We recently made a bunch of SETI folk aware of some of the oddities in the 2/23/1987 neutrino data. Is there any interest in getting some sort of collaborative project going? If you haven't, please review the info at: http://bigfc252.posterous.com/tag/About
While an interesting subject for SETI analysis, the decodings you do seem very speculative. At 168,000 light years away, there is just no way that sentient beings at SN 1987A could have learned of human codes like ASCII (just 50 years old) within the bounds of known physics.
If you have access to the raw data of the event, perhaps you could attempt to use some of the algorithms we've been working with on this forum - e.g. FFT, autocorrelation, kurtosis and KLT - and see if anything stands out.
>While an interesting subject for SETI analysis
On 2/23/1987, for the first time in earthling history, 6 teams of highly regarded scientists used neutrino and gravitational wave detectors to observe an apparent type-II supernova. Portions of the collected data are inconsistent with what's expected to occur during a conventional astrophysical event. We believe that these points alone make the phenomenon a worthwhile candidate for SETI exploration.
“The Mont Blanc event, which occurred about four and a half hours before the Kamiokande one, appeared surprising because it did not ﬁt most a la page theories, according to which a gravitational stellar collapse must occur in a very short time, i.e. of the order of a few seconds or even less.” (P. Galeotti and G. Pizzella, “Sn1987a Revisited after 20 Years: May the Supernova Bang More Than Once?,” Arxiv preprint arXiv:0706.2235, 2007.)
“The 5 LSD events, occured 4.5 hours before the main signal, cannot be accounted for." (G. Pagliaroli, F. Vissani, M.L. Costantini, and A. Ianni, “Improved analysis of SN1987A antineutrino events,” Astroparticle Physics, vol. 31, 2009, pp. 163–176.)
“A signiﬁcant event cluster in the LSD experiment was observed several hours earlier and had no counterpart in the other detectors and vice versa. It can be associated with SN 1987A only if one invokes very non-standard double-bang scenarios of stellar collapse." (G.G. Raffelt, “Supernova neutrino observations: What can we learn?,” Arxiv preprint astro-ph/0701677 [ SPIRES], 2006.)
“Statistical analysis of all four detections leaves most scientists believing that something happened to produce the events that the Mont Blanc experiment detected, but that it was not SN1987A. The mystery remains as to why the Mont Blanc experiment happened to detect a greater number of neutrinos than usual on the same day that the supernova was detected. The scientists involved in neutrino detection are prepared to accept this as not nearly so great a mystery as the notion that the supernova could have emitted two separate blasts of neutrinos, more than four hours apart. Hence, they conclude – at least outside Mont Blanc – that the supernova's neutrinos were recorded by the detectors in Ohio and Japan, and probably in Russia, but not in the Italian Alps." (D. Goldsmith, Supernova! : the exploding star of 1987, New York: St. Martin's Press, 1989.)
>the decodings you do seem very speculative. At 168,000 light years away, there is
>just no way that sentient beings at SN 1987A could have learned of human codes like
>ASCII (just 50 years old) within the bounds of known physics.
We've opted to demonstrate a form of SETI analysis which appears unintuitive from earths current cosmic vantage-point. In our view, "creative SETI" should be encouraged, especially during a crowd sourcing effort.
>If you have access to the raw data of the event, perhaps you could attempt to use
>some of the algorithms we've been working with on this forum - e.g. FFT, autocorrelation,
>kurtosis and KLT - and see if anything stands out.
Unfortunately, beyond the posted data summaries, such raw resources are not usually available to the general public. I'm sure team SETI, though, would have no trouble tracking down all the relevant material, and finding out exactly how it's been analyzed so far.
I have no problem with "creative SETI". What I am telling you is that even if those decodings signified something, without far less ambiguous evidence, we would not be able to convince a wider community of their relevance, no matter how indulgent we here on the forum were of your claims.
The Institute, though, is ideally positioned to help convince a “wider community” that the detections are “odd” and worth exploring.
We're particularly curious about (further?) opportunities for data abstraction, so that the insight of our fellow earthlings can be harnessed without having to worry about the underlying code / math / physics. Of course, we're also extremely excited to see how the data stands up to more traditional forms of SETI analysis.
