Skip to content

L. Corwin: MINOS Results

March 15, 2011

I will not discuss the detailed description of the detectors (near and far) that was provided by the speaker in this talk. Let me just briefly discuss what was shown, as far as the results are concerned.

MINOS achieved 1.7×10^20 protons on target in the antineutrino mode, and 7×10^20 in the neutrino mode. These data produce quite precise measurements of the muon neutrino oscillation parameters through their disappearance.

Oscillation of muon neutrinos to other flavours will manifest as a deficit at the far detector. Charged current interactions can be distinguished very well, using a well-tested paradigm (a hadronic shower with a long track from the generated muon).

The new result of MINOS has been  in the arxiv since a week ago. They have the World’s most precise delta_m^2 measurement:  2.32+0.12-0.08 x10^-3 eV^2.

Antineutrinos also undergo CC interactions yielding positive muons.  The antineutrino beam is not 93% pure as the neutrino beam: it is  only 40% pure. So they need to select events with positive reconstructed muon charge. Uncertainties are dominated by statistics.

According to the slides, “systematics were incorporated in the result with the Feldman-Cousins method” – this does not make sense to me, since FC is a unified method for constructing a confidence belt, while the way systematics are included in a measurement is a different topic altogether; maybe he got confused with the “Cousins-Highland” method of incorporating systematics in an upper limit by quasi-frequentist integration of the nuisance parameters…

Anyway, let us get back to the talk. What does MINOS find in their oscillation analysis from comparing neutrinos and antineutrinos ? Well, it appears that there is presently an interesting tension between neutrinos and antinenutrins, at the 2.3 sigma difference, in the oscillation parameters. Definitely something to watch closely in the future!

One bonus track after the main result: the far detector of Minos can also detect atmospheric neutrinos. The baseline is from 10 to 13000 km, and the L/E parameter spans 4 orders of magnitude. The data is divided in events with neutrino interactions in the detector and events in the rock producing upward going muons. A likelihood fit to the L/E distribution produces the oscillation parameters. The fit results have lower accuracy than the beam results, but they are still consistent.

In the future, new antineutrino analysis with an at least doubled dataset is predicted. New electron neutrino results will also be produced, as well as an improved atmospheric neutrino result.

4 Comments leave one →
  1. March 15, 2011 7:35 pm

    Cool, thanks for the report! So, we must wait a bit longer for the new antineutrino results. But good to hear about the 2.3 sigma tension.

  2. Minoan permalink
    March 18, 2011 12:51 am

    The mention of the use of the Feldman cousins method in our analysis was correct, and not a mistake on the part of the speaker.

    Consider; the Feldman cousins method generates and fits a large number of randomly generated experiments to build the correction tables. By including a set of randomly chosen (within our systematic limits) systematic shifts for each of these experiments, you can account for the effects of the systematics in the feldman-cousins correction.

    The tradeoff is that you have to generate a lot more experiments to get a statistically meaningful measurement out of this – luckily, computing power is cheap and the algorithm isn’t that expensive for us.

    • March 18, 2011 6:54 am

      Dear Minoan,

      I well know that you can insert nuisance parameters in a frequentist calculation, and as such Feldman Cousins is no exception. However, the slides say that the analysis accounted for systematics by means of the FC method. This is a misquoting error, because the Feldman-Cousins method has nothing to do with the marginalization of nuisance parameters, and all to do with how to construct meaningful and unified confidence intervals. As I already said, the technique of marginalization should cite another paper, Cousins and Highland (“Incorporating systematic uncertainties into an upper limit”). NIMA320 (1992) 331.


      • Minoan permalink
        March 18, 2011 1:05 pm

        To make sure we are on the same page, what exactly do you mean by “marginalization of nuisance parameters” – to me, that tends to imply that you are fixing for certain variables whilst allowing others to float to a minimum.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: