LHC Results and the Interpretation of Cosmic-Ray Data · 2014. 5. 27. · Petrukhin, NPB 151 (2006) 57 Barcelo at al. JACP 06 (2009) 027 Dixit et al. EPJC 68 (2010) 573 Petrukhin

LHC Results and the Interpretation ofCosmic-Ray Data

Ralph Engel, Tanguy Pierog, and Ralf Ulrich

Karlsruhe Institute of Technology (KIT)

Cosmic ray flux and interaction energies

2

Energy (eV/particle)1310 1410 1510 1610 1710 1810 1910 2010 2110

)1.

5 e

V-1

sr

-1 s

-2 J

(E)

(m2.

5Sc

aled

flux

E

1310

1410

1510

1610

1710

1810

1910

(GeV)ppsEquivalent c.m. energy 210 310 410 510 610

RHIC (p-p)-p)aHERA (

Tevatron (p-p) 14 TeV7 TeVLHC (p-p)

ATICPROTONRUNJOB

KASCADE (SIBYLL 2.1)KASCADE-Grande 2012Tibet ASg (SIBYLL 2.1)IceTop ICRC 2013

HiRes-MIAHiRes IHiRes IIAuger ICRC 2013TA SD 2013

Center-of-mass energy

Laboratory energy

Exotic models for the knee

3

New physics: scaling with nucleon-nucleon cms energy

E0

EX ~100 TeV

log(E)

log(Flux)Knee due to wrong energy reconstruction of showers?

Atmosphere

Cosmic ray

Threshold scales with E/A

Petrukhin, NPB 151 (2006) 57Barcelo at al. JACP 06 (2009) 027Dixit et al. EPJC 68 (2010) 573Petrukhin NPB 212 (2011) 235


)1.

5 e

V-1

sr

-1 s

-2 J

(E)

(m2.

5Sc

aled

flux

E

1310

1410

1510

1610

1710

1810

1910




ATICPROTONRUNJOB



LHC data probe the region beyond the knee

4

~20% of energy needs to be transferred to invisible channel

(d‘Enterria et al., Astropart. Phys. 35, 2011)

Problem of limited phase space coverage

5η

-15 -10 -5 0 5 10 15

ηdN

/d

0123456789

10

[GeV

]η

/d kdE

01002003004005006007008009001000

=14 TeVs charged→p+p all→p+p

tracking

calorimetry

ATLASCMSALICELHCbLHCf

ATLASCMS+TOTEMALICELHCbLHCf

Detector Collider setup

Beam Beam

Air showers: Particles of highest energy most important

h =� ln tanq2

q

LHC data probe the region beyond the knee

6!"#"$%&'()*#+,(-./('*01/(!2#$2#!! (((((((((((((((((((((((((((((((((((((((((((( ((((( ((((((((((((((((((((((((((((((((( (((((((34567(78*9:;<<64(='*01>

!(('?4<@;7A?47<B9(74:4(5C('0(DB7;ECF

(

!"#$%&'()*+,%-)-./$)'0&-.-'&!"#$%&'()*+,%-)-./$)'0&-.-'&

$GH(I;J "GK2(I;J LG$(I;J

!(H$$A.;J(74:4(M;EE(<;N<B7OP;7(=+'C(:O9;7(:B(QNNQ/(I;54:<B9>

!(R4<:6PE;(DOE:6NE6P6:S(9B:(PBDNE;:;ES(M;EE(N<;76P:;7(4:(LG$(I;J

!(Q6DNE;C:(DB7;EC(=T.QU*IA$!/(QV,W%%("G!>(C;;D(X;::;<(:?49(DB<;(<;P;9:(B9;C(GGG

LHC: Exotic scenarios for knee very unlikely, model predictionsbracket LHC data on secondary particle multiplicity

Protons: Elab = 3 x 1016 eVPseudorapidity distribution of charged particles

h =� ln tanq2

q

(D‘Enterria at al. Astropart Phys 35, 2011)

1.2 Interaction models for describing LHC data

7

Cosmic Ray Models and Recent LHC Data: LHCb

e.g.

Comparison on event generator level:

Forward energy flow

Forward Lambda production, strangeness

More in preparation...

Eur.Phys.J. C73 (2013) 2421

[email protected] CRMC, an interface to cosmic ray event generators 9

Cosmic Ray Models and Recent LHC Data: TOTEM

Forward charged multiplicities: Europhys.Lett. 98 (2012) 31002

[email protected] CRMC, an interface to cosmic ray event generators 11

(Europhys.Lett. 98 (2012) 31002)

Energy flow (CMS)Forward particle multiplicity (TOTEM)

(Eur.Phys.J. C73 (2013) 2421)

- Models used for p-p, p-Pb and Pb-Pb data- Interface to cosmic ray event generators CRMC (R. Ulrich et al.)

LHCf: forward photon production at 7 TeV

8

LHCf Collaboration / Physics Letters B 703 (2011) 128–134 133

Fig. 5. Comparison of the single photon energy spectra between the experimental data and the MC predictions. Top panels show the spectra and the bottom panels show theratios of MC results to experimental data. Left (right) panel shows the results for the large (small) rapidity range. Different colors show the results from experimental data(black), QGSJET II-03 (blue), DPMJET 3.04 (red), SIBYLL 2.1 (green), EPOS 1.99 (magenta) and PYTHIA 8.145 (yellow). Error bars and gray shaded areas in each plot indicate theexperimental statistical and the systematic errors, respectively. The magenta shaded area indicates the statistical error of the MC data set using EPOS 1.99 as a representativeof the other models. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this Letter.)

