I meant just older of course ... just older ... As for the virtues ... well there are some. And no I do not think of gradual loss of short term memory though I can see that that could be sometimes a benefit. Instead I think for instance of the fact that one has been around for some time. And in the case of LSND (Liquid Scintillator Neutrino Detector) that actually matters.
1995
I do remember quite vividly one of the first talks about the LSND detection of the neutrino oscillations in 1995 (or early 1996 have I mentioned something about memory going away ... just a short term ... eh ?!) by one of its lead researcher. Shoes shuffling, leafing through the papers, and this air of solemnity followed by the talk oscillating between rather apologetic "gosh look we have done the best we could to get rid of that but it still there" to more optimistic "gosh look we have cracked the particle physics as we know it wide open" ...
And it looks like 15 years on we are still in this kind of "either ... or ..." situation but then there is also a twist.
What the LSND claimed then was a detection of so-called neutrino oscillations on a length scale smaller than a hundred of meters. To do so the experimentalists produced a neutrino beam made of predominantly only muon anti-neutrinos (νμ) and looked for, and found, an excess of electron anti-neutrinos (νe;) over what was expected at a distance of ∼ 30 meters away from the production place. The standard explanation is that such a coordinated appearance of the neutrinos of some specific type, almost from "nowhere", is due to the neutrinos changing they flavor as a function of a distance from their source - a phenomenon referred to as neutrino flavor oscillations. The extra neutrinos, which were detected were in fact also produced at the source but were initially of a different type and were transformed on the way to our detector. The effect had been well expected by then and in fact confirmed earlier in neutrinos produced either in the Sun (so-called solar neutrinos) and the Earth atmosphere (i.e., atmospheric ones). What did not go well with the predictions was the oscillations length of less than a hundred of meters. Both solar and atmospheric neutrinos showed much longer oscillation lengths (∼ 10,000 and ∼ 1,000 times respectively). Given the fact that the oscillation length is inversely proportional to a square of mass difference between two oscillating neutrino species, these latter observations pointed towards much smaller mass differences than required by the LNSD detection. In fact these would have to be: Δ m2 solar ∼ 10 -5 eV2, Δ m2 atm ∼ 3 x 10 -3 eV2, and Δ m2 LSND ∼ 1 eV2 to fulfill the constraints of the solar, atmospheric, and LNSD neutrinos experiments, respectively. As we have only three neutrino families in the standard model, two different oscillations set the mass difference for all of the three and the third measurement, and one of its kind at that time as LSND happened to be, should fit the constraints due to the two other ones. Should but it did not.
A bummer then ?! Or a Nobel prize ?! That's a thin line of course ...
Two more facts are relevant here. First, that the solar neutrinos oscillations are seen as a deficit of electron neutrinos, (νe), with respect to the solar models predictions. Second, also for the atmospheric neutrinos what is observed is again just a deficit but this time of muon anti-neutrinos, (νμ). So a rather desperate idea was proposed to reconcile all three observations, which called up to live sterile neutrinos, (νs). A sterile neutrino is, by design, a neutrino-like species - so it could take part in oscillations with the other known neutrinos - but at the same time does not interact weakly - so it could bypass the limit on a number of neutrino families from electron-positron collisions ...
As stealthy a particle as they only could be then ...
The emerging picture of what could have been going on is as follows. Given that only LSND saw the event excess this would have to be a "standard" oscillation of a type νμ → νe, while at least in one of the other cases the observed deficit would have to be explained as one of the standard neutrinos turning into a sterile one. In such a scenario the mass difference between muon and electron neutrinos would have to be on order of 1 eV ...
The original experiment was clearly not without its problems (is any ever ?!). The LSND team kept however working publishing their final paper few years later. No essential changes to their claim had been made by then. The statistical significance of their results grew by then to 3.8 σ ... meaning than in a fewer than 0.1% cases one should see such a result as a consequence of a random fluke. Leaving some people in the community grinding their teeth ...
