Affiliation
Royal Astronomical Society of Canada (RASC)
Wed, 09/18/2013 - 17:15
Looking at the graphs, it seems to me that there is an up and down pulsation within the overall descending curve. Is this real, or a chance artifact? If real, what would cause it? Patrick McDonald (MDP) PS. If this is covered in another forum, please advise and accept my apologies.
Hello
Arne and I have discussed these pulsations. I assume you are talking about the ones that are and hour or so in time. What is really needed is several time series (at least 2) to confirm the behavior. It is in places on the light curve, and in other places, one can conclude that its an artifact. There are places where 2 time series overlap, but they do different things. Arne has been working with observers to eliminate the artifacts from various causes.
I am going to go most of the night tonight. Looks clear here in ACK. Any takers? Had a good clear night last night. Most of east coast look clear. It would be great to get several light curves supporting either pulsations or not.
Perhaps you are talking about the changes in slopes that occur days apart. The light curve does seem to do this also. Let us know which you are seeing.
Gary
WGR
I just uploaded my time series from last night (analysis from 560 images). I have started to box-average my measurements, e.g. here I averaged always 11 data points. This corresponds to about five minutes for every data point after averaging. I think the error bars (standard deviation) make more sense this way.
Regarding the "daily" fluctuations: I am wondering if the sparse measurements during half of the days emphasize the appearance of fluctuations. It would be really nice if some of our Asian observers would upload some time series, or even more observations :) .
I am observing tonight, too, weather is so nice here in the NE of US. But I have a hard time to keep up with analysis of all the data...
Cheers,
Helmar (AHM}
Hello Helmar
Very nice light curve from last night. I too observed tonight, and just uploaded the data. Observed until air mass = 2. Can't wait to see your data from tonight, and any other observers.
Gary
Gary,
Thanks for the compliment! Part of it is your fault, since you have taken time series for a while, too. I am a beginner in this field, and this nova is the first object I reported data of to the AAVSO.
In the beginning I compared my single data points to your data I was pretty close (except your "superhumps"...). That gave me confidence to continue
I had a number of e-mail exchanges with Arne, too, after his gracious offer to look at our data. I thought, too, that this is a great opportunity to learn more and improve (not only my own skills, but also the overall quality of the AAVSO data, if more people would take advantage of Arne's offer).
This week has been incredible in terms of weather here in the NE of USA. Several nights in a row with outstanding skies. I didn't get a lot of sleep, and at work people are wondering about my puffy and/or red eyes (I didn't fall asleep in a meeting, yet). But taking data and analysing them it is a lot of fun. Now I understand when people say, variable observations are addictive...
Well, enough said, I am taking data right now, another series during a clear night. Looking forward to see your (stacked?) data.
- Helmar (AHM)
Hello Helmar and all:
Can't wait to see your data today. I got a big surprise last night when I plotted it. Seemed like the night was about the same a prior night. Weather seemed the same.
The only difference is that I tried to stack 6 images at a time, rather than 10. Should not make a difference.
Gary
Hi Gary et all,
Just uploaded my data from last night (around JD 2456555.7), again box-averaging over 11 data points.
I always wish my data would show the same interesting features as yours... but I think they don't
I am wondering, though, why your data look so much smoother, why are the error bars so small (or the other way: why are my error bars so much larger)? I do know that I calculate the mean and the standard variation of the 11 points I am box-averaging over. Do you know what determines your error bars? Maybe Jim could chime in?
Richard (SRIC) submitted a time series with transformed data. His and my data "line up" quite nicely.
Would it make sense and be possible for you, Gary, to revisit your data set and use the same comp and check stars as Richard and I have used (i.e.80 000-BLC-955 and 98 000-BLC-945)? I would do a re-run myself with 98 and 112, respectively, but I think I do not have 112 in my field-of-view (by the way, I could not find 112 in the photometry tables, what star is that actually?).
My apologies for all my questions and requests, but I just try to understand...
