Tuesday, 5 May 2015

Pheno 2015 Liveblog: Day 2 Session 1

We start the second day of Pheno 2015 with some talks on flavour physics and QCD.

8:45 am: B Physics: The Next Generation, Tom Browder

I always preferred DS9, myself.

The CKM matrix has two irreducibly complex phases, which are conventionally taken to be Vub and Vtd.  There is also the well-known consistency relation, the unitarity triangle.  The best probes of flavour physics were until recently from the B-factories, but LHCb is now (as of this year) able to do comparably or better.

Consistency of SM
A recent new result is the first joint BaBar-Belle data analysis looking at CP violation in b decays to cud.  The experiments themselves do not have the data to resolve a signal, but in combination they can make a 5.4σ observation.  This measurement of the appropriate angle is consistent with b to ccs.

Despite all this, a 10 to 20% NP amplitude in Bd physics is consistent, indeed has a (non-significant) better p-value.

Mixing in Bs is also consistent with SM.  Likewise, penguin diagrams (loop-level decays) are all consistent; in this latter case, LHCb is still inferior to the B-factories.

Searches for New Higgses
As is well-known, these are most sensitive to new charged Higgses.  For example, additional contributions to B decaying to τν.  However, this is experimentally difficult, especially at a hadron detector; the best sensitivity comes from the tau decaying to an electron, i.e. the B decays to a single charged track and MET.  The full data results from Belle and BaBar give only 3σ evidence for the existence of this decay; it has not been observed.  Nonetheless, this is enough to place constraints on light charged Higgses.  Belle II or similar could have much higher mass reach than the LHC.

Another signal is B to Dτν.  This has recently been done by LHCb, a challenging measurement.  BaBar actually made a claim based on an anomaly to have found NP, but ruled out a 2HDM for any mass or tan β.  However, we are waiting on Belle and LHCb to present their results to properly judge this.

Rare B Decays
LHCb found rarest B decay, to dimuons, which is at 10-9.  There is an anomaly for B0, but statistics are still very low.

Historically, the Z was "discovered" this way (in ee to μμ).  Looking for asymmetries in decays to kaons plus leptons.  This is a three body final state, with a lot of angles to consider.  Asymmetry in SM is pure interference effect.  LHCb recently reported its results here, and found a 3.7σ anomaly in over 2000 signal events.

If true, why would this be the first NP hint?  Interference, so linearly sensitive.  Loops involve all heavy SM particles.

Why might it not be true?  Worries about form factors, tails of other decays, binning effects; examination under way.

Future Experiments
Belle II, SuperKEKB; competitive and complementary to LHCb.  Need ridiculous luminosity, which demands high focusing; beam widths at ~ 50 nm!

Actually, I lied.  I always preferred B5.

9:20 am: Flavour Physics in the New Era, Wolfgang Altmannshofer

The big questions in flavour physics can be split into three categories: the SM flavour problem, neutrino masses, and the NP flavour problem.

SM flavour problem
This is essentially the question for the large hierarchy in the SM Yukawa structure.  We see this both in the masses and in the mixings.  One possible approach is to take a Froggatt-Nielsen approach, with fermion masses forbidden by flavour symmetries.  With appropriate charges the light quark masses can then be suppressed by large powers of a ratio of scales.  The main alternative approach is the partial compositeness idea, as seen in 4D composite models or 5D RS.

Other ideas do exist, though.  By forbidding couplings of the light quarks to the Higgs, their masses come from loops of new particles.  Now we suppress the masses by loop factors.  This is an old idea (Weinberg '72), but a recent new implementation is based on the MSSM.  By having only one generation coupling to the Higgs, and anarchic sfermion masses, the observed spectrum is mostly fit; the muon is the big problem, which comes out too small.  The biggest problem here is that anarchic sfermion structure demands the sfermions be heavy, to avoid constraints.

New Physics flavour problem
Essentially, if NP is at the TeV scale, why have we not seen it in flavour physics yet?  There are essentially only two explanations: the NP scale is very high (104 TeV or more), or it has a non-trivial flavour structure (MFV or something similar).

At dimension-6, non-diagonal couplings of the Higgs are generically generated.  We could look for them in Higgs decays, but we already have stringent flavour constraints.

Flavour anomalies
Here's a handy list of anomalies from Wolfgang's talk:
All of these are tensions of varying degrees of significance.  Pick your favourite!

In the B to Kμμ anomalies, argued that the observed violations of lepton flavour universality can only be statistical on NP; other explanations (such as hadronic effects or parametric uncertainties) cancel in such ratios.

For B to Kμμ, done with light charm quark; can we then really trust the SM prediction here?  Related: how solid are those flavour anomalies?  As far as the latter, all could be hadronic but violations of LFU.

9:55 am: Pushing the Frontiers of Perturbative QCD, Jesse Thaler 
Four decades of jet study: 1982 work at UA2.  A big "recent" experimental development has been the LHC segmentation, five times better than prior experiments, which gives almost a one-to-one relation between calorimeter energy deposition and hadron particle flow.  This also fed theoretical developments in jet algorithms, substructure and computation.

Main focus of talk: standard lore is that IRC-safe observables are only observables that can be calculated in a controlled way.  However, this can be challenged; IRC-safe observables can still have order-one non-perturbative corrections.  Even more striking, IRC-unsafe observables can be somewhat calculable.

Review: Safe Observables
These are observables that do not change when you add an infinitely-soft (IR safe) or exactly collinear (C safe) parton.  These correspond to divergences that cancel with virtual diagrams; for safe observables this cancellation happens order by order.

Constructing a IRC-unsafe observable must be done on a case by case basis; IRC-safe observables are universal.

A Standard Candle for Jets
Textbook QCD: the Altarelli-Paresis splitting function.  The basis for parton showers, pdf evolutions, NLO subtractions, etc.  But can it be measured?  It is only universal in the singular limit.  But let's try anyway.

Cluster jet in angular ordered tree.  Decluster by one stage, and examine splitting.  Can be done but is clearly IRC unsafe.  To address the soft problems, "groom" jet; cut branches below some threshold in energy fraction.  Now IR safe, but still C unsafe.

To go further, two approaches.  First: use Sudakov form factors.  If we impose a cut on jet mass, and demand it be non-zero, we will always be C safe.  But in real situations, jets are never zero due to Sudakov effects!  The collinear singularity we see only occurs at fixed order in the perturbative expansion. the all-orders sum not only cuts off this divergence, it sets it to zero!  Note: this cancellation does not occur order by order, but only when we do that resummation.

Alternative idea is to use fragmentation functions.  This involves absorbing the singularities into universald functions (pdfs).

The interesting thing is that the resultant observable is roughly independent of αS, jet pT, and is even approximately the same for quarks and gluons; it is (nearly) universal on jets!

Two new QCD tools
First, if have an isolated singularity, try to regulate it with Sudakov form factors.  Example: N-subjettiness, where a ratio of two safe observables is actually unsafe due to a singularity at zero in the denominator.  But this singularity is never seen due to Sudakov factors.  This explains why these observables match data so well.

Second, if have collinear singularity, try using a fragmentation function.  Example: weighted jet charge.

Non-perturbative corrections?  Actually larger than in safe observables.
Relation between two methods in standard candle? Agree at leading order, but formal statement of accuracy still unclear.
Is there a relation to the IRC-unsafe experimental definition of photons? Most likely; nothing unique here to QCD.

No comments:

Post a Comment