Contacter l'auteur / Contact the author

Recherche dans ce site / Search in this site


Supernovae Search

Tim Puckett's 60 cm f/8 R-C scope.

Estimating the supernovae magnitude (I)

Each year between one and four supernovae (SN) are discovered in Virgo cluster by amateurs. Early observations of amateur SN magnitude observations suffered from large zero-point corrections (in some cases of up to 0.8mv, see, when compared to mean photometric data that appeared in various papers written at the time). This was possibly due to an inadequate source of published data pertaining to the subject matter ie. sequenced magnitudes from which to compare. 

However, this is not the only criteria which factors into making a viable estimate. I will attempt to list some of these variables, both systematic and random which the observer should consider...

Personal equation (Systematic)

This includes : the telescope (aperture, quality of optics, position angle equation[see below]), seeing conditions of the site used to make the observation (rural or urban; dark or light polluted), weather conditions (humid or cold), the ability of the observer (number of years observing, color determination, sensitivity, etc.), the age of the observer young or older and (this is not included in too many studies of personal equation), the attitude of the observer at the time of the observation (stressed, relaxed, hard day at the work place, mother-in-law in town for her annual visit, mother-in-law leaving town from her annual visit , etc.)

To read : Estimate your personal equation

Charts (random)

Currently there are many sources which can assist the observer in not only making magnitude estimates of a newly found supernova, but can also help determine if a suspect star is a possible supernova candidate. Other comparison charts can be obtained from the AAVSO, UK Supernova Patrol, and the M-1 Supernova Network in Madrid (Also see the next link for additional resources). If you have access to a university library, articles written over the years concerning past SNe events will also give very detailed information on photometry, light curves, and the mechanics involved concerning the event and will usually include many comparison stars.

Useful links on Supernovae

SN1993J in galaxy M81. It was the brightest (Mv +8) after SN1987A.

SN1995BW (yellow) and SN1997W (blue) in galaxy NGC664 recorded on feb 1, 1997 by Carl Hergenrother.

Other methods

Several other methods exist for obtaining reference material relating to starfields. This process has been used for years by the master visual supernova discoverer...The Rev. Robert O. Evans of Coonabarabran, Australia. In fact his inspiration has been passed on to Dana Patchick of California (discoverer, 1987L) who accomplished a similiar feat....this utility is the photographing of the Palomar Sky Survey (in Rev. Evans case additional surveys were also made available to him) into a slide format! 

The idea is to have access to an astronomical library that has a collection of the POSS (Palomar Sky Survey) plates, then construct a macro lens attached to a 35mm-camera to accomplish close-up views and snapshots of individual galaxies. In proceeding in this manner you cut down on the costs of actually having to purchase the charts (available from Cal-Tech for $36 for six-6x6 degree copies of the POSS plates. Note that this price may have gone up, but that was the price a few years ago.

Ed. Note: The acquistion of these plate copies for the entire Virgo area can be obtained in 4 sets (of six), and the work of enlarging to snapshot-size charts may then be accomplished at ones leisure, this then breaks down to pennies per galaxy. Rev. Evans has thousands of such slides, Dana has spent a year getting several hundred to a thousand images for use at the telescope and for reference. The main idea for this project was to have a ready reference available in the event some star was spotted that appeared "out of place" near or involved in some galaxy. 

Like any experienced observer these gentlemen have enough skill to be able to visually estimate [within a few tenths of a magnitude] a particular star. (that is why the Experienced observer would stand a better chance at this project than one who is not familiar with making magnitude estimates). While this regimen has some pitfalls it will give the observer some satisfaction that he is able to "weed" out any stars that might already exist near some galaxy, and that he will not have to call someone long distance on the telephone, or energize a verification team to verify some spurrious essence you then become somewhat self-sufficient in the verification department. 

Making estimates... a suggestion

There are many unavoidable criteria that must be considered in making a valid magnitude estimate, some will be suggested here, however the professional community will be better apt in determining some exacting standards of these values. 

One must consider that a SN suspect will, to the visual observer (or CCD imager) appear simply as an additional point of light in a particular galactic starfield. But to the professional astronomer Galactic Absorption or the amount of reddening along our line of sight to the host galaxy must be obtained, as well as Internal Absorption, or the amount of absorption contained within the various galaxies where these events takes place. 