If it’s preferred, we'd be more than happy to take the bigfc site(s) offline, and re-submit the content to a platform of the community's choosing. The current “free social media” solution is fairly functional, but there’s obviously tons of room for improvement.
As far as the recently posted oddities go, we'll be sure to emphasize even more strongly that they’re "observational" in nature. If the material was posted to a more suitable collaborative platform (voting?), the community could judge for itself what’s most interesting.
So, to get the ball rolling. Avinash, how do you feel about crowd sourcing the 2/23 data? Is the current incarnation of the setiQuest platform suitable for such a project?
The neutrino was discovered in 1953 (http://www.ps.uci.edu/physics/news/nuexpt.html) which predates the first SETI observation. I imagine that from the very beginning, SETI scientists have thought about using this particle for signalling across or beyond the galaxy. The neutrinos pass right through everything. Except for gravitational lensing, they pass almost unimpeded from them to us. The practical problem we face is that neutrino detectors are fantastically inefficient. It is quite remarkable that the measurements referred to by bigfc252 are even possible.
I read the document referred to in the original posting a couple months ago (?) and found it an interesting read. As suggested by both anders and bigfc252, the use of ascii codes to signal us would require faster-than-light messaging; something that is impossible in the most widely accepted present day theories. Ultimately, the proposition of faster-than-light signalling implies time-travel. If ET can build a time machine, then all bets are off!
That said, ET might be closer than we think, so there is a scenario where the signalling could be possible. Unfortunately, we're not in the happy situation where we can build an instrument for SETI based on neutrino observations. Ultimately, this is why I didn't respond to the paper on the first go around.
Speaking for myself, I don't discourage speculation and exchange of ideas. On the other hand, we all focus our energies on areas where we can make the greatest impact. Since I don't think I can make an impact in the area, I don't plan to make any special moves in response to the neutrino report. But I will follow the conversation.
My 2 cents.
Thanks to Anders and Gerry for responding.
At the risk of sounding like a broken record, we’d like to (re)state why we feel the 2/23 observations are suitable candidates for SETI research.
The data is:
For what it’s worth, should someone (pro or amateur) decide to investigate this phenomenon, our group has no expectation of involvement. We’d of course appreciate being kept informed of any discoveries, and we’re happy to lend a hand if it’s requested of us, but we have no problem watching from the sidelines.
Prior to posting this message, we took the liberty of making several members of the Mont Blanc collaboration aware of this thread. Assuming a few of you found your way here, do you have any thoughts? In the name of science, would you be willing to make your raw data available to the SETI community?
@gerryharp You mentioned that “we're not in the happy situation where we can build an instrument for SETI based on neutrino observations.” Could you please elaborate? What sort of technical challenges stand in the way? Perhaps the Mont Blanc team could share their insight here as well.
The challenge is clearly that neutrinos interact only weakly with other matter (the same reason they can pass through anything unimpeded) - they would simply pass right through the detector without anyone or anything noticing. Current detectors, I understand, actually detect other types of particles that only very rarely are produced because of neutrinos passing by, not the neutrinos themselves.
So, if an alien civilization sent an neutrino signal towards us, a) we would actually not be able to detect it, only its arbitrary after-effects, and b) even these after-effects occur only very rarely, so chances that we would detect any given bit sent at us are low. End result: if aliens beamed Encyclopedia Galactica at us, we would receive only at small fraction of its contents, and the contents we did receive would be garbled.
Any response to the above points? Aside from the technical challenges, is there any reason the data should not be analyzed?
We've posted a short list of items that we intend to focus on over at bigfc252. Any input, positive or negative, would be most welcome.
We hope that periodically updating this thread isn't considered too spammy? If so, please feel free to suggest an alternative. Perhaps moving the thread to a "non-EMR" subgroup of some sort would be a better option?
Lastly, we'd like to emphasize that anyone's welcome to contribute. Just dig in!
I've moved it to the 'Citizen Science' forum. Feel free to update the thread.
Great. Thanks again.
Hello bigfc252. Today is 10/10/10 and Gizmodo has this amusing article about Douglas Adams and the number 42:
It begins with pointing out that the number 101010 in base-2 is equal to 42 in base-10. The number 42 was a major plot element in the Hitchhiker series of books. Now I'm a huge Douglas Adams fan and I think the whole 101010b2 = 42b10 thing is cute but there are two major analysis flaws. I see similar problems with the 2/23/87 analysis. Let me explain.
