Carleton Universitylogan/talks/LHCILCpheno.pdf · Carleton University Pheno 2008 Symposium – LHC...

LHC/ILC Synergy

Heather Logan

Carleton University

Pheno 2008 Symposium – LHC Turn On!

1

The LHC physics menu is very rich and exciting.

CMS Physics Technical Design Report, Volume II: Physics Performance,

CERN-LHCC-2006-021

ATLAS detector and physics performance technical design report, Volume 2,

CERN-LHCC-99-15

+ many more recent updates

The ILC physics menu is also very rich and exciting.

A. Djouadi et al., “International Linear Collider Reference Design Report,

Volume 2: Physics at the ILC,” arXiv:0709.1893 [hep-ph]

Physics case well established for LHC and ILC separately in wide

range of BSM scenarios.

This talk: LHC/ILC synergy.

Heather Logan LHC/ILC Synergy Pheno 2008

2

Synergy, complementarity, interplay, . . .

Idea is not new.

LHC has high energy, high luminosity, large reach for direct dis-covery of new heavy particles.

ILC has high precision / low background, threshold scan capa-bility, control over initial state quantum numbers, . . .

First LHC/LC study group meeting St. Malo ECFA/DESY LCWorkshop, April 2002http://www.ippp.dur.ac.uk/∼georg/lhcilc

Report finished fall 2004; published in Phys. Rept. 2006G. Weiglein et al. [LHC/ILC Study Group], “Physics interplay of the LHC and

the ILC,” Phys. Rept. 426, 47 (2006) [hep-ph/0410364]

LHC/ILC studies continue.


3

Synergy, complementarity, interplay, . . . concurrency?

Obvious now that concurrency is not going to happen.

But early studies tried to make the case for it.

- Use ILC measurements in LHC analyses?

- ILC observation → new search in LHC data?

- Change trigger menu of ATLAS/CMS as result of ILC discovery?

Talk by Sally Dawson at Victoria ALCPG Meeting July 2004:

pointed out we had not made the case for concurrency.

- No excuse not to be smart about saving LHC data for later revisiting.

- No examples of trigger menu changes: anything new from the model-builders

gets built in to the ATLAS/CMS triggers!


4

This talk:

What can be learned from LHC + ILC

that cannot be learned from LHC alone

or from ILC alone?

This implies some kind of combined analyses

(or using input from one machine for analyses at the other).

Will give three examples:

- Z′

- Higgs

- SUSY


5

Z′


6

LHC and ILC have comparable “reach” for Z-primes.

7.3 Z′ studies at the LC

Usually, the sensitivity reach of indirect searches is given at the 95% C.L. assumingspecial models, e.g. E6 or left-right symmetric models. To compare these 2σ devia-tions from the Standard Model expectations with the direct search reaches one hasto consider at least a 5σ deviation. Figure 7.2 illustrates the discovery and sensitivityreaches of the LHC and LC. The sensitivity limits at LC would improve up to roughly12% if all systematic errors could be reduced to zero.

0 2 4 6 8 10 12

SSMLRη

ψχ

P-=0.8P+=0.6

P-=0.8P+=0.6

LC

LC

LHC

LHC

Tevatron

mZ' [TeV]

0.5TeV

0.8TeV

100fb-1

1ab-1

1ab-1

1ab-1

14TeV

14TeV

15fb-1

2σ

Figure 7.2: Comparison of the Z′ discovery reaches at LHC [25] and the 2σ and 5σ sensitivity

bounds at LC [11].

7.3.2 Distinction of models with the LCTo learn about the Z′ properties the measurement of the couplings is essential. Thecouplings and also the behaviour of the angular distribution of the observables mea-sured with polarized beams allow the study of the nature of new gauge bosons. First,the usual scenario is considered where the extra gauge bosons carry vector and axialvector couplings.

In the ideal case the new gauge bosons are light enough to be found with the LHC.Then both colliders provide the feasibility to determine the Z′ couplings.

At the LC at high energies, fermion pair production is a process with high statisticsand clear topologies. Moreover with the expected integrated luminosity of 1 ab−1

and with the possibility of polarisation of at least the electron beam. If the mass of a

415

S. Riemann, in ILC RDR, arXiv:0709.1893

But what does that mean?


7

One example: ZH in the Littlest Higgs model.

M (TeV)

1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 60

0.5

1

1.5

2

-

ATLAS

Figure 22: Plot showing the accessible regions for 5σ discovery of the gauge bosons WH and ZH asa function of the mass and cot θ for the various final states. The regions to the left of the lines areaccessible with 300 fb−1.

27

Azuelos et al, hep-ph/0402037: 5σ discovery reach w/ 300 fb−1


8

Comparable reach for 5σ signal at ILC:

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1 2 3 4 5 6 7 8 9 10

s

f (TeV)

5 σ contours with MAH→ ∞ at √s = 500 GeV

SEARCHREACH

BY LEP IIEXCLUDED

LHCILC s’= √3/5

s’ = 0.50s’= s

s’ = s/2

Figure 6: (Color online) Comparison of ILC and LHC search reach. The LHC data wastaken from Fig. 8 of [17]. The search reach lies to the left of and underneath the contours.

possible for Z ′ bosons with masses <∼ 2 TeV, absolute determination of the Z ′ couplings isnot possible. There are three main reasons for this: (i) the number of observables is limitedin the hadron collider environment. The observables are the number of events (i.e., crosssection times branching fraction), the forward-backward asymmetry, and the rapidity asym-metry) for leptonic final states only. Other final states are not detectable above background,(tt final states are a possible exception, but such events will be heavily smeared and thusnot useful for a coupling analysis). (ii) The observables are convoluted with all contributingparton densities. (iii) The statistics are insufficient for MZ′ >∼ 2 TeV. Here, in the case ofthe LH, our results show that LEP II essentially excludes the region MZH

<∼ 2 TeV, andthus we do not expect the LHC to contribute to the parameter extraction in any significantway. We note, however, that a very precise mass measurement for ZH will be obtained atthe LHC.

