Affiliation
American Association of Variable Star Observers (AAVSO)
Wed, 08/19/2020 - 16:49
The standard reference photometry for the NGC 7790 field now extends down to magnitude 18. Check it out !
Note that when you go to VSP to request the field you will need to set the maglimit accordingly, eg 18.5
The data is from the original CCD Henden data set for this field.
Cheers,
George
I notice that the formal errors on Arne's photometry here are roughly 0.02 to 0.04 mag. These seems kinda large for a 'standard' field. Peter Stetson has a similar list from similar numbers of nights (typically 20 to 30 nights) posted at the CADC site, but his errors (labelled 'sigma', so presumably rms errors) are 10x smaller than the AAVSO list. Could these two lists be compared to look for systematic differences?
https://www.canfar.net/storage/list/STETSON/Standards/NGC7790
\Brian
Brian:
Do you happen to know if the two catalogs use the same statistical measure of precision, i.e., standard error or standard deviation. Standard error would equal standard deviation / sqrt N.
Also, did Arne's photometry really result from 20-30 nights of images?
Arlo Landolt's precision measurements are standard error.
Ken
We would have to ask Peter Stetson what he means by 'sigma' --- to me, that is the rms scatter on the mean, not "mean error of the mean" i.e. sigma divided by root-n (or root-n minus 1), what you are calling "standard error". In Arlo's papers, the errors are noted specifically as mean-error-of-the-mean. Looking now at Stetson's ReadMe files, he specifically states the errors are 'standard errors'. So multiply his uncertainties by about 6 to get rms errors, which will be roughly 0.010 to 0.015 mag.
Refering to Arne's posts on NGC 7790 from 2014, he sez "about 11 nights that were of proper quality [for brighter stars]. For 14 < V< 18, the photometry is from longer exposure images on 7 nights."
https://www.aavso.org/ngc-7790-standards-field-uploaded
It could be that the means of the Henden/Stetson datasets are very close systematically -- if it weren't that both files are a nuisance to futz with, I'd do it myself, so I'm hoping someone else smarter will.
\Brian
The errors for the NGC7790 Henden dataset are standard deviation of a single measure, not standard deviation of the mean (mean error). The deeper photometry will show the rising Poisson error, especially in the redder filters. Note that the U-band photometry only goes to about V=13, as they were based on short exposures to follow CEab and CF Cas in the cluster.
It would be interesting to compare Peter's compiled results with my photometry, though there will be likely differences unless his original sources used the same 14arcsec diameter aperture that I did.
Arne
For standard photometry, I am pretty sure the error you want to use is SE = standard error = standard error of the mean, not SD (standard deviation). The former tells you what the precision of the measurement is, since what you are trying to get is the true magntude, or equivalently the "population mean magnitude", from the sample of the measurements you take of it.
SD, only tells you the spread of the measurements, how repeatable they are, which is of some value, but SE is what all standards should quote, so you can compare the precision of them directly, without needing to know the details of how many measurements were taken.
Mike
In the 2000 paper, Peter Stetson wrote (https://iopscience.iop.org/article/10.1086/316595/pdf):
However, I failed to notice the aperture size he used.
I'd ask for "impressions" or even advice with NGC 7790 AAVSO standards. Namely - people who are using it to find transformation relations should use as wide apertures - 14" in diameter. At the same time, some "important" (color-wise) stars are close to bright neighbors and many fainter stars have wings of neighbors in 14" aperture...
In the case of good seeing and sufficient plate scale (to sample those small stellar images the) situation probably is a pretty good one. But that's not very often the case for roughly sea-level amateur observatories, where seeing is typically worse than 2 arcseconds and even more so when plate scales should be 1 arcsec/pix or more. Could you give some guidelines how to deal with such rather common situation?
Best wishes,
Tõnis