The KASCADE-Grande experiment
description
Transcript of The KASCADE-Grande experiment
![Page 1: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/1.jpg)
1
The KASCADE-Grande experiment
Andrea CHIAVASSAUniversita` degli Studi di Torino
Cosmic Ray Physics Large Scale Experiments at the Second Decade of the 21st Century
Moscow, 16-18 May 2011
![Page 2: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/2.jpg)
2
Motivations for the KASCADE-Grande experimentThe range 1016 – 1018 eV is crucial for different reasons:- complete “knee” studies- investigate galactic-to-extragalactic transition- hadronic interactions- anisotropies
area required to detect~100 events/year
1 km20.01 km2
![Page 3: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/3.jpg)
3
Measurements of air showers in the energy range E0 = 100 TeV - 1 EeV
KASCADE-Grande= KArlsruhe Shower Core and Array DEtector +
Grandeand LOPES
![Page 4: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/4.jpg)
4
![Page 5: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/5.jpg)
DetectorDetector Detected Detected EAS EAS componecomponentnt
Detection Detection TechniqueTechnique
DetectDetector area or area (m(m22))
GrandeGrande Charged Charged particlesparticles
Plastic Plastic ScintillatorsScintillators
37x1037x10
KASCADE KASCADE array e/array e/
Electrons, Electrons,
Liquid Liquid ScintillatorsScintillators
490490
KASCADE KASCADE array array
MuonsMuons (E(Ethth=230 =230 MeV)MeV)
Plastic Plastic ScintillatorsScintillators
622622
MTDMTD Muons Muons (Tracking) (Tracking) (E(Ethth=800 =800 MeV)MeV)
Streamer TubesStreamer Tubes 4x1284x128
• Shower core and arrival
direction
– Grande array
• Shower Size (Nch number of
charged particles)
– Grande array
• Fit NKG like ldf
KASCADE-Grande detectors & observables
• Size (E>230 MeV)•KASCADE array detectors
•Fit Lagutin Function• density & direction (E>800 MeV)
•Streamer Tubes
![Page 6: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/6.jpg)
6
A standard event
KASCADE-Grande efficiency
Apel et al. NIMA 620 (2010) 202-216
SIM
DATA
1016eV
![Page 7: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/7.jpg)
7
Cross-check between KASCADE and Grande
Apel et al. NIMA 620 (2010) 202-216
■ KASCADE stations■ Grande stations
DATA
![Page 8: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/8.jpg)
8
Good agreement between the reconstruction accuracies of the 2 detectors
Apel et al. NIMA 620 (2010) 202-216
Arrival direction accuracy <1
Rayleigh fit
Nch accuracy: systematics < 10%Core position accuracy: < 8m(Nch
G-NchK)/Nch
K
■Mean value□RMS
DATA
![Page 9: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/9.jpg)
9
Muon reconstruction (from simulation QGSjet II & FLUKA)
Apel et al. NIMA 620 (2010) 202-216
(Nrec-N
tr)/Ntr < 20%
After applying muon correction functions
(Nrec-N
tr)/Ntr
SIMULATION
![Page 10: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/10.jpg)
10
• 1173 days of effective DAQ time.• Performance of reconstruction and detector is stable.• θ < 40° • 250 m < rKAS < 600 m
First Results:
•All particle Energy spectrum•Towards Mass Groups spectra
![Page 11: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/11.jpg)
11
Reconstruction of the energy spectrumWe use three different methods:•Nch as observable•N as observable•Combination of Nch and N as observables
•Cross check of reconstruction procedures•Cross check of systematic uncertainties•Test sensitivity to composition•Cross check of validity of hadronic interaction models
If not explicitly mentioned in the followingCORSIKA QGSjetII/FLUKA interaction model is used
![Page 12: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/12.jpg)
1) Nch,Spectra measured at different angles2) Nch,() Nch,(ref)
CIC to correct for atmosphericattenuation model independent
E = a + blog10(Nch,
Energy spectra model & composition dependent
![Page 13: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/13.jpg)
13
Nch – N technique5 angular bins treated independently
log10(Nch)
log10(Nch)
log
10(N
ch/N
)
log
10(E
/GeV
)
SIM
SIM
*error bars=RMS of distributions
log10(Nch/N) - log10(Nch/N)p
k = log10(Nch/N)Fe - log10(Nch/N)p
log10E=[ap+(aFe-a,p)]klog10(Nch)+[bp+bFe-bp)k]
log10Ep,Fe=ap,Felog10Nch+bp,Fe
log10(Nch/N )p,Fe=cp,Felog10Nch+dp,Fe
![Page 14: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/14.jpg)
14