To determine the accuracy of the parameter measurements, we perform some sample fits,using a χ-square analysis similar to the one described in the preceding section. We use thesame ILC observables as before. In some cases we also include data from a

√s = 1 TeV

run, for which we also take an integrated luminosity of L= 500 fb−1. We note that we canexchange MZH

for f , so we now take MZH, s, and s′ as our free parameters. We choose

a generic data point (s, s′, MZH) and use it to calculate the observables, which we then

fluctuate according to statistical error. We assume that the Large Hadron Collider wouldhave determined MZH

relatively well, to the order of a few percent for MZH<∼ 5−6 TeV; we

thus fix MZHand perform a 2-variable fit to s and s′. Scanning the s-s′ parameter space,

we calculate the χ2 at every point. We find the minimum χ2 point; the 95% CL regionsurrounding it is the region for which the χ2 is less than this minimum χ2 plus 5.99.

Figure 7 shows the results of this fit for two sample data points in the contrived scenario

13

Conley, Hewett & Le, hep-ph/0507198: 500 fb−1 at 500 GeV


9

But these are very different types of signals.

LHC: direct discovery of Z′ resonance in dileptons.

Can measure mass of Z′ very accurately.

Hard time measuring couplings:- get only a rate in dileptons- Forward/backward asymmetry requires rapidity cuts- Maybe tt, bb final states

ILC: sensitivity to Z′ through off-shell interference with Z, γ ex-change in e+e− → ff .

No direct measurement of mass.Some sensitivity if ILC runs at two different energies.

Sensitivity to left/right handed couplings to multiple fermionspecies, but coupling strengths are mixed up with Z′ mass fromthe propagator.


10

ILC coupling measurements need MZ′ input from LHC.

It’s obvious that this combination will be done.

Littlest Higgs ZH example: with mass from LHC, can fit cou-

plings and extract model parameters.

0.4

0.42

0.44

0.46

0.48

0.5

0.52

0.54

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

s

s’

Sample fits for 95% CL (MZH = 3.0 TeV, MAH

→ ∞, √s = 500 GeV)

s’= 0.5

0.4

0.42

0.44

0.46

0.48

0.5

0.52

0.54

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

s

s’


→ ∞, √s = 500 GeV)

s’=√3/5

0.4

0.42

0.44

0.46

0.48

0.5

0.52

0.54

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

s

s’


→ ∞, √s = 500 GeV)

Data point

Figure 7: (Color online) 95% CL sample fits to the data points (s = 0.5, s′ = 0.5) and (s =0.5, s′ =

√3/5), using e+e− → f f observables at a 500 GeV ILC, taking MZH

= 3.0 TeVand MAH

→ ∞.

with MAH→ ∞. For both of these points, the determination of s is very accurate. This

is due to the strong dependence of the ZHf f couplings on s, as discussed in the previoussection. The s′ determination is worse than that for s because of our choice s = 0.5. At thisvalue of s, the contributions from the ZL coupling deviations (which carry the s′ dependence)are smaller than the ZH contributions. The reason the s′ determination is better for s′ = 0.5than it is for s′ =

√3/5 is that the s′-dependent deviations in gVZLff

vanish for the lattervalue.

Figure 8 shows the results from a similar fit and illustrates how it can be improved withdata from a higher-energy run with

√s = 1 TeV at the ILC. Here, the s determination is

not much more accurate than the s′ determination, as the s′-independent ZH contributionsno longer dominate the fit for s = 0.9.

In Fig. 9 we show results from a fit with the full AH contributions. Not surprisingly,the parameter determination is much more precise, as the AH contributions, when present,dominate the χ2. Since the AH couplings depend only on s′, it is also no surprise that herethe s′ determination is in general much better than that for s.

If, for some reason, the LHC doesn’t provide a good measurement of MZH, we would

need to include that quantity, or equivalently f , in our fits to the data. In Fig. 10 we showthe results where we have set s′ =

√3/5 and we fit to s and f . Note that for one of the

data points, the allowed region doesn’t close. This highlights the importance of using boththe LHC and the ILC to reliably determine the model parameters.

14

0.6

0.65

0.7

0.75

0.8

0.85

0.9

0.95

1

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

s

s’


→ ∞)

Input point

0.6

0.65

0.7

0.75

0.8

0.85

0.9

0.95

1

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

s

s’


→ ∞)

s’=√3/5, √s = .5 TeVs’=√3/5, √s = 0.5,1 TeV (combined)

s’ = 0.5, √s = .5 TeVs’= 0.5, √s = 0.5,1 TeV (combined)

Figure 8: (Color online) Like Fig. 7 except MZH= 3.3 TeV and the data points are (s = 0.9,

s′ = 0.5) and (s = 0.9, s′ =√

3/5). Also shown for each point is an improved fit from addingdata from a

√s = 1 TeV, L= 500 fb−1run at the ILC.

0.6

0.62

0.64

0.66

0.68

0.7

0.64 0.68 0.72 0.76 0.8

s

s’

Sample fits for 95% CL (MZH = 3.3 TeV, √s = 500 GeV, light AH )

s’= 0.65

0.6

0.62

0.64

0.66

0.68

0.7

0.64 0.68 0.72 0.76 0.8

s

s’


s’=√3/5

0.6

0.62

0.64

0.66

0.68

0.7

0.64 0.68 0.72 0.76 0.8

s

s’


Input point

Figure 9: (Color online) Like Fig. 7, except MZH= 3.3 TeV and the data points are (s = 0.65,

s′ = 0.65) and (s = 0.65, s′ =√

3/5), and the full MAHcontributions are included.

15

Conley, Hewett & Le, hep-ph/0507198: 500 fb−1 at 500 GeV


11

Applies to general models: measure left/right fermion couplings.Studies assume mass of resonance as input from LHC.

lRC

-0.5 0 0.5

l LC

-1

0

1

LH

LR

KK

χSLH

bRC

-0.5 0 0.5b LC

-1

0

1

LH

LR

KK

χ

SLH

Figure 1: Resolving power (95% CL) for MZ′ = 1, 2, and 3 TeV and√

s = 500 GeV, Lint = 1ab−1. The smallest regions

correspond to MZ′ = 1 TeV and the largest to MZ′ = 3 TeV. The left side is for leptonic couplings based on the leptonic

observables σµP

e−P

e+

, AµLR, Aµ

F B. The right side is for b couplings based on the b observables σbP

e−P

e+

, AbF B, Ab

F B(pol)

assuming that the leptonic couplings are known and a b-tagging efficiency of 70%.

