To Infinity and Beyond !
S-LHCb
OR
“How I stopped worrying and
learnt to love what we can do
with 0.1 quadrillion* B’s.”
OR
“Who will be the true superhero
of 2015 flavour physics ?”
*
1014 – B’s produced at Super-LHCb over 5 years
1
Physics Opportunities at an Upgraded LHCb
Edinburgh Upgrade Workshop, 11/1/06
Guy Wilkinson, University of Oxford
with thanks for useful input to:
Robert Fleischer, Franz Muheim, Tobias Hurth,
Ulrik Egede, Vanya Belyaev
2
Contents
Why plan for B-physics beyond ~2015? Scenarios for discussion.
Definitions and assumptions
Selected topics:
• Super-precise angle measurements: Φd, Φs and γ
• Non-SM CPV phases in b→s Penguins
• b→sll, b→dll and b→(s,d)γ
• Bd,s→μμ
• Beyond the B: τ and charm physics
(Will try my best to motivate interest of each and quantify
performance – but there will be a fair amount of guesswork.)
Conclusions
3
Comparisons with Super-B
Won’t be showing plots like this →
(Unlike Browder @ CKM 2006)
SuperKEKB at 50 ab
ΔS(φKs)
+ ΔS(K K Ks)
ΔS(η’Ks)
ΔS(KsKsKs)
0
ΔS(π Ks)
sin2χ(Bs → J/ψφ)
*0
S(K γ)
Br(B → Xsγ)
ACP(B → Xsγ)
*+C9 w/ AFB(K l l )
*+C10 w/ AFB(K l l )
+ Br(Bs → μ μ )
+
+
Br(B → K νν)
+
Br(B → Dτν)
0
Br(B → Dτν)
sin2φ1
φ2(ππ isospin)
φ2(ρπ)
(*)
φ3(DK )
φ3(Bs → KK)
φ3(Bs → DsK)
|Vub|
-1
-1
LHCb (0.002 ab )
ΔS(φKs)
+ ΔS(K K Ks)
ΔS(η’Ks)
ΔS(KsKsKs)
0
ΔS(π Ks)
sin2χ(Bs → J/ψφ)
*0
S(K γ)
Br(B → Xsγ)
ACP(B → Xsγ)
*+C9 w/ AFB(K l l )
*+C10 w/ AFB(K l l )
+ Br(Bs → μ μ )
+
+
Br(B → K νν)
+
Br(B → Dτν)
0
Br(B → Dτν)
sin2φ1
φ2(ππ isospin)
φ2(ρπ)
(*)
φ3(DK )
φ3(Bs → KK)
φ3(Bs → DsK)
|Vub|
-0.5 -0.3 -0.1 0.1
0.3
0.5-0.5 -0.3 -0.1 0.1
no info
no info
no info
0.3
0.5
4
Comparisons with Super-B
Won’t be showing plots like this →
(Unlike Browder @ CKM 2006)
Several entries badly wrong,
and apart from this, misleading:
• obscures relative importance
of measurements, & requirements
on precision
LHCb (0.002 ab )
ΔS(φKs)
+ ΔS(K K Ks)
ΔS(η’Ks)
ΔS(KsKsKs)
0
ΔS(π Ks)
sin2χ(Bs → J/ψφ)
*0
S(K γ)
Br(B → Xsγ)
ACP(B → Xsγ)
*+C9 w/ AFB(K l l )
*+C10 w/ AFB(K l l )
+ Br(Bs → μ μ )
+
+
Br(B → K νν)
+
Br(B → Dτν)
0
Br(B → Dτν)
sin2φ1
φ2(ππ isospin)
φ2(ρπ)
(*)
φ3(DK )
φ3(Bs → KK)
φ3(Bs → DsK)
|Vub|
-0.5 -0.3 -0.1 0.1
0.3
no info
no info
no info
l!
fu
elp
• selective (eg. no Bs→ΦΦ)
ΔS(φKs)
+ ΔS(K K Ks)
ΔS(η’Ks)
ΔS(KsKsKs)
0
ΔS(π Ks)
sin2χ(Bs → J/ψφ)
*0
S(K γ)
Br(B → Xsγ)
ACP(B → Xsγ)
*+C9 w/ AFB(K l l )
*+C10 w/ AFB(K l l )
+ Br(Bs → μ μ )
+
+
Br(B → K νν)
+
Br(B → Dτν)
0
Br(B → Dτν)
sin2φ1
φ2(ππ isospin)
φ2(ρπ)
(*)
φ3(DK )
φ3(Bs → KK)
φ3(Bs → DsK)
|Vub|
-1
-1
tH
No
• don’t compare like-with-like
(5 years Super-B from 2015
with 1 year of LHCb in 2008)
SuperKEKB at 50 ab
0.5-0.5 -0.3 -0.1 0.1
0.3
0.5
I will however make case-by-case topic comparisons with Super-B.