Cheers,
Helmar (AHM)
P.S.: I noticed that Arne is getting the big guns out now (W30)
Hello All
Helmar's comments about keeping up with the data got me thinking. I hate to average the frames, because its a PITA to pull into Excel, and then manipulate, only to be rejected by WebObs for some format thing. But there is a better way, that I just tried. Long time AAVSO member Jim Jones has written a VBS script that stacks images in groups of x, where you just input x in a GUI. Its available on the Maxim page, as an "extra" at no charge. It does work only with Maxim as far as I know. JJ chime in here. Its fantastic.
I just took 386 images from last night, and boiled it down to 38 stacks of 10. I chose 10 at my option. Then redid the PT on the stacks. Saved as AAVSO extended format. Then I retracted my 386 observations and replaced them with the 38 from the stacks. Wah Lah. All done. It took about 1 minute for "StackFitsFrames" to produce the stacks. So its painless. Thank you Jim.
I usually am able to reduce the data from a long night by doing it while the camera is warming to room temp, and I am shutting down the equipment. This includes submitting via webobs.
Gary
Hello Gary et al.,
Just uploaded my time series from yesterday, see LCG, end date JD 2456555., with .6 days. Nice comparison!
I guess I started a new craze, "stacking"...
Emery (EEY) seems to have supplied a stacked time series, too.
Now I am glad that I asked Arne for his blessings when I uploaded my first stacked series. The data look so much clearer now - much easier to compare. I don't think that we need to look at fluctuations with time periods of smaller than five minutes, but that might be a question for discussion.
So, don't throw all your images away - we might be asked later for different stacking parameters...
My method is a little different than Gary's who is using Jim Jones' "stacking script" and Maxim. I don't own Maxim (and I am reluctant to spend the money for it), but I am using AIP4Win for image analysis. The latter creates "AAVSO Reports" in the extended format. I am not stacking the images, but the extracted magnitude values of the target, comparison and check stars (or ensemble averages). I agree, Excel is a bear to use. I wrote my own scripts using "IDL", a language that is actually used in the astronomy community with plenty of free astro-libraries available (which I am not using for these simple calculations, i.e. mean and standard deviation). IDL is not free, but I have access to it at work, and it allows me to make "pretty pictures".
I think it is instructive to look at the raw data and stacked series with various parameters, e.g. I box-averaged 3, 5, 11, 21, 51 etc. data points (at half minute intervals, my typical settings for image acquisition using CCDSoft), and you can still see major features of the light curve. If there is interest, I could share a compilation of these figures for a time series.
Best,
Helmar (AHM)
I had noticed the latter ones, both in V filter(green) and visual observations. They seem awfully persistent to be artifacts.
I agree, and I think Arne would also, that the changes in slope several days apart are probably real.
Gary
This thread belongs in the Nova forum; I'll have it moved sometime today. There are strong suggestions of brightness fluctuations on a multi-hour time scale. As Gary says, only a few observers are doing time series, and those rarely overlap, so some of the variation may be real, some may be instrumental, with no conclusions until we get good confirmation. What we can say is that the nova has faded about 0.25mag in the last 4 days, and is getting faint enough that larger telescopes can start playing an important role. If you can do time series, especially late in the western U.S. or points across the Pacific, those datasets would be very welcome.
BSM_Berry is currently not operational; we are in the midst of reconfiguring the hardware during full moon. Of course, it is clear. :-)
Arne
Hello Helmar
Your suggestion is a good one, about using the 80 and 98 comp/check stars. I cannot fit both in the same field for my setup. So I have to choose between them. The 112 is the next brightest star in the field when only the Nova and the 98 are in the field. Its a couple arc minutes west of the Nova. I use 112 to designate it, as I measured that value in the images.
Arne suggested that the 80 was flawed. So I have been using the 98 for the past week or so.
