Affiliation
American Association of Variable Star Observers (AAVSO)
Wed, 03/16/2016 - 22:31
My understanding is that observations around minima for the Bulletin stars is a higher priority. What would "around minima" mean? Could it be a percentage of days of the period? So for an LPV with a period of 200 days, one could observe for 10% of the period on either side of the predicted minima; 20 days before and 20 days after.
Thoughts?
Michael
[quote=CMJA]
My understanding is that observations around minima for the Bulletin stars is a higher priority. What would "around minima" mean? Could it be a percentage of days of the period? So for an LPV with a period of 200 days, one could observe for 10% of the period on either side of the predicted minima; 20 days before and 20 days after.
[/quote]
I believe the issue with minima of LPV's is they typically lack sufficient observations to accurately identify the minimum dates. Mainly because many of them get fairly faint, and observers are unable to make positive observations down in those levels with their equipment at hand.
That said, ANY and several positive observation(s) "around" the predicted minimum time is of great value. Precisely what percentage of time to do them is not important 10%, 20% or even the whole "lower half" of the light curve is valuable!
Mike
Yes I was thinking
Hi Mike,
Yes I was thinking of going where most visual observers don't; in the fainter regime. I am trying to set-up an automated survey without having to tweak CCD exposure times to avoid saturation or get enough SNR. If I can limit the range of the magnitude that I need to consider, it makes setting this up a little easier. Thanks for the feedback!
Michael
Hi Mike! I use 30s for CVs and 20s for Miras (which tend to be brighter). I can get up to mag 10.0 without saturating, or even well into 9th if my focus is off. :) Beyond that, I hacked ACP to adjust the number of exposures for each star (CV or Mira) based on measuring the brightness of the variable in the first frame, and then stack whatever exposures are actually taken.
You can see my system in action in three videos I posted on YouTube:
https://www.youtube.com/channel/UCFF-jyftqHUeQwsVMh99fXQ
Messed up the thread/quote
ACP knows nothing of variable stars of course. I added code to the acquireimages file so that it could locate and measure the brightness of the variable in the first frame and then change the number of exposures accordingly (changing the exposure was more problematic, so I let that go).
Hacking ACP. Of course! What was your hack?
Good videos. You sound like the NASA launch narrator - "Go for throttle up".
[quote=MDW]
Hi Mike! I use 30s for CVs and 20s for Miras (which tend to be brighter). I can get up to mag 10.0 without saturating, or even well into 9th if my focus is off. :) Beyond that, I hacked ACP to adjust the number of exposures for each star (CV or Mira) based on measuring the brightness of the variable in the first frame, and then stack whatever exposures are actually taken.
You can see my system in action in three videos I posted on YouTube:
https://www.youtube.com/channel/UCFF-jyftqHUeQwsVMh99fXQ
[/quote]
Hi Michael C.:
I know you are a ccd observer so I would look at "when to observe" in a similar way to what Mike L noted, but propose a different protocol. Using Table 3.1 – Typical limiting magnitudes from the Visual Observing Manual, you could identify those LPVs that most visual observers cannot follow near minimum. They are too faint as Mike L. mentioned. Search (filter) the Bulletin for a full list of the LPVs that get below say 12 (or 13?). This could be your starting list of your variables of interest. Use LCG or VStar to look at recent light curves for these stars. Note how much time they spend below this magnitude. You can note specific durations around minimum or just make some general observations. Although this is tedious at best, it might help provide a more accurate answer to your question. You could choose to monitor some of these LPVs (or all) for this period of time. Of course, is it worth that much effort compared to a "guesstimate"?
I think the other question is how frequently to measure them? Although the norm is a frequency of weekly or so, I propose 1% of the period (every 2-3 days or so). Taking weather into account perhaps you could try every clear night? Very few ccd observers are doing what you propose! Thus, we have the lack of good light curves at minimum for most of these targets.
As Mike L stated "even the whole "lower half" of the light curve is valuable!". The other side of the question is what new information is gained by monitoring the LPVs so often. Historical completeness versus astrophysical info?
Any help? I suspect you have asked yourself the same questions?
