Affiliation
American Association of Variable Star Observers (AAVSO)
Wed, 05/28/2014 - 10:37
A simple and basic question about WebObs page: the variable mag. to be put into the 3rd window must be corrected by the difference between catalogue and observed check magnitude, or not? Sorry, but I could not find the answer in the helps. I'm using ensembles and not-transformed magnitudes. Thanks
Hi Giuseppe,
The variable's magnitude is NOT corrected by the catalog/observed check magnitude. Your ensemble solution should already produce a magnitude that has been adjusted by the catalog magnitudes of the stars in your ensemble, so that is sufficient. The check star is a separate star, reported using the same ensemble that you used for the variable, so that it may be possible to adjust your data in the future if better photometry becomes available for the ensemble stars.
Arne
Thank you, Arne. That's the way I used to do, but I wished a confirmation.
An ignorant but related question
For years WebObs files I have submitted contain the raw instrumental magnitudes of the comp (for single comp observations) and the raw instrumental magnitude of the check. My stupid question is whether I should be reporting the (un-transformed) standardized magnitude of the check star. It seems to me this is a matter of convention rather than substance since if you identify the comp and provide its raw instrumental magnitude then if you report the the check star AUID and raw instrumental magnitude you can directly calculate its standardized magnitude, and if you report the standardized magnitude, instead, you can directly calculate the the raw instrumental magnitude. So it doesn't seem to matter which you report, except is one from is used by convention and perhaps making it easier for a data user comprehend when the check and comp stars' magnitudes are both raw instrumental magnitudes.
Is there a convention? does AAVSO care if the check magnitude is a raw instrumental magnitude or an un-transformed standardized magnitude?
As a second issue: if you are doing ensemble photometry and want to provide users with the same ability to analyze your results as with single comp observations, don't you need to identify all of the comps, their individual instrumental magnitudes and the individual weights assigned to each comp (or state they are all equal) in the Notes field of each image? If that is correct and we want to make our ensemble photometry as useful as possible to data users, then don't we, as Ricky Ricardo might say, "got a lot of writin' to do" in the notes, to an extent that becomes impractical if you are using more than a few comps in your ensemble? Any suggestions how to do this better?
Brad Walter, WBY
Brad,
The documentation for the Extended Format Record that is submitted to AID via WebObs ( here at https://www.aavso.org/aavso-extended-file-format ) states:
So, instrumental mags they should be.
------------------------------------------- EDIT
I apologize, for I did not read the documentation closely enough. If you are doing ENSEMBLE then in fact the definition of the data for KMAG is no longer instrumental mag of the check star but should be that's star's magnitude standardized in the same fashion as the target star, with the ensemble.
This matches Arne's description below and in the https://www.aavso.org/content/new-sequence-reanalyze-v1687-cyg thread.
----------------------------------------------
Now, as for ensembles, you have pointed out their weakness in AID: they can't be adjusted for changes in photometry. Or this can be seen as a weakness of the Extended Format record: it doesn't allow for the description of the ensemble. What ever. I would believe this is enough of a problem that ensembles should be deprecated. When we pushed the use of transforms a few years back Arne stated a clear preference for transforms over ensembles ( https://www.aavso.org/transforming-ensembles ).
Is there a way to extend the AEF record to hold ensemble definitions? Sure, there is a way if we are prepared to do a whole lot of writing. The Notes field can be extended with subfields. We could document a convention something like this: number the comps and create subfields |C01NAME= |C01INSTMAG= |C01ERR= and so on for all the stars in the ensemble. But still, there are many ways to do an ensemble with many kinds of weighting schemes. These would all have to be defined and then noted in a field (eg |EMETHOD= ESIMPAVG). I think this path would quickly breakdown. Too much is being put on the researcher to verify the observation.
A simpler fix for ensembles might be to put enough information into the Notes subfields so that a researcher can ignore the ensemble. If you provided the |VMAGINS= |CMAG= |CNAME= (the target's instrument mag, one comp star's instrumenal mag and name) then the researcher could quickly get a verifiable single-comp measurement.
TransformApplier provides this kind of information in the Notes field for exactly this reason: to let the researcher access the raw data and skip over any post processing.
George
Hi Brad,
Ensemble was deprecated in TA because that program can not properly transform the observation without knowledge of what stars were used for the ensemble. TA works best with the traditional technique of target, comparison star and check star.
For my own photometry, I first transform the target and comparison stars, then I perform ensemble photometry using the transformed values. This works, because I do ensemble in "magnitude space" rather than "flux space" . I then apply weights according to many factors. Other ensemble techniques require that you first do ensemble using fluxes (so the brightest comp star gets the most weight in the solution). Those two methods require different techniques for transformation, and we decided not to deal with the issue in TA.
If you do use ensemble techniques (and there are many reasons why ensemble can yield better accuracy; see my CCD school), the ideal would be to give all information necessary for the researcher to understand the observation and potentially redo it or make adjustments. This is a very complex problem IMHO and is why I avoided the issue. :-) You could list all of the stars in the ensemble in the "comments" section, for example, but you would have to list their AUIDs and instrumental magnitudes, as well as the technique you used for performing ensemble and the transformation coefficients you used. Perhaps a table for each observer could be created and maintained, but that is not a simple task - it involves a database, an archival method, and a sophisticated program to maintain that database and provide an interface with the observer and researcher. So my philosophy was to assume that anyone using ensemble techniques knew what they were doing and not try to second-guess them.
As I've mentioned before, and as described in the Extended Format description, both comparison star and check star should be raw instrumental values for tradiational target-comp-check photometry. These photometric values are not necessary, but when reported, need to be consistent in order to be used for quality-checking by the researcher. In all cases, the AUIDs for the comp and check DO need to be reported.
Arne
Thanks for this precision! I started contributing data only recently and it seemed strange to me that the KMAG should be instrumental magnitude. I did not see that it was different if I use an ENSEMBLE solution (which I do). So, I always specified the standardized magnitude of the check star (calculated in the exact same way as the variable) in the comments field (because that seemed like the important value to report). But I guess now I'll have to update all my submissions. Should keep me busy while the skies are cloudy!
jf