2006
Well not all of them. Some were trying to validate, or more likely disvalidate - I presume given what a nice bunch of chaps scientists are, the LSND result. The MiniBooNE (mini Booster Neutrino Experiment), conceived some time around 1997, started investigating potential oscillations of type νμ → νe in 2003. They had optimized the experiment to target the same mass range where LNSD saw the signal. They published their results in 2007 finding no excess in observed numbers of electron neutrinos, (νe), on a 3σ level over what is expected from the background signal. This all looked like the old LSND result is finally dead, relegated to - at the very best - a statistical fluke status. The things seemed to have got back to normal as happily acknowledged with a sigh of relief by some main stream physicists.
And with that the sterile neutrinos looked soundly buried once and for ever ...
2010
And then they came back ...
The first round of the MiniBooNE experiment did not really reproduce the experimental circumstances of the original LSND run as they focused on neutrinos rather anti-neutrinos. A small difference, but nevertheless a difference. In a demonstration of rather unusual scientific ardency the miniBoone team, against, one would think, all odds, decided to repeat their tedious investigation but this second time focusing on the anti-neutrinos. The results came out in mid-2010 after three long years of data collecting. And lo and behold, a surprise came with it. The new results seem tantalizingly consistent with the old LSND result pointing towards a high mass difference between two involved (anti)neutrino species. But wait ! The older MiniBooNE result still stands as well ! So in place of one unexpected and unwelcome result, we now have three and not all three straightforwardly consistent ones ... So it looks like something has to go ... or to come ...
Recapitulating we started from something odd and have arrived at something really weird... First, it has turned out now that neither one nor even two extra species of sterile neutrinos will do if standard interactions are assumed. Though if the latter are allowed to be more exotic, one species may still do. Moreover taken at their face value the discrepancy between neutrino and anti-neutrino results seems to point to a CP violation needed to explain why particles behave so differently than their anti-companions. The CP breaking has been observed before in some quark interactions (specifically manifesting itself in neutral meson K decays) but so far was always found to be very weak.
If the effect is confirmed its consequences for fundamental physics then are nothing short of revolutionary. And it will reverberate beyond it as well. The CP violations was one of Sakharov's conditions to explain the matter-antimatter asymmetry seen in the observable Universe. The neutrino masses on order of ∼1 eV will have significant consequences for galaxy clustering in the Universe in particular on the smallest angular scales, (even if so-called mixed (cold+hot) dark matter models are nobody's favorite at the moment). As will have the presence of extra relativistic species ... Interestingly the latest WMAP results seem to prefer 4 rather than 3 families of massless (or nearly so) neutrinos. A coincidence ... very likely given that 3 are perfectly allowed by their fits, but still amusing to note ... Now how the limit from the nucleosynthesis, which also restricts the number of relativistic species to 3 in the early Universe (plus photons of course), could be avoided may seem as more of a challenge but then why sterile neutrinos would have to follow an equilibrium distribution ?
much ado about nothing ?
As of now the latest MiniBooNE results stands on 2.7σ. Meaning that still there is as high as 0.5% chance that it is just due to the background fluctuations. Not much one has to concede. (Note that in cosmology 2σ results are considered worthy staking a claim, though there is by definition a 5% chance that they are nothing more than a fluke.) This is however less than 3.8σ quoted by the final analysis of LSND. Taken together the both results would make for rather an incontrovertible detection. But then shall we combine them together ? Well, that depends but at least from one point of view, the one suspecting that LSND is not a fluke but some systematic issue, we should abstain from doing so. But then we are left only with that 2.7σ delivered by MiniBooNE ... Suggestive - yes, but is it really pinning it down yet ? I reckon only time (and more data) will show ...
And just as a side remark:
Does not it look like that tantalizing, though statistically not necessarily convincing, detections have suddenly got more common (recall CDMS last year) ? A sign of us starting probing the new physics frontier, or more of a cultural change ?
No comments:
Post a Comment