Other values also include where in the galaxy the event occurs. Absorption values might, for instance be higher if the event takes place within an HII region, etc. (The recent SN 1994I in M51 [NGC 5194]) displayed a more subluminous posture, possibly due to this fact absorption was ~1.8 higher along our line of sight to the event). 

(Ed. Note: SN 1994I was designated a type Ic SN, which as noted in, followed a slower decay than the mean type Ia, could absorption have been the culprit, or where different explosive scenarios at work here?).(Additionally, one should keep in mind that when making magnitude comparisons using CCD imaging to adhere to the following suggestion from B.Skiff [Lowell Observatory].."it should be noted once again that in principle there is no way to transform data for emission-line objects like novae and supernovae to any standard photometric scale based on ordinary stars.  The spectra are simply too dissimilar to avoid systematic errors between different filter/detector combinations.  

The problem is most acute toward the blue, but workable with broadband filters in the red---as long as each system is well calibrated..."(message to Novanet, 3/7/95). So what criteria should the visual/CCD observer take into consideration? Firstly, (although perhaps not of too much consequence). Atmospheric Extinction can be factored in determining the displacement in degrees from the zenith...using the formula :

cosZ= (sin a)(sin b) + (cos a)(cos b)(cos H)    


Z = distance from the zenith to geometric horizon nearest object

a = observer's latitude on earth

b = declination of object being measured

H = hour angle of the object being measured.

Once the correct zenith distance has been determined, the amount of atmospheric extinction can be determined by the formula :

M=0.35 (sec z1 - sec z2)


M = extinction coefficient

z1 = zenith angle of highest star

z2 = zenith distance of star closest to horizon.

To read : Tim Puckett’s Award-Winning Ambition, Sky & Telescope, 2012

SN1998BU in NGC3368 recorded on May 27, 1998.

SN2000CJ (blue) in galaxy NGC6753 recorded by Nick Suntzeff. Documents ISN.

Position angle equation

Angle error in making visual estimations arise from the physiology of human vision. Color error can be minimized by selecting comparison stars similiar to the variable color index. Angle error results change in the apparent difference between a variable and a comparison star when their orientation changes relative to the line between the observers eyes.

Observations made using altazimuth instruments with fixed eyepiece positions are particularly prone to angle error - as are bonocular and naked-eye observations because the observer normally faces in the direction of the variable and uses his/her body as an altazimuth mounting, the eyes remaining parallel to the horizon. 

The observer will face east to view a star field rising in the east, then face west to observe the same starfield setting. As a result, the star field will appear completely reversed between the two observations, east changing from down to up, north from left to right.

To confirm this suspicion, I performed an observational experiment. Reclining horizontally, I made one series of estimates with my feet toward the east and a concurrent series of estimates with my feet toward the west. I tilted my head back beyond the zenith when necassary".

Ed. Note: My Williams' estimates indicated a displacement or magnitude differance from the above methods of ~0.25m while observing the variable star BX AND. He further states: "....These observations were made on the same night, by the same observer, with the same telescope and comparison star values.....This should convince visual observers of the need to maintain the same orientation of the star field for all estimates of a variable star..." 


The Argelander method is the most popular for estimating magnitudes. By assigning a step to each comparison star in a pair that brackets the brightness of the star in question, the magnitude of the latter can be inferred. The step is chosen from a universal scale of steps. 

An obvious extension of the Argelander method, is the use of additional pairs of comparison stars. More observers now use this method because it leads to an improvement in the accuracy by reducing random errors, with very little increase in effort and observing time. Not only are random errors reduced with this practice, but sesitivity problems (comparison stars being red) are somewhat diminished, as well as problems introduced by erroneus magnitudes. The Argelander scale of steps is, in principle, universal, but each observer applies it in a different way. Usage of comparison stars with wrong magnitudes contributes in an important way to obtaining false light curves. 

On another note some observers will find inconsistancies between magnitudes and their observations. The problem here is that reference charts used may have been drawn up decades ago. When these inconsistancies become apparent, and lacking any recent photoelectric measurement updates, estimates of the magnitudes should be attempted visually (this is where experienced observers stand a better chance). This may be accomplished from any two or more comparison stars whose magnitudes can be determined. 