LHCf detectors by two methods; first by using the distribution ofparticle impact positions measured by the LHCf detectors and sec-ond by using the information from the Beam Position Monitors(BPMSW) installed ±21 m from the IP [24]. From the analysis ofthe fills 1089–1134, we found a maximum ∼4 mm shift of thebeam center at the LHCf detectors, corresponding to a crossing an-gle of ∼30 µrad assuming the beam transverse position did notchange. The two analyses gave consistent results for the locationof the beam center on the detectors within 1 mm accuracy. Inthe geometrical construction of events we used the beam-centerdetermined by LHCf data. We derived photon energy spectra byshifting the beam-center by 1 mm. The spectra are modified by5–20% depending on the energy and the rapidity range. This isassigned as a part of systematic uncertainty in the final energyspectra.

The background from collisions between the beam and theresidual gas in the vacuum beam pipe can be estimated from thedata. During LHC operation, there were always bunches that didnot have a colliding bunch in the opposite beam at IP1. We callthese bunches ‘non-crossing bunches’ while the normal bunchesare called as ‘crossing bunches.’ The events associated with thenon-crossing bunches are purely from the beam-gas backgroundwhile the events with the crossing bunches are mixture of beam-beam collisions and beam-gas background. Because the event rateof the beam-gas background is proportional to the bunch inten-sity, we can calculate the background spectrum contained in thecrossing bunch data by scaling the non-crossing bunch events. Wefound the contamination from the beam-gas background in the fi-nal energy spectrum is only ∼0.1%. In addition the shape of the

energy spectrum of beam-gas events is similar to that of beam-beam events, so beam-gas events do not have any significant im-pact on the beam-beam event spectrum.

The collision products and beam halo particles can hit the beampipe and produce particles that enter the LHCf detectors. Howeveraccording to MC simulations, these particles have energy below100 GeV [10] and do not affect the analysis presented in this Let-ter.

5. Comparison with models

In the top panels of Fig. 5 photon spectra predicted byMC simulations using different models, QGSJET II-03 (blue) [22],DPMJET 3.04 (red) [21], SIBYLL 2.1 (green) [25], EPOS 1.99 (ma-genta) [20] and PYTHIA 8.145 (default parameter set; yellow) [26,27] for collisions products are presented together with the com-bined experimental results. To combine the experimental data ofthe Arm1 and Arm2 detectors, the content in each energy bin wasaveraged with weights by the inverse of errors. The systematic un-certainties due to the multi-hit cut, particle identification (PID),absolute energy scale and beam center uncertainty are quadrati-cally added in each energy bin and shown as gray shaded areas inFig. 5. The uncertainty in the luminosity determination (±6.1% asdiscussed in Section 2), that is not shown in Fig. 5, can make anenergy independent shift of all spectra.

In the MC simulations, 1.0 × 107 inelastic collisions were gen-erated and the secondary particles transported in the beam pipe.Deflection of charged particles by the D1 beam separation dipole,particle decay and particle interaction with the beam pipe are

(LHCf Collab., Phys. Lett. B 703, 2011)

Re-tuning of models needed, size of effect still unclear, distributions for neutrons needed

Forward production of gamma-rays with low pt pp ! g X

η>10.15

η>10.15

8.77 <η<9.46

8.77 <η<9.46

Arm 2

Arm 1

Shower as cascade process: primary energy

9

Decay of neutral pions feeds em. shower componentDecay of charged pions (~30 GeV) feeds muonic component

0 10000 20000 30000 40000 50000 60000particle number

atm

osph

eric

dep

th (g

/cm

2 )

electrons

muons (x 10)

hadrons(x 10)

0

1

3

5

7

10

20

0

200

400

600

800

1000

altit

ude

(km

)

Air shower ground arrays: Ne and Nµ

102

103

104

105

106

107

108

102 103 104 105 106 107 108 109 1010

Num

ber o

f muo

ns N

µ (E

µ>1

GeV

)

Shower size Ne (Ee>1 MeV)

QGSJet, protoniron

1014 eV

1015 eV

1017 eV

1016 eV

10

lnE = a · lnNe + b · lnNµ

Energy conservation

KASCADE and KASCADE-Grande

1014 eV

1015 eV

1017 eV

1016 eV

102

103

104

105

106

107

108

102 103 104 105 106 107 108 109 1010

Num

ber o

f muo

ns N

µ (E

µ>1

GeV

)

Shower size Ne (Ee>1 MeV)

Sibyll, protonQGSJet, proton

Sibyll, ironQGSJet, iron

11

lnE = a · lnNe + b · lnNµ

Energy conservation

KASCADE and KASCADE-Grande

Air shower ground arrays: Ne and Nµ

12

The energy spectrum from surface detector data (I)

slant d

epth [g/cm2

]

1000

500

40

30

dE/dX [P

eV/(g

/cm2)]

20

10

r [m]500 1000 1500 2000 2500

Sign

al [V

EM]

1

10

210

310

410

r [m]500 1000 1500 2000 2500

Sign

al [V

EM]

1

10

210

310

410

/eV)FD

lg(E18.5 19 19.5

/VEM

)38

lg(S

1

1.5

2

2.5795 eventsEmax = 6× 1019 eV

C. DiGiulio (0142), this conf.

3 / 23

The energy spectrum from surface detector data (I)

slant d

epth [g/cm2

]

1000

500

40

30

dE/dX [P

eV/(g

/cm2)]

20

10

r [m]500 1000 1500 2000 2500

Sign

al [V

EM]

1

10

210

310

410

r [m]500 1000 1500 2000 2500

Sign

al [V

EM]

1

10

210

310

410

/eV)FD

lg(E18.5 19 19.5

/VEM

)38

lg(S

1

1.5

2

2.5795 eventsEmax = 6× 1019 eV

C. DiGiulio (0142), this conf.

3 / 23

Calorimetric measurement at very high energy

Example: event observed by Auger

Lateral distribution

Longitudinal p

rofile

Ecal

=Z ✓

dEion

dX

◆

obs

dX

E = fcorr

·Ecal

Cross-check:Muons at ground

Energy spectrum of cosmic rays

13


)1.

