What say ye, Resident Physicist?
In which I ask Kelseymh his assessment of the experiment which they think may have detected neutrinos travelling faster than light. (Holy run-on sentence, Batman.)
(Yes, I should have checked his orangeboard first...I'm a bad stalker. :P )
(Yes, I should have checked his orangeboard first...I'm a bad stalker. :P )
26
comments
|
Add Comment
|
sort by:
active |
newest |
oldest
Sep 25, 2011. 7:03 PMSpaceProds
says:
I agree, if neutrinos were found to move faster than light that would be amazing, not to mention something good to think about when I'm bored. How recent was this study?
Sigh. Thursday. It didn't make CNN or Fox News, presumably because there weren't any climate scientists involved for Murdoch's henchmen to bash. Please see all of the links I included in my posting for details.
(Have you posted on this topic? If so, it's caught in the filters...)
ARGH, ARGH, ARGH. I posted at 1:13 pm yesterday, shortly after you created the topic. Four paragraphs of cogent text, along with links to both the science reporting and blog discussions. I can see it, but obviously no one else can :-(
I'm sorry, but I can't take those words seriously out of the mouth of an Englishman (or anyone else with an interesting accent). :D
I guess only you would insert dramatic pauses between each word and recite it in iambic pentamenter. No wire hangers.
It's usually punctuated by a thump on the ear, or a twang with a ruler.
Perhaps the SPAMmers are now trying to sell us neutrino beam enhancement products (herbal, of course), and Robot has added "neutrino" to the filters?
Talk like a pirate Day has come and gone. Please resume the regular sophisticated scientist talk.
Avast, ye scurvy Yank! Thou best be readin' the topic title, matey, before tossin' about ye scurrilous banter...
Just a follow up. Something I had not fully realized when I read the OPERA paper is that they do not detect anything travelling at all, neither faster nor slower than light.
What OPERA has done is to look at the distribution of event timestamps for their ~16,000 neutrino detections, compare that distribution to the measured time profile of the proton spills from the CERN SPS (which is what produces the neutrinos), and fit for the time offset between the spills and the events at Gran Sasso.
Most of that time offset is the 2.43 millisecond flight time from CERN to LNGS. When you subtract that away, you're left with about a microsecond of residual delay. Most of that microsecond involves all the stages of detector response, readout, etc. (the OPERA preprint goes through this in detail).
In the end, what they claim is that the residual offset between the near detector trigger distribution and the far detector time profile is 60 ns less than it "ought to be." My money is on the idea that they haven't measured all their signal delays properly.
What OPERA has done is to look at the distribution of event timestamps for their ~16,000 neutrino detections, compare that distribution to the measured time profile of the proton spills from the CERN SPS (which is what produces the neutrinos), and fit for the time offset between the spills and the events at Gran Sasso.
Most of that time offset is the 2.43 millisecond flight time from CERN to LNGS. When you subtract that away, you're left with about a microsecond of residual delay. Most of that microsecond involves all the stages of detector response, readout, etc. (the OPERA preprint goes through this in detail).
In the end, what they claim is that the residual offset between the near detector trigger distribution and the far detector time profile is 60 ns less than it "ought to be." My money is on the idea that they haven't measured all their signal delays properly.
(I wonder what word you keep using that gets filtered...)
Thank you for translating the outline! Most of the paper (of course) flies right over my head, so it's nice to have an explanation that's simpler than that, but provides more information than a headline.
One last question, if you don't mind - what is the reason this experiment is not being heavily replicated both at the same accelerator and at others? Is it that it takes a lot of time/resources to do so? Or that the only other facilities that could attempt to replicate it are unavailable? Or it's just too soon to start doing that/they are in the process?
Thank you for translating the outline! Most of the paper (of course) flies right over my head, so it's nice to have an explanation that's simpler than that, but provides more information than a headline.
One last question, if you don't mind - what is the reason this experiment is not being heavily replicated both at the same accelerator and at others? Is it that it takes a lot of time/resources to do so? Or that the only other facilities that could attempt to replicate it are unavailable? Or it's just too soon to start doing that/they are in the process?
There are three accelerators in the world which can produce high energy neutrino beams -- CERN, Fermilab, and J-PARC in Japan.
Because neutrinos barely interact with matter at all (the sterotypical quote is that a neutrino can travel through a light-year of lead with just a 50-50 chance of hitting something), the detectors have to be extremely massive. OPERA is 625 tons of active detector, and many more hundreds of tons of support structure. You can imagine the length of time, and budget, required to build something like that.
The MINOS experiment at Fermilab already did this once, in 2007, and got a result just 1.8 sigma above 1. (that is, slightly high but consistent with c. They are currently upgrading their electronics, and should be able to do a higher-precision test when that is completed in 2012/2013.
The T2K experiment uses the Superkamiokande detector to detect neutrinos from the J-PARC accelerator at Tsukuba. However, it's baseline is less than 300 kilometers, making it much more difficult to do a velocity measurement (you'd need better than 5-ns uncertainty to get a result equivalent to OPERA's).