Fig. 1(a) shows the resolving power of the lepton couplings assuming lepton universality and using the three

observables: σµP

e−P

e+, Aµ

FB and AµLR for MZ′ = 1, 2 and 3 TeV. As noted by Riemann there is a two-fold ambiguity

in the signs of the lepton couplings since all lepton observables are bilinear products of the couplings. The hadronic

observables can be used to resolve this ambiguity since for this case the quark and lepton couplings enter the

interference terms linearly. Fig. 1(b) shows the resolving power for b-quark couplings based on the b-quark observables

σb, AbFB, Ab

FB(pol) assuming that the leptonic couplings are accurately known from other measurements and a b-

tagging efficiency of 70%. One could gain additional information by studying other observables with hadron final

states such as Rhad, AhadLR , and observables involving the c-quark.

We next consider the importance of polarization. In Fig. 2 we show results for the cases of no polarization, only

the electron is polarized, and both the electron and positron are polarized. The results are shown for MZ′ = 2 TeV,√s = 500 GeV and Lint = 1ab−1 using the three observables σµ

Pe−P

e+

, AµLR, Aµ

FB . Note that the appropriate values

of Pe− and Pe+ are used in eqn. 1 and for the unpolarized case ALR does not contribute. Clearly polarization will be

important for measuring couplings and disentangling models if a Z ′ were discovered although positron polarizaton

does not appear to be an important factor for these measurements.

In Fig. 2 we assumed a Z ′ mass of 2 TeV. But the LHC has the potential of discovering a heavy neutral gauge

boson up to 5 TeV or higher. Supposing that this is the case, can the ILC still give us useful information? In Fig. 3 we

show the resolving power for Z ′’s with MZ′ = 1, 2, 3, and 4 TeV, again using only the three µ observables assuming

the e− and e+ polarizations given above. Reasonably good measurements can be made for the MZ′ = 2 TeV case.

For MZ′ = 3 TeV the resolving power deteriorates but the measurements can still distinguish between many of the

currently popular models. At MZ′ = 4 TeV it becomes quite difficult to distinguish among the models although

some models could still be ruled out.

In Fig. 4 we examine possible improvement in the resolving power by including more observables. In the previous

figures we only included three observables with final state muons. If τ leptons could be observed with reasonable

efficiency an additional five observables (στP

e− P

e+, Aτ

LR, AτFB, Pτ the τ polarization, and Aτ

FB(Pol)) can be included

ALCPG0108

Godfrey, Kalyniak & Tomkins, hep-ph/0511335: 1 ab−1 at 500 GeV, MZ ′ = 1, 2, 3 TeV.

- Extended gauge group- TeV-size extra dimensions- Compositeness (Technicolor, Randall-Sundrum)


12

Higgs


13

If the Higgs is sufficiently Standard-Model–like, its discovery isguaranteed at LHC.

)2

(GeV/cHm

)-1

dis

cove

ry lu

min

osi

ty (

fbσ

5-

1

10

210

100 200 300 500 800

CMSjjνl→WW→qqH, Hννll→ZZ→qqH, H

, NLOννll→WW*/WW→H4 leptons, NLO→ZZ*/ZZ→H

-τ+τ, γγ →qqH, H inclusive, NLOγγ→H

bb→H, WH, HttCombined channels

CMS Physics TDR, via ILC RDR

Measure the mass...


14

... and measure rates in each channel.

LHC, 200 fb−1 (except 300 fb−1 for ttH, H → bb, WH, H → bb). Zeppenfeld, hep-ph/0203123

Given MH, rates are completely determined in SM.

Check if rates are consistent with SM predictions!


15

Ratios of rates give ratios of partial widths.Adding mild theory assumptions allows to fit Higgs couplings.

[GeV]Hm110 120 130 140 150 160 170 180 190

(H,X

)2 g

(H,X

)2

g∆

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1(H,Z)2g

(H,W)2g

)τ(H,2g

(H,b)2g

(H,t)2g

HΓ

without Syst. uncertainty

2 Experiments-1 L dt=2*300 fb∫-1WBF: 2*100 fb

[L] 200 fb−1 (except 300 fb−1 for ttH(→ bb), WH(→ bb)). Zeppenfeld, hep-ph/0203123[R] Duhrssen, Heinemeyer, H.L., Rainwater, Weiglein & Zeppenfeld, hep-ph/0406323


16

ILC: High-precision measurements of Higgs production couplings,

decay branching ratios.

HIGGS PHYSICS

and to check the prediction of the Higgs mechanism that they are indeed proportional tofermion masses. In particular, BR(H → τ+τ−) ∼ m2

τ/3m2b allows such a test in a rather

clean way. The gluonic branching ratio is indirectly sensitive to the ttH Yukawa couplingand would probe the existence of new strongly interacting particles that couple to the Higgsboson and which are too heavy to be produced directly. The branching ratio of the loopinduced γγ and Zγ Higgs decays are sensitive to new heavy particles and their measurementis thus very important. The branching ratio of the Higgs decays into W bosons starts to besignificant for MH >∼ 120 GeV and allows measurement of the HWW coupling in a modelindependent way. In the mass range 120 GeV <∼ MH <∼ 180 GeV, the H → ZZ∗ fractionis too small to be precisely measured, but for higher masses it is accessible and allows anadditional determination of the HZZ coupling.

TABLE 2.1Expected precision of the Higgs branching ratio measurements at ILC for MH = 120 GeV and a luminosityof 500 fb−1. Ranges of results from various studies are shown with c.m. energies of 300 GeV [8], 350 GeV[93, 94, 95] and 350/500 GeV [96].

