Stories
Slash Boxes
Comments

SoylentNews is people

posted by mrpg on Monday March 18 2019, @09:11AM   Printer-friendly
from the brilliant! dept.

ATLAS Observes Light Scattering off Light — New result studies photons interacting at high energies:

[...] Today, at the Rencontres de Moriond conference (La Thuile, Italy), the ATLAS Collaboration reported the observation of light-by-light scattering with a significance of 8.2 standard deviations. The result utilises data from the most recent heavy-ion operation of the LHC, which took place in November 2018. About 3.6 times more events (1.73 nb−1) were collected compared to 2015. The increased dataset, in combination with improved analysis techniques, allowed the measurement of the scattering of light-by-light with greatly improved precision. A total of 59 candidate events were observed (see Figure 2), for 12 events expected from background processes. From these numbers, the cross section of this process, restricted to the kinematic region considered in the analysis, was calculated as 78 ± 15 nb.

Curiously, the signature of this process – two photons in an otherwise empty detector (see the event display in Figure 1) – is almost the opposite of the tremendously rich and complex events typically observed in high-energy collisions of two lead nuclei. Observing it required the development of improved trigger algorithms for fast online event selection, as well as a specifically-adjusted photon-identification algorithm using a neural network, as the studied photons have about ten times less energy than the lowest energetic photons usually measured with the ATLAS detector. Being able to record these events demonstrates the power and flexibility of the ATLAS detector and its event reconstruction, which was designed for very different event topologies.

This new measurement opens the door to further study the light-by-light scattering process, which is not only interesting in itself as a manifestation of an extremely rare QED phenomenon, but may be sensitive to contributions from particles beyond the Standard Model. It allows for a new generation of searches for hypothetical light and neutral particles.

[...]Links:

So how does this fit in with the wave/particle duality of light?


Original Submission

This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(1)
  • (Score: 1) by shrewdsheep on Monday March 18 2019, @03:49PM (2 children)

    by shrewdsheep (5215) on Monday March 18 2019, @03:49PM (#816447)

    Just glancing through the first link, there seems to be strong selection going on with respect to the events analyzed. Not only are certain energy ranges selected but a "learning algorithm" seems to have been employed as well. In both experimental and observational studies, selection of subgroups can be highly problematic and can lead to biased results. Maybe somebody in the know can elaborate on as to why the selected events would represent a physical phenomenon? As opposed to: hey, interesting pattern here, let us train a net to filter events from the data showing similar patterns.

    • (Score: 2) by HiThere on Monday March 18 2019, @04:04PM (1 child)

      by HiThere (866) Subscriber Badge on Monday March 18 2019, @04:04PM (#816461) Journal

      I'm not sure quite what your question is. Are you saying "Wouldn't instrument noise have generated this signal?", then I can see it as a valid question that they should answer. Or are you saying "Wouldn't this effect arise by random emissions from the background layer of virtual particles?", then it could also deserve an answer, although the answer would almost certainly be no.

      OTOH, if you're asserting "If you go looking for any pattern, you'll find it." the answer is no without asking any further. You can only find patterns that exist, and expected noise is determinable. The "filter" you're describing selects candidates, and SOME of those may be expected from noise, but the number that should be expected for that reason should be determinable. When you get more than expected, it becomes interesting, and if you repeatedly get more, it becomes quite likely that the effect is real, and not just noise.

      --
      Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
      • (Score: 1, Interesting) by Anonymous Coward on Monday March 18 2019, @05:36PM

        by Anonymous Coward on Monday March 18 2019, @05:36PM (#816527)

        You can only find patterns that exist

        That's decidedly not true if you are applying filtering methods. I can't currently find it, but there was a series of experiments by different groups that found some electron emission line, apparently confirming each other, except for some unexplained shift in energies. Well, it turned out that the line didn't actually exist, but was entirely fabricated by the filtering/search methods. As was found later, splitting the data in two, employing the filtering/search technique on the one half, and then looking for the "found" peak in the other half turned up nothing.

  • (Score: 0) by Anonymous Coward on Monday March 18 2019, @04:38PM

    by Anonymous Coward on Monday March 18 2019, @04:38PM (#816484)

    "[...] of searches for hypothetical light and neutral particles."
    hehehe i SEE what you did there.

    totally unscientific but one must laud the people with their limited time span dedicating themselfs to re-learning (and hopefully understanding) previous (dead and buried) generations knowledge only to be left with a short life span of their own to contribute a little morsel to the canon.
    and yet, still dinner does not cook itself, traffic jams are for everyone, the leaky faucet doesnt care if you have phd in theoretical physics and the leather jack with permanent wet hair on the motorcycle STILL gets the hotty pregnant first ... "alas said atlas".

(1)