The first flaw is a bias towards the base-10 number system. Humans have 10 digits but the dinosaurs did not. So base-10 is no more valid than base-6 or any other base except maybe base-2. Gizmodo makes another mistake of treating 10 in base-10 also as 10 in base-2 but that is besides the point. Base is arbitrary.
The second flaw is that of units. Units are extremely important in science. Feet per second and kilometers per hour are both velocity, but they just have different units. One thing every physics student learns is that it is important to keep track of all your units and cancel them correctly. The mistake Gizmodo made, other than concatenating numbers, is that months is different than days which is different than years. You can't treat them all the same. Then another flaw is that these units are completely arbitrary. I mean what is the significance of the year? Or seconds in the 2/23/1987 analysis? What relevance is Greenwich mean time or some other base timezone? And then what about leap days and leap seconds? Units are all completely arbitrary.
Now the units problem can be solved by making the units cancel such as in a ratio. In my http://baudline.blogspot.com/search/label/SETI blog I frequently talk about "suspicious numbers" that I have found in the setiQuest data. These all are unitless ratios and most are created by the division of two frequency values that I have measured to about 8 significant digits. These suspicious numbers turn out to be non-fractional integer values which I suspect are caused by some DSP artifact that is internal to the ATA. My point is that if you are going to extract some supreme significance out of a number then it had better be unitless, otherwise it is just some arbitrary value.
Gerry Harp made a similar point a couple weeks ago in this [no-glossary]thread[/no-glossary] when he said the conclusion in the 2/23/1987 analysis "would require time travel." This is basically the same thing that I am saying. Gerry based his response on what the conclusion would require. My base-10 and units response looked at the path of how the 2/23/1987 conclusion was reached. Basically, ET would have to know things that they could not possibly know.
Just so people don't think I'm a downer on the whole 10/10/10 thing let me say that I am currently wearing my Novell/SuSE "10" baseball hat I got at an old LinuxWorld conference and I am also downloading the Ubuntu 10.10 Linux distribution that was just released today!
Thanks for your thoughts Sigblips, they’re very much appreciated.
Regarding the choice of units (as well the usage of base 16, actually), suffice to say they weren’t pulled out of thin air. For a taste of how we stumbled upon the currently posted SETI obs, and 2/23 in general, please review the following:
Expect a slightly more complete "cryptanalysis" of the coded document, as well as an open letter to Fermilab, to be posted to bigfc252 soon.
Lastly, with respect to the time travel issue that keeps (understandably) popping up, we remind you that many astrophysicists believe that the early neutrino and g-wave detections could not have been produced by SN1987A. So, as Gerry touched upon, there’s a very real possibility that they originated somewhere closer to home. On the other hand, even if they were produced by Sanduleak -69° 202a, we think it would be a major blunder to disregard a potential SETI finding solely because it doesn’t align with our current understanding of the cosmos. In other words, we humbly suggest attempting to focus on the data, and not the physics.
PS: bin 101010 = dec 42 was indeed amusing. We tend to prefer last years 9/9/9, though. That may have something do do with it being the 252nd day of the year, which is part of Mont Blanc time ;) (emphasis on the wink)
Focusing on the data, this 'signal' is not significant. Coincidences like this happen all the time in nature - the odds of it having occured by chance are much higher than, for instance, the odds of a long-lived narrowband RF signal being generated by chance. Therefore, betting even a small amount of resources or credibility on it is not a good gamble. The threshold of interest is much higher in SETI - otherwise the field would quickly go bankrupt.
The nice thing about the use of ASCII is the tangible foundation it establishes, which allows one to formulate a basic probability.
For each letter in the Latin alphabet, it's lower/upper case ASCII codes can be concatenated in two possible ways, creating 26*2 permutations.