When measurements from various observing sessions are averaged and the new magnitudes implemented, better light curves can be obtained.  For an observer sensitive to red, the use of a red comparison star makes the variable appear fainter than it really is, therefore it seems convenient to use a magnitude for that comparison star that agrees with what the observer sees. From this point of view each observer is a particular "photometer" and the question arises as to wheather it makes sense, strickly speaking, to collect data from slightly different "photometers".  Finally, there are the effects related to different instruments and observing conditions. These effects are less important. Some time ago (the authors), we observed a possible aperture-dependance. We undertook a study of different apertures. 

The observing method included the techniques outlined above and also the out-of-focus technique, which is useful in estimating stars that are close in brightness or red.. To avoid position-angle problems as much as possible, it is also good observational practice to observe from one side of the telescope and then the other - if using a reflector - to compensate for the uneven response of the retina. We assumed a linear relation between observed magnitude and aperture (borrowed from an idea by Bobrovnikoff and Morris...[note....some possible flaws involving associated observational methods, involving comets, were mentioned in the colloquium proceedings:S.Edberg, pg.95-99], however studies by Bateson {IAU Colloquium #46} did not confirm this for variable stars...ed]. thus :

m(obs) = m(stand) + a[alpha](A -5)


m(obs) = is the estimated magnitude        

m(stand) = is the standard magnitude defined to be the magnitude corresponding to a previously chosen standard aperture (in this case 5 cm)               

A = is the aperture of the instrument in cm.        

a[alpha] = is obtained from a least-squares fit. The convention plots magnitude correction - that is estimated magnitude, minus standard magnitude - verses aperture.

The authors found a definate trend: the larger the telescope, the fainter the star. This convention incorporated several hundred observations made, spanning about a year. Some observations taken, where not intended for use in this particular study. We (the authors) ruled out red stars in order to avoid color problems,and used the observational techniques quoted above. However, we think that a more sensible parameter is the limiting magnitude of the telescope, which depends on both aperture and observing conditions, thus including several factors in a single parameter. The limiting magnitude is a barrier near which stars are more difficult to observe, but at the same time differances in brightness are more easily detected. 

Much brighter stars may not be so clearly distinquished (when plotting Argelander step methods versus distance) from limiting magnitude. For a given differance in brightness between a pair of stars, say 0.5m, the step assigned to this separation tends to be higher as we approach the limiting magnitude. In theory, at the limiting magnitude, the step would be infinite if one of the stars were to lie on the border and the other below it. 

Furthermore, a straight line with a non-zero slope is evident for values less than four magnitudes above the limiting magnitude..." [a figure was included in their article, however it was not reproduced here] (ed. note: The information revealed here was shortened, and indicates the authors EV and PV's intent....This method could be considered and is used here as a point of conversation....the authors might have already provided a more updated and/or current result). 


With all considerations factored, the visual observer should have had some previous magnitude estimating experience. Comparison stars should be well sequenced, and stars with varying color indices (if possible) should be used as the event decays (Note: Comparison stars within the same field of view will have almost non-existent differances in atmospheric extinction from that which might be present)(See Light and Color Curves). 

The ideal situation to help eliviate scatter found in many event estimations, would be a standarized comparison sequence that all SN enthusiasts could have access to! The Guide Star Catalog is perhaps the largest, most readily available source, but unfortunetely this work suffers from some disparity amongst its magnitude sequences. Hope for the future might include transformation formulae, or a better more accessible method for standardizing magnitudes around various galaxies, however this work would probably approach biblical proportions.  

In summation, The visual observer should attempt to bracket magnitudes by following V band reference star magnitudes compared to the SN (if possible). Some attention should be given to the color index (B-V) model given in the section on LIGHT AND COLOR CURVES to possibly eliminate some of the magnitude scatter.  

It is important to remember that accurate magnitude measurements need not equate a stress-free, non-smoking, non-coffee drinking individual, who owns a perfect telescope, observes from the right position from a sub-arc second photometric site at 72 degrees, who is not color sensitive, and ranges in age from 22-35 years old, who was weaned at birth with a just takes a bit of determination and practice, practice, practice.

Second part

Supernovae Light Curves

Page 1 - 2 -

Back to:


Copyright & FAQ