Decay mode Relative precision (%) References

bb 1.0–2.4 [8][93] [94][97]

cc 8.1–12.3 [8][93] [94][97]

τ+τ− 4.6–7.1 [8] [93] [94]

gg 4.8–10 [8] [93] [94][97]

WW 3.6–5.3 [8][93] [94] [95]

γγ 23–35 [94] [96]

There are several studies on the sensitivity of the Higgs branching ratios for a light SMHiggs boson at ILC. Although each analysis is based on slightly different assumptions ondetector performance, center-of-mass energy, and analysis method, overall consistent resultsare obtained. The accuracies of the branching ratio measurements for a SM Higgs bosonwith a mass of 120 GeV are listed in Tab. 2.1, while for MH =120, 140 and 160 GeV fromthe simulation study of Ref. [93], they are shown in Fig. 2.12. For MH >∼ 180 GeV, theavailable decay modes are limited as the Higgs boson predominantly decays into two gaugebosons. In such cases, the measurement of at least one Higgs–fermion coupling is importantfor establishing the fermion mass generation mechanism. The H → bb branching ratio canbe determined with a 12%, 17% and 28% accuracy for, respectively, MH = 180, 200 and 220GeV, assuming an integrated luminosity of 1 ab−1 at

√s = 800 GeV [98].

Note that invisible Higgs decays can also be probed with a very good accuracy, thanksto the missing mass technique. One can also look directly for the characteristic signature ofmissing energy and momentum. Recent studies show that in the range 120 GeV <∼ MH <∼160 GeV, an accuracy of ∼ 10% can be obtained on a 5% invisible decay and a 5σ signal canbe seen for a branching fraction as low as 2% [92].

II-24 ILC-Reference Design Report

MH (GeV/c2)S

M H

iggs

Bra

nchi

ng R

atio

10-3

10-2

10-1

1

90 100 110 120 130 140 150 160

ILC RDR, 500 fb−1 at 350-500 GeV, for mH = 120 GeV Battaglia & Desch

Enables model-independent extraction of Higgs couplings,

constraints on non-SM Higgs.


17

However:The Higgs boson in the Standard Model

FIGURE 2.7. Production cross sections of the SM Higgs boson at the ILC as a function of MH for√s = 500 GeV (left) and

√s = 1 TeV (right); from Ref. [33].

femtobarns. The ttH final state is generated almost exclusively through Higgs–strahlungoff top quarks and the process allows thus the determination of the important gHtt Yukawacoupling in an almost unambiguous way. The electroweak and QCD corrections are knownand are moderate [76], except near the production threshold where large coulombic correctionsoccur and double the production rate [77]. For MH <∼ 140 GeV, the main signal ttH →W+W−bbbb is spectacular and b–quark tagging as well as the reconstruction of the Higgs masspeak are essential to suppress the large backgrounds. For larger Higgs masses, MH >∼ 140GeV, the process leads mainly to Htt → 4Wbb final states which give rise to ten jets if all Wbosons are allowed to decay hadronically to increase the statistics.

The cross section for double Higgs production in the strahlung process, e+e− → HHZ, isat the level of ∼ 1

2 fb at√

s = 500 GeV for a light Higgs boson, MH ∼ 120 GeV, and is smallerat higher energies [75]. It is rather sensitive to the trilinear Higgs–self coupling λHHH : for√

s=500 GeV and MH =120 GeV for instance, it varies by about 20% for a 50% variation ofλHHH . The electroweak corrections to the process have been shown to be moderate [78]. Thecharacteristic signal for MH <∼ 140 GeV consists of four b–quarks to be tagged and a Z bosonwhich needs to be reconstructed in both leptonic and hadronic final states to increase thestatistics. For higher Higgs masses, the dominant signature is Z + 4W leading to multi–jet(up to 10) and/or multi–lepton final states. The rate for double Higgs production in WWfusion, e+e− → νeνeHH, is extremely small at

√s = 500 GeV but reaches the level of 1

2 fbat 1 TeV; in fact, at high energies, only the latter process can be used.

Finally, future linear colliders can be turned to γγ colliders, in which the photon beamsare generated by Compton back–scattering of laser light with c.m. energies and integratedluminosities only slightly lower than that of the original e+e− collider. Tuning the maximumof the γγ spectrum to the value of MH , the Higgs boson can be formed as s–channel reso-nances, γγ → H, decaying mostly into bb pairs and/or WW ∗, ZZ∗ final states. This allowsprecise measurement of the Higgs couplings to photons, which are mediated by loops possibly

ILC-Reference Design Report II-19

ILC RDR, arXiv:0709.1893

ttH kinematically limited at 500 GeV ILC

Why is top so heavy? Special role in EW symmetry breaking?


18

Combined fit using LHC + ILC500: precision mostly dominatedby ILC. No ILC ttH measurement included.

ttH coupling better than LHC alone due to ILC input to LHC fit.

[GeV]Hm110 120 130 140 150 160 170 180 190

(H,X

)2 g

(H,X

)2

gΔ

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1(H,Z)2g(H,W)2g

)τ(H,2g(H,b)2g(H,t)2g

HΓ



[GeV]Hm110 120 130 140 150 160 170 180 190

(H,X

)2 g

(H,X

)2

gΔ

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1(H,Z)2g

(H,W)2g

)τ(H,2g

(H,b)2g

(H,t)2g

HΓ



Figure 1: Relative accuracies of squared Higgs-boson couplings and the total Higgs-boson width achievable at the LHC alone

with mild theory assumptions [8] (left plot) and in a combined analysis using input from the LHC and ILC [11] (right plot).

squared coupling ∆g2tth

/g2tth

rather than the coupling itself), which corresponds to a relative improvement of about

30% compared to the analysis in Ref. [10].

The complementarity of the LHC and the ILC in both the e+e− and the photon collider mode for determin-

ing Higgs-boson couplings has been investigated in Ref. [12]. The decays H → WW, ZZ have been analysed for

200 GeV <∼ MH

<∼ 350 GeV in a Two Higgs Doublet Model (II) with CP violation. It has been found that the

measurements at the photon collider are complementary to the ones at the LHC and the ILC, since they are sensitive

to different combinations of Higgs-boson couplings. A combined analysis of the LHC and the ILC in its e+e− and

photon collider modes is necessary in order to precisely determine the CP-violating H–A mixing angle φHA.

