AAVSO Supernova Search Manual

Part one: A brief survey of supernova searching to April 1993


The first supernova in modern times to be discovered visually was SN 1885A in M31 (the Andromeda galaxy), the brightest of the modern era until the advent of SN 1987A. Although its discoverer is subject to debate, it is usually attributed to Hartwig, a professional astronomer at the Dorpat Observatory. Its discovery was an accident, not part of any search effort.

The first visual discovery of a supernova by an amateur also occurred by accident. Jack Bennett of South Africa stumbled across an 11th-magnitude supernova next to the nucleus of M83 in NGC 5236 in July 1968. He was using a five-inch comet-seeking refractor. Supernovae that bright are rare; thankfully, Jack was alert enough to recognize that the galaxy did not "look right." This supernova was of Type II (plateau), and was studied for nearly two months, after which it could no longer be distinguished against the bright nucleus of the galaxy.

April 1979 witnessed the second amateur discovery, which must also be looked upon largely as an accident. Certainly, Gus Johnson, of Maryland, U.S.A., did a little supernovae searching, but it was only one of his many interests. Using his eight-inch Newtonian reflector, Gus discovered a supernova in the outskirts of M100 shortly before it reached maximum light. Rising to the 11th magnitude range, it was one of the prototype examples of the "linear" variety of Type II. This particular discovery was studied extensively, including ultraviolet and radio observations. With modern VLBI radio techniques, this supernova will probably become one of the major yardsticks for measuring the distance to the Virgo cluster of galaxies.

Between the two amateur discoveries, the AAVSO developed a supernova search activity as part of their nova search section, and observers like Ernst Mayer and Tom Fetterman made some galaxy observations. A small number of charts were also prepared. Also, James Bryan of Texas undertook his chart-making and observations of a handful of face-on spirals, publishing two articles about it. Like many amateurs, he lived and worked in a city at that time, and could not observe easily with great regularity. In spite of a poor observing site he made an independent discovery of the 1980 supernova in NGC 6946, some days after the professionals found it. This was also an 11th magnitude "linear" Type II, and was extensively studied.

Similarly, before 1981, Gregg Thompson of Brisbane, Australia, began making a substantial set of about 230 charts of bright galaxies. Steve Lee, a night assistant at the Anglo-Australian Telescope, had made enlargements from the old northern and new southern surveys. Thompson's charts were drawn on the basis of these photos, but showed stars at their visual magnitude as seen with a telescope. During 1980, some of the charts became available (in a rather primitive form) for Gregg and his friends to use. The ultimate aim was to publish them, and thus foster supernova hunting. The final product (1990) is far superior to these early versions.

Late in 1979, when Sky & Telescope magazine carried news of the supernova in M100 to Australia, I began once more to tackle the many obstacles which had brought my search to a standstill some years before. A phone call in search of help changed the course of the story rapidly. Arthur Page was then leader of the variable star section of the Astronomical Association of Queensland. He put me into contact with Gregg, who made some of the charts available to me. He also suggested that I get to know Tom Cragg at the Anglo-Australian Observatory at Siding Spring, Coonabarabran, Australia. Tom was both willing and able to check out my many false alarms, and he gave me the opportunity to start making my own photographs of galaxies, copied from the surveys. I began building a large collection of galaxy photos, both of galaxies covered by the charts, and of many others that I would want to observe. Naturally, it took me a lot of time, effort and money to learn how to make these negatives, and then to build the collection. It did not happen overnight, or even in one or two years. Through 1980 I slowly intensified my search, and steadily improved my "bits and pieces" portable ten-inch telescope. The year culminated with an independent discovery of SN 1980N in NGC 1316 (Fornax A), which had been discovered a day or two earlier by M. Wischnjewsky on a plate taken by J. Maza of the University of Chile.

