Tabby's star has been in the geek media because of its inability to maint ain a constant light output like other stars of its type:
formatting link
.html
That's not weird enough. Now, some astronomers have found 234 stars in th e Sloan Digital Survey archives (of 2.5 million stars) that apparently emit periodic light pulses with about the same frequency/period*:
formatting link
The authors say it's *probably* not instrument glitches- if it were it sh ould apply to all the stars in the survey database, so it's either somethin g those few stars are doing or some sort of data mis-massaging that weirdly only works on those stars' spectra. Thus they recommend other investigator s look more closely with other instruments to confirm what they think they' re seeing.
The LGM option is discussed and more or less dismissed.
I say "frequency/period" because the data as presented is somewhat opaque to my "limited skill set". Anybody here with more math than me able to dec oncolvulate it for simple minds like mine? Guesses as to the cause are welc ome but not really relevant.
ain a constant light output like other stars of its type:
ry.html
the Sloan Digital Survey archives (of 2.5 million stars) that apparently em it periodic light pulses with about the same frequency/period*:
should apply to all the stars in the survey database, so it's either someth ing those few stars are doing or some sort of data mis-massaging that weird ly only works on those stars' spectra. Thus they recommend other investigat ors look more closely with other instruments to confirm what they think the y're seeing.
ue to my "limited skill set". Anybody here with more math than me able to d econvolute it for simple minds like mine? Guesses as to the cause are welco me but not really relevant.
formatting link
picks up the same story. The take-away message is that the data needs to be looked at more closely. The neutrinos that seemed to be travelling faster than light come to mind, though little green men are a trifle more plausibl e than super-luminal neutrinos.
I'm not entirely sure what the astronomers are claiming they see in the spectrum, either -- it seems to be some modulation of the light at
1.6THz, though.
I suspect that the most impressive result that'll come out of this is that they've found some new kind of star, and not one that happens to be colonized by the Orion Syndicate or any other green men, little or otherwise.
--
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
I'm looking for work -- see my website!
I am a little suspicious that the bin where this artefact sometimes appears is awfully close to 768 which could easily be a result of some quirk in the FFT algorithm and interpolation that they are using.
If real it could mean that the absorbtion lines in the star are being moved in a peculiar way or that sometimes the instruments monochromator is somehow vibrating. It would be interesting to check that the stars they have found show the same behaviour everytime and also when observed by an instrument using a completely different technology.
If it is real then it is certainly a curiousity that will need explaining. My money is on an instrumental artefact of some sort.
I have no idea how the Sloan survey is compiled, but it would certainly be worth investigating whether those particular stars are the only ones being imaged by the same instrument.
Yes, thats an issue. The only way of detecting this seems to be to do the Fourier analysis of the spectrum. You certainly cant detect individual pulses. And that means you cant see if the pulses are being modulated with any 'intelligent', ie non-random, recognisable signal. So this is a means of detecting a 'beacon' without being able to read a signal off it. Interesting, but remains to be proven whether its being generated by anything intelligent, and that is even after all the possible detection artifacts are eliminated.
OK. I'll play then with my pointy headed theorist hat on.
It is really hard to see how something can vary at terahertz rates given the normal heuristic that to vary measurably on a timescale of t an incoherent luminous object must be roughly of size ct. That still leaves the option of coherent emission of some sort and beats.
So my suggestion as an explanation is that the F2 stars come in multiple sorts super giants like Polaris, main sequence like Procyon and maybe the left wing of the giant branch. I never paid much attention to stellar stuff beyond the HR diagram so I'm no expert.
I reckon it is only the right sort of supergiant showing this.
It strikes me that the possible get out of jail free card here is that the super giants will have huge deep chromosphere full of extremely processed material with relatively high metallicity and insane numbers of absorbtion and emission lines. If there was some way that an optically pumped laser could form in the atmosphere of such a star with the right composition then the variation seen could be the result of beats between two very similar frequency laser lines.
I hesitate to suggest this option partly because there is a lot of crank stuff on the internet about laser stars which mostly gets hits on completely mad cranks or SETI rather than astrophysical processes. The only one a quick search came up with was a neutral hydrogen near IR one in ngc7027. OTOH OH masers in star formation regions are two a penny.
There is a precedent for some F class stars being a bit weird too with plenty of strong emission lines excited in the stellar atmosphere. Some are rich in strontium and others in chromium & europium (Brexiteers cover your ears). From the online catalogue of spectra the normal ones:
formatting link
To make life more interesting it is also known that their properties vary a lot with intrinsic luminosity:
formatting link
And then there are the real oddballs:
formatting link
(OK they are F0 stars but I reckon it still holds)
And we know that certain lanthanides will lase rather well under the right conditions. Since these stars are shedding atmosphere in intense solar wind the conditions might make laser emission possible sometimes.
It is no more than a guess.
But perhaps I was a bit too keen to dismiss it as an instrumental artefact. The obvious thing they should do is look at the spectra of the brightest ones and use another sampling rate for their observations.
