Author Topic: Boyajians Star Updates And Discussion  (Read 295758 times)

Offline Star One

  • Senior Member
  • *****
  • Posts: 13996
  • UK
  • Liked: 3974
  • Likes Given: 220
Re: Boyajians Star Updates And Discussion
« Reply #140 on: 04/02/2017 09:29 am »
The rumour goes that there will be results from other observatories which will end this kind of speculation, as that's been the inference when others have made similar mistaken speculations.

Can you clarify what the rumours are saying?

Look on the first link I posted in my post directly before the one you've quoted here. Before you ask yes it's very vague but that's not surprising in the circumstances. You'll also need to follow the link within the link.
« Last Edit: 04/02/2017 09:31 am by Star One »

Offline TakeOff

  • Full Member
  • ***
  • Posts: 392
  • Liked: 85
  • Likes Given: 115
Re: Boyajians Star Updates And Discussion
« Reply #141 on: 04/02/2017 12:32 pm »
Okay, with v-shape I mean the sharp bottom. The shadowing never remains constant. The asymmetries around is another big headache. The light curve has sharp dips that are suddenly cut off by a sharp rise. Is this characteristic for signals generated by electrical glitches?


The very clean dip two years before the mess at the end of the light curve must really make it all more challenging to explain. So it wasn't a one off event, but something that happens every other year or so. Once as a clean signal and later repeated with similar shape and amplitude but embedded in noise.


What confidence intervals should one have in mind when looking at the light curve in Boyajian's paper? How good is the time resolution of the dips, could it be that they went down to zero for a few seconds?


Day 1559 the light curve in the paper is suddenly cut by half of a percent, when it reaches its highest level and during an otherwise calm period. Apparently some kind of recalibration. It makes me wonder about the 100 year and 4 year dimmings, how well calibrated they are. If one suddenly shifts the entire light curve by half of a percent every other year one might make up any kind of funny correlations. Or maybe there's no data because the curve looks flat before that drop? There are a couple of longer periods with flat curves, is there no data about potential dips there too? Like around day 805 just after the first clean big one.


What about the upside here? After day 1528 the star is up to almost 1% brighter than the normalized average. And not in any transiting manner. What could brighten a huge F-star like that by half of a percent steadily for two weeks? Or is the normalization off?

Online jebbo

  • Full Member
  • ****
  • Posts: 940
  • Cambridge, UK
  • Liked: 608
  • Likes Given: 309
Re: Boyajians Star Updates And Discussion
« Reply #142 on: 04/02/2017 01:37 pm »
Okay, with v-shape I mean the sharp bottom. The shadowing never remains constant. The asymmetries around is another big headache. The light curve has sharp dips that are suddenly cut off by a sharp rise. Is this characteristic for signals generated by electrical glitches?

No, as has been said repeatedly, instrumental errors have been pretty much eliminated as an explanation.  It was looked at very seriously, and the curve matches no conceivable fault.

The space craft rotated 90 degrees every quarter, such that each target star is on a different CCD module, in a different orientation. Which means not only that different pixels are collected, but the optimal aperture for wach quarter is different as well. So any fault would have to affect *only* the pixels in the optimal aperture on 4 different CCDs and not affect any adjacent pixels.

Quote
What confidence intervals should one have in mind when looking at the light curve in Boyajian's paper? How good is the time resolution of the dips, could it be that they went down to zero for a few seconds?

No. The CCD pixels collect flux for 30 minutes (in long cadence). Then an overall flux value is created from a set of aperture pixels selected from the current pixel centroid and point spread function.

Quote
Day 1559 the light curve in the paper is suddenly cut by half of a percent, when it reaches its highest level and during an otherwise calm period. Apparently some kind of recalibration. It makes me wonder about the 100 year and 4 year dimmings, how well calibrated they are. If one suddenly shifts the entire light curve by half of a percent every other year one might make up any kind of funny correlations. Or maybe there's no data because the curve looks flat before that drop? There are a couple of longer periods with flat curves, is there no data about potential dips there too? Like around day 805 just after the first clean big one.

The processing of the curve (PDCSAP pipeline) is intended for transit detection and produces artefacts like that as a natural consequence of the processing - essentially the algorithm corrects for pointing errors and removes a moving-median from the curve (it is much more complex, but that is the essence), then normalises the values to "1".