But it was 1981 which proved to be the real watershed for organized, intensive supernovae searching by amateurs. Within a mere fifteen days, commencing on February 24, I made two outright discoveries, and missed the opportunity for a third. The two "hits" were SN 1981A in NGC 1532, a 13th magnitude star of Type II (plateau), and SN 1981D in NGC 1316 (Fornax A), a 12th magnitude classical Type I supernova found shortly before maximum. The "one that got away" was SN 1981B in NGC 4536.

Because Tom was so wary about the possibility of making a mistake with the first discovery, it was several days before he felt confident enough about SN 1981A to announce the new find. Also, the star varied in brightness in a way that we had not expected.

SN 1981D was completely different. Found with the portable ten-inch from the front driveway at Tom's home on Siding Spring Mountain, Coonabarabran, Australia, in just a few hours it was discovered, verified and reported to the Central Bureau for Astronomical Telegrams. As it happened, the Chilean professionals had been taking a number of photographs of their previous supernova in NGC 1316, SN 1980N. When they received the discovery notice about SN 1981D they reviewed their pictures to find that it was present on pictures made 9, 8, 7 and 6 days before discovery, showing the supernova rising from magnitude 20.5 up to 15.0.

The rest of 1981, and all of 1982, yielded no discoveries, although I did try hard to observe the supernova in NGC 1187. However, further improvements to the telescope, and an even wider range of galaxies on my observing list, gave more room for success in the future.

Eight more discoveries followed in 1983 and 1984. Three of these supernovae were of special interest, because they led to the recognition of a completely new class of supernova, first called "peculiar Type I", but which became known as Type Ib. Indeed, some believe that this group also includes another identifiable group called Type Ic.

The first of these three, supernova 1983N in NGC 5236 (Messier 83) was discovered over a week before it reached maximum brightness. It reached eleventh magnitude, and was very extensively studied. Observers with the Very Large Array radio telescope recorded a radio curve different from any previously observed object.

The appearance of an identical supernova the following year in NGC 991, SN 1984L, showed that these supernovae formed a class of their own, and were not simply peculiar objects. Especially, studies of the spectra of supernovae in this new class six months or more after maximum brightness show radical differences between the new class, and the classical Type I's.

These two supernovae, 1983N and 1984L, are the recognized prototypes of Type Ib.

Supernova 1983V in NGC 1365 was not studied so well, but became identified as an example of Type Ic, along with a small number of other objects which were found by the professionals.

Since that time, opinion has fluctuated as to whether Type Ib and Type Ic supernovae are two separate types, or are variations within the one basic type. Type Ib is deficient in helium, whereas Type Ic is not, and several other differences exist, although there are many similarities.

Large professional telescopes follow observing schedules which are arranged months in advance; it is not always easy for an observer to alter plans to accomodate the discovery of a transient object like a supernova. Often, the equipment attached to such a telescope is unsuitable for this purpose. But other times coincidence leads to golden opportunity. Important steps in our quest for knowledge can sometimes arise in this way. Dr. Marshall McCall happened to be using the Anglo-Australian Telescope on two such occasions; he was able to perform spectropolarimetry on two of the brighter supernovae, to determine whether the exploding shells were spherical or not (an important point in some theoretical studies). Type Ia supernovae appeared to be spherical, whereas the Type Ib's appeared not to be spherical. Another "chance" study led to the discovery of intergalactic gas clouds between the Milky Way galaxy and nearby galaxies in Centaurus.

Yet another study arising from an amateur discovery led to the first measurable evidence of iron in the late- time spectra of a supernova, made by observing infra-red spectral lines from a Type Ib about one year after maximum light. Several other supernovae showed unusual spectral features, which still await explanation.