But remember my selection is LHCb inspired. (But it is not narrow!)
5
Why B-physics beyond 2015, & why at LHCb ?
Testing the flavour structure of the SM has taken many years. Elucidating the
flavour structure of the NP found at the LHC will surely not be a simple task.
Either: LHCb will find exciting NP signal (eg. Bs→μμ or Φs enhancement,
UT non-closure at ~5o level). In this case there will be a host of follow up
measurements, all of which will benefit from more statistics/better experiment.
→ script will write itself – so don’t consider further
Or: ATLAS & CMS will find NP, but no big signatures in flavour sector.
For instance Minimal Flavour Violation (MFV) might give identical CPV to
SM, but give (possibly small) deviations in b→(d,s) Penguins and Bd,s→μμ
→ important to test SM predictions in CPV down to the level of the
theoretical predictions (and the theory in 2020 [and beyond?], not now! )
→ make improved studies of b→(d,s) Penguins and rare decays
in measurements which will be statistically limited at LHCb.
Both motivate flavour physics beyond 2015. But why at LHCb? This talk !
6
Definitions and Assumptions
Unless stated otherwise names and numbers will refer to the following:
‘LHCb’ 10 fb-1, ‘Super-LHCb’ 100 fb-1, ‘Super-B’ 50 ab-1
For Super-LHCb I merely extrapolate present statistics assuming
identical trigger and reconstruction performance (unlikely to be true).
For Super-B numbers I either take values from recent talks (eg. flavour
workshop) or reports, or I extrapolate from existing B-factory results.
For Super-LHCb I do not (cannot) estimate the contribution of experimental
systematics. Bear this in mind, but also recall that we intend to determine our
main systematics (eg. tagging, τ resolution…) from data. Also at LHCb:
σ calibration signal >>> σ physics signal
and so it is reasonable to hope that experimental systematics can be controlled.
Theory errors I discuss where relevant, although some of my extrapolations
are arbitrary and mostly for illustration purposes. But improvements in theory
estimates (esp. lattice QCD) are crucial aspects of any upgrade argument.
7
Progress with theory
Increase in computing power is very likely to lead to significant
improvements in any quantity which can be calculated on the lattice
Pre-LHCb
2010
2014
6 TFlop year
40 TFlop year
1 PFlop year
11%
5%
4%
2%
13%
5%
4%
2%
ξ
5%
3%
2.5%
1.5%
Vub-excl.*
11%
6%
5%
3%
Vcb-excl.*
4%
2%
1.5%
1%
Now
B̂ K
f Bs B̂ Bs
(table courtesty of Vagnoni talk, CKM 2006 Nagoya)
In addition, there will be a wealth of precise BR measurements against
which theorists can tune and test these and other calculations.
→ Expect significant improvements in many areas before 2015-20
8
Why measure β to < 1% ?
Surely post LHCb, β will be known as well as we need to. Why go further ?
“Because it’s there!”
George Mallory
(lost on Everest 1924)
sin2β
LHCb-upgrade project leader
It is the parameter of the UT we can measure most precisely, so we should.
Also we should note that the (final) indirect constraint will be very stringent !
9
Measuring β at the sub-percent level (< 0.2o)
Direct measurement of Φd (=2β in the SM) is tested against indirect prediction
which is fixed by Rb (the Vub/Vcb side)
Rb now known to 7% from Vub inclusive
→ translates into β indirect of 1.6o
Compare with (Super-)LHCb stat error:
(Super-)LHCb 2 fb-1
σ (stat)
0.66o
10 fb-1
100 fb-1
0.30o
0.09o
Rb
β
Looks OK… but precision on Rb will
improve as lattice improves Vub exclusive.
Must plan for Rb precision of ~2% → β indirect error then 0.5o
Also here experimental systematics will be a major consideration. Maybe
we will have to be stricter in choice of events (eg. tagging) to keep σsyst ≤ σstat ?
All good arguments for > 10 fb-1 !
10
Controlling Penguins in sin2β
If we plan a very precise measurement of β we must worry about the Penguin
pollution. (This probably mandatory at LHCb, not just Super-LHCb).
