameters.

16

1.5 History of work on the search The supernova search operates under the umbrella of ‘PARG’, the Perth Astronomy Research Group. PARG is composed of members from Perth’s universities, and Perth Observatory (an independent institution owned by the State Government), and holds regular meetings to discuss ongoing astronomical research, and plan new programs.

The earliest work towards setting up the supernova search began around 1986, when initial work began on adapting the image processing package Vista, using photographic plates digitised using a video camera and ‘frame grabber’ to test image processing techniques. From that time to the present, there have been three people responsible for working on the various components of the supernova search. These are Malcolm Evans, a UWA Computer Science student; Ralph Martin, an observatory employee and parttime Curtin University Physics MSc. student, and myself. In addition, technical staff at the Observatory designed and built all of the telescope and dome automation hardware, and the mechanical workshop staff at the University of WA built the dewar for the CCD camera. The progress of the search and contributions of the people involved are summarised in Table 1-2.

Table 1-2 Work on supernova search components 1988-1996 1 deg. Communicated by A. Beresford. G. Pizarro (European Southern Observatory). 1.0-m Schmidt. Rough position by R. M. West. 7’ coma, two wavy tails in p.a. 145 deg, length 3 deg. P. Bouchet, S. Benetti and others confirmed the comet, which was visible with the naked eye; 2 deg tail. J. Drummond (Gisborne, New Zealand). Communicated by W. Orchiston.

SUPERNOVA 1995W IN NGC 7650 A. Williams and R. Martin, Perth Observatory, report the discovery on Aug. 5.65 UT by the Perth Astronomy Research Group’s automated supernova search of an apparent supernova (mag about 16) located 9" west and 22" south of the center of NGC 7650 (R.A. = 23h25m.3, Decl. = -57d48’, equinox 2000.0). J. Greenhill and K. M. Hill, Canopus Observatory, University of Tasmania, write: "Instrumental CCD magnitudes measured with the Canopus Observatory 1-m telescope on Aug. 10.6 UT show that the object was 1.0 mag fainter in v and much hotter (b-v = -0.26) than the closest bright star, located about 1’.6 south and west of the supernova. S. Benetti and T. Augusteijn, European Southern Observatory; and A. Sarajedini, Kitt Peak National Observatory, communicate: "Inspection of a preliminarily-reduced CCD spectrum (range 396-793 nm, resolution 1 nm) obtained on Aug. 17.29 UT with the Danish 1.54-m telescope (+ DFOSC) at La Silla confirms this object as a supernova of type II, about 1 month after maximum. Strong H gamma, H beta and H alpha lines with P-Cyg profiles are superimposed on a fairly blue continuum; Ti II 455.0-nm and Fe II 501.8-nm and 516.9-nm lines are also present with P-Cyg profiles. The expansion velocity deduced from the H gamma and H beta absorption is about 8000 km/s. We derive a velocity for the H alpha absorption of about 9600 km/s. The Fe II lines show a somewhat slower expansion velocity (about 5800 km/s). The velocity shift of a narrow H alpha emission component is consistent with the recession velocity of NGC 7650 (3300 km/s)."

1995 August 18

(6206)

Brian G. Marsden

113

Circular No. 6351 Central Bureau for Astronomical Telegrams INTERNATIONAL ASTRONOMICAL UNION Postal Address: Central Bureau for Astronomical Telegrams Smithsonian Astrophysical Observatory, Cambridge, MA 02138, U.S.A. [email protected] or FAX 617-495-7231 (subscriptions) [email protected] or [email protected] (science) Phone 617-495-7244/7440/7444 (for emergency use only)

SUPERNOVA 1996N IN NGC 1398 A. Williams and R. Martin, Perth Observatory, report their discovery, during the Perth Astronomy Research Group’s automated supernova search on the Perth-Lowell 0.61-m reflector, of a supernova (R about 16) on Mar. 12.538 UT. SN 1996N is located 46" east and 12" north of the center of NGC 1398 (R.A. = 3h38m.9, Decl. = -26o20’, equinox 2000.0). Nothing was visible at this location on Feb. 16. L. Germany and B. Schmidt, Mount Stromlo and Siding Spring Observatories; and R. Stathakis and H. Johnston, Anglo-Australian Observatory, report that a spectrogram (range 550-900 nm) obtained on Mar. 23.4 UT at the Anglo-Australian Telescope shows SN 1996N to be a type-Ib/c supernova, about 2 weeks past maximum.