2.2. Higgs physics in the NMSSM

While the search for the Higgs boson of the SM has been well studied at the LHC and ILC, physics beyond the

SM can give rise to very different Higgs phenomenology. This can be due to modified couplings, mixing with other

states, non-standard production processes or Higgs decays into new particles. Even in the Minimal Supersymmetric

Extension of the Standard Model (MSSM) the Higgs-boson properties can be rather different from those of a SM

Higgs boson. The Next-to-Minimal Supersymmetric Model (NMSSM) has recently found considerable attention as

an attractive extension of the MSSM, since it allows to avoid the fine-tuning and “little hierarchy” problems of the

CP-conserving MSSM. While the NMSSM is theoretically well motivated, its Higgs phenomenology at the LHC can

be very challenging [13]. This is due to the fact that over a large part of the parameter space the SM-like CP-even

Higgs boson of the NMSSM dominantly decays into two light CP-odd Higgs bosons, h → aa. Confirmation of the

nature of a possible LHC signal at the ILC would be vital. For example, the WW → h → aa signal, as well as the

usual e+e− → Zh → Zaa signal, will be highly visible at the ILC due to its cleaner environment and high luminosity.

PSN 0011

G. Weiglein, hep-ph/0508181 [Preliminary]


19

However:

10010–5

10–4

10–3

10–2

10–1

10

bb

τ τ

gg

Z

Z

0Z 0W +

+ –

µ µ

γ

γ γ

+ –

W –

ss

cc t t

0

150 200 250

Higgs Mass (GeV)

Bra

nchi

ng R

atio

300 350 400

HDECAY

H → γγ is a rare mode: statistics limited at ILCBR measurement ∼ 20–35% precision at ILC

Does anything new run in the loop?


20

LHC only Combined fit including ILC500Compare old and new results for gHγγ:

[GeV]Hm110 120 130 140 150 160 170 180 190

G(p

redi

cted

)G

(new

)

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1H / Ginv.G

(W,t)γ(new) / GγG

(t)g(new) / GgG


[GeV]Hm110 120 130 140 150 160 170 180 190

(pre

dict

ed)

Γ(n

ew)

Γ

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1(W,t)γΓ(new) / γΓ

(t)gΓ(new) / gΓ


Sven Heinemeyer, LHC/ILC meeting, SLAC, 23.03.2005 19Talk by S. Heinemeyer, LHC/ILC mtg, SLAC, March 2005

Sensitivity comes from LHC, but need ILC to nail down othercouplings (especially at low MH).Note also ggH coupling measurement at higher Higgs masses.


21

However:

10010–5

10–4

10–3

10–2

10–1

10

bb

τ τ

gg

Z

Z

0Z 0W +

+ –

µ µ

γ

γ γ

+ –

W –

ss

cc t t

0

150 200 250

Higgs Mass (GeV)

Bra

nchi

ng R

atio

300 350 400

HDECAY

H → µµ is an extremely rare mode: BR ∼ 3× 10−4

Light Higgs production: ∼ 60 fb from ZH, ∼ 80 fb from WBF⇒ ∼ 20 H → µµ events in 500 fb−1

Best you could do is 4σ ↔ 25% meas. before any cuts

Do second-generation fermions behave the same as third-generation?


22

H → µµ at LHC, from inclusive production:

y axis: enhancement factor needed over SM rate

100 110 120 130 140 150 1600

1

2

3

4

5

6

7

8

9

10

��

��

��

Han & McElrath, hep-ph/0201023

LHC, 300 fb−1 × 2 detectors; solid = 3σ, dashed = 5σ.

Reach similar in WBF Cranmer & Plehn, hep-ph/0605268


23

H → µµ as a benchmark measurement for Super LHC:

10× LHC luminosity, 3000 fb−1 per experiment.

mH (GeV) S/√

B δσ×BR(H→µµ)σ×BR

120 7.9 0.13130 7.1 0.14140 5.1 0.20150 2.8 0.36

Gianotti, Mangano & Virdee, et al, SLHC report, hep-ph/0204087; numbers extrapolated

from Han & McElrath, hep-ph/0201023

Again, fold together with ILC precision on production couplings

and total width to isolate Hµµ coupling.


24

The point:

- LHC has mass reach: ttH

- (S)LHC has high lumi: Hγγ, Hµµ?

- ILC has low background: precision measurements of decay BRs,

including dominant, hadronic bb

- ILC has clean, precisely-calculated production: decay-independent

precision measurements of production cross sections

Combined: get more than either machine alone


25

SUSY


26

LHC:

- Kinematic access to heavy (colored) SUSY particles (squarks,

gluinos)

- Mass differences from decay chains

ILC:

- Access to uncolored SUSY particles (within kinematic reach)

- Precision mass measurements, including LSP mass

- Precision measurements of couplings of light charginos, neu-

tralinos → gaugino/higgsino composition


27

Measuring the SUSY particle mass spectrum

Cascade decaysg qR

q q

N1

(a)

g qL

q q

N2 f

f f

N1

(b)

g qL

q q′

C1 f

f ′ f

N1

(c)

g qL

q q′

C1 W

N1 f ′

f

(d)

Figure 8.2: Some of the many possible examples of gluino cascade decays ending with a neutralinoLSP in the final state. The squarks appearing in these diagrams may be either on-shell or off-shell,depending on the mass spectrum of the theory.

8.5 Decays to the gravitino/goldstino

Most phenomenological studies of supersymmetry assume explicitly or implicitly that the lightest neu-tralino is the LSP. This is typically the case in gravity-mediated models for the soft terms. However,in gauge-mediated models (and in “no-scale” models), the LSP is instead the gravitino. As we saw insection 6.5, a very light gravitino may be relevant for collider phenomenology, because it contains as itslongitudinal component the goldstino, which has a non-gravitational coupling to all sparticle-particlepairs (X,X). The decay rate found in eq. (6.32) for X → XG is usually not fast enough to competewith the other decays of sparticles X as mentioned above, except in the case that X is the next-to-lightest supersymmetric particle (NLSP). Since the NLSP has no competing decays, it should alwaysdecay into its superpartner and the LSP gravitino.