5 e

V-1

sr

-1 s

-2 J

(E)

(m2.

5Sc

aled

flux

E

1310

1410

1510

1610

1710

1810

1910




ATICPROTONRUNJOB



Energy spectrum well-measured, sys. uncertainties need to be reduced further

Key targets: mass composition and anisotropy

14

6

Primary Energy, E [GeV]310 410 510 610 710 810 910 1010 1110

]-1

s-1

sr

-2 m

1.6

dN

/dE

[GeV

× 2.

6E

1

10

210

310

410

Pamela ProtonPamela HeCreamII ProtonCreamII HeCreamII CCreamII OCreamII MgCreamII SiCreamII Fe

Proton_totalHe_totalC_totalO_totalFe_totalZ=53 groupZ=80 group

FIG. 4: Overview of the spectrum from below the knee to the ankle with the fit of Table III. Air shower data shifted as inFigs. 2 and 3. Left: lines showing individual groups of nuclei from all populations compared to data from PAMELA [9] andCREAM [7] at low energy. Right: shaded regions show the overlapping contributions of the three populations.

the all-particle spectrum is given by

φi(E) = Σ3j=1 ai,j E

−γi,j× exp

!

−E

ZiRc,j

"

. (3)

The spectral indices for each group and the normaliza-tions are given explicitly in Table II. The parameters forPopulation 1 are from Refs. [7, 8], which we assume canbe extrapolated to a rigidity of 4 PV to describe the knee.In Eq. 3 φi is dN/dlnE and γi is the integral spectral in-dex. The subscript i = 1, 5 runs over the standard fivegroups (p, He, CNO, Mg-Si and Fe), and the all-particlespectrum is the sum of the five. This model is plotted asthe solid line in Figs. 2 and 3.

B. An alternative picture and global fit

Spectra for the second fit are given by the same Eq. 3but with qualitatively different parameters, as given inTable III. In particular, the first population has a muchlower cutoff of Rc = 120 TV. This description is relatedto the significantly harder spectra assumed for the firstpopulation. Each component in the first population is fit-ted only above Rc = 200 GV, after the spectra hardeningnoted in Refs. [8] and [9]. With these harder spectra (ascompared to Table II), the heavy components cannot beextended past the knee region. It is interesting to notethat Rc ≈ 100TV is the classical result for the expectedmaximum energy of supernova remnants expanding intothe interstellar medium with an un-amplified magneticfield of a few µGauss [44].

The spectrum with the parameters of Table III isshown in Fig. 4 from below the knee to the ankle. Thecontributions of individual groups of nuclei are shown,as well as the spectra of nuclei from CREAM [8]. Wenote that the bump in the spectrum around 1017 eV cor-responds with the “iron knee” reported by KASCADE-Grande in their electron rich sample [45] and also notedby GAMMA [37]. A tendency for increasing mass abovethe knee has been noted for a long time (for example byCASA-MIA [46]), which seems now to be confirmed withhigher resolution.Another noteworthy feature is the possibility illus-

trated in this fit of explaining the ankle as a Peters cy-cle containing only protons and iron. This possibility isalso suggested in Ref. [32] as an example of their “disap-pointing” model [47] of the end of the cosmic-ray spec-trum. Such a picture is disappointing because the end ofthe spectrum would correspond to the highest energy towhich cosmic-ray acceleration is possible, rather than tothe Greisen-Zatsepin-Kuz’min effect in which higher en-ergy particles lose energy in interactions with the cosmicmicrowave background [48, 49].

C. Comments on fitting with several populations

In both fits above we refer to three populations of par-ticles, with spectral indices for each nuclear componentand a single characteristic maximum rigidity for eachpopulation. The latter assumption has the effect of mak-ing the composition become heavier as each population

6

Primary Energy, E [GeV]310 410 510 610 710 810 910 1010 1110

]-1

s-1

sr

-2 m

1.6

dN

/dE

[GeV

× 2.

6E

1

10

210

310

410

Pamela ProtonPamela HeCreamII ProtonCreamII HeCreamII CCreamII OCreamII MgCreamII SiCreamII Fe

Proton_totalHe_totalC_totalO_totalFe_totalZ=53 groupZ=80 group

FIG. 4: Overview of the spectrum from below the knee to the ankle with the fit of Table III. Air shower data shifted as inFigs. 2 and 3. Left: lines showing individual groups of nuclei from all populations compared to data from PAMELA [9] andCREAM [7] at low energy. Right: shaded regions show the overlapping contributions of the three populations.

the all-particle spectrum is given by

φi(E) = Σ3j=1 ai,j E

−γi,j× exp

!

−E

ZiRc,j

"

. (3)

The spectral indices for each group and the normaliza-tions are given explicitly in Table II. The parameters forPopulation 1 are from Refs. [7, 8], which we assume canbe extrapolated to a rigidity of 4 PV to describe the knee.In Eq. 3 φi is dN/dlnE and γi is the integral spectral in-dex. The subscript i = 1, 5 runs over the standard fivegroups (p, He, CNO, Mg-Si and Fe), and the all-particlespectrum is the sum of the five. This model is plotted asthe solid line in Figs. 2 and 3.