Clear Skies
Gary
[quote=WGR]
Your suggestion is a good one, about using the 80 and 98 comp/check stars. I cannot fit both in the same field for my setup. So I have to choose between them. The 112 is the next brightest star in the field when only the Nova and the 98 are in the field. Its a couple arc minutes west of the Nova. I use 112 to designate it, as I measured that value in the images.
Arne suggested that the 80 was flawed. So I have been using the 98 for the past week or so.
[/quote]
When Gary was showing flares on the nova a few days ago, I asked him to send me a few images. Those images demonstrated that the culprit was the 80 star on Gary's images, which were showing the mirror effect (the 80 star was fading rapidly, then exponentially recovering, while other stars in the field were constant). I suggested to him that either the 80 star was variable, or there was some instrumental effect, and I was pretty sure the latter was correct as the inverted flares are not a common variable-star feature. On a subsequent night, I had BSM-Berry data at the same time Gary was showing flares, and my photometry of the 80 star was constant within its uncertainty. Gary then switched to the 98 star for his comp, as there was nothing obvious in the images and the 98 star avoids the problem.
The two-night calibration that we did using W30 also show the 80 star as constant, and it closely matches the color of the nova. The 98 star is considerably different (B-V about 0.9), and so if I had a choice, I'd use the 80 star as my comparison. Be aware of its two companions to the east and use an appropriate aperture size.
Gary's recent data has low scatter because he is now using the 61cm telescope at Maria Mitchell, and is stacking multiple frames. The bumps and wiggles shown by several observers (including BSM Berry!) are suggestive of some low-level variation, but there is not enough time series overlap, nor geographical diversity, to confirm this behavior. It has a similar time scale to airmass variation or GEM flips, and so needs to be looked at very carefully.
Gary, can you point out the 112 star better for me? There is a star west of the variable by 3-4 arcmin, but it has a magnitude of 11.09 using the W30 data. It is also red, with B-V = 1.37 and would not be recommended as a comparison, though possibly ok as a check.
Arne
Hello Arne
The star that I call 112, can be identified by the image and sequence values you sent a couple of days ago as the 10.98, Its between the Nova and the 9.829. Not exactly in line.
Tonight I am back on Cape. Have to suffer with the 12 inch. No 24 inch tonight!
Gary
Hello, Arne mentioned:
Gary's recent data has low scatter because he is now using the 61cm telescope at Maria Mitchell, and is stacking multiple frames.
This is true, but having 5 photometric nights in a row, with no clouds from dusk until dawn, probably helped a lot. I expect 6 weeks of clouds to average out this spectacular NE weather. I did not have to dump a single image, most nights.
Gary
RE:"But there is a better way, that I just tried. Long time AAVSO member Jim Jones has written a VBS script that stacks images in groups of x, where you just input x in a GUI. Its available on the Maxim page, as an "extra" at no charge. It does work only with Maxim as far as I know. JJ chime in here. Its fantastic."
Thanks Gary! Just tried this with my triple set of 35x8sec images at 1x1 bin. I'm using a 200mm telephoto lens at F/8 with a full frame (24x35mm) CCD so I'm able to get 11 comp stars to the 10.1 Vp magnitude. (I enclose a picture of my set-up on the attachment.)
Using this stacks .vbs program I made 21 images composed of 5 stacked 8-sec 1x1 bin Vp filter shots of the area centered on V0339 Del. Now that I'm getting more "smooth" data for each set, I'll shot this object at least a couple more nights before oct new moon.
James
It's a pitty that weather here is not permitting any time series or measurements at the moment. Next week it will probably clear up again, so maybe I can try to catch another time series again after some break...
RE"It's a pitty that weather here is not permitting any time series or measurements at the moment."
Superb clear night in SoCal tonight! I adjusted my 200mm lens to F/11 and am going for BVBV images for 2.5 hours on V0339 Del. Time exposure for these will be 20sec at B:5X20sec & V:150X20sec sequences; too bad my download times are 14 seconds each image!
