Return to HomePage
Return to the Index of VSNET discussion topics
> I'm a little uneasy about the way amateur astronomy is headed at the moment. > There seems to be a great desire to find new things - minor planets, > supernovae and new variables - but little desire in some cases to do > anything more than find these objects. I rather liked it in the days when > not only did you find a new object, but then felt responsible to investigate > it in as much depth as you could manage.Your comment just clearly illustrates my thoughts, too. Ar least in our country, there is a concept of a "new object hunter", and some people indeed call themselves with this title in popular astronomy magazines. My feeling is that many of these "new object hunters" have little interest in finding out the nature of their newly discovered objects. What matters more is whether the object is called "NOVA" "SUPERNOVA" etc. on IAUCs, or when these discovery announcements are issued. Once the object has given a "NOVA" designation, they seem to feel that their work is complete. I remember there was a comment on the appoaching Comet Hale-Bopp several years ago, in which a comet hunter expressed that he is not interested in already discovered comets. Considering the rarity of Hale-Bopp-class comets, his comment must have been true for any class of "known" comets.
One of my concerns is that I have too frequently seen amateur astronomers, who originally aimed to investigate it in as much depth as they could manage, feeling happy with newly acquired data and result, someday turned to a specialized way to a "new object hunter" or its variety. I even feel there must be an external force to selectively turn renowned observers to head toward a specialized way. Or would it be a natural consequence of human genes?
Regards, Taichi Kato(vsnet-chat 6033)
The problem is not limited to amateurs. I keep track of all the supernovae mentioned in the IAU Circulars. As you may know, the number of discoveries has grown tremendously in the past decade:
1995: 57 SNe
1998: 161
2001: 296
Let's look at the year 1998. How many of those 161 SNe have been
studied in detail? Let me do a quick search through the ADS system
for papers published on these 161 events ...
1998bu, 1998S, 1998bw, 1998aq, 1998de, 1998A
That's 6 out of 161. Hmmm.
To be fair, most of the supernovae discovered are very faint,
many are long past maximum light, and some are immersed within
their host galaxies. But it still seems a bit of waste, in some
respects.
Michael Richmond
(vsnet-chat 6049)
> And people > have different tastes and drives that lead them to do different things. It's > the same in variable stars - Rod Stubbings finds an outburst visually (good > on him, I couldn't stand doing that), Peter Nelson and others then do long > time-series CCD photometry on it (a million miles from what turns Rod on).Yes, there are very significant exceptions. There are many people who are doing their ways and greatly contribute variable star astronomy (yes, here I speak of "contribution to variable star astronomy"; there are naturally much larger variery of astronomy one can enjoy and contribute to). However, even in the field of variable star astronomy, I have encountered a number of people, who had been steadily observing variable stars or monitoring outbursts, change their fashion once they become renowned in some aspect. The net cosequence has been a continous removal of (a significant fraction of) variable star observers, who have been indispensable to the variable star community. Think of the number of observers who have been monitoring outbursts; there were much more people even ten years ago. Have CCD monitoring substituted their role? -- probably not yet. I have once discussed this matter with Janet Mattei, and she also worried about this, and encouraged to "keep observing variable stars". This is truly a complex issue, and is a global matter for the variable star community. If all people are heading in the same direction at "more important things", "more science", or "more efficiency", there will be left unfilled holes (no longer a niche) of our century-long activities in variable star science. On the contrary to your supposed diversity of astronomy, the interests of the people look like to be becoming increasingly homogenized.
Regards, Taichi Kato(vsnet-chat 6073)
While writing a comment on NMO Campaign (not directly relevant), an idea comes to my mind. Present-day (amateur) CCD observations heavily rely on automated CCD photometry programs. I have a concern whether the available features in such programs have forced the style of CCD observers into a very specialized form. For many CCD photometry programs, time-series analyses of a long run on a single object is an easy task, and people becoming familiar with CCD observations are getting more involved in this type of observation than other forms of observation, but this is not the only method of variable star observing. With the present-day automatic pointing of the telescope, one can reasonably get images over 100 objects within a single night. These observations will enable us to detect many outbursts of CVs, long-term light curve of long- or intermediate-period variable stars, both of which are very traditional style of variable star observing, but are not reasonably supported by current CCD photometry programs.
