Tommaso Dorigo promises that the paper “is guaranteed to have you fastened to the chair until you are done with its 70 pages”, but really, not so much – it’s actually terribly written, and you will most likely find yourself lost by page 28 or so.  Some of this is presumably because the paper wasn’t allowed to make a positive statement of a discovery, and so could only discuss things that look strange without developing much of an argument.  The paper presents a number of categories of odd events without too much of an obvious connection between them (except for having muon candidates at large impact parameter), and there’s an obvious selection bias in that you don’t hear about distributions that look normal.  It’s chock-full of detector jargon, and as a friend said to me, “it reads like someone dumped an internal CDF note on the arxiv.”

One serious objection I have with the writing style is that it plays somewhat subtle linguistic tricks on the reader.  For example, the events that they do model well are termed “QCD”, while those they don’t are “ghosts.”  Now there is every possibility that the “ghost” events are, in fact, due to particles produced via QCD, and they do discuss this, but after they’ve repeated “we subtract the QCD background” enough times you begin to believe that they know how to subtract all their Standard Model backgrounds perfectly based on data, which is not correct.

There are also amazingly few estimates of systematic uncertainty in the paper, which makes the quality of the numbers somewhat hard to judge.  Almost all the uncertainties you see are statistical.  Well, I assume they are – the paper doesn’t really say.

Do I have any physics comments?  Papers like these are exactly why we HEP experimentalists say that we should have exclusive control of the data – only the collaboration has the full knowledge of the detector, and their reputation is on the line.  In general this means that if a collaboration claims that something is inconsistent with what they know about their detector modeling, you take that extremely seriously.  On the other hand, if a paper explicitly says that there’s the possibility that the effect they’re seeing is not real, you also take that very seriously, and a large fraction of the collaboration seems uncomfortable with even that level of hedging.  Though it’s extremely hard for an outsider to judge (especially given the style of the paper, as I said before), my guesses would be

• this isn’t a non-Standard Model effect;
• their modeling of light hadron production is wrong;
• their understanding of the secondary interactions of particles in the detector material is incomplete.

But, again, I have no special knowledge of any kind, so this is merely a layman’s opinion (albeit one who wasted a bunch of time looking at tracks with large impact parameters in a different detector…)

What would convince me that the effect is real?  First, observation of some electrons in the multi-muon events (it’s certainly possible for new physics to prefer muons to electrons, but not generic); secondly, confirmation by D0.

I should comment (though don’t take this as gospel, I’m just an experimentalist) that there are very strong constraints on whatever this could be already, should it be a new particle.  The proposed production rate is huge – comparable to that of b quarks – and to see those kinds of cross sections at a hadron collider suggests production in a QCD process, and that the new particle is fairly light (much lighter than the top, for example).  If it carries electromagnetic or weak charges, and is lighter than half the Z mass, it would have been seen already at electron-positron colliders; but it needs to couple to muons somehow…

Just for laughs, you can read arxiv:0810.5730, which manages to cite the CDF paper 39 times in six pages while attempting to give a phenomenological explanation for the anomalous events.  In short: there are particles in a mass ratio 4:2:1, which apparently have no SM couplings other than to the tau, and by the way are produced at a 100 nb rate with a huge boost.  Funny!

9:05 – I’m here in the back row of the main auditorium, which is absolutely full. There are supposed to be two viewing rooms in building 40, the main ATLAS/CMS building, but unfortunately those don’t open until 9:30 apparently.

9:07 – eww, relativistic mass.

9:09 – bizarre low frequency sound effects. Reminds me of those supernovas in planetarium shows which always scared me as a kid.

9:10 – live feed begins.

9:13 – they will try to send beam 1 (clockwise) around, one sector at a time.

9:17 – Lyn Evans doesn’t know how long it will take, but he hopes less than 12 hours.

9:19 – Robert Aymar thanks everyone.

9:22 – It’s like a Mars landing! But we get to try again and again.

9:29 – in building 40, where the chairs are comfy.

9:34 – some beam went in! Everyone laughs when a commentator asks “did you see it???”