When looking at absolute flux, this curve (though easily accessible) is NOT reliable, and you need to calibrate against the full frame images (and other stuff).

Quote
What about the upside here? After day 1528 the star is up to almost 1% brighter than the normalized average. And not in any transiting manner. What could brighten a huge F-star like that by half of a percent steadily for two weeks? Or is the normalization off?

As above, this is a direct consequence of the correction pipeline and is not real (think how a moving median removal behave when there is a massive prolonged dip).

As hop suggested, if you are serious, you need to learn how the instrument actually works and how it can fail.

--- Tony
« Last Edit: 04/02/2017 01:54 pm by jebbo »

Offline Star One

  • Senior Member
  • *****
  • Posts: 13996
  • UK
  • Liked: 3974
  • Likes Given: 220
Re: Boyajians Star Updates And Discussion
« Reply #143 on: 04/02/2017 02:56 pm »
Okay, with v-shape I mean the sharp bottom. The shadowing never remains constant. The asymmetries around is another big headache. The light curve has sharp dips that are suddenly cut off by a sharp rise. Is this characteristic for signals generated by electrical glitches?

No, as has been said repeatedly, instrumental errors have been pretty much eliminated as an explanation.  It was looked at very seriously, and the curve matches no conceivable fault.

The space craft rotated 90 degrees every quarter, such that each target star is on a different CCD module, in a different orientation. Which means not only that different pixels are collected, but the optimal aperture for wach quarter is different as well. So any fault would have to affect *only* the pixels in the optimal aperture on 4 different CCDs and not affect any adjacent pixels.

Quote
What confidence intervals should one have in mind when looking at the light curve in Boyajian's paper? How good is the time resolution of the dips, could it be that they went down to zero for a few seconds?

No. The CCD pixels collect flux for 30 minutes (in long cadence). Then an overall flux value is created from a set of aperture pixels selected from the current pixel centroid and point spread function.

Quote
Day 1559 the light curve in the paper is suddenly cut by half of a percent, when it reaches its highest level and during an otherwise calm period. Apparently some kind of recalibration. It makes me wonder about the 100 year and 4 year dimmings, how well calibrated they are. If one suddenly shifts the entire light curve by half of a percent every other year one might make up any kind of funny correlations. Or maybe there's no data because the curve looks flat before that drop? There are a couple of longer periods with flat curves, is there no data about potential dips there too? Like around day 805 just after the first clean big one.

The processing of the curve (PDCSAP pipeline) is intended for transit detection and produces artefacts like that as a natural consequence of the processing - essentially the algorithm corrects for pointing errors and removes a moving-median from the curve (it is much more complex, but that is the essence), then normalises the values to "1".

When looking at absolute flux, this curve (though easily accessible) is NOT reliable, and you need to calibrate against the full frame images (and other stuff).

Quote
What about the upside here? After day 1528 the star is up to almost 1% brighter than the normalized average. And not in any transiting manner. What could brighten a huge F-star like that by half of a percent steadily for two weeks? Or is the normalization off?

As above, this is a direct consequence of the correction pipeline and is not real (think how a moving median removal behave when there is a massive prolonged dip).

As hop suggested, if you are serious, you need to learn how the instrument actually works and how it can fail.

--- Tony

Also this has been explained to the OP at various times to no avail.

Offline M.E.T.

  • Senior Member
  • *****
  • Posts: 2311
  • Liked: 2912
  • Likes Given: 508
Re: Boyajians Star Updates And Discussion
« Reply #144 on: 04/02/2017 03:53 pm »
So regarding the obvious lack of a steady repetition of the dips. Does it suggest (as seems the implication to me), that whatever is orbiting the star is speeding up, slowing down, and/or changing orbits between one orbit and the next?

And also,  is the star "lifting" hypothesis as presented by one theorist a few months ago still a potential solution, given the data as it stands?