Back in the 1960's, the Japanese amateur, Kaoru Ikeya, had made a name for himself by sharing the discovery of a very bright comet. By the 1980's, Ikeya had begun to change his interest to the visual search for supernovae. His first discovery took place in late 1984, when he found SN 1984R in NGC 3675. Regretably, this discovery took place at a time when no professional observers were able to, or were keen enough, to follow it up (a fate which befell a number of other objects), and so no spectra were made of this star, and its type is unknown. His second discovery, SN 1988A in NGC 4579 (M58), received much better treatment. Ikeya used a ten-inch reflector with an alt-asimuth mounting, situated at his home, which is in an area with bad light pollution -- his two discoveries are quite an achievement.

One of the most interesting visual discoveries was SN 1986G, in the dust lane of the nearby radio galaxy NGC 5128, which is commonly called "Centaurus A". This classical Type I supernova reached 11th magnitude although highly obscured by the famous dust lane. If not for this obscuration, it probably would have reached magnitude 8.5, making it one of the brightest of modern supernovae. Maximum brightness occurred over a week after discovery, which is fortunate, for the light curve of this supernova showed a much faster evolution than many others of Type Ia.

SN 1987A in the Large Magellanic Cloud (LMC) must also be classed, in part, as an amateur visual discovery. It certainly was found accidentally. Two of the three official discoverers saw it visually, and one was an amateur. In my opinion, the amateur, Albert Jones of New Zealand, should have received more notice for his part in the discovery than he did. Albert is a prolific and proficient variable star observer, and he was making estimates of a star in the LMC when he saw the supernova. Supernova 1987A has been studied at very great length, which has helped to rewrite our knowledge of supernovae in many ways, and will continue to do so for many years to come.

Later that year, in August 1987, Dana Patchick, of Culver City, California, found SN 1987L in NGC 2336 with his 17.5-inch Dobsonian telescope, as part of a concentrated search that he was developing. It was a classical Type I supernova, found some weeks after maximum light. Another recent supernova of outstanding brightness to be discovered visually was SN 1989B in NGC 3627, (Messier 66). It was discovered about a week before maximum light, and reached magnitude 11.8, despite being somewhat affected by dust clouds in M66. This supernova was extensively studied by many professional astronomers. It belonged to the Type Ia class of supernovae. It was also independently discovered by an Italian amateur, Federico Manzini.

SN 1991T in NGC 4527 was a discovery which perhaps shows the "coming of age" of amateur visual supernova hunting, because five amateurs from four different locations around the world were involved officially in its discovery. These included Stephen Knight and Wayne Johnson of the U.S.A., and Mirko Villi and Giancarlo Cortini of Italy.

Another Japanese amateur, Reiki Kushida, made her first visual discovery of a supernova in December, 1991, by finding a star of mag. 14.5 in the giant Virgo galaxy, M84. This was a Type Ia supernova, but with a number of peculiar features about it.

N. Brown, of Perth, Western Australia, is also listed as an official visual independent discoverer of SN 1992A in NGC 1380.

SN 1992H in NGC 5377 was discovered in February by William Wren, suing one of the smaller professional telescopes at McDonald Observatory. He works there in the visitor's center, and hunts for supernovae visually in his spare time, usually with his own telescope. He has now obtained some official observing time with one of these telescopes at McDonald Observatory. In this way, the Texas astronomers recognize the value of amateur visual searching.

Two more discoveries came my way later in 1992. They were both Type II (plateau) supernovae. One was in the central Virgo galaxy NGC 4411B. The other was in the far southern galaxy NGC 2082.

March 1993 saw the visual discovery of one of the brightest and most important of recent supernovae. Members of the "M1" group, which is part of the Madrid Astronomical Association, had the galaxy NGC 3031 (Messier 81) under heavy surveillance. The supernova appeared during a gap of only two days between observations, thus pin-pointing the time of explosion very well. Francisco Garcia Diaz made the lucky observation, and found the supernova. It was of Type II plateau, but had some very unexpected variations. This supernova was the brightest northern one for quite a number of years, and it received an enormous amount of attention from both amateur and professional observers.