In principle, any of the MSSM superpartners could be the NLSP in models with a light goldstino,but most models with gauge mediation of supersymmetry breaking have either a neutralino or a chargedlepton playing this role. The argument for this can be seen immediately from eqs. (6.58) and (6.59);since α1 < α2,α3, those superpartners with only U(1)Y interactions will tend to get the smallestmasses. The gauge-eigenstate sparticles with this property are the bino and the right-handed sleptonseR, µR, τR, so the appropriate corresponding mass eigenstates should be plausible candidates for theNLSP.

First suppose that N1 is the NLSP in light goldstino models. Since N1 contains an admixture ofthe photino (the linear combination of bino and neutral wino whose superpartner is the photon), fromeq. (6.32) it decays into photon + goldstino/gravitino with a partial width

Γ(N1 → γG) = 2 × 10−3 κ1γ

( mN1

100 GeV

)5( √〈F 〉

100 TeV

)−4

eV. (8.9)

Here κ1γ ≡ |N11 cos θW + N12 sin θW |2 is the “photino content” of N1, in terms of the neutralinomixing matrix Nij defined by eq. (7.33). We have normalized m

N1and

√〈F 〉 to (very roughly)minimum expected values in gauge-mediated models. This width is much smaller than for a typicalflavor-unsuppressed weak interaction decay, but it is still large enough to allow N1 to decay before ithas left a collider detector, if

√〈F 〉 is less than a few thousand TeV in gauge-mediated models, orequivalently if m3/2 is less than a keV or so when eq. (6.31) holds. In fact, from eq. (8.9), the mean

85

Use kinematic edges to get mass differences in decay chain

104 pb-1

105 pb-1

900

800

700

600

500

400

300

200

100

0 50 100 150 200 250 300 350 400 450 500

m0 (GeV)

m1/

2 (G

eV) tanβ = 2, A0 = 0,

µ < 0

D_D_

1061

n

THLEP2 + Tevatron (sparticle searches)

TH

EX

103 pb-1

Figure 5: Reach for observing dilepton endpoints in SUGRA models with 1 fb−1, 10 fb−1

and 100 fb−1. Theory (TH) and experimental constraints are also indicated [4].

0

50

100

150

200

0 50 100 150mll (GeV)

dσ/d

mll (

Even

ts/1

00fb

-1/0

.375

GeV

)

(a)

0

100

200

300

400

0 200 400 600 800 1000mllq (GeV)

dσ/d

mllq

(Eve

nts/

100f

b-1/5

GeV

)

(b)

0

100

200

300

400

0 200 400 600 800 1000High mlq (GeV)

dσ/d

mlq

(Eve

nts/

100f

b-1/5

GeV

)

(c1)

0

200

400

600

0 200 400 600 800 1000Low mlq (GeV)

dσ/d

mlq

(Eve

nts/

100f

b-1/5

GeV

)

(c2)

0

50

100

150

0 200 400 600 800 1000mllq (GeV)

dσ/d

mllq

(Eve

nts/

100f

b-1/5

GeV

)

(d)

0

20

40

60

80

100

0 200 400 600 800 1000mhq (GeV)

dσ/d

mhq

(Eve

nts/

100f

b-1/5

GeV

)

(e)

Figure 6: Dilepton + jet distributions for mSUGRA Point 5 as described in the text.

illustrated in Figure 5. In particular, a large part of the mSUGRA parameter space withacceptable cold dark matter has light sleptons and hence enhanced !+!− decays.

104 pb-1

105 pb-1

900

800

700

600

500

400

300

200

100

0 50 100 150 200 250 300 350 400 450 500

m0 (GeV)

m1/

2 (G

eV) tanβ = 2, A0 = 0,

µ < 0

D_D_

1061

n


TH

EX

103 pb-1



0

50

100

150

200

0 50 100 150mll (GeV)

dσ/d

mll (

Even

ts/1

00fb

-1/0

.375

GeV

)

(a)

0

100

200

300

400

0 200 400 600 800 1000mllq (GeV)

dσ/d

mllq

(Eve

nts/

100f

b-1/5

GeV

)

(b)

0

100

200

300

400

0 200 400 600 800 1000High mlq (GeV)

dσ/d

mlq

(Eve

nts/

100f

b-1/5

GeV

)

(c1)

0

200

400

600

0 200 400 600 800 1000Low mlq (GeV)

dσ/d

mlq

(Eve

nts/

100f

b-1/5

GeV

)

(c2)

0

50

100

150

0 200 400 600 800 1000mllq (GeV)

dσ/d

mllq

(Eve

nts/

100f

b-1/5

GeV

)

(d)

0

20

40

60

80

100

0 200 400 600 800 1000mhq (GeV)

dσ/d

mhq

(Eve

nts/

100f

b-1/5

GeV

)

(e)



104 pb-1

105 pb-1

900

800

700

600

500

400

300

200

100

0 50 100 150 200 250 300 350 400 450 500

m0 (GeV)

m1/

2 (G

eV) tanβ = 2, A0 = 0,

µ < 0

D_D_

1061

n


TH

EX

103 pb-1



0

50

100

150

200

0 50 100 150mll (GeV)

dσ/d

mll (

Even

ts/1

00fb

-1/0

.375

GeV

)

(a)

0

100

200

300

400

0 200 400 600 800 1000mllq (GeV)

dσ/d

mllq

(Eve

nts/

100f

b-1/5

GeV

)

(b)

0

100

200

300

400

0 200 400 600 800 1000High mlq (GeV)

dσ/d

mlq

(Eve

nts/

100f

b-1/5

GeV

)

(c1)

0

200

400

600

0 200 400 600 800 1000Low mlq (GeV)

dσ/d

mlq

(Eve

nts/

100f

b-1/5

GeV

)

(c2)

0

50

100

150

0 200 400 600 800 1000mllq (GeV)

dσ/d

mllq

(Eve

nts/

100f

b-1/5

GeV

)

(d)

0

20

40

60

80

100

0 200 400 600 800 1000mhq (GeV)

dσ/d

mhq

(Eve

nts/

100f

b-1/5

GeV

)

(e)



Paige, hep-ph/0211017


28

Because LHC sensitivity comes from mass differences, mass un-certainties are correlated.

