Measurement of the proton-air inelastic cross section with ARGO-YBJ

1
Measurement of the proton-air inelastic cross
section with ARGO-YBJ
21 European Cosmic Ray Symposium 9-12 September
2008 Kosice, Slovakia

• A. Surdo
• Istituto Nazionale di Fisica Nucleare
• Lecce Italy
• on behalf of ARGO-YBJ Collaboration

2
Introduction

• Inelastic proton-air (and total p-p) cross
  section can be derived from
• cosmic ray measurements through several methods
• Direct method
• measure the distribution of X1, the 1st
  interaction point of p-air
• collisions, to directly measure the mean free
  path lp-air
• ? feasible (in principle) at relatively low
  energies only
• Indirect methods
• (a) for fixed primary energy and zenith angle,
  measure the exponential
• tail of the shower maximum depth (Xmax)
  distribution

(b) for fixed primary energies, measure the
exponential zenith angle distribution of
the shower intensity ? slope parameter (L)
connected to lp-air
ARGO-YBJ approach
3
The ARGO-YBJ experiment

• Collaboration between
• Istituto Nazionale di Fisica Nucleare (INFN)
  Italy
• Chinese Academy of Science (CAS)
• Site Cosmic Ray Observatory _at_ Yangbajing
  (Tibet), China

High Altitude Cosmic Ray Laboratory _at_
YangBaJing Site Altitude 4,300 m a.s.l. , 600
g/cm2 Site Coordinates longitude 90 31 50
E, latitude 30 06 38 N
4
ARGO-YBJ physics objects
5
ARGO-YBJ detector
8 Strips 1 Pad (56 62 cm2)
Central carpet 130 Clusters 1,560 RPCs 124,800
Strips (?5,600m2 active area)
99 m
74 m
10 Pads 1 RPC (2.80 1.25 m2)
78 m
111 m
RPC
Strip spatial pixel (6.5 x 62 cm2)
Pad time pixel Time resolution 1 ns
Analog RPC charge read-out 0.5 cm lead
converter (2009)
6
ARGO-YBJ detector
Currently completed and in data taking
Inclusive trigger Npad gt 20 (shower mode
trigger) on the central carpet. Trigger rate ?4
kHz and data flow ?7 MB/s.
7
Measurement of the Flux attenuation
Use the shower frequency vs (secq -1)
for fixed energy and shower age (h0vert. depth).
The absorption length L is connected to lint
through
?
sp-Air (mb) 2.41104 / lint(g/cm2)
L k lint

• The parameter k takes into accounts how the
  primary energy is dissipated in the shower
• k determined by simulations, depends on
• interaction model
• shower fluctuations
• actual set of experimental observables
• ..

• Warning
• Constrain XDO Xdet X0 or
• better XDM Xdet Xmax
• Select deep showers (large
• Xmax, i.e. small XD0 or XDM)

8
Analysis general criteria

• Exploit peculiar detector features
• space/time granularity,
• full-coverage technique,
• high altitude (h0 ?600 g/cm2)

detailed space-time pattern for unique EAS
reconstruction

• Select deep showers (large XMax, i.e. small XD0
  or XDM) in order to
• minimize the impact of shower development
  fluctuations

9
Event selection

• Basic cuts
• Shower fully reconstructed (0ltqzenlt40) through
  conical fit
• Shower detected size (Nstrip) gt 400
• Core reconstructed inside a fiducial area (64 x
  64 m2)
• more specific cuts, based on extension of the
  detected shower,
• hit density near the reconstructed core and
  lateral profile,
• in order to
• reduce the contamination of external core
  events
• put a constraint on the maximum XDM value
• Finally

10
Monte Carlo simulation

• CORSIKA showers, by p and He primaries
• Energy ranges p (0.3-3000)TeV
• He (1-3000)TeV
• Zenith angle range 0ltqlt45
• QGSJET interaction model
• Use of information on the longitudinal shower
  profile (Xmax,)
• Full detector response simulation based on GEANT
  package
• Proper choice of the sampling area including the
  detector
• Same analysis chain as for real data

11
Cuts in-dependence on the zenith angle
Energy
XDM XDet XMax
No significant zenith angle dependence below 30
degrees. A slight shift might be seen above 40
degrees. In this analysis we stop at 40 degrees
12
The energy scale

• Use of the strip multiplicity (Nstrip)
• for the estimation of shower size
• ? up to ?100 TeV primary energy

• gt For DNstrip ? fold the MC energy distribution
• with parametrized sp-air ? distribution of
  lint.
• gt Get the energy corresponding to ltlintgt (?ELog).

