As I discussed in the Director's Report from the SAS/AAVSO meeting, XZ Cet is an interesting star. Available photometry, for example, shows considerable variation at the peak brightness, yet the rest of the light curve falls pretty much into place. That would imply to me some emission mechanism occurring near max. Doug Welch also called for continued monitoring of the candidate anomalous cepheid.
Therefore, I ask observers to either continue their observations of XZ Cet, or to begin observing this star. Details are given in a prior post. Also, I expect to continue my public discussion of each observer so that we can all learn from each other's mistakes or excellent techniques.
If you do a single nightly visit to XZ Cet, then I ask that you (a) use at least two filters; (b) do at least 3 sets of observations (for example, BVBVBV), take the average and standard deviation, and report that rather than the 3 individual sets. A new Windows program called Boxster is available for download on the transformation page; you should try this out or develop your own technique.
If you do time series, the best uncertainty estimator is (K-C). You can do a running average and report the standard deviation as the uncertainty for the target.
If you can transform, please do so. Again on the transformation page:
http://www.aavso.org/transforms-everything-you-need-transform-your-ccd-…
are programs to download that both calculate your coefficients and apply your coefficients.
I recommend that you use a single comparison star for now, and not ensemble, so that we can learn about transformation and apply it easiest. You can always go back and update your submissions with ensemble results if you want to later.
I will also be monitoring the EE Cep campaign, soon to be announced. Let's get some really excellent photometry this year!
Arne
You suggested that folks use a single comp star. Why not pick one, so that we can all use the same one, to eliminate that variable from the equation? Just tell us which one.
Gary
[quote=WGR]
You suggested that folks use a single comp star. Why not pick one, so that we can all use the same one, to eliminate that variable from the equation? Just tell us which one.
Gary
[/quote]
Hi Gary,
I would suggest the closest comparison star that is reasonably bright: 000-BLF-059 at 10.7V, about 10arcmin to the NE. For those with small fields of view, try to offset your field center so that you include both XZ Cet and this comparison star. That automatically gives you the 116 comparison as well, so we can use that star as the check star.
Arne
Arne:
Just to help others avoid an unnecessary question. There are two 107 comps at about the same radius in the FOV. The one near the 99 comp is not the correct comp star you selected. The correct comp is NNE of the target.
BTW, the magnitudes on the VSP/VSD chart are not the same as I got previously from APASS. They differ by a few 0.01 mags! Too bright for APASS?? All this is a good reason to select the comp for this campaign so we use the same comp and same mags! However, can you look at the wright scope and apass comps to confirm mags?
Ken
Just curious about a couple of things re the XZ Cet comps:
1. The magnitude starndard deviations for 000-BLF-059 in B&V are larger than those for 000-BLF-055 (99 label star) or 000-BLF-056 (the other 107 label star). Is the new APASS release likely to reduce the STDEVs for the 000-BLF-059 magnitudes?
2. Is the new APASS release going to include Rc & Ic data for 000-BLF-055 and 000-BLF-056?
Brad Walter
Arne wrote:
" I will be monitoring the EE Cep campaign, soon to be announced..."
Arne, I'm sending data to Piotr Wychudzki at the Nicolaus Copernicus Astronomical center in Poland, is this the campaign you're refering to, or will the AAVSO be running it's own campaign?
Douglas.
Hi Ken,
You are right; the comparison star I was suggesting is the closest of the 107 stars, located NE of the variable and not next to the 99 star.
I'll get the comparison star magnitudes updated with what I think is the best set of merged values in the next week or so.
Arne
Hi Dougles,
The AAVSO will be issuing an Alert Notice shortly regarding observing EE Cep. While we will be requesting observations to be sent to the AAVSO International Database, we are also working with the Polish team on their specific campaign. So sending observations to both places is ok, or just to the AAVSO and let them retrieve them from our database, or whatever.
Arne
The Nicolaus Copernicus Astronomical Center suggest to use the following chart and sequence for the EE Cep campaign:
https://sites.google.com/site/eecep2014campaign/for-observers
Btw.: should'nt we move this topic to a separete thread?
Thanks Robert, Ill start a new thread.