ILC input: precision measurement of LSP mass collapses thedegeneracy!

SPS1a, LHC/ILC study

M. Chiorboli et al., in Physics Interplay of the LHC and ILC, hep-ph/0410364

SPS1a; dots are LHC, vertical bands are ILC measurement σ = 0.2%


29

ILC can see lighter states: precision mass measurements,

chargino/neutralino sector parameters.

0102030

L B G I C J HMA E F K D 0102030

L B G I C J HMA E F K D

0102030

L B G I C J HMA E F K D 0102030


0102030


Post-WMAP BenchmarksLHC

Nb.

of O

bser

vabl

e Par

ticle

s LC 0.5 TeV

gluino squarks sleptons χ H

LC 1.0 TeV LHC+LC 1 TeV

CLIC 3 TeV CLIC 5 TeV

0102030


Figure 3: Summary of the numbers of MSSM particles that may be detectable at variousaccelerators in the updated benchmark scenarios. As in [3], we see that the capabilities of theLHC and of linear e+e− colliders are largely complementary. We re-emphasize that massand coupling measurements at e+e− colliders are usually much cleaner and more precisethan at hadron-hadron colliders such as the LHC, where, for example, it is not known howto distinguish the light squark flavours.

14

Battaglia et al., hep-ph/0306219

Need LHC for the heavier states: complete the mass spectrum

with ILC input for decay chain reconstruction.


30

Weak-scale SUSY parameters + RGEs → run up to high scale.

Probe grand unification, SUSY breaking pattern.

(a) 1/Mi [GeV−1]

Q [GeV]

(b)

(c) M2j

[103 GeV2]

Q [GeV]

(d) M2j

[103 GeV2]

Q [GeV]

Figure 2: mSUGRA: Evolution, from low to high scales, of (a) gaugino mass parame-

ters, and (b) unification of gaugino mass parameter pairs; (c) evolution of first–generation

sfermion mass parameters and the Higgs mass parameter M2H2

; (d) evolution of third–

generation sfermion mass parameters and the Higgs mass parameter M2H1

. The mSUGRA

point probed is defined by the parameters M0 = 200 GeV, M1/2 = 250 GeV, A0 = -100 GeV,

tanβ = 10, and sign(µ) = (+). [The widths of the bands indicate the 1σ CL.]

11

(a) 1/Mi [GeV−1]

Q [GeV]

(b)

(c) M2j

[103 GeV2]

Q [GeV]

(d) M2j

[103 GeV2]

Q [GeV]






. The mSUGRA



11

(a) 1/Mi [GeV−1]

Q [GeV]

(b)

(c) M2j

[103 GeV2]

Q [GeV]

(d) M2j

[103 GeV2]

Q [GeV]






. The mSUGRA



11

Blair, Porod & Zerwas, hep-ph/0210058 [mSUGRA]

Requires ILC precision meas. of lighter masses, couplings.

Requires LHC reach for heavier masses.


31

Testing SUSY coupling relations

Gauge couplings ↔ gaugino Yukawa couplings:

Gluon-quark-quark coupling ↔ gluino-squark-quark coupling.

q

q* q’*

q’

g

q,q*

g q,q*

g

q

g

g

g

g

g

q

q,q*

q

q,q*

g

q

g

g

q

q

q

q*

g

g

q

Figure 1: Some examples for Feynman diagrams for partonic squark and gluino productionin hadron collisions. Dots indicate the gauge coupling gs, while squares stand for the Yukawacoupling gs.

dominantly produces u and d squarks, with smaller admixtures of sea-flavoured squarks, indirect proportion to the quark content of the proton at the relevant x and Q2 values.

Due to the flavour locking, only the first two generations of squarks are thus relevant,for which mixing effects are small and we can take mass and current eigenstates to beidentical to good approximation. That is, the heavier q mass eigenstate is pure qL (weakisospin doublet), and the lighter one pure qR (weak isospin singlet). Nominally, the lighterone would be the better target for a high-statistics study, simply due to phase space, butsince it doesn’t couple to weak interactions, it decays almost exclusively via the hyperchargecoupling to a same-flavour quark and the LSP. Since charge tagging for light-flavour jets isexceedingly difficult, this decay mode effectively obscures the fact that we had same-flavoursquarks to begin with. Moreover, since it only contains a jet and missing energy, the modewould be extremely challenging to separate from the background. The only feasible avenuethus appears to be to use flavour/charge tagging modes of the heavier mass eigenstates, theqL.

For qL, the charge of the squark can be tagged through a chargino decay chain,

uL → d χ+1 → d l+ νl χ

01, dL → u χ−

1 → u l− νl χ01, (3)

u∗L → d χ−

1 → d l− νl χ01, d∗

L → u χ+1 → u l+ νl χ

01, (4)

and similarly for sL and cL. For a given squark flavor, the sign of the final-state lepton isrelated to the charge of the (anti-)squark. The production of same-sign squarks throughthe diagram in the lower left corner of Fig. 1 with this decay channel will therefore leadto same-sign leptons in the final state, while other direct squark production processes willtend to produce opposite-sign leptons in the final state. At this level, the signal is thuscharacterized by two same-sign leptons, two hard jets and missing transverse energy in thefinal state.

3

Freitas & Skands, hep-ph/0606121

LHC analysis, but requires ILC input for squark decay BRs.


32

MSSM vs. NMSSM

NMSSM is the MSSM plus one extra singlet superfield: 5th neu-

tralino, extra Higgs scalar and pseudoscalar.

Case study: choose parameters so that masses and cross sections

of accessible neutralinos & charginos can look identical between

MSSM and NMSSM at LHC and at ILC500.

How do we check?