Because neutrinos barely interact with matter at all (the sterotypical quote is that a neutrino can travel through a light-year of lead with just a 50-50 chance of hitting something), the detectors have to be extremely massive. OPERA is 625 tons of active detector, and many more hundreds of tons of support structure. You can imagine the length of time, and budget, required to build something like that.
The MINOS experiment at Fermilab already did this once, in 2007, and got a result just 1.8 sigma above 1. (that is, slightly high but consistent with c. They are currently upgrading their electronics, and should be able to do a higher-precision test when that is completed in 2012/2013.
The T2K experiment uses the Superkamiokande detector to detect neutrinos from the J-PARC accelerator at Tsukuba. However, it's baseline is less than 300 kilometers, making it much more difficult to do a velocity measurement (you'd need better than 5-ns uncertainty to get a result equivalent to OPERA's).
I'm extremely confident that there's an unexpected, and subtle, systematic error in either the raw data or in OPERA's analysis
(preprint arXiv:1109.4897). I've done some back-of-the-envelope calculations for the obvious things (e.g., surface distance vs. line-of-sight) and the obvious things are the wrong order of magnitude. So I don't believe this group has made any obvious or elementary mistake.
Neutrino detection is a very hard problem. Doing coincidences with widely separated detectors is a very hard problem. Doing both at the same time, with a timing precision of nanoseconds, is hard squared!
I'm especially reassured by the three-dimensional GPS measurements of the L'Aquila (Gran Sasso) site (Fig. 7, page 10). The very small uncertainty in the measurements is clear from the scatter along the trend lines, and is just a few cm in each direction. The beautiful systematic trend of plate tectonics, with the theta-function from the earthquake in the middle, both give me confidence that they have good data in hand.
There are references in the preprint for how the geodetic surveys at each end were done, and how they translate the two sets of local coordinates into a common frame of reference (to get the baseline distance measurement).
You can read some of the scientific journals' reports:
You can find other discussions of the paper and its (non-)implications in several of the good science blogs, including
Thank you very much for that response! I appreciate your putting the effort into it. I now have a reading list. :)
A question - forgive me if I fail at reading comprehension, but to clarify - when you say
>I'm especially reassured by the three-dimensional GPS measurements...
you're saying that, based on the accuracy of their data, you're sure the problem is quite subtle/hard to see? Or have I misread your point?
A question - forgive me if I fail at reading comprehension, but to clarify - when you say
>I'm especially reassured by the three-dimensional GPS measurements...
you're saying that, based on the accuracy of their data, you're sure the problem is quite subtle/hard to see? Or have I misread your point?
Yes, exactly. Their coordinate precision and accuracy at each end is well documented, and supported by the data. They have made public one of their internal technical notes (OPERA note 132) where they describe in detail how they connect the two coordinate systems to make the baseline distance measurement.
See my follow up, by the way. What they are really measuring is a residual delay between the proton-spill trigger at the CERN end, and their neutrino event trigger at the LNGS end. My guess is that the 60 ns discrepancy is going to end up being buried somewhere in that very complex time budget.
See my follow up, by the way. What they are really measuring is a residual delay between the proton-spill trigger at the CERN end, and their neutrino event trigger at the LNGS end. My guess is that the 60 ns discrepancy is going to end up being buried somewhere in that very complex time budget.
(Ummm, if you posted a followup, I hate to tell you, but that has also been filtered. Maybe we should move to ScienceFile?)
So, n00b question - if there is an extremely subtle problem with _this_ experiment, why doesn't that cast similar doubt on _other_ experiments out of the same lab/group/facilities?
So, n00b question - if there is an extremely subtle problem with _this_ experiment, why doesn't that cast similar doubt on _other_ experiments out of the same lab/group/facilities?
Because this is the only measurement which depends on the detailed timing. Their other analysis are based on processing the individual events, and extracting quantities like energy, angular distribution, etc., whch don't depend on the DAQ timing to nanosecond level.
That's why they wrote such a detailed paper, and why they put it out for the larger physics community to critique.
That's why they wrote such a detailed paper, and why they put it out for the larger physics community to critique.
The general assumption seems to be that there is a systematic error somewhere. The original team can't find it, so they've released the raw data (research papers tend to have the results of the analysis, graphs & what-not), with the specific request for other physicists to dig through and find the mistake.
Indeed. It would, however, be infinitely more interesting if they don't find a mistake. :p
 |
Vancouver Mini Maker Faire 2012
Rebuilding NordicTrack ski machine drive rollers
Looking for New Zealand-based Instructables authors for conference on August 27 in Wellington
Call to makers - Brighton Mini Maker Faire
Milk Crates - not as green as you think
TEDxBaghdad - Iraq - violence, dust storms and open sourced manufacturing
UK Mini Maker Faire - The Derby Silk Mill - New Poster to Share!