Lanthanides aren't going to be abundant anywhere. Their atomic numbers run from 57 to 71, and anything heavier than iron (atomic number 26) is formed by neutron capture in a supernova. They aren't called rare earths for nothing.
A little knowledge is dangerous. You are way out of date. Rare earths are only "rare" in the sense of mineable ore deposits being uncommon.
formatting link
Heavier elements are also formed inside massive luminous red giant stars by slow neutron capture in significant enough amounts to show up in the stellar spectra hence the strong strontium and europium lines.
formatting link
Or a more technical lecture PDF which looks OK at a first glance
formatting link
Super giants do quite a good line in generating elements that require endothermic reactions as they burn through lighter fuel towards iron.
It makes strontium, barium and lead a lot more common than they would otherwise be since they are locally more stable end points.
Rare earths aren't actually all that rare either - just hellishly difficult to separate. Uranium is about as common as Tin on Earth and 3x more common than Tungsten whilst the Cerium abundance is between Cobalt and Nickel neither of which are considered "rare".
Rare earths with odd numbers of protons are somewhat rarer.
What a pity that your link is to a text about abundance in the earth's crus t. For information that has more to do with the abundance of about elements in stellar atmospheres see
formatting link
Note that the scale is logarithmic
If you'd got through to page five of that documents you'd have got a rather better looking graph of exactly the same information as I posted above.
There aren't all that many iron atoms around and a hell of lot less of anyt hing as heavy as a lanthanide.
But there still isn't all that much of any of them.
The surface of the earth isn't the atmosphere of a star (which is what you were talking about). Nickel isn't much heavier than iron (28 protons versus 26), and not much less abundant in the solar system as a whole, but cobalt is much less abundant than either, and cerium is as much less abundant tha n cobalt as cobalt is less abundant than nickel.
True. That does seem to be driven by nucleosynthesis.
I tried to read the article. I think what they are doing is converting light wavelength to frequency, becuase it makes the Fourier bins work out better, ie. linear, instead of logarithmic. So, the 1.6 THz isn't the modulation, it is the frequency of the light. Assuming I was getting that right (not at all clear) I was not able to find the frequency of the modulation. But, since they mentioned comets or manufactured structures orbiting the stars, it seems the modulation is at a MUCH lower frequency, like hours or days. But, I couldn't find any reference to that in the paper.
It falls into the category of explaining a misunderstanding of what they are claiming. I thought they were claiming temporal modulation of the starlight at the stated timescale. I read the article more carefully this time instead of skimming it. What they are actually saying is that they find a sinusoidal variation of the spectrum amplitude with frequency (though the data is collected proportional to wavelength). They express this as a time (since their raw data is in frequency space). TBH I'd prefer the discrete FFT bin number.
Looks to me now more like it an artefact induced by however they are subtracting off the continuum baseline. Why it only affects some stars is a mystery. Averaging all the spectra showing this signal together and processing it might shed some light on what is going on.
In summary what they are claiming is that in frequency space the amplitude of the spectrum is modulated by a small sinusoidal variation at a very specific delta T. That is an FFT of the amplitude-frequency graph gives a spike in roughly the same place bin 765 every time there is a detection. This means the spectrum is amplitude modulated so that
I(f) = 1 + eps.sin(a+f/F')
It worries me a bit that their independent analysis of the red and blue spectrographs seems to produce a much weaker result than the two used together. Some quirk in how they are combined might also be a factor.
In their position I would compute these Fourier components in original wavelength space and correlate them with the data since they only need to look at about 11 discrete bins of the FFT and 11.N will beat N.lg2N and avoid doing any awkward interpolations.
I can see how a Fabry Perot etalon or bad antireflection coating might result in something that generated amplitude interference for N.lambda a la Newtons rings but I am at a complete loss as to how you get systematic errors in sensitivity at lambda/N aka N.f.
Any ideas for a systematic error in a fibre based spectrograph that could introduce this sort of frequency dependent amplitude response?
I find the idea that all ETI's communicate by broadcasting bursts of
700+ Dirac delta functions at the same deltaT distinctly uncompelling. There is something odd going on - but what is it?
On Oct 28, 2016, Martin Brown wrote (in article ):
Their explanation is difficult to follow. Judging from their address, their native language is likely to be French.
What they are looking for is amplitude modulation of a optical carrier. I haven?t done the math, but I?d bet that if you apply their math to a sinewave carrier modulated by a square wave at 1% of the carrier frequency, you?ll get something like their results. A random telegraph data signal will give similar results.
The strange math is necessary because they are trying to pull a very weak signal out of measured spectral data expressed in uniform wavelength increments, which will not be uniform when transformed into the frequency space needed for the Inverse FFT that yields the plots. Although not clearly stated, it appears that they are calculating the autocorrelation by the FFT route. The details may be better explained in Borra (2010).
I?d suggest these options to Borra, whose email address is on the first page of the article. The answer should be illuminating.
The Introduction provides references to the genesis of their method, most likely with a full mathematical analysis. Borra (2010) seems to be the place to start.