→ combined analysis.


33

ILC: Measure masses, cross sections, branching ratios of lighercharginos/neutralinos; assume MSSM and extract M1, M2, µ,tanβ.

Assume MSSM and predict mass, gaugino content of neutralinos3 and 4.

Compare LHC measurement: contradiction!

Figure 2. Predicted masses and gauginoadmixture for the heavier neutralinos χ0

3 andχ0

4 within the parameter ranges consistentat the ILC500 analysis in the MSSM anda measured mass mχ0

i= 367 ± 7 GeV of

a neutralino with sufficiently high gauginoadmixture in cascade decays at the LHC. Wetook a lower bound of a detectable gauginoadmixture of about 10% [9].

without spin correlationswith spin correlations

AFB [%]

mνe [GeV]2500200015001000500

20

15

10

5

0

√s = 350 GeV

Figure 3. Forward–backward asymmetryof e− in the process e+e− → χ+

1 χ−1 , χ−

1 →χ0

1"−ν, shown as a function of mνe . For

nominal value of mνe = 1994 GeV theexpected experimental errors are shown. Forillustration only, the dashed line shows thatneglecting spin correlations would lead toa completely wrong interpretation of theexperimental data [13].

[5] P. Bechtle, K. Desch and P. Wienemann, Comput. Phys. Commun. 174 (2006) 47 [hep-ph/0412012];R. Lafaye, T. Plehn and D. Zerwas, hep-ph/0404282.

[6] G. Weiglein et al. [LHC/LC Study Group], Phys. Rept. 426 (2006) 47 [hep-ph/0410364].[7] B. C. Allanach et al., Eur. Phys. J. C 25 (2002) 113 [eConf C010630 (2001) P125] [arXiv:hep-ph/0202233].[8] G. A. Moortgat-Pick et al., hep-ph/0507011, to be published in Physics Reports.[9] G. A. Moortgat-Pick, S. Hesselbach, F. Franke and H. Fraas, JHEP 0506 (2005) 048 [arXiv:hep-ph/0502036];

G. A. Moortgat-Pick, S. Hesselbach, F. Franke and H. Fraas, hep-ph/0508313.[10] M. Drees, Int. J. Mod. Phys. A 4 (1989) 3635; J. R. Ellis, J. F. Gunion, H. E. Haber, L. Roszkowski and

F. Zwirner, Phys. Rev. D 39 (1989) 844; U. Ellwanger, M. Rausch de Traubenberg and C. A. Savoy, Phys.Lett. B 315 (1993) 331 [hep-ph/9307322]; Z. Phys. C 59 (1993) 575; B. R. Kim, A. Stephan and S. K. Oh,Phys. Lett. B 336 (1994) 200; F. Franke and H. Fraas, Int. J. Mod. Phys. A 12 (1997) 479 [hep-ph/9512366];U. Ellwanger, J. F. Gunion and C. Hugonie, JHEP 0502 (2005) 066 [arXiv:hep-ph/0406215].

[11] P. N. Pandita, Z. Phys. C 63 (1994) 659; U. Ellwanger and C. Hugonie, Eur. Phys. J. C 5 (1998)723 [arXiv:hep-ph/9712300]; S. Hesselbach, F. Franke and H. Fraas, Phys. Lett. B 492 (2000) 140[hep-ph/0007310]; F. Franke and S. Hesselbach, Phys. Lett. B 526 (2002) 370 [hep-ph/0111285];S. Hesselbach, F. Franke and H. Fraas, hep-ph/0003272; Eur. Phys. J. C 23 (2002) 149 [hep-ph/0107080];S. Y. Choi, D. J. Miller and P. M. Zerwas, Nucl. Phys. B 711 (2005) 83 [hep-ph/0407209]; V. Barger,P. Langacker and G. Shaughnessy, Phys. Lett. B 644 (2007) 361 [arXiv:hep-ph/0609068].

[12] G. Moortgat-Pick and H. Fraas, Acta Phys. Polon. B 30 (1999) 1999 [hep-ph/9904209]; G. Moortgat-Pick,A. Bartl, H. Fraas and W. Majerotto, Eur. Phys. J. C 18 (2000) 379 [hep-ph/0007222].

[13] K. Desch, J. Kalinowski, G. Moortgat-Pick, K. Rolbiecki and W. J. Stirling, JHEP 0612 (2006) 007[hep-ph/0607104]; K. Rolbiecki, K. Desch, J. Kalinowski and G. Moortgat-Pick, hep-ph/0605168.

[14] G. Moortgat-Pick, H. Fraas, A. Bartl and W. Majerotto, Eur. Phys. J. C 7 (1999) 113 [hep-ph/9804306];G. Moortgat-Pick and H. Fraas, Phys. Rev. D 59 (1999) 015016 [hep-ph/9708481].

[15] K. Kawagoe, M. M. Nojiri and G. Polesello, Phys. Rev. D 71 (2005) 035008 [hep-ph/0410160].[16] B. K. Gjelsten, D. J. Miller and P. Osland, JHEP 0506 (2005) 015 [hep-ph/0501033].[17] H. U. Martyn and G. A. Blair, hep-ph/9910416; J. A. Aguilar-Saavedra et al., hep-ph/0106315;

K. Abe et al., hep-ph/0109166. T. Abe et al., hep-ex/0106056.

Moortgat-Pick et al., hep-ph/0508313


34

Summary

LHC will open up the TeV-scale frontier.- Electroweak symmetry breaking?

- Dark matter?

- New forces?

- Supersymmetry?

- Extra dimensions?

- Something we haven’t thought of yet??

ILC will provide the precision measurements needed to under-stand the new physics.

ILC analyses will not be done in isolation: important inputs fromLHC physics.

? LHC analyses should not allow information to be “lost”: beprepared to come back with ILC data and break correlated un-certainties.


35

Carleton Universitylogan/talks/LHCILCpheno.pdf · Carleton University Pheno 2008 Symposium – LHC...

Documents

Transcript of Carleton Universitylogan/talks/LHCILCpheno.pdf · Carleton University Pheno 2008 Symposium – LHC...