B. An alternative picture and global fit

Spectra for the second fit are given by the same Eq. 3but with qualitatively different parameters, as given inTable III. In particular, the first population has a muchlower cutoff of Rc = 120 TV. This description is relatedto the significantly harder spectra assumed for the firstpopulation. Each component in the first population is fit-ted only above Rc = 200 GV, after the spectra hardeningnoted in Refs. [8] and [9]. With these harder spectra (ascompared to Table II), the heavy components cannot beextended past the knee region. It is interesting to notethat Rc ≈ 100TV is the classical result for the expectedmaximum energy of supernova remnants expanding intothe interstellar medium with an un-amplified magneticfield of a few µGauss [44].

The spectrum with the parameters of Table III isshown in Fig. 4 from below the knee to the ankle. Thecontributions of individual groups of nuclei are shown,as well as the spectra of nuclei from CREAM [8]. Wenote that the bump in the spectrum around 1017 eV cor-responds with the “iron knee” reported by KASCADE-Grande in their electron rich sample [45] and also notedby GAMMA [37]. A tendency for increasing mass abovethe knee has been noted for a long time (for example byCASA-MIA [46]), which seems now to be confirmed withhigher resolution.Another noteworthy feature is the possibility illus-

trated in this fit of explaining the ankle as a Peters cy-cle containing only protons and iron. This possibility isalso suggested in Ref. [32] as an example of their “disap-pointing” model [47] of the end of the cosmic-ray spec-trum. Such a picture is disappointing because the end ofthe spectrum would correspond to the highest energy towhich cosmic-ray acceleration is possible, rather than tothe Greisen-Zatsepin-Kuz’min effect in which higher en-ergy particles lose energy in interactions with the cosmicmicrowave background [48, 49].

C. Comments on fitting with several populations

In both fits above we refer to three populations of par-ticles, with spectral indices for each nuclear componentand a single characteristic maximum rigidity for eachpopulation. The latter assumption has the effect of mak-ing the composition become heavier as each population

Example: three differentsource populations

Extragalactic sources

Composition not well-known at high energy,

needed for astrophysical interpretation

(Gaisser, Stanev, Tilav 1303.3565)

Tuning of interaction models to LHC data (i)

15

0255075

100125150175200225250

10 2 10 3 10 4 10 5 10 6

√s (GeV)

σ (m

b) p + pTotal

Inelastic

Elastic

EPOS 1.99QGSJETII-03QGSJET01SIBYLL 2.1TOTEM

0255075

100125150175200225250

10 2 10 3 10 4 10 5 10 6

√s (GeV)

σ (m

b) p + pTotal

Inelastic

Elastic

EPOS LHCQGSJETII-04TOTEM

Figure 6: Total, inelastic and elastic p-p cross section calculated with EPOS 1.99 (solid line),QGSJETII-03 (dashed line), QGSJET01 (dash-dotted line) and SIBYLL 2.1 (dotted line) on leftpanel, and EPOS LHC (solid line) and QGSJETII-04 (dashed line) on right panel. Points are datafrom [5] and the stars are the LHC measurements by the TOTEM experiment [6].

0

1

2

3

4

5

6

7

-10 -5 0 5 10 pseudorapidity η

(1/N

) dN

/ dη

7 TeV

900 GeV

ALICE p + p → chrg Inel>0

EPOS 1.99QGSJETII-03QGSJET01SIBYLL 2.1

0

1

2

3

4

5

6

7


(1/N

) dN

/ dη

7 TeV

900 GeV


EPOS LHCQGSJETII-04

Figure 7: Pseudorapidity distribution dN/dη of charged particles for events with at least onecharged particle with |η| < 1 for p-p interactions at 900 GeV and 7 TeV. Simulations withEPOS 1.99 (solid line), QGSJETII-03 (dashed line), QGSJET01 (dash-dotted line) and SIBYLL 2.1(dotted line) on left panel, and EPOS LHC (solid line) and QGSJETII-04 (dashed line) on rightpanel, are compared to data points from ALICE experiment [7].

3 Progress due to LHC measurements

3.1 Phase space coverage

Phase space plot in η vs. p⊥ of the different LHC experiments

3.2 Model comparison to LHC data

Old and new models side-by-side:

• Cross section p-p (total, elastic)

• pseudorapidity distribution

• multiplicity distribution

• Antibaryon production rate, discussion of comparison Tevatron vs. LHC

6

0255075

100125150175200225250

10 2 10 3 10 4 10 5 10 6

√s (GeV)

σ (m

b) p + pTotal

Inelastic

Elastic

EPOS 1.99QGSJETII-03QGSJET01SIBYLL 2.1TOTEM

0255075

100125150175200225250

10 2 10 3 10 4 10 5 10 6

√s (GeV)

σ (m

b) p + pTotal

Inelastic

Elastic

EPOS LHCQGSJETII-04TOTEM

Figure 6: Total, inelastic and elastic p-p cross section calculated with EPOS 1.99 (solid line),QGSJETII-03 (dashed line), QGSJET01 (dash-dotted line) and SIBYLL 2.1 (dotted line) on leftpanel, and EPOS LHC (solid line) and QGSJETII-04 (dashed line) on right panel. Points are datafrom [5] and the stars are the LHC measurements by the TOTEM experiment [6].

0

1

2

3

4

5

6

7


(1/N

) dN

/ dη

7 TeV

900 GeV



0

1

2

3

4

5

6

7


(1/N

) dN

/ dη

7 TeV

900 GeV


EPOS LHCQGSJETII-04

Figure 7: Pseudorapidity distribution dN/dη of charged particles for events with at least onecharged particle with |η| < 1 for p-p interactions at 900 GeV and 7 TeV. Simulations withEPOS 1.99 (solid line), QGSJETII-03 (dashed line), QGSJET01 (dash-dotted line) and SIBYLL 2.1(dotted line) on left panel, and EPOS LHC (solid line) and QGSJETII-04 (dashed line) on rightpanel, are compared to data points from ALICE experiment [7].