One method which has been proposed (Fleischer, Eur.Phys.J. C10 (1999) 299)
is to measure the CP asymmetries in Bs→J/ψKS and relate these to Bd→J/ψKS
through U-Spin (good approximation for
1st order correction ?)
BR for Bs mode ~ 2 x 10-5, so we will
have 80 x fewer events than for Bd→J/ψKS
Previous LHCb attempt to show feasibility
(1999 workshop) had B/S>>1. But we should
be able to do better…
Worth evaluating precision of approach at LHCb and Super-LHCb
11
A precision measurement of Φs
Initial hope in measuring Φs (= -2ηλ2 in SM) with, say, J/ψΦ, is that NP will
manifest itself through significantly larger value than v. small SM expectation.
But situation could be as in Bd system where any NP phase < few degrees
Prepare for precise measurement, noting
that in SM Φs is very well constrained:
Φs indirect = -0.037±.002
Compare with (Super-)LHCb statistical reach:
(Super-)LHCb 2 fb-1
10 fb-1
100 fb-1
0.021
0.009
0.003
σ (stat)
10% measurement (modulo systematics)
Note however that here also we must worry about Penguins, and their relative
effect could be large compared with this precision. Control with Bd→J/ψρ
(Fleischer, Phys.Rev. D60 (1999) 073008) .
12
Goal for the ultimate γ measurement
Geometry of UT is such that the
mixing side (Rt) dominates
in indirect prediction of angle γ .
Thus tree-level γ measurement
in conjunction with knowledge
of Rt can be used to probe for new
physics contributions to the latter.
γ
Rt
Present situation already tells
us that a precise test is needed !
Uncertainty on γ
from sides (UTFIT)
= 6o
Lattice improvements will lead to shrinking of this error – let us guess by a
factor of 5 to 10 at t=∞. So is it possible to measure γ to 0.5-1 degree ?
13
γ expectations at LHCb
Many ways to measure γ through tree-level processes at LHCb. Include:
1. Bs→DsK, precision of ~14o / 2 fb-1
2. Bd→D(*) π, requires external input on rB or U-Spin
3. B±,Bd→D(*)K(*) , with the D0 decaying to:
2 bodies: πK, KK, ππ
3 bodies: Ksππ, KsKK, KsKπ
4 bodies: Kπππ, KKππ
best modes offer 5-10o
precision each for 2 fb-1
A not outrageous expectation is that with 10 fb-1 we could achieve a
precision of ≤ 3o. Based on approaches 1. and the best understood modes
of 3., together with guesses about likely evolution of Dalitz modelling errors.
This would make a major impact on today’s UT, but does not meet our
ultimate goal of 0.5-1o . What then can we hope for at Super-LHCb ?
14
Prospects for a sub-degree error on γ
In extrapolating to 100 fb-1 it is only reasonable to consider strategies which
are theoretically clean &/or where performance and each is well-understood.
In the LHCb era we will be able to benefit from the combined information of
many DK modes. Some will have associated (eg.) modelling errors which will
(probably) prove limiting in ultra-precise regime. The √N law will not apply !
So today focus on 3 modes:
LHCb
(10 fb-1)
Super-LHCb
(100 fb-1)
Super-B
(50 ab-1)
DsK
27 k
270 k
/
D(Ksππ)K
≤ 25k
≤ 250k
46k
D(Kπ)fav K
280k
2.8 M
89k
GW extraps
from published
B-factory
analyses
One immediate observation: Super-B will struggle to match LHCb results…
15
Prospects for a sub-degree error on γ
What γ sensitivity could we hope for with these modes at Super-LHCb ?
1. Bs→DsK: here use naïve √N → 2o for 100 fb-1
2. D(KSππ)K Dalitz
Likely that statistical precision will be ~ 2o (better than Super-B !)
Error associated with knowledge of Dalitz space is at present
crippling (10o), but it is certain to reduce, in particular due to
CP-tagged data gathered at CLEO-c and BES-III (→). Down to 2-4o ?
(Similar trick will be possible with KSKK/KSKπ and 4-body modes)
3. D(two body)K
D→hh modes are clean and statistics huge, but with these alone
the problem is barely constrained. Either: a) add 4-body channel (Kπππ) ,
b) add CP-odd eigenstate (eg. ΦKs), c) or exploit D*→D0(π,γ) trick.