James
p.s. Graphs of last weeks data attached
I observed the nova last night (Sept 22, 10:15 to about 12:15 EDT) with my SSP-5a on a 10" LX200 taking 21 observations about 6 minutes apart. I have uploaded my data in BV&R and you can see it using the LCG for last night
In V, the magnitude is 8.222+/-0.004 using all 21 data points with a p-p variation of 0.017. The plot looks pretty much like I would expect with normal astronomical seeing in Michigan.
Thanks, Gerry! If you can continue your time series for a couple of nights, that would be great. The last night or two have not had much fluctuations from other observers, so I don't know whether we are in a quiet phase or a different mix of observers. For the next few nights, the northeast looks clear, and I expect several eastern time series to be made.
Arne
Just uploaded last nights' observations (2013-09-23, after JD 2456559.4) , time series of about 550 images every 1/2 minutes, box-averaged over 11 data points. Now I got some curious variations (steps and pulsations?) that might need some "careful critique".
Comparison, check star and differences look good.
LCG-link: http://www.aavso.org/lcg/plot?auid=000-BLC-933&starname=NOVA%20DEL%2020…
Cheers,
Helmar (AHM)
Hello
Just posted on the Photometry Forum on how Maxim aligns using the "1 star Method" with and without "use Centroid.
Its probably best to concentrate the discussion in one place, so I won't post the answer here.
Gary
The night of 25-26th was pretty good and I observed about 3 hours in BVR with my SSP-5a. The R data is the best naturally since seeing has less effect on that passband. P-P variation is about 0.01 mag in R and no sign of any pulsation. After doing 8 hours of this over three nights, I'm going back to my regular stars and doing just a single observation of the nova.
Thank you very much for all your efforts on V339 Del!
Your work is potentially very useful for addressing a general question that might have particular relevance for V339 Del: What happens to the accretion disk after a nova explosion? More specifically: 1) When does the accretion disk reform (assuming that it was destroyed by the eruption)? and 2) Is an accretion disk responsible for the bipolar structure that is seem from some novae?
Since cataclysmic-variable-like optical flickering is one signature of an accretion disk, the appearance of such flickering could be a crucial piece of information for understanding how and when the disk reforms. For V339 Del, *if* gamma-ray production ends up having something to do with bipolar structure or jets, it would be invaluable to know whether the bipolar outflows could have been produced by an accretion disk (as in active galaxies, protostars, etc.). Recording the turn-on of optical flickering could thus be very important for V339 Del.
Since V339 Del is well past the "fireball" stage, and our view down to the accretion region should now be clear, any detection of optical flickering -- or constraints on the amplitude of such flickering -- is now meaningful. But since many types of systematic error can mimic stochastic brightness variations like those from an accretion disk, however, it would be best to have simultaneous light curves from multiple observers. In other words, time series from two different observers on the same night would be more valuable than two nights with time series from only one person each night. Since I don't know if or when fast variations might be expected to appear, I cannot really argue that people should be spending their time searching for them every night. Perhaps looking for fast variations once every week or two would be more manageable (and still very useful).
Finally, in addition to CV-like disk flickering, any periodic variations would of course be excellent for revealing whether V339 Del is magnetic (i.e., a polar or intermediate polar) and uncovering instabilities in the system.
Thanks again!
- Jeno
Hello Jeno,
Where it is understandable that you cannot predict the time when any "stochastic variations" might appear, could you give an estimate of the size/dynamic range we should be looking for (e.g. in V: How many magnitudes, tenth of magnitudes)? During what time frame: Hours, 15 minutes, minutes?
Most of us doing time-series have now pretty decent error bars and - after stacking - seem to be able to reduce seeing effects to small variations (e.g. look at last night's data in V for AHM, EEY, HQA, MZK and OAR - in alphabetic order
), as LCG-curve (JD2456563.45+) :
http://www.aavso.org/lcg/plot?auid=000-BLC-933&starname=NOVA+DEL+2013&l…
The "waviness" seems to be real, but is probably much smaller in variation than what you are referring to?