Ken
Yes, I have Bulletin
Hi Ken,
Yes, I have Bulletin 79 in a dB and I have filtered as you mentioned. I had looked at a couple dozen of light curves in LCG with a cross-section of periods (100, 200, 300, 400+ days). That helped generalize what I think I need to set-up in terms of observing and CCD exposure parameters. I've already characterized what exposure I need in which band-pass and for the colour of the star (B-V).
As I mentioned to Mike, I building an automated system to survey the Bulletin near minima without having to continually manually tweak too many things. I want to set it and forget it, and let the observatory go off and do its thing. I'm glad you mentioned the weather constraint. Given my so-so clear skies in the southern Ontario, I'd just let the system get observations any clear night it can. If I observe an LPV with a 200-day period around minima at 10% of its period, that would mean I would get an observation for 20 days on either side of minima, but probably way fewer days due to weather. Generalizations? Yes. Hopefully some reasonable compromises
Thanks for your thoughts.
Michael
Michael:
You mentioned the other annoying parameter - exposure length. I monitor the LPVs (Miras) too. I never wanted to spend the time to find out where in the cycle the Mira was so I punted! I selected two exposure times (1 long; 1 short) to cover their typical full range of magnitude. I want to transform the data so I always use two filters. Over the years I have changed my choice of filters and modified the exposure lengths.
I have ended up with "one" scheme (I use CCD Autopilot) that works for me. I started with B;V filters. I used B at 105s and 10 s. I used V at 90s and 10s. In the beginning I was running duplicates of each. It was taking a long time to do all the nightly LPVs. I changed to V;I. Same time as mentioned for V BUT the Miras are so red that the only I filter duration that worked was about 5s and even then they were often saturated, even at minimum!! Impacted by scintillation too? So, I got angry with the I filter run because it was often saturated. Annoying!
Again I punted. Even though the I filter made more sense to me astrophysically, I switched to an R filter. All my filters (BVRI) have transform coeffs so I switched to V;R. I only need to run the R at 7s. No longer R duration. I really wanted the V transformed magnitude. Not so interested in the transformed R magnitude but I do report both.
So this is now my regular protocol. It works nearly all the time and I never have to "think" about it any more, no matter what the magnitude of the LPV (max or min)! One of the V image durations (90s or 10s) and the 7s R duration nearly always work, so the run is "successful" no matter what the phase. And, the additional / necessary short durations don't add much total time to the runs.
Ken
Hi Ken,
Good info, and strategies to consider, thanks!
I wonder if Ic/ND filters are available? Or one can stack a ND filter with the Ic filter in the same cell? My Astrodon Ic is 3mm thick and a Baader ND is 2mm thick. I think my filter cell can accommodate 5mm (just) with the retaining ring hanging on by one thread.
Maybe the AAVSO should just have a Bulletin CCD program for fainter stars.
[quote=MZK]
Michael:
You mentioned the other annoying parameter - exposure length. I monitor the LPVs (Miras) too. I never wanted to spend the time to find out where in the cycle the Mira was so I punted! I selected two exposure times (1 long; 1 short) to cover their typical full range of magnitude. I want to transform the data so I always use two filters. Over the years I have changed my choice of filters and modified the exposure lengths.
I have ended up with "one" scheme (I use CCD Autopilot) that works for me. I started with B;V filters. I used B at 105s and 10 s. I used V at 90s and 10s. In the beginning I was running duplicates of each. It was taking a long time to do all the nightly LPVs. I changed to V;I. Same time as mentioned for V BUT the Miras are so red that the only I filter duration that worked was about 5s and even then they were often saturated, even at minimum!! Impacted by scintillation too? So, I got angry with the I filter run because it was often saturated. Annoying!
Again I punted. Even though the I filter made more sense to me astrophysically, I switched to an R filter. All my filters (BVRI) have transform coeffs so I switched to V;R. I only need to run the R at 7s. No longer R duration. I really wanted the V transformed magnitude. Not so interested in the transformed R magnitude but I do report both.
So this is now my regular protocol. It works nearly all the time and I never have to "think" about it any more, no matter what the magnitude of the LPV (max or min)! One of the V image durations (90s or 10s) and the 7s R duration nearly always work, so the run is "successful" no matter what the phase. And, the additional / necessary short durations don't add much total time to the runs.
Ken
[/quote]