Not possible to give headline number today, but no one mode will dominate
16
CP tagged D0 decays at the ψ(3770)
Asner, CKM 2006
17
CP tagged D0 decays at the ψ(3770) (ctd)
Asner, CKM 2006
18
New Physics phase in b→s Penguins ?
sin2βeff measurements that involve b→s Penguins show intriguing bias
w.r.t. charmonium results. Naïve average gives 2.6 sigma discrepancy.
Consistent sign of bias throughout
channels is tantalising (theory correction
generally has other sense…)
Still, it is perhaps more correct to
average only 3 cleanest channels,
→ ΦKS, η’KS and KSKSKS
19
Size of NP hint and Timescale for Discovery
GW average of present ΦKS, η’KS and KSKSKS results (~1 ab-1)
‘Discrepancy’ = 0.13 ± 0.08
Ergo, with present central value there
will be no 5σ discovery at Belle/BaBar.
Super-B projections for 50 ab-1 :
ΦKS error ≈ 0.030
η’KS error ≈ 0.025
(includes systematics)
Only ΦKS really possible at LHCb. Here
statistical precision of 0.32, 0.14 and 0.05
expected with 2, 10, 100 fb-1 (extrap. from Yuehong).
(Although could Super-LHCb be optimised with this decay in mind ?)
Apparent conclusion: Super-B a better bet for discovery.
20
BS→ΦΦ at (Super-)LHCb
Best way to probe b→s Penguin phase at LHCb is through Bs→ΦΦ.
W−
Accesses same physics as B-factory
Bd modes and should be theoretically
very clean (how clean ?)
⎧b
⎪
0⎪
Bs ⎨
⎪
⎪⎩ s
t
g
s⎫
⎬φ
s⎭
s⎫
⎬φ
s⎭
Reconstruction studies (Bruno de Paula) predict 4000 events / 2 fb-1
(with 60% uncertainty due to B.R.) and 0.4 < B/S < 2.1 (90% C.L.)
Sensitivity studies (Xie) based on full angular analysis gives σ (ΦNP) ≈ 0.10
for 2 fb-1 , where ΦNP can be considered as NP phase driving sin2β ‘anomaly’
Precision post LHCb
0.044 – good, but no cigar! (‘only’ 2.9σ)
Precision post Super-LHCb 0.014 – as good as we need? (~ ΦKS theory limit)
Extremely promising – looks a very good reason for an upgrade, and is (more
than) competitive with Super-B. But more work needed to quantify potential.
21
B→K*μμ at Super-LHCb and Super-B
Interest and potential of K*μμ well established. Headline measurement
is the New-Physics-sensitive 0 point in the forward-backward asymmetry.
Recently, attention focused on
problems introduced by non-resonant
background. Solution may dilute
statistical sensitivity… but this will
not be a concern at Super-LHCb
2 fb-1, gives 7.7k events. Fit AFB →
LHCb, 2 fb-1
( s0 true = 4.10 )
s0 = 4.11 ± 0.52 GeV2 (13% error)
And 10 fb-1 gives 39k events (7% error on s0)
So in 100 fb-1, 385k events (2.1% error on s0)
Mμμ2 (GeV2)
Extrapolate current Belle numbers to 50 ab-1 → 16k events
22
How clean is s0 in K*μμ?
Super-LHCb would make very precise measurement of s0 (and Super-B ≈
2 years of LHCb). But what is the theoretical uncertainty ?
NLO contribution calculated
within QCD factorisation
(Beneke, Feldmann, Seidel)
~ 9% uncertainty
Similar size, but hard to estimate,
error from 1/mb power corrections
So present error is in 10-15% ballpark. So s0 alone is not a very strong
argument for Super-LHCb… Two caveats:
- treatment of non-resonant background may degrade precision / fb-1
- it is surely reasonable to hope for some improvement in error by 2020 ?!
23
Inclusive vs exclusive AFB in B→sμμ
Nice advantage of (Super-)B environment is opportunity of making inclusive
measurements. Such an approach attractive since theory error < exclusive case.