Offline TakeOff

  • Full Member
  • ***
  • Posts: 392
  • Liked: 85
  • Likes Given: 115
Re: Boyajians Star Updates And Discussion
« Reply #145 on: 04/02/2017 05:30 pm »
The space craft rotated 90 degrees every quarter, such that each target star is on a different CCD module, in a different orientation. Which means not only that different pixels are collected, but the optimal aperture for wach quarter is different as well. So any fault would have to affect *only* the pixels in the optimal aperture on 4 different CCDs and not affect any adjacent pixels.
There are other possible failure modes. Without any astrophysical explanation or independent observation, a technical malfunction must be the base case conclusion. Both big anomalies occurred just after the telescope was rotated, and rotated to the same orientation. But then the anomalies continue after it is turned again, so it's not about failing elements in the CCD, but other things can fail when the spacecraft is rotated.
Quote
Quote
What confidence intervals should one have in mind when looking at the light curve in Boyajian's paper? How good is the time resolution of the dips, could it be that they went down to zero for a few seconds?

No. The CCD pixels collect flux for 30 minutes (in long cadence). Then an overall flux value is created from a set of aperture pixels selected from the current pixel centroid and point spread function.
So the bottom flux of the light curve is unconstrained? Could it bottom out on  -100%?
Does the light curve really bounce from crashing to soaring within a 30 second period? From a star of about 2 million kilometer diameter.

Quote
What about the upside here? After day 1528 the star is up to almost 1% brighter than the normalized average. And not in any transiting manner. What could brighten a huge F-star like that by half of a percent steadily for two weeks? Or is the normalization off?

Quote
As above, this is a direct consequence of the correction pipeline and is not real (think how a moving median removal behave when there is a massive prolonged dip).

As hop suggested, if you are serious, you need to learn how the instrument actually works and how it can fail.

--- Tony

So, does the normalization uncertainty explain the brightening of about 1% above the normalized level, is it not considered to be astrophysical?
« Last Edit: 04/02/2017 05:33 pm by TakeOff »

Offline hop

  • Senior Member
  • *****
  • Posts: 3352
  • Liked: 553
  • Likes Given: 891
Re: Boyajians Star Updates And Discussion
« Reply #146 on: 04/02/2017 08:41 pm »
So the bottom flux of the light curve is unconstrained? Could it bottom out on  -100%?
Does the light curve really bounce from crashing to soaring within a 30 second period? From a star of about 2 million kilometer diameter.
Detailed descriptions of the pipeline and data products are available from https://keplergo.arc.nasa.gov/

However, it's important to remember that the output of the standard pipeline is not the only thing anyone has looked at. Instrumental error or pipeline artifacts was the most obvious candidate when this thing was initially discovered, and the original group looked at the data at various levels to rule that out. Others have subsequently re-analyzed the data in various ways.

Offline notsorandom

  • Full Member
  • ****
  • Posts: 1740
  • Ohio
  • Liked: 438
  • Likes Given: 91
Re: Boyajians Star Updates And Discussion
« Reply #147 on: 04/03/2017 02:18 pm »
So regarding the obvious lack of a steady repetition of the dips. Does it suggest (as seems the implication to me), that whatever is orbiting the star is speeding up, slowing down, and/or changing orbits between one orbit and the next?

And also,  is the star "lifting" hypothesis as presented by one theorist a few months ago still a potential solution, given the data as it stands?
Whatever blocked the light doesn't have to be in orbit of the star. It could be something like a gas or dust cloud in the interstellar medium. It could also be a comet swarm or something similar passing through the system or having been knocked onto a hyperbolic trajectory from the far reaches of the system by a passing star. It also doesn't have to be one single thing. A swarm of objects all in different orbits would make a pattern of dips that would look random until observed long enough to detect a pattern.

Offline TakeOff

  • Full Member
  • ***
  • Posts: 392
  • Liked: 85
  • Likes Given: 115
Re: Boyajians Star Updates And Discussion
« Reply #148 on: 04/03/2017 02:39 pm »
So regarding the obvious lack of a steady repetition of the dips. Does it suggest (as seems the implication to me), that whatever is orbiting the star is speeding up, slowing down, and/or changing orbits between one orbit and the next?

And also,  is the star "lifting" hypothesis as presented by one theorist a few months ago still a potential solution, given the data as it stands?
Whatever blocked the light doesn't have to be in orbit of the star. It could be something like a gas or dust cloud in the interstellar medium. It could also be a comet swarm or something similar passing through the system or having been knocked onto a hyperbolic trajectory from the far reaches of the system by a passing star. It also doesn't have to be one single thing. A swarm of objects all in different orbits would make a pattern of dips that would look random until observed long enough to detect a pattern.
Two Earth years apart explaining both the single very clean asymmetrical first deep dip, and then two years later also explain the pretty chaotic light curve including a couple of deep asymmetric dips too? Somehow anything passed between us and it, but never between us and any other star (as in 160,000 stars in the same telescope), in order to cause this light curve? Who can blame me for not buying that. I might've bought one of them, but not both.