Cheers,
Helmar (AHM)
Wow! Your data are phenomenal. It is possible that the variations you are seeing are exactly the signatures of resumed accretion that I was mentioning.
Overall, the amplitude and time scale of the rapid variations you have detected are reasonable for accretion. You are right that in an accreting white dwarf that has not just experienced a nova explosion, the typical rms amplitude of the flickering would be somewhat larger than what you have found (maybe around tenths of a magnitude). But for a recent nova, any flickering light from the accretion disk probably makes up only a small part of the optical signal. You are also detecting emission from the hot ejecta, which is unlikely to be variable on a time scale of hours. If this is what is going on in V339 Del, I would expect the rapid variations to *increase* in fractional amplitude (i.e., in magnitude units) as the nova continues to fade. A change in the character of the rapid variations from 'wavy', with a typical time scale of hours, to more jagged, with detectable variations on a time scale of minutes, is something that could indicate that the disk is in the process of being rebuilt (e.g., Sokoloski & Kenyon 2003, ApJ, 584, 1027; arXiv:astro-ph/0211041). Fantastic!
Congratulations on this nice finding (and for a beautiful demonstration of the power of having high time-resolution observations from multiple observers).
Cheers,
Jeno
I'm doing a time series right now, would of course be interesting to get someone joining in so I have something to compare later :-). This is my first observation of V339 Del thru a telescope (before I did it with telephoto lenses), I hope I didn't mess up something.
Cheers
HB
Well, during the ca 1.5 h observation that I did on the 30th, I could not confidently detect any non-random variations/pulsations, at least none that exceeded the scatter in (checkstar - compstar). I did 13 sec exposures with a cadence of 1 exposure every 20sec, and the (variable - compstar) values have a std dev of ca 0.017, not that good actually, but comparable to the std dev of (compstar - checkstar). I used the nearby stars 80 and 90 respectively as comp and checkstar.
I made another, longer time series last night which I will reduce tomorrow.
Cheers
HB
Hello Bikeman,
How did your longer time series come out? There should be a number of time series measurements on the days of your observations, see LCG. I am not sure about your time zone, though, Europe-Germany? What camera are you using (DSLR?)?
Best,
Helmar (AHM)
Timezone is GMT+2, Central European DST . During the working week I usually try to do exposures from around 21:00 to midnight local time, weather permitting.
The Oct 1st data is still sitting on my hard disk waiting to be reduced. I'm currently using an Olympus e420, which served me reasonably well with 40k+ shutter actions so far, but I'll get a Canon EOS 1100D in one or two weeks. Should have MUCH lower dark current and readout noise. No sensor cleaning tho .... (The sensor cleaning on the Olympus is nearly perfect).
I've got a question: do you guys deal with 2nd order , color-dependent, extinction effects?
Cheers
HB
Hello Jeno,
Thank you very much for the compliments! Your positive feedback made me smile - and I am sure some others, too. Here is another nice set from yesterday (JD 2456568.45+) from David, Gary and Helmar (including some B values):
http://www.aavso.org/lcg/plot?auid=000-BLC-933&starname=NOVA+DEL+2013&l…
And thanks for the reference! That and your explanation make it clearer what you/we are looking for.
Best,
Helmar (AHM)
Yes, the
Hi Helmar,
Yes, the brightness variations were very obvious last night in your and my data. I am observing it now and should continue until its two low several hours from now. I will post the data in the morning.
-- Dave LDJ
I just had to stop due to incoming clouds
.
Keep it rolling, the variations appeared to become really clean lately! Hopefully that will continue.
Good luck!
Cheers, Helmar (AHM)
Unfortunately the stretch of nice "photometric" weather seems to come to an end in the NE of USA. Hopefully some others will be able to continue the time-series coverage for possible pulsations/fluctuations of Nova Del 2013.