Log(E/TeV)
lint (g/cm2)
l(MC)int
Average lint also used to evaluate k factor k
L(MC)obs/l(MC)int
13
Analysis of MC data secq distributions (?L(MC)ob
sh0/Slope) and k factors k L(MC)obs/l(MC)int
Nstrip Elog (TeV) l(MC)int (g/cm2) k
400 1000 4.0 78.53 2.1 2.01 0.06 0.05
1000 2000 8.3 76.15 1.8 1.53 0.02 0.04
3000 4000 19.8 73.45 1.5 1.59 0.04 0.03
6000 8000 38.7 71.44 1.3 1.68 0.06 0.03
8000 12000 53.5 70.51 1.2 1.71 0.07 0.03
gt 8000 76.7 69.50 1.6 2.05 0.06 0.05
14
Analysis of real data secq distributions
from MC
sCR-air (mb) 2.41104/l(exp)int (g/cm2)
15
Heavy primaries contribution
proton

• Hoerandel AP 19 (2003) 193
• taken as reference.
• JACEE and RUNJOB for the
• evaluation of systematic error

helium
16
Heavy primaries contribution
Correction on reconstructed s at 80 TeV
Above 1 TeV primary Helium fraction 40
After analysis cuts 15-20 Heavier
primaries can be neglected
Nstrip Elog (TeV) Helium correction spAir (mb)
400 1000 4.0 1.00 0.04 0.01 261 13 8
1000 2000 8.3 1.00 0.02 0.01 278 7 7
3000 4000 19.8 1.00 0.04 0.01 303 15 7
6000 8000 38.7 0.96 0.05 0.03 288 19 11
8000 12000 53.5 1.00 0.05 0.03 289 19 10
gt 8000 76.7 0.95 0.04 0.04 322 17 16
17
Inelastic p-air cross section
(statistical errors only in the plot)
18
From p-air to p-p cross section

• Several available models to obtain stotp-p from
  sinelp-air
• Glauber Matthiae theory
• Durand Pi
• Wibig Sobczynska
• .
• Models agree within few in our energy range
• ? systematic error 5

Nstrip Elog (TeV) spAir (mb) spp (mb)
400 1000 4.0 261 13 8 38 3 3
1000 2000 8.3 278 7 7 42 2 3
3000 4000 19.8 303 15 7 49 4 3
6000 8000 38.7 288 19 11 44 5 4
8000 12000 53.5 289 19 10 45 5 4
gt 8000 76.7 322 17 16 55 6 5
19
Systematic uncertainties

• Variations of the atmospheric depth (pressure)
• h0MC 606.7 g/cm2 (4300m a.s.l. standard
  atm.),
• h0MC / lth0gt 0.988 0.007 ? impact on
  cross section analysis ?1
• Uncertainty on l(MC)int RMS of MC distribution
  ? 23
• Uncertainty on the contribution of heavy neclei
  comparing slopes
• from different (pHe) fluxes (Hoerandel,
  Jacee, RunJob) ? 14
• Uncertainty on sp-air to sp-p comparing
  different models ? 5

Statistical and systematic errors independently
propagated
20
Total p-p cross section
21
Total p-pbar cross section
22
Summary and outlook

• The flux attenuation technique has been
  successfully used in
• ARGO-YBJ experiment, by exploiting the
  detector features
• and location.
• The inelastic proton-air (and the total p-p)
  cross section has been
• measured in a scarcely explored energy
  region and results are in
• agreement with previous ARGO results.
• More checks on systematics are in progress
  (shower fluctuations, interaction models, heavy
  primaries contribution, ).
• Shower age and energy determinations will be
  improved by the use
• of timing (rise time, front curvature,..) and
  topological information
• In the future, the analysis will be extended to
  higher energies
• (up to 1PeV), thus covering a region with few
  experimental data,
• by exploiting the analog RPC readout.