I thereby request to authors of these programs to implement a function to make photometry of many targets (i.e. many variable stars) within the same night, without bothering to issue many independent commands or time-consuming eye examination of individual images. There will be two approaches. One will be to equip present variable star charts or sequences (or photometric standards), with which one can automatically measure any variable star which is included in these chart/sequence collections once the observer get a CCD image of the field. This is an extention of the traditional style of variable star observing. For CVs, Henden/Sumner/Misselt sequences (and future extensions) will be sufficient. We may need a similar public database of sequences for other types of variable stars until we can get all-sky reliable photometric catalogs.
The other is an extention of time-series studies. Once the observer selects a set of comparisons for a certain object, the program automatically uses this set for the same object on later observing nights. The set of comparison stars can be redefined later when the initial set becomes inadequate. After operating with this style, the observer can get a fair collection of personal sequences for many variable stars with necessary precision and consistency what the observer expects for time-series photometry. The second approach would be similar to what RoboScope has been doing for long-term activities of CVs.
Writing such functions would be, by themselves, a difficult task, but once they are implemented, they will be a far greater tool for all CCD variable star observers, and I would surely recommend such a tool has already proposed by other variable star organizations. Any further suggestions or comments are welcome.
Regards, Taichi Kato(vsnet-chat 6074)
Kato-san provided two possible modes of operation, both of which involve differential photometry: one starts with a catalog of known stellar magnitudes in some field, and then measures other stars relative to these known ones; and the other starts with a target star, and picks comparison stars (of unknown magnitude) against which to compare it repeatedly.
The TASS Mark IV reduction pipeline
http://spiff.rit.edu/tass/pipeline/pipeline.html
combines these two approaches. It follows a sequence of steps which include
a) measure the instrumental magnitudes and (x, y) positions
of all stars in every image
b) use an astrometric catalog (Tycho-2) to determine the (RA, Dec)
positions of all stars
c) identify stars of known magnitude by matching to a photometric
catalog (again, Tycho-2, but this works much less well
for photometry than for astrometry; alas, it is one of
the few catalogs which covers the whole sky)
d) transform the instrumental magnitudes to standard magnitudes
for all stars in all fields
The result is a list of positions and magnitudes for all the stars
detected on all the frames. No human intervention is required.
The procedure is far from perfect, however. It depends upon having images with a wide enough field that there are 20-50 astrometric and photometric reference stars in each one. I have found that _automatic_ procedures require many stars to yield a fair solution; a human may be able to calibrate images with 6-10 reference stars, but my software cannot do so reliably :-(
The Tycho-2 catalog is not a very precise photometric reference. However, it does provide many stars in every one of our images (which are 4x4 degrees on a side). Observers at clear, dark sites may be able to perform traditional, all-sky photometry: take some images of fields with photometric reference stars, and other fields of target stars; then make a photometric solution for the night to transfer the calibration to the target stars. The pipeline contains a software package which can handle this job. However, the skies at our sites are often partly clear, on and off throughout a night, so that we are forced to perform differential photometry within each image instead (using the same software package, but with different options). I suspect that many other observers suffer from the same problem. Those with small fields of view may have a difficult time finding stars of known magnitude within each field.
The pipeline is broken up into different pieces to handle the different stages. Perhaps some observers might use one or two of the pieces which fit their needs. Reading through the documentation might give other observers ideas for their own work. I warn everyone that my software is not designed for inexpert use -- and I don't mean 'inexpert' in a perjorative sense! -- and is far, far less easy to use than AIP4WIN or other professional packages.
Michael Richmond
(vsnet-chat 6075)
I have done almost exactly what Taichi suggests and my reasons were almost the same. There is a semi-automatic telescope at Brno observatory capable to get images of 50 stars in 4 hours, 3 frames in each VRI passband. Leaving aside that processing such a heterogenous dataset (each triplet of images needs different combination of dark frame, flat field a comparison stars) is _very_ time consuming, there is a lot of place for user errors. The process is highly repetitive and you can easily mistake comp magnitudes, filter designation etc. About 95% of images are processed without any error written on the screen.
Concerning difficulty of writing, it was quite easy. Because I am not any kind of highly skilled computer programmer and I realised that the worst point are photometrical routines, I "looted" a great piece of software - MUNIPACK - for FITS operations, aperture photometry and catalogue matching and wrote Perl scripts to operate on inputs and outputs. The whole code constitutes of approx. 1700 lines with commentary and it took me some 50 hours of time to write it (including beta testing). (there is a Czech proverb "Measure twice, cut once" so I spent hours just thinking about the problem without doing actually anything :-))
The software fits perfectly needs of me and my collegues. However, I have serious doubts if it would fit needs of other people. Everyone's got a different camera, different style of observing, different requests on output format. Creating such a system for general use is almost impossible. Maybe there should be some general rules and suggestions (or even pieces of code) so everyone is able to write his own software?