Finally, a type Ia supernova of magnitude 13.6 was found by me visually on April 30 in the southern galaxy IC 5270. This galaxy belongs to the Gruz group which is being studied for distance measurements purposes.

To the best of my knowledge, the amateur visual discovery tally at the end of April 1993 is:

Robert Evans 26 (including 6 shared)
Kaoru Ikeya 2 (including 1 shared)
Jack Bennett 1
Francisco Garcia 1
Gus Johnson 1
Reiki Kushida 1
Dana Patchick 1
W. Wren 1
N. Brown (Perth) 1 (shared)
Giancarlo Cortini 1 (shared)
Albert Jones 1 (shared)
Stephen Knight 1 (shared)
F. Manzini 1 (shared)
Mirko Villi 1 (shared)

A number of observers have made independent discoveries, a few of which have been mentioned earlier in the story. The above tally includes only those discoveries which are officially recognized.

From the beginning of 1983 up until mid-1988, out of a total of 99 supernovae discovered worldwide, 18 have been discovered (alone or shared) visually by amateurs. Clearly, 18% is an enormous improvement on the contribution made by amateurs earlier in the story. With the advent of improved professional automatic searching, the amateur percentages may decline a little in the future. But an examination of the list of these amateur discoveries from 1983 to mid-1988 shows that they include many of the brightest and most important recent supernovae.


The great majority of supernovae discovered so far have been found photographically. The history of this type of search can be divided into several periods:

1885 - 1920 About a dozen supernovae were found during this period, mainly on Mount Wilson plates. No real progress was made in understanding their true nature.

1920 - 1933 Edwin Hubble established both the nature and great distances of galaxies, but there was a strange lack of interest in studying the novae which had appeared in the spiral nebulae, and which, therefore, must have been very much brighter than the other galactic novae.

1933 - 1942 Dr. Fritz Zwicky of the California Institute of Technology and Dr. Walter Baade of the Mount Wilson Observatory gave the name "supernova", when Zwicky began a systematic search for them. Not until the eighteen-inch Palomar Schmidt was put into operation after September 1936 did the search succeed. A few discoveries were made each year, and these established the basic classical knowledge of supernovae which stood until the mid-1980's. Rudolph Minkowski of Mount Wilson pioneered spectroscopy of supernovae.

1942 - 1950 The Second World War disrupted all work in this area. Only a few supernovae were found (by accident) during this period with large telescopes; no progress was made in understanding them.

1950 - 1975 Chiefly through the efforts of Zwicky, a co-operative search began to be organized in 1950, involving a number of observatories worldwide which either operated moderate sized Schmidt telescopes or intended to do so. A good number of brighter supernovae were discovered, credit shared among the various observatories involved in the search.

Palomar Observatory's great contribution had to await 1958, when the forty-eight inch Schmidt telescope became available for supernova hunting. Before then, the big Schmidt was used to make plates for the first Palomar Observatory Sky Survey. Unfortunately (for supernova studies at least) the observatory authorities did not allow these plates to be searched for supernovae promptly. Over fifty supernovae were eventually found by Zwicky and others on the Survey plates, and on rejected plates. But they were all "dead" before they were discovered, becoming little more than statistics. All that is known about them is that they existed.

After 1958, the Schmidt became a powerful tool in the search, revealing a sizeable fraction of all known supernovae. Many of these stars were important discoveries, but, naturally, many of them were very faint. Nearly all of the faint ones, and some of the bright ones, lacked spectra to determine their type, or to find out anything else about them. So, many of these discoveries became little more than statistics, also. What could have been learned from these discoveries was lost.

Zwicky retired in the late 1960's, and his search was taken over by Charles Kowal. In 1975, the big Schmidt was withdrawn from supernova searching, soon after Zwicky's death in 1974. This marked the end of an era in searching for supernovae. Throughout this period, a search of a few dozen spiral galaxies had continued with the eighteen-inch Schmidt, also, although very few supernovae were found. This searching was also discontinued a few years later.

