The description of the AAVSO Extended File Format indicates that for the Comp and Check stars of an observation the instrumental magnitudes should be entered. I am hoping someone involved in the definition of this format can answer two questions for me:
1. Is the instrumental magnitude the extra-atmospheric magnitude which would be corrected for first order extinction? (and second order extinction if in the B band)
2. What is the rational for using instrumental magnitudes?
I admit I am guilty of putting in the catalog magnitude of the Comp star, and the reduced/transformed standard magnnitude of the check star into the database. This makes sense to me since then a quick look at the check magnitude compared to the catalog magnitude gives some idea of the quality of the observation, and may illuminate a data reduction or star identification issue. Using the catalog magnitude for the comp is somewhat redundant as that information should be in the 'Chart Field' at least indirectly. I do PEP observations so my chart field is always 'PEP'. A researcher probably can't use PEP directly to find the magnitude I used for the comp star, unless they know that there is a file in the PEP section of the Web site which lists the catalog data for the stars in the PEP program. However, this file only covers the B and V bands, so again listing the catalog magnitude of the comp star as part of the data submittal makes sense to me. Also there has not to my knowledge been strict version control of the PEP star catalog data, so for older observations it may not be clear if the file has been updated and a researcher may have difficulty tracing back any changes to the PEP comp star catalog. Listing the values explicitly in the comp magnitude field would eliminate this issuel. Also the PEP star catalog/list is not directly tied into the chart database to my knowledge, so again there could be differences in the information pulled from the chart database, when compared to the PEP star catalog file.
Any history or rationale for the choice of instrumental magnitudes for the comp and check star would be appreciated.
You knew I'd respond along the way.
The Extended Format was designed to be as compact as possible while delivering as much information to the researcher as possible. For just reporting the magnitude of the target star, all you need is the final magnitude, the labels of the comp and check stars, and the chartID so that the researcher knows what values were used for the comp and check. Putting the chart magnitudes of the comp and check into the report is redundant (and a possible source of user error anyway) since that information is already available through the chartID.
Therefore, I attempted to make the information content higher by requiring the observer to report the instrumental magnitudes of the comp and the check. These should be the raw magnitudes coming out of your software, without extinction corrections. Subtracting the check from the comp then gives a rough measure of the true uncertainty of the estimate, as well as a possible crude way of determining the transformation coefficient. More importantly, tracking the instrumental magnitude of the comp for a time series shows the extinction and when the GEM flip might occur. As you mention, the magnitude difference between comp and check also can be used to ensure the correct stars were measured.
The PEP stars actually have pretty good version control, though it is all internal at HQ. As you mention, they only include B&V values, so if you also did R or I measures, the research could never recover your estimate exactly unless you provide the values you used in the comment/note field (which I highly recommend) Use the comment/note field as your friend. When doing ensemble, for example, you might list the IDs of all comps that formed the ensemble, or notes about a particular instrument setup, etc. At one point, we were going to enter the PEP comparison stars into the VSP/VSD database, and add another clickbox to VSP charts to only show PEP comparisons, but that was never finished.
The Extended Format has worked pretty well, considering that it was developed over a decade ago and hasn't seen major changes since. We were able to get all of the major software vendors on-board, and it made life much easier for those maintaining the AID. I'm sure it could be improved, however!
You are right I was hoping you would respond, and I appreciate the rationale and history. The PEP community is working on adding some additional metadata in the comment field, and as you said it is our friend and has plenty of space available. There is still a bit of a gap for I, R, J and H bands (although there is some J and H comp star information on the IR PEP page).
Metadata proposed includes the first (and for B) second order extinctions used, along with the transformation coefficient, and the catalog magnitude of the comp star. For V only observations we would also include the catalog delta BmV between variable and comp. This along with the instrumental magnitudes would be sufficient for someone to reduce the data and really does not take up too much space. Other thoughts along this line would be appreciated.
When I proposed the Extended Format, all that was currently available was a very limited format that was used for visual observations, slightly modified for CCD use. At that time, I felt it was already a big stretch to require observers to fill in all of the information in the Extended Format, and so left out several parameters that could have been useful for some investigations and some submissions. For example, equipment or transformation coefficients could be valuable. One option we considered was to have the observers submit a table of information for their site, so that you could use the obscode and obtain the site information. However, we disguarded this concept because of human traits - it is hard to get people to keep such tables current, and as information changed, you needed a change log in order to find out what was in use at the time of a specific observation. Some of that information could, of course, have been submitted with each measure, with lots of redundancy, extra effort, more chance of error, etc.
Another Extended Format that is specific to PEP or DSLR could be developed. Or, the current one could be extended again. Those are HQ decisions and not mine, so I just recommend to use the comment field to include anything that you feel might help the researcher in the mean time.
I"m a CCD observer using VPHOT and transformation coefficients. It's a simple cut and paste step to add my coefficients to the VPHOT report comment field. I will make it a part of my process. Thanks again for your help in making our data more valuable.
I'm trying to use Transform Applier (TA) to get transformed magnitudes of my T CrB data. Trouble is that I have only one check star near my object and no other star with an AUID listed. I ended-up using a 13 Vmag star (GSC 2037:996), but when I run this in TA, it gives me a "No KREFMAG available. Possibly bad chart reference." error; the AASVO chart reference is correct. I know I can cheat and use a star with and AUID on the chart (but outside my FOV) and TA works. I just worry that by "lying" about the check star's true ID it may cause an error with the transformed magnitude. Looking at almost 2 years of data, this star GSC 2037:996 seems non-variable. Is this ok or any other options?
Well I found a workaround. Since Transform Applier (TA) only takes AUID stars in the chart Im using (X25642LL) to reduce the BVRI magnitudes of T CrB, I used the AUID of star 105, which is 000-BJS-901 as a substitute for the actual check star I used, GSC 2037:996. TA like this and it worked like a charm. After saving my Transformed BVRI data, I did a "Replace All" in notepad and changed 000-BJS-901 back to the actual "check star" I used, GSC 2037:996. WebObs seemed to treat this as "acceptable" so hopefully I haven't done something "un-ethical" in the photometry world! Webobs has over 2500 new transformed T CrB BVRI magnitude data from 13Jun18-17July20 deposited!