SUPERNOVA 1996B IN NGC 4357 L. Wang and J. C. Wheeler, University of Texas at Austin, report: "SN 1996B was observed again using the Imaging Grism Instrument at the 2.1 m telescope of the McDonald Observatory on Mar. 20 UT. Inspection of a preliminarily-reduced CCD spectrum (range 424.7-813.5 nm, resolution 2.0 nm) shows that the H-alpha feature that was strong in our Jan. 22 spectrum (IAUC 6292) has now decreased in strength sharply. The residual H-alpha feature is flat-topped (with FWHM about 10~000 km/s) and resembles that of SN 1993J in its late nebular phase. A weak absorption feature is still detectable at the blue side of the flat-topped H-alpha emission feature. There are also indications of the existence of the various helium lines that were detected in SN 1993J, but the spectrum is too noisy to establish this firmly. The redshift of the host galaxy, as measured from the neighboring H II region, is 4440 km/s. SN 1996B is therefore likely to be another type-IIb supernova similar to SN 1993J. Observations in other wavelengths are encouraged."

COMET 29P/SCHWASSMANN-WACHMANN 1 Further total visual magnitude estimates (cf. IAUC 6320): Feb. 24.39 UT, 11.2 (C. S. Morris, Lockwood Valley, CA, 0.26-m reflector); Mar. 8.85, 11.1 (K. Hornoch, Lelekovice, Czech Republic, 0.35-m reflector); 13.29, 10.8 (R. Keen, Mt. Thorodin, CO, 0.32-m reflector); 17.25, 10.9 (Morris); 23.19, 11.3 (A. Hale, Cloudcroft, NM, 0.41-m reflector). (C) Copyright 1996 CBAT 1996 March 23 (6351) Daniel W. E. Green

114

Circular No. 6558 Central Bureau for Astronomical Telegrams INTERNATIONAL ASTRONOMICAL UNION Mailstop 18, Smithsonian Astrophysical Observatory, Cambridge, MA 02138, U.S.A. [email protected] or FAX 617-495-7231 (subscriptions) [email protected] or [email protected] (science) URL http://cfa-www.harvard.edu/cfa/ps/cbat.html Phone 617-495-7244/7440/7444 (for emergency use only)

SUPERNOVA 1997Z IN NGC 3261 R. Martin, A. Williams, and S. Woodings, Perth Observatory, report the discovery of a supernova (mag R about 15.5) from a CCD frame taken on Feb. 10.64 UT. SN 1997Z is located 58" east and 10" south of the center of NGC 3261 (R.A. = 10h29m.0, Decl. = -44o39’, equinox 2000.0). S. Benetti and M. Turatto, European Southern Observatory (ESO); and E. Cappellaro, Osservatorio di Padova, communicate: "Inspection of a reduced CCD spectrogram (range 380-800 nm, resolution 1.1 nm) obtained on Feb. 12.12 UT with the ESO/MPI 2.2-m telescope (+ EFOSC2) confirms SN 1997Z as a type-II supernova, about 100 days after explosion. In the spectrum, several lines with typical P-Cyg profiles are superimposed on a relatively red continuum. Most of the lines arise from H I, Na I, Fe II, and Sc II. The expansion velocities (in km/s) deduced from the minima of the strongest lines are 7200 (H-alpha), 6150 (Na I); 4800 (H-beta), and 3550 (Fe II). With the same instrument, we obtained the following magnitudes for the new object: B = 19.0, B-V = +1.4, and V-R = +1.3. The offsets from the galaxy nucleus are 55".1 east and 8".7 south."

V1333 AQUILAE A. M. Levine, Massachusetts Institute of Technology (MIT), and B. Thomas, Goddard Space Flight Center (GSFC), on behalf on the RXTE ASM team at GSFC and MIT; M. Garcia and P. Berlind, Smithsonian Astrophysical Observatory; and P. Callanan, University College Cork, report that the low-mass x-ray binary Aquila X-1 (= V1333 Aql) appears to be in a moderate-intensity outburst in quicklook results from the RXTE All Sky Monitor. Recent daily-average flux levels in the band 2-12 keV are as follows: Jan. 27, about 60 mCrab; 28, 120; 29, 190; 30, 260; Feb. 1, 310; 2-5, about 360. Observations of the optical counterpart V1333 Aql from the Whipple Observatory 1.2-m telescope on Feb. 6.543 UT yield V = 16.9. Observations at other wavelengths are encouraged.

COMET 46P/WIRTANEN Total visual magnitude estimates: Jan. 3.09 UT, 12.6 (A. Hale, Cloudcroft, NM, 0.41-m reflector); Feb. 1.75, 11.8 (M. Meyer, Frauenstein, Germany, 0.25-m reflector); 4.09, 11.3 (R. Keen, Mt. Thorodin, CO, 0.32-m reflector); 6.74, 11.4 (K. Sarneczky, Raktanya, Hungary, 0.44-m reflector). (C) Copyright 1997 CBAT 1997 February 12 (6558) Daniel W. E. Green

115