9:38 – beam went to point 3.

9:40 – instant replay!!!

9:42 – they’ve actually managed to go to point 5 before, so so far nothing completely new.

9:45 – “It’s not just like switching on your mobile phone.” Beam at point 5.

9:51 – “They’re very young! 30s and 40s!”

9:53 – CMS seems to be showing a trigger rate. They are now going to test the beam dump at point 6.

9:59 – LHC status page here.

10:01 – Lyn Evans tempts fate by saying everything might be done in an hour.

10:07 – Point 7. Orbit has lots of excursions between 6 and 7 so they need to fix it.

10:12 – Point 8. One more before ATLAS!

10:18 – Point 1!!!! Loud clapping in ATLAS viewing room.

10:23 – next step is full circle.

10:25 – full circle.

The two red dots in the leftmost image show successive passages of a proton pulse through a monitor at point 2.

10:34 – They’re talking with ALICE. The ATLAS logbook shows that we did indeed see signals in the liquid argon calorimeters as the beam went through.

10:42 – we have ATLAS event displays! And very ugly they are, too.

10:43 – current plan seems to be to switch to beam 2 in an hour and a quarter.

10:49 – Everyone’s gone to do other things; room is much emptier now.

10:52 – Maiani dreams of a linear collider at CERN.

10:57 – Commenter points out that cost of LHC ~ cost of Beijing olympics. But only one will tell you about dark matter.

11:25 – Will Young-Kee Kim wear pajamas??? Only videoconferencing will tell!

11:29 – so enthusiastic, the commentators are. And so soothing.

11:30 – Fermilab promo movie.

11:33 – Umm… it’s unclear if the FNAL directorate are wearing pajamas, although they’re certainly color-coordinated. They also seem to have large stickers of some kind.

11:35 – Pier Oddone brings insane US liquor licensing laws to the attention of the world.

11:37 – The DOE is proud of us! Yay!

11:41 – as is the NSF.

11:43 – in the great American tradition, many people talk.

13:05 – after an unannounced lunch break for me, time to get back to work.  Everything on the accelerator side was very impressive this morning, and now the detectors have to follow that act.  I leave you with an ATLAS event display from when the beam was stopped on the collimators just upstream of point 1:

Your friendly local Geneva morning paper wishes to reassure you that CERN will not kill you on Wednesday.  At any rate, we wouldn’t be making black holes tomorrow; there will be no colliding beams, and the beam energies will be below the Tevatron’s.  So when the end (doesn’t) come, it will most likely be on a totally unheralded day in October.

UPDATE: as promised, available here (Sun 15:30).

(* last = last confidently-identified fully-reconstructed D_s D_s^* event)

CLEO-c stopped taking data earlier this week.  We looked in the last data run (7:38 am to 8 am) of just over 61 thousand events for collisions that produced D_s mesons, and we actually found one.  Even better, we found an event where you could see both the D_s⁺ and D_s^-, and where the photon from the D_s^*+ → γ D_s⁺ transition was visible.  Above, you can see the event display, with all the tracks labeled; the event is consistent with the following sequence of events:

• e⁺ e^- → γ → D_s^*+ D_s^-
  • D_s^*+ → γ D_s⁺
    • D_s⁺ → K^- K⁺ π⁺ π⁺ π^-
  • D_s^- → K_S K^-
    • K_S → π⁺ π^-

It was actually unlikely that we’d find such a nice event.  For the amount of data in the last run, we would expect roughly 75 D_s^* D_s events.  Our full reconstruction efficiency (getting both D_s candidates) is somewhat less than 1%, so we had a good chance of winding up with zero events like this.  It’s nice to be lucky though.

The latest Heavy Flavor Averaging Group results on D⁰ mixing are to be found in Alan Schwartz’s writeup for the recent BES-Belle-CLEO workshop.  The continual drip of new results has pushed the mixing significance to 6.7σ.

Last CLEO run

There’s still a lot of work to do – we need to decommission CLEO and reconstruct and analyze the last data!