In AFB zero-point analysis:
Inclusive theory error ~ 5 %
Exclusive theory error ~ 10-15%
(Numbers courtesy Tobias Hurth)
Situation at present – could
improve before 2015-2020
Whereas experimental (statistical) precision on s0 is:
Super-LHCb - 100 fb-1 (exclusive) 2 %
Super-B – 50 ab-1 (semi-inclusive) 7 % (extrap. from SLAC-R-709)
(semi-inclusive in sense that 60% of decays are reconstructed)
So inclusive analysis has marginal advantage, with present theory errors…
24
Other opportunities in K*μμ
Other important observables in K*μμ system will almost certainly benefit
from Super-LHCb statistics, eg. those considered in transversity angle
analysis in which decay is described
in terms of 4-variables:
s
θl
θK*
φ
μμ mass squared
AFB angle
equivalent K* angle
angle between K* and μμ decay planes
In this basis can construct several asymmetries which at leading order depend
on short range information only, in particular:
25
AT
(1)
and AT (2) vs Mμμ
Variables appear theoretically rather clean ! (Kreuger, Matias, hep-ph/0502060)
(no 1/mb
error
included)
Moreover, very sensitive to New Physics
effects through eg. RH currents in C7eff
AT(2) with such contributions
(note scale)
A powerful way to probe chiral structure
of b→s transitions (c.f. b→s gamma)
26
Initial LHCb studies of AT(2) sensitivity
Initial LHCb studies (Egede) has demonstrated that AT (2) is extractable,
using a toy-MC study in which Mμμ divided into 4-bins
This result with 2 fb-1
(for more information see Oct. `06
Flavour Workshop talk and RD WG)
For Super-LHCb shrink error
bars by 7 (Super-B, by 1.4). Then one
will be sensitive to the very small SM
evolution, while still being an order of
magnitude away from theory uncertainty.
A very sensitive probe of b→s in which theory error is promisingly small !
27
Studies of B→Xd l+l- ?
All measurements we are interested in performing with B→sμμ are
presumably equally interesting with B→dμμ, eg. as important test of MFV.
b→s mode
b→d mode
b→d yield /
2 (→100) fb-1
Bd→K*μμ
Bd→ρ(ω)μμ
10-100
(→ 500-5000)
Bs→Φμμ
Bs→K*μμ
10-30
(→ 500-1500)
(self-tagging mode!)
(Yields are GW guesses. Range expresses further guess as to effect of
additional cuts w.r.t. b→s analogue required to further suppress background)
Only possible at Super-LHCb ? Worth a more detailed look !
28
Prospects with radiative Penguins
Super-LHCb will accumulate very large samples of the ‘standard’ b→sγ
events (eg. 3.7 million K*γ , 0.5 million Φγ). But we can also hope for:
B→ωγ 2000 events in 50 ab-1,
B/S < 3.5 (extrap. from existing study)
B→ρ0γ : existing K*γ selection would
give 50k events, but tighter cuts will be
needed – awaiting DC06 studies
W
b
t
d
t
γ
(Super-B expects around 3200 events in ρ0γ channel, but will also collect ρ±γ)
Many interesting tests can be carried out with these events, eg. Vtd/Vts .
Note that again the theory error on this exclusive measurement is not so bad:
7% (Ball, Jones and Zwicky hep-ph/061208), and there are also competitive
lattice based results (Bechirevic, Lubicz and Mescia, hep-ph/0611295)
Higher granularity/resolution ECAL would help, plus open other possibilities…
29
Hunting for Bs→K*γ
b→dγ Penguin in the Bs system is Bs→K*γ. Cannot separate from the Bd
decay with present ECAL, but maybe possible with improved resolution ?
LHCb ECAL
Factor 2 better energy resolution
Bd→K*γ
Bs→K*γ
(Must also consider problem of combinatorial background)
30
Bs→μμ and Bd→μμ at Super-LHCb
However, SM signal at 5σ still
lies within reach of LHCb,
although not comfortably so…
Going further still:
BR (x10-9)
Hope for significant NP enhancements in Bs→μμ (but not in Tevatron range!)
10
9
8
7
6
5σ
5
4
SM prediction
• Within SM and CMFV, signal
3σ
BR is predicted to 10%. (Buras,
hep-ph/0604057). Testing this
cannot be done with 10 fb-1.
• Rate of (Bd/Bs)→μμ is even more
Integrated Luminosity (fb-1)
tightly constrained in SM & CMFV.
“ Bs→μμ/Bd→μμ = 32.4 ± 1.9 , one of magic numbers of CMFV ” (Buras)
Matching such precision beyond even Super-LHCb (50-100 events?), but
Bd→μμ observation could be, provided lepton id. can suppress Bd→ππ.
3
2
1
0
1
2
3
4
5
6
7
8
9
10
31
Accumulating Rare Charm Decays
Super-LHCb has opportunity to collect a huge sample of valuable
D0 decays (x 20 what could be accumulated at Super-B)
LHCb (10 fb-1)
Super-LHCb (100 fb-1)
Super-B (50 ab-1)
D0→K+K-
5 x 106
2.5 x 108
1.2 x 107
D0→K+π-
2 x 105
1 x 107
5 x 105
+
D0 from B decays only. With present trigger scheme
we expect same number again from primary charm.