3 Progress due to LHC measurements

3.1 Phase space coverage

Phase space plot in η vs. p⊥ of the different LHC experiments

3.2 Model comparison to LHC data

Old and new models side-by-side:

• Cross section p-p (total, elastic)

• pseudorapidity distribution

• multiplicity distribution

• Antibaryon production rate, discussion of comparison Tevatron vs. LHC

6

Tuning of interaction models to LHC data (ii)

16

10-4

10-3

10-2

10-1

50 100 150 multiplicity n

P(n

)

ATLAS p + p → chrg √s = 7 TeV | η| < 2.5 Nch> 1 pt>0.1 GeV


10-4

10-3

10-2

10-1


P(n

)

ATLAS p + p → chrg √s = 7 TeV | η| < 2.5 Nch> 1 pt>0.1 GeV

EPOS LHCQGSJETII-04

Figure 8: Multiplicity distribution of charged particles with pt >100 MeV and |η| < 2.5 forp-p interactions at 7 TeV. Simulations with EPOS 1.99 (solid line), QGSJETII-03 (dashed line),QGSJET01 (dash-dotted line) and SIBYLL 2.1 (dotted line) on left panel, and EPOS LHC (solidline) and QGSJETII-04 (dashed line) on right panel, are compared to data points from ATLAScollaboration [8]

0

0.02

0.04

0.06

0.08

0.1

0.12

0 25 50 75 100 Nch

ratio CMS p + p √s = 7 TeV

| y| < 1.

(ap+p)/(π++π-)


0

0.02

0.04

0.06

0.08

0.1

0.12

0 25 50 75 100 Nch

ratio CMS p + p √s = 7 TeV

| y| < 1.

(ap+p)/(π++π-)EPOS LHCQGSJETII-04

Figure 9: Ratio of anti-proton and proton yield to charged pion yield as a function of charged par-ticle multiplicity for |y| < 1 for non-single diffractive (NSD) p-p scattering at 7 TeV. Simulationsare done with EPOS 1.99 (solid line), QGSJETII-03 (dashed line), QGSJET01 (dash-dotted line)and SIBYLL 2.1 (dotted line) on left panel, and EPOS LHC (solid line) and QGSJETII-04 (dashedline) on right panel. Points are data from CMS experiment [9].

7

10-10

10-9

10-8

10-7

10-6

10-5

0 1000 2000 3000 Energy (GeV)

dNγ /

dE

LHCf p + p √s = 7 TeV 8.81 < η < 8.99 ∆φ=20


10-10

10-9

10-8

10-7

10-6

10-5

0 1000 2000 3000 Energy (GeV)

dNγ /

dE

LHCf p + p √s = 7 TeV 8.81 < η < 8.99 ∆φ=20

EPOS LHCQGSJETII-04

Figure 10: Energy spectrum dN/dE of single photons with 8.81 < η < 8.99 for p-p interactionsat 7 TeV. Simulations with EPOS 1.99 (solid line), QGSJETII-03 (dashed line), QGSJET01 (dash-dotted line) and SIBYLL 2.1 (dotted line) on left panel, and EPOS LHC (solid line) and QGSJETII-04 (dashed line) on right panel, are compared to data points from LHCf experiment [10].

• Very forward photon production (LHCf, Feynman-x)

3.3 Predicted air shower properties

Old and new models (two stacked plots):

• Xmax vs. shower energy

• Muon number vs. shower energy

• Muon energy spectrum

8

Predictions for depth of shower maximum

17

Energy (eV)1710 1810 1910 2010

)2>

(g/c

mm

ax<X

550

600

650

700

750

800

850

900

p

FeEPOS 1.99SIBYLL 2.1QGSJETII-03QGSJET01

HiRes-MIAHiRes (2005)Yakutsk 2001Fly’s EyeYakutsk 1993Auger (2013)

Energy (eV)1710 1810 1910 2010

)2>

(g/c

mm

ax<X

550

600

650

700

750

800

850

900

p

Fe

EPOS LHCQGSJETII-04

HiRes-MIAHiRes (2005)Yakutsk 2001Fly’s EyeYakutsk 1993Auger (2013)

New models favour interpretationas heavier composition than before

pre-LHC models

post-LHC models

Predictions for muon number at ground

18Energy (eV)

1510 1610 1710 1810 1910 2010

)-0

.9

(G

eV0.

9/Eµ

N

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0.05

0.055

p

Fe

EPOS LHCQGSJETII-04

Energy (eV)1510 1610 1710 1810 1910 2010

)-0

.9

(G

eV0.

9/Eµ

N

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0.05

0.055

p

Fe

EPOS 1.99SIBYLL 2.1

QGSJETII-03QGSJET01

pre-LHC models

post-LHC models

New models favour interpretationas lighter composition than before

Further improvement: p-O collisions at LHC

19

0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

10 2 10 3 10 4 10 5 10 6

√s (GeV)

Ine

lasti

city p + p

EPOS LHCQGSJETII-04

0.50.550.6

0.650.7

0.750.8

0.850.9

0.951

10 4 10 6 10 8 10 10 10 11

Ekin (GeV)

ine

lasti

city p + O

EPOS LHCQGSJETII-04

Figure 15: Inelasticity as a function of center-of-mass energy for p-p interactions on left panel andp-O interactions on right panel. Predictions are from EPOS LHC (solid line) and QGSJETII-04(dashed line).