A, 9765, 6597
B, 9866, 6698
C, 9967, 6799
D, 10068, 68100
E, 10169, 69101
F, 10270, 70102
G, 10371, 71103
H, 10472, 72104
I, 10573, 73105
J, 10674, 74106
K, 10775, 75107
L, 10876, 76108
M, 10977, 77109
N, 11078, 78110
O, 11179, 79111
P, 11280, 80112
Q, 11381, 81113
R, 11482, 82114
S, 11583, 83115
T, 11684, 84116
U, 11785, 85117
V, 11886, 86118
W, 11987, 87119
X, 12088, 88120
Y, 12189, 89121
Z, 12290, 90122
11987 (lower/upper W) is the only permutation which contains a four digit, 20th century year (1987) as a plain text sub-string. The first Mont Blanc event was detected at .11987 days during 1987. For any given day during that year, there was a 1 in 100,000 chance of an event randomly occurring at that very unique and intriguing moment. Interestingly, an alignment of this sort will not be possible again until 2088 (lower/upper X).
Note that the 10^-5 figure does not take into account the fact that the neutrino burst was not guaranteed to occur during 1987 (not even close). Nor does it factor in the next two digits of plain text (.1198702), which denote the month, creating a partial timestamp in "ISO 8601" format. It also doesn’t take into consideration any of the encoded date/time associations, which include the month, day, time of self, as well as potential predictions of the time of the later neutrino burst (All written forward as well as backward). It should, however, provide a basic sense of rarity, at a point before the decoding process has even begun.
A probability of 1 in 100,000 is not very low in SETI. Even when discounting interference, ATA observes events on that order literally all the time (as in countless times per day). Does this mean that ATA sees ET countless times per day?
"It should, however, provide a basic sense of rarity, at a point before the decoding process has even begun."
This statement is incorrect. What you have done is 1) obtain data, 2) select a message that fits the data and only then 3) determine its rarity. The decoding process begins already at step 2. This is bad SETI for all but highly exceptional messages of hundreds or thousands or millions bits of length.
This is an interesting exercise in scientific method, however, and cases like this one will be helpful in building up the knowledge and formalisms we need to moderate the flood of data we will get from citizen science.
There is a name for the approach I outlined above, and it is "data dredging":
"A key point in proper statistical analysis is to test a hypothesis with evidence (data) that was not used in constructing the hypothesis. This is critical because every data set will contain some patterns due entirely to chance. If the hypothesis is not tested on a different data set from the same population, it is impossible to determine if the patterns found are chance patterns." (From: http://en.wikipedia.org/wiki/Data_dredging)
Fair enough. Plain text analysis might be considered part of the decoding process.
As for the rest of your posts....we don’t blame you one bit for your skepticism.
Here's my personal favorite description of the aforementioned "neutrino SETI" oddity.
"The first neutrino was measured in the Mont Blanc Detector at 2:52:36.792 (± 2 ms) 
2:52:36.792 = 0.119870277.. days
119 (the first three digits after the decimal) corresponds to the ASCII character 'w.' 87 (the subsequent two digits) corresponds to the ASCII character 'W.'
The letter 'w' is the 23rd letter in the Latin alphabet.
98 (the third and fourth digit after the decimal) corresponds to the ASCII character 'b.'
The letter 'b' is the 2nd letter in the Latin alphabet.
The first, second, third, fourth, fifth, sixth and seventh digit after the decimal represents the combined number 1198702.
The neutrino was detected on the 23rd day of the 2nd month (February) in the Holocene* year 11987" 
 The Signal Recorded by the Mont-Blanc Underground Neutrino Detector on 1987 FEB 23
 IAU Circular 4332
 “2:52:36.792” confirmed with the Mont Blanc collaboration through private communication.
 SETI oddity found in section 2 of the Fermilab coded letter.
 Description composed and edited by the SETI community.
*Note: The Holocene calendar was not invented until 11993 HE.
The problem isn't your choice of units. The problem is that you have units.
Like I said above, units are arbitrary. Maybe this units example will help explain what I'm trying to say. Imagine that you are in a physics lab conducting an experiment. You make all your measurements using imperial units (feet, pounds, gallons, etc.). When you analyze the data you get an amazing result. Now you convert your data to metric units (meters, grams, liters, etc.) and this time when you [no-glossary]re[/no-glossary]-analyze the data you get a different conclusion. So your choice of units has influenced the outcome of the analysis. I'm sorry but that just isn't science.