Regarding sequences and comparison stars - it took me another 50 hours to select comparison stars for some 60 fields. And I will spent further time to check it - due to my past experiences I won't dare to send data without checking comp stars and identifications.
There are other things, which should be included in such software. One might measure all stars in the FOV and discover a lot of new variable stars. Or it is possible to measure sky brightness and follow light pollution in your area (there are not many such measurements - astro-ph/0301115).
Best regards,
Ondrej Pejcha
I understand the process. I presume that most of professional photometrists have their own ways of doing the more-or-less similar process, but the application is usually limited, or the software would assume some extra skills for operation. This is why I requested to popular/commercial software writers. Usually there is no necessity for human intervention to get raw magnitudes, but there is usually necessity for extracting magnitudes of specific variables (either with additional software operation or with script writing) -- this process could be more facilitated with the current advent of the software technology.
The other (more important?) thing is that there is undoubtedly prevailing tendency that "the modes of software operation regulate human modalities". [Do humans use software, or software uses human?] When there are many varieties of software that enable different modes of operation, this would not be a serious problem. But in actuality, automatic photometric software is getting more confined to a single-mode operation, which, in my words, stems from an old fashion established decades ago (this mode is a "CCD synthetic aperture" extension of photoelectric photometry). Although this method is proven to be efficient to some special applications (which are also a fashion among professional astronomers in some 20-30 years ago) to precision photometry of low-amplitude, short-period stars, this is not the sole application of CCD photometry, neither this application illustrates the full potentiality of (amateur) activities with CCDs.
For example, this method is suitable for low-amplitude objects, such as novalike variables, short-period pulsators and eclipsing binaries. The amateur applications of CCDs are getting more and more restricted to these objects (see recent IBVS issues), while once the object deviates from this category, there are only few applications. Think how different classes of objects (e.g. from the AAVSO targets) have been continuously observed with CCDs. There are only few exceptions; even the most peculiar object V838 Mon was observed so infrequently that amateur CCD observations were not as efficient as visual observations to fully describe the object's behavior, let alone novae (think how many telescope tracked U Sco or CI Aql outbursts during which CCD time-resolved observations of low-amplitude objects were plenty). The same is true for (more fashionable?) dwarf nova observations. Once the object becomes undetectable with certain software, the obsrervation is usually discontinued even if the original images have extractable information. The result is that there is very restricted coverge of a particular phase of the full development of the dwarf nova behavior. Think how many people stopped observing, or even encouraged to stop observing, after EG Cnc once faded from the 1996 outburst, and the star experienced unexpected series of post-outburst rebrightenings once CCD observers turned their eyes to different objects. If the object had been tracked even after the object faded, there must have been full CCD coverage of at least some of rebrightenings. It was as if the main cause of the rebeightenings was from an instability between the encouraging and discouraging forces after the initial fading ;-) [I leave to readers identifications which were encouraging/discouraging forces; just look back the web/e-mail log]. Such a sad cosequence may be avoided, if the coming software enables meaningful photometry (not simple non-detection) of faint objects, and if the discouraging force is removed.
Back to other objects, the lack of observations does not mean the less degree of expected scientific return for the objects which are outside the scope of low-amplitude objects. My experience and suggestion are the rather contrary. The regime of low-amplitude, short-period objects has been so well established in the past decades that research on these objects has been generally confined to (of course, there are exceptions) aim at fine analysis, longer baseline observation, or statistics -- all of them are concepts established dacades ago, and what one can do is an incremental refinement requiring progressively more observing time. If the same amount of telescope time were devoted to different classes of objects, there would have been much more scientific return (of course, there is necessity of different science to interpret the data).
I have encountered such classes novelties almost every time I turned the telescope to a different class of object, or an object of timely interest (usually a transient object), since the early 1990's, and would like share the coming wonders with more generalized CCD observers. From my experience, I would warn against the current prevailing fashion of too specialized (and not sufficiently novel) application of CCD photometry.
Regards, Taichi Kato
Return to HomePage
Return to the Index of VSNET discussion topics
Return to Daisaku Nogami's page