0

1

2

3

4

5

6

7


(1/N

) dN

/ dη

14 TeV

p + p → chrg Inelastic

EPOS LHCQGSJETII-04

0

2

4

6

8

10

12

14

-10 0 10 pseudorapidity η

(1/N

) dN

/ dη

10 TeV

O + p → chrg Inelastic

EPOS LHCQGSJETII-04

Figure 16: Pseudorapidity distribution dN/dη of charged particles for inelastic events for p-pinteractions at 14 TeV on left panel and O-p interactions at 10 TeV on right panel. Predictions arefrom EPOS LHC (solid line) and QGSJETII-04 (dashed line).

10-4

10-3

10-2

10-1

200 400 multiplicity n

P(n

)

p + p → chrg √s = 14 TeV

EPOS LHCQGSJETII-04

10-4

10-3

10-2

10-1


P(n

)

O + p → chrg √s = 10 TeV

EPOS LHCQGSJETII-04

Figure 17: Multiplicity distribution of charged particles for inelastic events for p-p interactions at14 TeV on left panel and O-p interactions at 10 TeV on right panel. Predictions are from EPOS LHC(solid line) and QGSJETII-04 (dashed line).

13

0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

10 2 10 3 10 4 10 5 10 6

√s (GeV)

Ine

lasti

city p + p

EPOS LHCQGSJETII-04

0.50.55

0.60.65

0.70.75

0.80.85

0.90.95

1

10 4 10 6 10 8 10 10 10 11

Ekin (GeV)

ine

lasti

city p + O

EPOS LHCQGSJETII-04

Figure 15: Inelasticity as a function of center-of-mass energy for p-p interactions on left panel andp-O interactions on right panel. Predictions are from EPOS LHC (solid line) and QGSJETII-04(dashed line).

0

1

2

3

4

5

6

7


(1/N

) dN

/ dη

14 TeV

p + p → chrg Inelastic

EPOS LHCQGSJETII-04

0

2

4

6

8

10

12

14

-10 0 10 pseudorapidity η

(1/N

) dN

/ dη

10 TeV

O + p → chrg Inelastic

EPOS LHCQGSJETII-04

Figure 16: Pseudorapidity distribution dN/dη of charged particles for inelastic events for p-pinteractions at 14 TeV on left panel and O-p interactions at 10 TeV on right panel. Predictions arefrom EPOS LHC (solid line) and QGSJETII-04 (dashed line).

10-4

10-3

10-2

10-1

200 400 multiplicity n

P(n

)

p + p → chrg √s = 14 TeV

EPOS LHCQGSJETII-04

10-4

10-3

10-2

10-1


P(n

) O + p → chrg √s = 10 TeV

EPOS LHCQGSJETII-04

Figure 17: Multiplicity distribution of charged particles for inelastic events for p-p interactions at14 TeV on left panel and O-p interactions at 10 TeV on right panel. Predictions are from EPOS LHC(solid line) and QGSJETII-04 (dashed line).

13

Currently predicted uncertainty in most optimistic case

Expected sensitivity of shower observables

20

]2

[g/c

mm

axM

ean

X

740

760

780

800

820

840

860

880

900

920

940

LHC-pOf0.5 0.6 0.7 0.8 1 2 3

]2

[g/c

mm

axRM

S X

30

40

50

60

70

80

90

100

110

]2

[g/c

mm

axM

ean

X

670

680

690

700

710

720

730

740

750

760

cross sectionmultiplicityelasticitycharge ratio

LHC-pOf0.5 0.6 0.7 0.8 1 2 3

]2

[g/c

mm

axRM

S X

12

14

16

18

20

22

24

26

28

Figure 3: Impact on Xmax.

Edec =E0

(ntot)n after n interactions. Since one muon is produced in the decay of each charged

particle, we get

Nµ = nnch =

!

E0

Edec

"α

, (3)

with α = ln nch/ ln ntot = 1 + ln R/ ln ntot ≈ 0.82 . . . 0.95 [3] where R = nch/ntot < 1. Thenumber of muons produced in an air shower depends not only on the primary energy andair density, but also on the total particle multiplicities and in a much more sensitive way [4]of the charged over all particle ratio of hadronic interactions.

It should be kept in mind that the parameters of the model are only effective quantitiesand are not identical to the respective quantities measured at accelerators. In particular, theapproximation of all secondary particles carrying the same energy is only motivated by thefact that it allows us to obtain simple, closed expressions. The well-known leading particleeffect, typically quantified by the (in)elasticity of an interaction, can be implemented in themodel [2] but will not be considered here.

• Elongation rate theorem and scaling violations

• Results of study with re-scaled model features

• Importance of baryon-antibaryon pairs

4

LHC-pOf0.5 0.6 0.7 0.8 1 2 3

QG

SII

µ/N

µRe

lativ

e N

0.7

0.75

0.8

0.85

0.9

0.95

1

1.05

1.1

LHC-pOf0.5 0.6 0.7 0.8 1 2 3

QG

SII

µ/N

µRe

lativ

e N

1.1

1.15

1.2

1.25

1.3

1.35

1.4 cross sectionmultiplicityelasticitycharge ratio

Figure 4: Impact on muon number predictions.

) e(N

10M

ean

log

10.02

10.04

10.06

10.08

10.1

10.12

10.14

10.16

10.18

10.2

10.22

10.24


LHC-pOf0.5 0.6 0.7 0.8 1 2 3

) e(N

10RM

S lo

g

0.04

0.05

0.06

0.07

0.08

) e(N

10M

ean

log

9.88

9.9

9.92

9.94

9.96

9.98

10

10.02

10.04

10.06

LHC-pOf0.5 0.6 0.7 0.8 1 2 3

) e(N

10RM

S lo

g

0.025

0.03

0.035

0.04

0.045

Figure 5: Impact on shower size at ground level.