How I would go about analyzing your data is to try to extract a modulated signal that has informational content in it. Then you could search for special numbers, ASCII, or whatever. The problem with this approach is that you don't have enough data of the 2/23/1987 "event" to do this.
Thoughts on the timestamps decoded from the Fermilab letter?
The units would be time. You need to get the units to cancel somehow. A ratio is one way.
We've experimented with ratios, but found nothing that really stands out so far. With respect to time, the focus has been on Mont Blanc / IMB, as they had fairly accurate clocks and were mentioned in the letter.
From our perspective:
1) Fermilab posts a coded letter on its official site, asking for code crackers.
2) The letter is found to contain the timestamps of the two neutrino bursts.
3) We discover Mont Blanc, Maryland, Rome, and the header.
4) Docs are posted detailing the timestamps, 2/23, and the header. Tweets.
5) Site constructed.
6) Here we are now.
Surely that sequence of events is worth investigating?
A good place to start might be getting in touch with your colleages in Batavia. Surely they receive tons of mail from citizen scientists. What stood out about this particular letter? Something must of for them to crowd source it.
It very clearly contains the date (2/23) and times of the neutrino bursts (2:52:36, 316451(7), 7:35:4) ...though the posted format is not perfect.
Here's a few more timestamps for our setiQuest pals to ponder.
FNAL Letter, Section 2, Attempted walkthrough: http://www.symmetrymagazine.org/breaking/2008/05/15/code-crackers-wanted...
For your convenience: http://www.fnal.gov/pub/about/communication/staff.html
(Updated) Analysis of the neutrino timestamps found in section 2: http://goo.gl/EGeDj
This coincidence is too cool not to mention.
Section 2: 30, B, 31, E, Q, 28, W, 29, L, O, L, O, G, I, N, I, D, 31, F
>>> for x in [30, 31, 28, 29, 31]: print chr(int(str(int(str(x), 16)), 16))
Has anyone found anything new in the letter? New timestamps? Or perhaps a better format?
http://api.twitter.com/1/users/show.xml?screen_name=bigfc252 (first tweets/emails sent shortly after the account creation date)
...perhaps "ET" isn't a big fan of human religion?
Oh, and for the record, I'm not implying in any way, shape or form that I'm in contact with "ET." :)
To the SETI community:
We’ve made you aware of an astrophysical anomaly of great importance, and we’ve alerted you to a series of potentially intelligent artifacts in its resulting data summary. It’s now up to you to take action.
Please, acquire the raw 2/23 data, and make it available to the general public. You don’t need to reallocate any major resources if you don’t want to; that’s the beauty of crowd sourcing. Simply track it down, make arrangements to acquire it under a suitable license, back it up in a safe place for future generations, and release it. That’s all we’re asking.
Anders said: "While an interesting subject for SETI analysis (...header skepticism...) If you have access to the raw data of the event, perhaps you could attempt to use some of the algorithms we've been working with on this forum - e.g. FFT, autocorrelation, kurtosis and KLT - and see if anything stands out."
Sigblips said: "(...header skepticism...)How I would go about analyzing your data is to try to extract a modulated signal that has informational content in it. Then you could search for special numbers, ASCII, or whatever. The problem with this approach is that you don't have enough data of the 2/23/1987 "event" to do this."
It’s been 23+ years. Many of the original scientists are nearing the end of their career. Some have even passed away. Let’s secure this potentially planet-altering resource while we still have the opportunity.
Here is an interesting article on Gizmodo called "The World's Largest Magnet is Being Built in India"
The magnet is going to be 50,000 tons and they are going to use it as a Neutrino Observatory. The supernova SN1987A event is even mentioned. The device should be much more sensitive and have modern data collection equipment. So next time a neutrino burst hits Earth there might be enough data to do some interesting analysis on.
I'm curious, did anyone at the SETI Institute / SetiQuest end up doing anything with this data?
I'm the original poster, btw.
Sorry if this will sound (and perhaps even turn out to be) stupid, but are you talking about the same supernova as these guys?
..they've been trying to crack a coded message sent to Fermilab for ages, now. The supernova seems to be a confirmed constant.
Maybe the coders are actually bored ET-s who are fed up with SETI ignoring this neutrino message. :)