Clear skies,
Helmar (AHM)
I have now submitted my Oct 1st observations as untransformed, Tri-color-green measurements, with a zero point from Tycho2 VT measurements. Blocks of 10 individual measurements with 13 sec exposure time and 20 sec cadence were averaged to a single reported measurement.
Unfortunately there is again very little overlap with other observers, but I'm confident that at least the first "hump" that you see in the LC is real. In general the amplitude and timescale of variations look similar to what others have observed here.
Whether you want it or not, you are always making measurements in 3 filters with a DSLR, and I also looked at the lightcurves from the blue and red filters. The blue filter results look very much like the green filter LC, so in other words the color (TB-TG) is almost constant within the uncertainty of the meaurements. The red light curve looks quite a bit smoother, and for example the first "hump" is not visible in the TR lightcurve. I guess this would mean that the "hump" is in the continuum part of the spectrum, not the H-alpha line?
Cheers
HB
The Nova Del 2013 light curve seems to show some interesting structure in my time series data taken on Oct. 6th. The comp - check star plot appears to be fairly flat except for some scatter over the same period.
More info on my web site: https://sites.google.com/site/kseobservatory/
-Emery (EEY)
Great Light
Hello Emery
Great Light curve. Thanks for observing for a long time. This behavior has been hinted at in prior nights, but not as dramatic.
Its clear in New England tonight, and I encourage everyone to observe as much as possible. It would also be great to get a West Coast and a Hawaii observer to go all night, to define what the late light cure looks like. Its also necessary to have a couple of obserers at the same time.
I have to travel tomorrow at 8am, butI plan to observe all night.
Gary
Hello Gary,
Thanks for the encouraging comments. I've sent in some more observations made on Oct. 8th which show some clearly visible fluctuations in the V339 light curve. These are the most dramatic yet and it would be great if other observers could confirm that these are real and not some systematic errors in my equipment.
I've posted the "quick look" image that AIP4Win generates from the MMT here: https://sites.google.com/site/kseobservatory/
The weather doesn't look favorable for observing over the next few nights here in the San Diego area but I'll continue to cover V339 Del as long as I can.
-Emery
Emory,
I have recorded pulsations of Nova Del 2013 from the nights of Oct. 6-7 EDT and Oct. 7-8 that seem to overlap with your observatons and they clearly show pulsations - at least one period a little more than an hour and a hints of a longer period. Graphs attached.
Taken with a 14 inch telescope with a focal reducer, CCD camera: SBIG ST7xme, Green filter, 20 sec integrations, dark and flat field compensated, but not transformed to V yet.
The first graph is broken-up because of trying a lower CCD temperature midway through, soon to be succombed by frost. The second graph was interrupted by an unexpected program/camera "crash" after only 2 hours while unattended.
Another run with exceptionally clear skies again tonight (Oct8-9).
Don Collins (CDK)
Hello Don,
Thanks for posting your light curves! It looks like the amplitude and period of the fluctuations correspond with my observations. Have you sent in any of your recent data? It would be interesting to see if our observations overlap in VStar.
I know how frustrating equipment problems can be. I operate my ST-7E at -10C and I usually have to wait over an hour for it to cool down and "defrost" before I can use it. Baking out the desicant plug every other month helps. I use CCDSoft V5 and it never crashed on my PC (the PC is mounted below the scope and also runs the scope remotely). It's old software but seems to be reliable.
It looks like we'll have cloudy nights here in San Diego until Friday!
Best wishes,
-Emery
The pulses that I observed on Oct. 5-6, Oct. 7-8 are not visible on Oct. 8-9 (last night). For comparison here are the links for the graphs for Oct. 5-6, Oct. 7-8 :
http://www.aavso.org/sites/default/files/NovaGraph20131006.png
http://www.aavso.org/sites/default/files/NovaGraph20131007.png
and last nights flat graph is attached.
Don