Hi Arne,
I plan to observe XZ Cet when it is conveniently placed later in the year. In a July 3rd post you said:
"If you do time series, the best uncertainty estimator is (K-C). You can do a running average and report the standard deviation as the uncertainty for the target."
How many observations should be used for the running average, would 5 be sufficient (i.e. the current observation plus 2 before and 2 after)?
You also suggested using only one Comp star initially. I use a variation of the Citizen Sky Intermediate Spreadsheet to analyse DSLR time series. My spreadsheet requires an ensemble of at least 6 Comp stars to work out transformation and extinction corrections.
There are not enough bright comps in the photometry table for XZ Cet, even for my 6x4 degree FoV. So I plan to use V and B-V values from Tycho 2 for six bright, non-variable, stars for the ensemble. I see that different souces of Tycho 2 data have slightly different catalogue values for the same star. The difference is usually only a few millimag and shouldn't make a significant diference at the 0.01 mag accuracy you are looking for in the XZ Cet project. But which source would you recommend for Tycho 2 data? Cheers,
Mark
Hi Mark,
I like 5 points for the running average. Those points are therefore close in time to the current observation, and so reflect well the current uncertainty. More data points tend to average out transparency fluctuations; fewer points don't measure the underlying "population" in a statistical sense. You can use 3, or 9, or whatever; 5 is just a general guideline. Usually you pick an odd number of points to average so that the current observation is centered.
The reason I asked for a single comparison star, at least at the beginning, was to make it easier to compare one observer with another; find and correct discrepancies. Ensembles work fine, but they include more parameters that need to be checked, such as the underlying published photometry, how you combine the ensemble stars, how you transform each one, how you handle first and second order extinction, what affect your aperture size or sky annulus size makes, etc. That makes it more difficult to compare apples and apples. However, I'm happy to look at ANY set of data and make (hopefully) useful comments!
Arne
Hi Arne,
calculating the standard deviation of the running average of K-C is easy for the situation where one Check and one Comp star are used. But how do I do it with an ensemble of Comp stars? I thought perhaps the following would be ok.
Calculate the K-C standard deviation for each Comp star:
STDEV(K-C1), STDEV(K-C2), ...,STDEV(K-Cn)
K-C standard deviation for the Comp star ensemble would then be:
STDEV(K-Ce) =SQRT(STDEV(K-C1)^2, STDEV(K-C2)^2, ...,STDEV(K-Cn)^2)
Does that make sense? Cheers,
Mark
I've already collected about 5 nights of data on this object. Because its so bright I had to limit my exposures to I=60sec, B&V=45sec, and R=30 sec. I've been sequencing them BBBIIIVVVRRR, but I see in the earlier portion of this post that a sequence BVBVBV is preferred.... Should I sequence the run as a BVRIBVRIBVRI? I'm going to use the Transform Coefficients derived from my NGC 7790 data to this object using 107 as a comp and 116 as the check star.
James
James,
Arne was just giving giving a two color example. If you are doing more than two colors, may I suggest the following sequencing using 4 colors as an example:
IRVBBVRI
This makes the mid exposure time and airmass of the averages of the pairs of images of each color in each sequence as close to the same as possible and keeps pairs of images in the bands most affected by airmass differences closest together in time and airmass. If you average multiple sets sequences together, the midpoints of the averages of the colors are still as close to the same airmass and time as possible. I average multiple sets rather than taking a longer sequence (IIRRVVBBBBVVRRII, for example) because, depending on your hour angle, sky conditions, focus change, integration times and purpose, you may not want to average over such a long time span / airmass range. This method also avoids any autocorrelation that can occur from using moving averages of photometric measurements.
MaximDL requires unique suffixes for each image in a sequence in an autosave dialog. Other software probably does as well. That is easy to do by indexing the suffixes, I1, R1, V1, B1, B2, V2, R2, I2, for example. Oh, and don't forget to turn off "Group by slot" in the autosave options. I have left that turned on more times than I care to admit after taking calibration images.
Brad Walter
Thanks Brad,
I use CCDAutopilot, controlling MaximDL, to do my CCD imaging, so sequencing induvidual filter shots is a little more problematic. I only have 8 slots per object to shot so I can only do 2 sequences of 4 filter induvidual shots at a time. As a work around, I just name the second CCDAP image XZ cet2, XZ cet3, etc...i usually have the system do this while I'm asleep, so using MaximDL alone is not an option ;).