Offline Star One

  • Senior Member
  • *****
  • Posts: 13996
  • UK
  • Liked: 3974
  • Likes Given: 220
Re: Boyajians Star Updates And Discussion
« Reply #149 on: 04/03/2017 02:48 pm »

Offline as58

  • Full Member
  • ****
  • Posts: 835
  • Liked: 300
  • Likes Given: 186
Re: Boyajians Star Updates And Discussion
« Reply #150 on: 04/03/2017 04:04 pm »
So regarding the obvious lack of a steady repetition of the dips. Does it suggest (as seems the implication to me), that whatever is orbiting the star is speeding up, slowing down, and/or changing orbits between one orbit and the next?

And also,  is the star "lifting" hypothesis as presented by one theorist a few months ago still a potential solution, given the data as it stands?
Whatever blocked the light doesn't have to be in orbit of the star. It could be something like a gas or dust cloud in the interstellar medium. It could also be a comet swarm or something similar passing through the system or having been knocked onto a hyperbolic trajectory from the far reaches of the system by a passing star. It also doesn't have to be one single thing. A swarm of objects all in different orbits would make a pattern of dips that would look random until observed long enough to detect a pattern.
Two Earth years apart explaining both the single very clean asymmetrical first deep dip, and then two years later also explain the pretty chaotic light curve including a couple of deep asymmetric dips too? Somehow anything passed between us and it, but never between us and any other star (as in 160,000 stars in the same telescope), in order to cause this light curve? Who can blame me for not buying that. I might've bought one of them, but not both.

Isn't it just as hard to explain why this purported observational error happens (many times, even!) only with this particular star?
« Last Edit: 04/03/2017 04:08 pm by as58 »

Online jebbo

  • Full Member
  • ****
  • Posts: 940
  • Cambridge, UK
  • Liked: 608
  • Likes Given: 309
Re: Boyajians Star Updates And Discussion
« Reply #151 on: 04/03/2017 04:16 pm »
Isn't it just as hard to explain why this purported observational error happens (many times, even!) only with this particular star?

Harder!

This supposed error only affects pixels in the optimal aperture for the star, and only when those pixels are observing that star (one quarter a year) ... and the pixels are on 4 different CCDs in different orientations, and the aperture is a different shapes in each of the different quarters. Read-out errors are precluded due to this as well ... very much special pleading.

--- Tony
« Last Edit: 04/03/2017 04:50 pm by jebbo »

Offline as58

  • Full Member
  • ****
  • Posts: 835
  • Liked: 300
  • Likes Given: 186
Re: Boyajians Star Updates And Discussion
« Reply #152 on: 04/03/2017 04:59 pm »
Isn't it just as hard to explain why this purported observational error happens (many times, even!) only with this particular star?

Harder!

This supposed error only affects pixels in the optimal aperture for the star, and only when those pixels are observing that star (one quarter a year) ... and the pixels are on 4 different CCDs in different orientations, and the aperture is a different shapes in each of the different quarters. Read-out errors are precluded due to this as well ... very much special pleading.

--- Tony

So it's aliens. Not building a Dyson sphere around the star, but instead they're messing up with Kepler.

Offline Star One

  • Senior Member
  • *****
  • Posts: 13996
  • UK
  • Liked: 3974
  • Likes Given: 220
Re: Boyajians Star Updates And Discussion
« Reply #153 on: 04/19/2017 07:36 am »
Have Aliens Built Huge Structures around Boyajian's Star?

https://www.scientificamerican.com/article/have-aliens-built-huge-structures-around-boyajian-rsquo-s-star/

Unfortunately it's behind a paywall for now.

Offline ChrisWilson68

  • Senior Member
  • *****
  • Posts: 5266
  • Sunnyvale, CA
  • Liked: 4992
  • Likes Given: 6459
Re: Boyajians Star Updates And Discussion
« Reply #154 on: 04/19/2017 08:15 am »
I think we should split this thread into two separate threads.

Thread A: For the one poster who thinks it's an observational error to argue with everyone else.