During this general period, several observers at co-operating observatories ran up interesting lists of discoveries, in some cases continuing to the present day. Paul Wild of Berne Observatory can boast about 36 discoveries up to SN 1985A. M. Lovas of Konkoly Observatory logged 31 up to SN 1987M. L. Rosino of Asiago Observatory discovered about 20 before his retirement. Most of the tallies were at Palomar, however, where Milton Humason ran up a total of about 30 discoveries after his retirement from the Mount Wilson staff in 1959. Kowal has over 75 official discoveries, and may have many others which have not been officially recognized, and Zwicky himself can claim a total around 125.

First Amateur Photographic Discoveries

It was during this period, from 1950 to 1975, that the first amateur photographic discovery of a supernova was made.

G. Romano, a young Italian amateur astronomer who later became professional, made the first amateur discovery using the photographic mode. It is now known as SN 1957B, and it was found in NGC 4374 (Messier 84), reaching a maximum of about magnitude 12.5, and was of Type I. This galaxy is one of the largest and brightest of the elliptical galaxies in the central part of the Virgo cluster.

Romano subsequently made two other photographic discoveries, but these were both part of the professional supernova search being conducted at Asiago Observatory. So, they cannot be classed as amateur discoveries. The first of these was in a small elliptical galaxy in central Virgo, NGC 4564. This supernova was also of Type I, and reached a maximum brightness of magnitude 11.2, which means it was a clear magnitude brighter than its entire parent galaxy! This supernova was called SN 1961H. His other discovery was SN 1970O, which was a thirteenth magnitude supernova in an anonymous galaxy.

1975 to the Present

The period 1976 to 1979 was a very thin time with very few discoveries. With the close of the Palomar search it seemed that the urge to find supernovae had temporarily evaporated. From 1979 to 1984 J. Maza in Chile led the temporary operation of a new southern search. As with the Palomar search, the many very faint discoveries were not studied further, the few bright ones receiving most of the attention. The group reported 39 discoveries.

Since 1983 theoretical knowledge has been in constant flux. The classical distinctions between Type I and Type II are being modified, and may disappear. Type Ib has now been recognized, along with differences between "linear" and "plateau" Type II's. The appearance of SN 1987A altered the scene still further in a number of ways, and other supernovae with strange spectra still remain to be explained.

New digital spectrographs have made it easier to decipher the spectra of supernovae, and modern supercomputers enable us to construct theoretical models of the atmospheres of exploding stars, then to develop synthetic spectra to compare with those observed. As a further benefit, supernovae have now become important as possible cosmic distance indicators.

In the last few years, other observers have also run up large tallies of discoveries. These include Jean Mueller at Palomar Observatory, Christian Pollas in southern France, Robert McNaught at the Anglo- Australian Observatory, and a revived Chilean group using a telescope at Cerro Tololo. All are instances where access is possible to photographs made with a large Schmidt telescope.

More Amateur Discoveries

From 1983 to 1985, four supernovae were discovered photographically by amateurs in Japan: two apiece by K. Okazaki and S. Horiguchi. Okazaki's discoveries, made with a ten-inch Wright-type Schmidt telescope, were both shared with me, his second also shared with a Russian professional. They are SN 1983G (NGC 4753), and SN 1984E (NGC 3169). Horiguchi's two finds were SN 1985B (NGC 4045), and SN 1985G (NGC 4451). These four by the Japanese observers, added to the three mentioned previously by Romano, total seven amateur discoveries by photographic means which have been officially recognized.

The first two months of 1992 contained two amateur photographic discoveries. The first of these was SN 1992A in NGC 1380, made by Bill Liller in Chile. In some respects Bill is a professional astronomer, and his past history has certainly been in professional astronomy. But he also does much amateur work with modest gear from his backyard observatory. He has found a number of novae by the "Problicom" method devised by Ben Mayer. And this supernova discovery was part of that program. As already mentioned, this supernova was also discovered visually some hours later by N. Brown of Perth, Western Australia.