Could be enhanced with track/vertex trigger at L0 ?
-
These are events ‘on tape’. Tight cuts needed
to suppress background and to partially reconstruct
B to find D0 birth vertex for time dep. studies
z res. of parent B
(Patrick Spradlin)
(lose factor of 3
and still have tails)
32
Opportunities in charm physics
Sample sizes will allow for improving limits on
mixing and CPV by 2 orders of magnitude
w.r.t existing limits, eg. CPV in KK down to 10-4
Mixing exclusion plot 2006
The cynic:
“ LHCb already has potential to see mixing & CPV
at the level when it might be a NP signal (eg.
10-2 for CPV). Below that it is SM and dull ”
The visionary:
• duty to push on until mixing and CPV observed !
• any observation would be a new beginning for charm physics.
Inevitably interest would focus on elucidating nature of phenomenon,
and extending measurement programme: CPV in the mixing,
triple products correlations in ρ0ρ0, CPV in other modes…
Super-LHCb would be the experiment to answer this call !
33
Lepton Flavour Violation Searches in τ decays
An attractive feature of Super-B is its sensitivity to lepton violating τ decays
τ→μγ
τ→μμμ
3 x 10-8
3 x 10-9
-5
10
PDG2005
Belle
BaBar
-6
10
-7
10
-8
10
-9
10
μ−γ
e−γ
μ−π0
e−π0
−
μη
e−η
μ−η′
e−η′
e−e+e−
e−μ+μ−
e+μ−μ−
− + −
μee
μ+e−e−
μ−μ+μ−
e−π+π−
e+π−π−
− + −
μππ
μ+π−π−
− + −
eπK
e−π−K+
+ − −
eπK
− + −
eKK
e+K−K−
μ−π+K−
− − +
μπK
+ − −
μπK
− + −
μKK
μ+K−K−
e−ρ0
−
e KS
*
e−K
−–
e K*
e−φ
μ−ρ0
−
μ KS
− *
μK
−–
μ K*
μ−–φ
pγ
–
pπ0
Λπ−
−
Λπ
In SM with neutrino mixing,
LFV τ decays occur at ≤ 10-14
But other models predict
values in range 10-10 – 10-7
Any observation of LFV in μ
decays elsewhere would give
this work extreme topicality
Current B-factory limits
are generally at a few 10-7,
Super-B potential for 10 ab-1
(sensitivity to UL at 90% CL):
Super-B with 10 ab-1
34
τ→μμμ at (Super-)LHCb
Obvious LFV channel for us to search for is τ→μμμ. First looked at in
2000 (Bartalini) and is being re-visited now in the RD WG (Shapkin, Belous)
Main source of τ’s at LHCb will be Ds→τυ. Total number of τ produced
within acceptance for 1 year at Super-LHCb ~ 1012 (extrap. from Bartalini)
So guessing a likely trigger & reconstruction efficiency of 5% (similar to
Bs→μμ ?) one would expect ~50 τ→μμμ events / yr with a B.R. of 1 x 10-9
However, present LHCb studies have large background retention: 105 events
in normal year of 2 x 1032 operation. Not good ! My questions:
• What is this background? Guess mainly combinatoric (but decays such
as B+→ω(π+π-π0)μ+υ will eventually be an irreducible problem)
• What other cuts could be deployed? (Isolation cut not used?)
Important topic which deserves more attention (even at LHCb) – we should
understand our potential. Nice test-case for Super-LHCb vertex detector ?
35
Conclusions
Clear need for an upgraded heavy flavour programme beyond 2015.
Lattice improvements will motivate better γ and (to an extent) Φd (β)
measurements. S-LHCb is the best place to do this.
NP in the Bs sector may be small → precise Φs measurement valuable.
Bs→ΦΦ looks to be a very sensitive probe of NP phases in b→s,
and competitive with Super-B possibilities.
There are NP sensitive studies in exclusive b→sμμ which will be statistics
limited at LHCb. Doubly so for b→dμμ and d γ . (Super-B inclusive?
Evolution of exclusive theory + statistical power need to be considered.)
Confirming/ruling out SM/(C)MFV in Bs,d→μμ requires S-LHCb
D physics may well take centre-stage at some point soon. S-LHCb can
improve existing mixing/CPV reach by two orders of magnitude.
Potential for a few x 10-9 reach in τ→μμμ. More studies needed!
36