I'll change my sequence order to match yours to avoid the AM problem. Since I start this at its meridan point, I'll go 3 sets of BBBBVVRRIIBBBBVVRRII. Thanks for the advice!
James
I use shorter sequences and more sets but essentially the same thing. Each of my sequence is only 8 images and I do multiple sequences. IRVBBVRI - IRVBBVRI - IRVBBVRI . . .. There is no reason I can't make them longer and I often average two sequences together. I have a steel tube telescope so I often can't go more than two sequences without re-focusing.Each sequence gets in incremental number assigned by the software. So I end up with file names like XZ CET-001I1H45, XZ CET-001R1H20, XZ CET-001V1H45, XZ CET-001B1H90, XZ CET-001B2H90, XZ CET-001V2H45, XZ CET-001R2H20, XZ CET-001I2H45, and the 3 digit numbers increment automatically with each 8 image set. The H stands high res (1x1 bin = H, 2x2 = M, 3x3 = L) and the final number is the integration time. I really don't need the resolution in the file name since I do all of my photometry binned 1x1. It is a habit I got into when I was doing astrophotography and sometimes binned R G B frames. I used to include the date to the file name prefix along with the object, but it makes the name long and all the files go into a directory with the date in the directory name anyway.
Brad Walter
Hello James
I assume that you use CCDAutopilot to direct your scope to several targets thru the night while you sleep or do you observe a single object?
Gary
To: Gary,
RE:"I assume that you use CCDAutopilot to direct your scope to several targets thru the night "
That's correct. I usually shoot 20-25 objects a night in BVI filters. Here is my mechanical/software system:
Mech: Ap1200GTO mount, 13" F/7.5 classical cassegrain (2,442mm FL), Robofoucs on AP2.7 focuser, SBIG ST-10XME with Lumicon off-axis guider and SBIG remote guide head.
Software: CCDAP5, FocusMax, Robofocus, Bisque Sky6, DD Pinpoint 5, ASCOM 6.1, AP GTO drivers, Maxim DL or CCDSoft(when using AO-L device).
You need a telescope on a Go-to type mount with the abilty to plate solve your CCD images so CCDAP can frame your variable star filed to within 30 arc seconds. Since my focal length is relatively high and CCD Fov relatively small, I need to have this accuracy for proper object aquisition. The trick to get automation to work is having CCDAP control all these other software through ASCOM or CCDAP's drivers that can "hook-up' to them. Another useful technique is to center the variable star field and confirm the guiding device has a adequetly bright (ADU count >1000) guidestar. You can pre-plan this by using the Sky's FOV inicator and rotating the virtual guide box so the orientation matched your physical telescopic/guide FOV system. This can be proplematic (difference btw 2000 coordinates and present day equatorial coorinates) so I always take the time to confim I have:
1.) Variable star in field; 2.) adequate number of comparison stars in FOV; and 3.) a bright guide star within 25% of the center for the guider's FOV.
Once these critiria are met, I do a Pin-Point plate solve of the variable star field and copy and paste these equatorial coordinates to CCDAP's target screen. You can schedule bias, darks and flats before or after the image run. You can delay the times between imaging to make sure it clears meridan or any nearby pesky trees. You can even control rotation (if you have a mechanical rotator) or have it park you scope and close-up your observatory if certain (bad) weather parameters are met.
Without this software, I would have only done a small fraction of the almost 40K variable star submissions I've made in the last few years. I have more data than I can reduce and submit most of the time! For more information on CCDAp and its requirements see: http://www.ccdware.com/products/ccdap5/
This automation is most useful at aquiring data for XZ Cet, as it comes to meridian around 3:25am local time and at least 4 hours before I wake-up and go to work!
James
Hello James
Thanks for the info on your use of CCD autopilot. That's a real nice setup your have. You must have spent lots of time getting it to all work together.
Being retired, I have a little more flexibility with respect to waking up in the morning and going to work. My work consists of doing PT on the observations from last night, on most mornings. I also Part Time at Maria Mitchell Observatory, but I don;t call that work. Most of what I do there is related to observing.
Gary