The bulk of the article is devoted to analyzing and ruling out all such possible alternate explanations. This analysis is required for any surprising observational astronomy article to be taken seriously.
The fact that only a few stars (but not galaxies et al), falling into a very definite range of spectral types, and so on, pretty much rules out local instrument problems. The many surrounding stars and galaxies provide the control -- those instrumental problems would affect everything in the picture, not just a single star.
If the ETIs were using an optical laser beam (~ 10^14 Hz) that was amplitude modulated by a Terahertz (~ 10^12 Hz) digital data signal, one would get such a spectrum. We cannot produce such a signal at such a power level with anything like today?s technology, but that will solve itself over the next century or two.
The real counterargument is that the power level is about one hundred thousandth of the total optical output of a sun-like star -- 10^-5 of a star is still a hell of a lot of power.
I don?t buy the theory that the ETIs are beaming the energy directly to us, as they would have no way to know where to send it - we are thousands of light years away and don?t radiate obviously unnatural signals with sufficient power to be detectable at such a distance. So, the modulated light is radiated isotropically, which suggests that it?s the star itself that?s doing this. In any event, the Astronomy community will soon chime in, and they know well the pitfalls of astronomical measurements.
Thank you Martin Brown. This is exactly the kind of "dumbed down" analysi s I was looking for.
ir
At the risk of seeming bigoted, that would explain a few things.
r math to a
y,
signal
But as you say below, not in a background of similar signals that don't s how the correlation.
rly
FT
Well now. Did they arm themselves with "strange math" to go looking for s omething strange, or did they stumble over it while trying out their "stran ge math"?
d
first
ace
a
sing
ery
This is the sole reason I posted my appeal for simplification. If it hadn 't been for that I'd have just waited for other researchers to take it seri ously enough to do the cross-checking.
ude
such
the next
That's the other simplification I was hoping for; if I had access to the appropriate instruments, how would I set them up and what would I see on wh ich instrument? They could have just said that (grumble).
Thank you.
tar
tly to us,
th
ight
lf
Hrm. We have such ideas as launch lasers that have made it from science f iction to NASA's blue-sky drawing boards that could easily have equivalent ERP in a very narrow beam, but not AFAIK modulated.
OTOH if you're using such a laser to propel an unpowered (and presumably uncrewed) instrument package (a la Starwisp) I can see modulating it to get return Doppler data on the spacecraft's acceleration/velocity profile.
Okay, that's way past blue-sky (into black-sky?). One ET civilization sho oting their version of Starwisp at us at random I could look at with only o ne raised eyebrow, but more than two hundred of them at wildly different di stances, not so much. It almost certainly has to be something new in stella r physics along the lines of the large-scale gas lasers mentioned elsethrea d.
ell
Yup. I just hope it's not "soon" as defined by astronomers and geologists ...
On Oct 28, 2016, snipped-for-privacy@bid.nes wrote (in article):
Borra?s 2010 article is very clear, as it turns out.
Having read Borra?s 2010 article, it?s all clear. The issue is finding the modulated light despite the glare from the local star, as suspected. The objective is to detect ETIs who are trying to be noticed in the course of astronomical observations.
The proposed signal consists of pairs of femtosecond laser pulses separated by picoseconds, modeled on how one generates a frequency comb. The shorter the individual pulses, the wider the bandwidth. The spacing between the pulses governs the spacing between the teeth of the frequency comb.
In practice, the comb is buried under the light from the local star, so the trick is to extract it anyway.
Step one is to convert the spectrogram from the wavelength scale to the frequency scale, because the comb teeth are uniformly spaced only in frequency space.
Step two is to take in inverse FFT of the spectrogram in frequency space, yielding the autocorrelation of the original data. The axis is time offset. Any frequency comb will generate a peak at the time spacing of adjacent pulse pairs.
Autocorrelation? You will not find this word used by Borra, but that?s what he has computed. Recall the two routes to compute autocorrelation, direct computation using the definition, or by use of the fourier transform.
In the fourier transform, one computes the complex spectrum, converts it to a power spectrum, and then computes the inverse fourier transform of the power spectrum to yield the autocorreclation (neglecting correction factors).
Astronomical measurements of the spectrum give the power spectrum, so an inverse fourier transform yields the autocorrelation directly.
Now that I?ve read Borra?s papers, I see that they did expect that the ETIs are intentionally directing beams at promising nearby stars (there are about a million stars within 1000 light years), hoping to be noticed. Borra goes through a rough analysis that shows that one can reach 1,000 light years using present-day Earth technology, albeit at great expense, so Borra feels justified in assuming that the ETIs (who are assumed to be advanced far beyond Earth) can have done this.
But this would require for more attention span than Humans are likely to muster - any signal detected today must have been sent 1,000 years ago, and any reply would not be received for 1,000 years more.
Given that we cannot go visiting, it?s not clear what the point would be. If faster than light travel is invented, there would be no reason to bother with a laser beacon - just go visiting.
I think we are safe - 1,000 years to get here, plus 1,000 years for the return trip.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.