The second discovery was made by Shunji Sasaki of Japan, SN 1992G in NGC 3294. It was found at mag. 14., but it rose in brightness to mag 12.7 by the later part of February.

To the best of my knowledge, these seven supernovae are the only amateur photographic discoveries of supernovae to April 1993.


The story of electronic searching may be divided up as follows:

(1) Before the Second World War, Zwicky and a colleague from Radio Corporation of America (RCA) planned a search method which combined large Schmidt telescopes with live television detectors. This plan was to involve amateurs who would search the fields for supernovae, but it never happened.

(2) During the 1960's, J. A. Hynek and his colleagues, Powers and Dunlap, developed a twenty-four inch telescope with remote computer controls, and a television detector. Observers compared galaxies on the television screen with reference photographs. Fourteen supernovae were found this way, according to Stirling Colgate. This project lost its funding after only two years, although Dunlap's name appears in the list of supernova discoverers over a fifteen year period.

(3) Around 1970, Stirling Colgate began to work on a fully automatic system. The entire operation would be computer controlled, requiring only occasional human intervention. Many major problems in developing this system nearly defeated the project but it became operational in February, 1987. Despite many thousands of galaxy observations, no supernovae were discovered by June 1988. Major problems and equipment failures persist to this day.

(4) During the early 1980's, the Berkeley Automatic Supernova Search began to mount a totally automatic search operation. The frustrations and failures in this effort prompted a decision to modify the system so that the telescope produced a picture of the galaxy on a television screen, and an observer made the comparison with a computer file photograph. This time, however, the detector was a charge-coupled device (CCD).

The Berkeley group's first discoveries were made with this kind of system. Still, the group persisted with its ultimate aim. Before mid-1988 it was rewarded with several discoveries using the totally automatic mode. Their tally of discoveries reached a total of twenty during 1991. By 1990, the project had involved the expenditure of over five million dollars.

They bought a new robotic telescope, planning to install the telescope in a location like Hawaii, and process the pictures in Berkeley, probably by satellite. This telescope would be much faster than the first one, although it was optically of a similar size, and faster computers would also be used. The "remote" nature of this project has proved too difficult, and temporary defeat has been admitted.

So, the search with 76cm telescopes is in temporary recess, but the search for very faint supernovae with large telescopes continues. This search for very distant supernovae is a much more difficult task than many had realized, and discoveries will probably be few and far between. And this search will probably be much more costly, per discovery, that the 76cm search has been. Yet such a quest is of vital importance in answering many cosmological questions.

The Berkeley group has big plans involving many years of development. Their future will be as interesting as the ups and downs of their past.

Other groups are planning automatic searches with various types of equipment, and in various parts of the world. Some of these telescopes may not be so suitable as the kinds used so far. A mistake often made in developing these programs, even by professional astronomers, is that the capacities of the proposed search are planned using as a basis the theoretical possibilities of a particular CCD. But, many practical difficulties are not recognized, or are ignored.

Time will reveal the true value of these projects.


The first amateur discovery of a supernova using a CCD as a camera was made by Eric Thouvenot of France. The supernova was SN 1990N in NGC 4639. Eric was one of several amateurs who were making CCD pictures with one of the telescopes of the Pic du Midi Observatory. Eric compared the CCD picture with the Palomar Survey some hours after the picture had been taken, and found the supernova. The purpose of taking the pictures was linked to the development of the Buil-Thouvenot CCD Atlas of Galaxies, and so was only partly a search for supernovae.

The second amateur discovery using a CCD has also been made by a Frenchman. Christian Buil discovered SN 1992I in NGC 2565 at the end of February, using a new large amateur club telescope.

These observers are now mounting a major search by this method, and numbers of other amateurs are now entering this scene, as well.