Thread B: For all other discussions of Boyajian's Star exception debating whether it's likely to be observational error.

I think the vast majority of people here are more interested in B than A and it's unfortunate all those people have to wade through so many posts that belong in thread A.

Offline baldusi

  • Senior Member
  • *****
  • Posts: 8356
  • Buenos Aires, Argentina
  • Liked: 2539
  • Likes Given: 8273
Re: Boyajians Star Updates And Discussion
« Reply #155 on: 04/19/2017 12:26 pm »
I think we should split this thread into two separate threads.

Thread A: For the one poster who thinks it's an observational error to argue with everyone else.

Thread B: For all other discussions of Boyajian's Star exception debating whether it's likely to be observational error.

I think the vast majority of people here are more interested in B than A and it's unfortunate all those people have to wade through so many posts that belong in thread A.
There's also the "Ignore List" and the "Report to moderator" functionality.

Offline LouScheffer

  • Senior Member
  • *****
  • Posts: 3381
  • Liked: 6109
  • Likes Given: 836
Re: Boyajians Star Updates And Discussion
« Reply #156 on: 04/19/2017 02:58 pm »
Here's a horribly non-scientific, ad hominum observation that has changed my mind about whether the long term dimming is real.

There is debate over this issue, with Schaefer convinced of long term dimming, and Hippke arguing that it's statistical fluctuations in a difficult-to-calibrate data set.  Recently I had cause for an email exchange with Hippke, on a different topic.  From these I get the impression of a smart, enthusiastic, big picture guy.   But conversely, at least in my interaction, he was not the kind of scientist who obsessively worries about the finest of details.

Extracting the long term dimming, if any, from these difficult data sets, needs extreme attention to detail.  I've never met Schaefer, but the impression I get from his papers is that's right up his alley. 

So before, there were two pieces of evidence for dimming (Schaefer long term, and Montet's dimming during Kepler).  There was one against, which was Hippke.  Classically, science should be only about the evidence, and not about the scientist, but in a true Bayesian analysis all information should be considered (especially since I don't have the time, inclination, or data to do the analysis myself).  My personal guess as to whether the dimming is real has risen from about 2/3 to about 9/10, based upon my new found skepticism of the competing analysis, based only on the apparent personality of the proposer.  Scientifically ugly, but that's the way I feel....

Offline Star One

  • Senior Member
  • *****
  • Posts: 13996
  • UK
  • Liked: 3974
  • Likes Given: 220
Re: Boyajians Star Updates And Discussion
« Reply #157 on: 04/23/2017 09:13 pm »
Should be a new SETI related paper from Jason Wright published Tuesday.

https://mobile.twitter.com/Astro_Wright/status/856208078282178560

Offline as58

  • Full Member
  • ****
  • Posts: 835
  • Liked: 300
  • Likes Given: 186
Re: Boyajians Star Updates And Discussion
« Reply #158 on: 04/25/2017 07:01 pm »
Should be a new SETI related paper from Jason Wright published Tuesday.

https://mobile.twitter.com/Astro_Wright/status/856208078282178560

This is the paper: https://arxiv.org/abs/1704.07263

(Not directly related to Boyajian's star, though)

Offline Star One

  • Senior Member
  • *****
  • Posts: 13996
  • UK
  • Liked: 3974
  • Likes Given: 220
Re: Boyajians Star Updates And Discussion
« Reply #159 on: 04/25/2017 07:05 pm »
Should be a new SETI related paper from Jason Wright published Tuesday.

https://mobile.twitter.com/Astro_Wright/status/856208078282178560

This is the paper: https://arxiv.org/abs/1704.07263

(Not directly related to Boyajian's star, though)

That's a pretty way out there topic for him to be writing a paper on!

Tweeted exchange by him about Doctor Who and the paper.

Quote
Jonathan McDowell‏ @planet4589
Prior technological species. Jason doesn't address Dr Who's Silurians specifically, though.
Jason Wright‏ @Astro_Wright
Replying to @Astro_Wright and @AdamFrank4
(link: https://arxiv.org/abs/1704.07263) arxiv.org/abs/1704.07263

Jason Wright‏ @Astro_Wright

Replying to @planet4589
Only because none of my Whovian tweeps told me about them when I asked for examples! ;-)
« Last Edit: 04/25/2017 07:29 pm by Star One »

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
0