5

LHC-pOf0.5 0.6 0.7 0.8 1 2 3

QG

SII

µ/N

µRe

lativ

e N

0.7

0.75

0.8

0.85

0.9

0.95

1

1.05

1.1

LHC-pOf0.5 0.6 0.7 0.8 1 2 3

QG

SII

µ/N

µRe

lativ

e N

1.1

1.15

1.2

1.25

1.3

1.35

1.4 cross sectionmultiplicityelasticitycharge ratio

Figure 4: Impact on muon number predictions.

) e(N

10M

ean

log

10.02

10.04

10.06

10.08

10.1

10.12

10.14

10.16

10.18

10.2

10.22

10.24


LHC-pOf0.5 0.6 0.7 0.8 1 2 3

) e(N

10RM

S lo

g

0.04

0.05

0.06

0.07

0.08

) e(N

10M

ean

log

9.88

9.9

9.92

9.94

9.96

9.98

10

10.02

10.04

10.06

LHC-pOf0.5 0.6 0.7 0.8 1 2 3

) e(N

10RM

S lo

g

0.025

0.03

0.035

0.04

0.045

Figure 5: Impact on shower size at ground level.

5

250300350400450500550600650700

102

104

3s (GeV)

inel

astic

cro

ss s

ectio

n (m

b) p+Air

050

100150200250300350400450500

103

105

106

3s (GeV)

mul

tiplic

ity p+Air | d| < 2.5

QGSJETII-03

EPOS 1.99

EPOS LHC

QGSJETII-04

0.10.15

0.20.25

0.30.35

0.40.45

0.50.55

0.6

103

105

106

3s (GeV)

ela

stic

ity p+Air

200

300

400

500

600

700

103

105

106

3s (GeV)

ine

last

ic c

ross

sec

tion

(mb)

/+Air

050

100150200250300350400450500

103

105

106

3s (GeV)

mul

tiplic

ity /+Air | d| < 2.5

0.10.15

0.20.25

0.30.35

0.40.45

0.50.55

0.6

103

105

106

3s (GeV)

ela

stic

ity /+Air

(Energy/eV)10

log12 13 14 15 16 17 18 19 20

Cro

ss s

ectio

n (p

roto

n−ai

r)

[mb]

200

300

400

500

600

700

SIBYLL 2.1 1.2× p−airmSIBYLL 2.1, 0.8× p−airmSIBYLL 2.1,

Energy [eV]1210 1310 1410 1510 1610 1710 1810 1910 2010

[GeV]ppsEquivalent c.m. energy 210 310 410 510

TevatronLHC

accelerator data (p−p) + Glauber

Ad hoc modelmodification

(Ulrich et al., Phys. Rev. D83, 2011)

proton at 1019 eV

p

He

CNO

Fe

Si

How reliable are tuned model predictions ?

21

- Energy-momentum and charge/flavour conservation

- Non-trivial correlations: model dependent

- Tuning by adjusting internal model parameters

Proposal, Helmholtz Young Investigator Group, Dr. Ralf Ulrich 7

η-10 -8 -6 -4 -2 0 2 4 6 8 10

)c (G

eV/

η/d T

dp

-110

1

10

210

310

410

ALIC

E

ALIC

E / L

HCb

CAST

OR,

T2

CAST

OR,

T2 HF

/FCa

l, T1

HF/F

Cal,

T1

ATLAS,CMS

)η/2 exp(-s = maxT

pp-p @ 14 TeV

ZDC,

LHC

f

ZDC,

LHC

f

TOTE

M R

PsAL

FA R

PsFP

420

TOTE

M R

PsAL

FA R

PsFP

420

η-10 -8 -6 -4 -2 0 2 4 6 8 10

)c (G

eV/

η/d T

dp

-110

1

10

210

310

410

~140m-240m-420m

~14-17 m

~11 m

Figure 5: Phase space coverage of the detectors installed at the LHC. The dotted lines indicate therange of the central detectors, where particle type identification and impulse measurements are optimal.

detectors it is possible to reliably distinguish particle types, resolve high densities of particletracks and measure the particle momentum to high precision. However, as demonstrated inFigure 6 (right), by far most of the beam energy is located at much larger pseudorapidities inthe very forward direction along the beam axis. The large LHC experiments do have signif-icant capabilities reaching far into the forward event region. The sub-detectors positioned inforward direction are mostly segmented calorimeters, which offer good energy reconstructionseparated into e.m. and hadronic energy. At CMS (Compact Muon Solenoid) these are HF(Hadronic Forward calorimeter), CASTOR (Centauro And STrange Object Research) and ZDC(Zero Degree Calorimeter). In addition to the calorimeters there are also some forward trackingdetectors: T1 (Tracker 1), T2 (Tracker 2) and TOTEM (TOTal Elastic and diffractive cross sectionMeasurement), where the latter is physically located far from the main part of CMS. In partic-ular the dedicated forward calorimeter CASTOR at 5.2 < η < 6.6, which is an integral part ofthe main CMS detector system, has unique capabilities that are extremely valuable to improveair shower modelling. In the phase space covered by CASTOR there is significant secondaryparticle production combined with a large energy flow.

