Thu, 03/02/2023 - 00:40
Just reaching out to ask for clarification regarding the "period" indicated in VSX, specifically for V0552 Aur. In VSX, it says magnitude range of "11.2 - 14.5 p" and period is "0.060868 d (87.650 min)". I did observe the variable for 1 1/2 hours, and the magnitude was steady at an average of mag 13.078 (CV filter) with std dev of 0.016.
What does the period of 0.060868 d (87.650 min) mean in VSX? Still waiting for my observer code to be connected to my account though, so not able to submit data to AAVSO.
If you use the…
If you use the Light Curve Generator to view the current data (choose all data), you see that this star does indeed vary by over a magnitude on a short time scale. I'm wondering if you somehow picked the wrong star as V552 Aur when doing your analysis.
I used ASTAP to plate solve and annotate the variables and check stars in the images. I also used the RA/Dec from VSX to mark the variable, and it's exactly where ASTAP annotated it as well. I plotted the light curve (but can't upload in the forum), and no significant variation over the 1 1/2 hour observation time... Double checked with VS Plotter, and it's the variable I analyzed.
You should do two things in a case like this:
- look carefully at the VSX listing for V552 Aur. Therein, it states that it is a suspected DQ Her cataclysmic variable, and pressing the "?" link gives you information about that class of variable star. Usually, cataclysmic variables have infrequent outbursts, which may be the cause of the large amplitude listed (though anytime you see the range listed as a "p" bandpass, you should be suspicious, as this indicates "photographic" and highlights that the range hasn't been updated in the past century). Further down in the listing, the Remarks section says that the period refers to low-amplitude (0.03mag) variations. With many types of variable stars, there are multiple periods and multiple types of variability. This is especially true of the cataclysmic variables, which often exhibit rotational modulation, hot spot eclipses, etc. in addition to their aperiodic outbursts. VSX only picks out one of the periods, often missing some interesting other features.
- don't just use the default date window of the light curve generator. The date window usually defaults to a few hundred days. Go to the "plot another curve" window and extend the date range. I often take the lower limit of, say, 2459300, and change it to 2457300 (a single keystroke change that extends it another 2000 days in this example). By doing do, you often pick up more photometry on a poorly studied star like this. Then concentrate on only one photometric band, like V. You can see that any normal variation is small.
Unless you have excellent observing techniques, you are unlikely to monitor the small variations this star normally exhibits. There are lots of equally interesting stars that can be more easily studied!
I'm not sure what you mean by "normal variation is small". What you suggest for looking at the light curve is exactly what I did before responding to Raymond. I see four time series in V and all of them have variations between 1 mag and 0.2 mags. I wouldn't call these small.
Raymond saw essentially no variation. It is an open question why that is.
What time series has 1mag variation? I don't see it; maybe I didn't look back far enough in time. The recent time series by BPO shows random noise, especially when you take into account the uncertainty of each measurement. Even that one shows extreme points at 13.22 and 13.63, or only 0.4mag (and remember, these points have reported errors of 0.1mag). The GCVS Remark indicates some flickering with this star as well. Basically, the VSX "amplitude" of 11.2-14.5p does not correspond to the star's typical behavior. So don't read too much into the VSX amplitude - trust your own photometry more.
I was looking at the time series by AAM around JD 2454185.5045. It ranges from 12.5 to 13.52. There is another on JD 2455830.5052 by SJNO with a range of 13.26 - 13. 69. This seems more than random variation.
Ths SJNO dataset says that it is using a DSLR (Canon 300D), reports as V even though the observations are untransformed, and does not give any error estimate. I would discount that dataset.
The one from AAM is harder to discount. He gives errors in the 0.03 range, but with large scatter. This is the only dataset that shows evidence of variation, but it is only one dataset with no confirmation. Even there, only 3 data points out of 50 lie more than +/- 0.1mag from the mean. I don't see any evidence of periodicity, though I have not phased those points with the VSX period. I don't know if the observations were at high airmass, or with long enough exposures to reach saturation, or what. That is why it is difficult to prove or disprove any theory based on a limited dataset. You need to remember that CVs do flicker, and that flickering is usually random, so unless you can phase the points with some period, I call that random. If you look at the Kozhenikov paper, you will see that they did several nights of precision time series and did not see any variation more than a few hundredths of a magnitude.
Again, monitor the star yourself and see. I strongly recommend high-quality observations to prove/disprove any variation. For example, take your data in multiples of 3-5 images, combining each set to create a mean and standard deviation. Don't just take a time series and assume that the scatter indicates variability. If possible, get a friend to monitor the star at the same time, so that you have overlapping datasets and can look for changes that appear in both sets. My guess is that any variation is going to be very small, and you will need high quality data to see it. This is one of those stars that started out as an NSV object (New Suspected Variable) and then for some reason got assigned a GCVS catalog name. The original photographic variation was probably not real, as 0.03mag periodic behavior would not have been visible with photographic techniques.
Regarding extracting mean and standard deviation from 3-5 images, I did sharing of frames across data points to have an effect of moving average (total exposure/stack time is greater than data point interval), and the result is much better than extracting data point from a single image without sharing/integrating frames from previous and next data midpoints.
This is the method I used for observing BX Leo (https://www.aavso.org/bx-leo), and the difference in linear regression slopes for the target and check stars are noticeable, and allowed me to observe the 0.08 mag difference - even 0.04 to 0.05 mag difference is already slightly noticeable.
By no variation, I…
By no variation, I didn't mean the data points are the same for the data set. There is variation with a standard deviation of 0.016, but the variation is not close to the 11.2 to 14.5 range that I thought the variable is behaving. I'm sure though it was V0552 Aur that I was observing.
What is the standard deviation of your check star during the same period of time?
Standard deviation of check star is 0.0297 or approx. 0.03. SNR ratio for the variable has an average of 93.41 and std dev of 9.41.
Your check star should be about the same magnitude as the target so you can assess whether the scatter in measurements is real or not? Was that the case here?
IF the check star was of similar magnitude and had a larger standard deviation than your target, do you think the target's variation is real?
Check star has an…
Check star has an average of 11.379 and std dev of 0.03. I think the variation is just random for the target's 13.052 to 13.108 range. I was checking if the variation was 11.2 to 14.5. But Arne explained about the variable type, and I now see the note "The period is for small-amplitude variations, flickering is also observed; the combined amplitude of these variations is 0.03." in VSX.
Ah, I see. So the different readings do vary in a non-random fashion, just not with the same range as VSX suggests.
I think the…
I think the variation is just random for the target's 13.052 to 13.108 range for my data. The period of 0.060868 d (87.650 min) in VSX is related to the Remarks section from GCVS Team: "The period is for small-amplitude variations, flickering is also observed; the combined amplitude of these variations is 0.03.".
Thanks for the insights, Arne! Bert Pablo was able to fix my account and observer code, and I was finally able to upload my data and generate a light curve.
Examining my light curve, I think it's similar to BPO's data from 2458463.298817 to 2458463.5114, which also does not show the 11.2 to 14.5 range - and we approximately have the same range.
I did trust my photometry and VSX data. I was doubting how I was interpreting the period field though.