ηPseudorapidity −10 −5 0 5 10

ηPa

rtic

le fl

ow d

N/d

0

1

2

3

4

5

Tevatron

CMS CASTOR ZDC

QGSJetIIQGSJet01EPOS1.60SIBYLL2.1NEXUS3.97


[G

eV]

ηEn

ergy

flow

dE/

d

0

200

400

600

800

1000

Tevatron

CMS CASTOR ZDC

Figure 6: Secondary particle production (left panel) and energy flow (right panel) in proton-proton inter-actions at 14TeV. The coverage of the CMS detector in pseudorapidity is indicated. The CASTOR andZDC calorimeters of CMS are indicated separately. QGSJetII [31, 32], QGSJet01 [33, 34], EPOS [35],SIBYLL [36] and NEXUS [37,38] are hadronic event generators that are used for cosmic ray data anal-ysis (NEXUS is considered outdated and is not used in data analysis any more).

Proposal, Helmholtz Young Investigator Group, Dr. Ralf Ulrich 7

η-10 -8 -6 -4 -2 0 2 4 6 8 10

)c (G

eV/

η/d T

dp

-110

1

10

210

310

410

ALIC

E

ALIC

E / L

HCb

CAST

OR,

T2

CAST

OR,

T2 HF

/FCa

l, T1

HF/F

Cal,

T1

ATLAS,CMS

)η/2 exp(-s = maxT

pp-p @ 14 TeV

ZDC,

LHC

f

ZDC,

LHC

f

TOTE

M R

PsAL

FA R

PsFP

420

TOTE

M R

PsAL

FA R

PsFP

420

η-10 -8 -6 -4 -2 0 2 4 6 8 10

)c (G

eV/

η/d T

dp

-110

1

10

210

310

410

~140m-240m-420m

~14-17 m

~11 m

Figure 5: Phase space coverage of the detectors installed at the LHC. The dotted lines indicate therange of the central detectors, where particle type identification and impulse measurements are optimal.

detectors it is possible to reliably distinguish particle types, resolve high densities of particletracks and measure the particle momentum to high precision. However, as demonstrated inFigure 6 (right), by far most of the beam energy is located at much larger pseudorapidities inthe very forward direction along the beam axis. The large LHC experiments do have signif-icant capabilities reaching far into the forward event region. The sub-detectors positioned inforward direction are mostly segmented calorimeters, which offer good energy reconstructionseparated into e.m. and hadronic energy. At CMS (Compact Muon Solenoid) these are HF(Hadronic Forward calorimeter), CASTOR (Centauro And STrange Object Research) and ZDC(Zero Degree Calorimeter). In addition to the calorimeters there are also some forward trackingdetectors: T1 (Tracker 1), T2 (Tracker 2) and TOTEM (TOTal Elastic and diffractive cross sectionMeasurement), where the latter is physically located far from the main part of CMS. In partic-ular the dedicated forward calorimeter CASTOR at 5.2 < η < 6.6, which is an integral part ofthe main CMS detector system, has unique capabilities that are extremely valuable to improveair shower modelling. In the phase space covered by CASTOR there is significant secondaryparticle production combined with a large energy flow.


ηPa

rtic

le fl

ow d

N/d

0

1

2

3

4

5

Tevatron

CMS CASTOR ZDC

QGSJetIIQGSJet01EPOS1.60SIBYLL2.1NEXUS3.97


[G

eV]

ηEn

ergy

flow

dE/

d

0

200

400

600

800

1000

Tevatron

CMS CASTOR ZDC

Figure 6: Secondary particle production (left panel) and energy flow (right panel) in proton-proton inter-actions at 14TeV. The coverage of the CMS detector in pseudorapidity is indicated. The CASTOR andZDC calorimeters of CMS are indicated separately. QGSJetII [31, 32], QGSJet01 [33, 34], EPOS [35],SIBYLL [36] and NEXUS [37,38] are hadronic event generators that are used for cosmic ray data anal-ysis (NEXUS is considered outdated and is not used in data analysis any more).

Example: generic LHC detector coverage

22

Relevance of Collider Experiments

central

forward

Central (|⌘| < 1)

Endcap (1 < |⌘| < 3.5)

Forward (3 < |⌘| < 5), HF

CASTOR+T2 (5 < |⌘| < 6.6)

FSC (6.6 < |⌘| < 8)

ZDC (|⌘| > 8), LHCf

How relevant are specificdetectors at LHC for airshowers?

! Simulate parts of showerindividually.

[email protected] UHECR and their interactions 17

Lateral Particle Density on Ground Level

Distance [m]0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000

]2De

nsity

[1/m

-410

-310

-210

-110

Central Endcap Forward CASTOR+T2 FSC ZDC

Muon Density

• Air shower models so far only tuned to about 10% !• Forward detectors are crucial.


Longitudinal Shower Development

]2Depth [g/cm0 200 400 600 800 1000

Num

ber

510

610

710

Electron Profile

• Air shower models so far only tuned to about 10% !• Forward detectors are crucial.


More than 50% of all measured secondaries from particles of η > 8(Ulrich, DPG meeting 2014)

Summary and outlook

23

- Particle physics explanation of knee strongly disfavoured

- All-particle spectrum can be measured with small model uncertainty

- Composition measurements depend directly on interaction models

- Progress made thanks to LHC data (and also fixed-target data, NA61)

Longitudinal shower profile: cosmic ray composition heavier than before

Muon number at ground: cosmic ray composition lighter than before

- LHC run with p-O would be very helpful (low luminosity, forward detectors)

- Some uncertainty will remain due to limited phase space coverage

Documents

LHC Results and the Interpretation of Cosmic-Ray Data · 2014. 5. 27. · Petrukhin, NPB 151 (2006) 57 Barcelo at al. JACP 06 (2009) 027 Dixit et al. EPJC 68 (2010) 573 Petrukhin