Pro & cons of the methods
Nch or Nalone:
• Constant intensity cut method• Correction for atmospheric attenuation is model independent• Calibration function QGSjet II: shower size (Nch or Nμ ) vs E• Composition dependent
Nch & Ncombined:
• Composition independent• Correction for atmospheric attenuation is model dependent• Calibration function QGSjet II: Nch-N vs E
![Page 15: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/15.jpg)
Check of resolutions and systematic errorsusing MC simulations
Reconstructed/True flux
log10(E/GeV)
flu
x rec
/flu
x tru
e
SIM/QGSjet II
Effect of Hadronic interaction model: EPOS data treated as exp. data and analyzed using QGSjet II
![Page 16: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/16.jpg)
16
Table of systematics on the flux
Source of uncertainty 1016eV
(%)
1017eV
(%)
1018eV
(%)
Intensity in different angular bins (attenuation) 10.2 9.3 13.0
Calibration & composition 10.8 7.8 4.4
Slope of the primary spectrum 4.0 2.0 2.1
Reconstruction (shower sizes) 0.1 1.3 6.6
TOTAL 15.4 12.4 14.7
Other uncertainties % % %
Sudden knee structures (extreme cases) <10
Hadronic interaction model (EPOS-QGSjet) -5.4 -12.3 -9.5
Statistical error 0.6 2.7 17.0
Energy resolution (mixed primaries) 24.7 18.6 13.6
![Page 17: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/17.jpg)
17
Experimental data
log10(E/GeV)
dN
/dE
x E
2.5
(m-2s-1
sr-1eV
1.5 )
DATADATA
The measured fluxesare different EASdevelopment in atmospehre not correctly described by the simulation?
0°<<16.7°16.7°<<24.0°24.0°<<29.9°29.9°<<35.1°35.1°<<40.0°
![Page 18: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/18.jpg)
18
Comparing the 3 methods (dI/dE x E3)
DATA
Spectrum measured with the Nch-N technique lies in the allowed region by the CIC analysis performed both with Nch and with N
![Page 19: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/19.jpg)
19
Residual plot
DATA
![Page 20: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/20.jpg)
20
2single power law / ndf =2.97
22 power laws / ndf =0.49
Ftest = 6.09Variance = 0.62Significance = 7.7
1=-3.0150.010
2=-3.2440.0772 power laws
Fermi function
Ftest = (2single power law / m) / (2
function / n), with m,n = ndf single power-law, function
Variance = 2n2(m+n-2) / m(n-2)2(n-4)
Significance in units of the standard deviation = Ftest / √Variance
2single power law / ndf =2.97
2fermi / ndf =0.65
Ftest = 4.56Variance = 0.51Significance = 6.4
![Page 21: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/21.jpg)
21
○ KASCADE QGSjet01▼ EAS-TOP ● KASCADE-Grande QGSjet2
Comparison with KASCADE & EAS-TOP
![Page 22: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/22.jpg)
22
The all-particle energy spectrum
DATA
![Page 23: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/23.jpg)
•The Energy spectrum shows structures. •The composition analysis is crucial to try to understand their origin.
•The composition studies are approached with different techniques (all based on Nch-N
observables and QGSjet model) like we did for the Energy spectrum to have a cross check of the results and in order to study the systematic errors of each technique:
– N/Nch distributions in bins of Nch (2)– unfolding– k parameter– YCIC= logN/logNch
![Page 24: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/24.jpg)
• The goal of the YCIC and “k parameter” algorithm is to separate the events into samples originated by primaries belonging to different “mass groups”– YCIC = log N(ref) / log Nch (ref)
– k = (log(Nch/N) - log(Nch/N)p) / (log(Nch/N)Fe- log(Nch/N)p)
• 2 and unfolding algorithm have the goal of measuring the spectra of single mass groups through statistical analysis of the two dimensional (Nch, N) spectra
24
![Page 25: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/25.jpg)
YCIC k paramater
25
Performances of the YCIC and “k parameter” algorithm. Calculated in the frame of CORSIKA full EAS simulation with QGSJet II interaction model.
Log(E/GeV)
k
Log(E/GeV)
![Page 26: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/26.jpg)
vs. Primary Energy
Vs. Distance from KASCADE
26
NY<0.84/NH & NY>0.84 /NFe
YCIC
![Page 27: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/27.jpg)
2 methodN/Nch distributions in bins of Nch
intermediate
![Page 28: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/28.jpg)
N/Nch distributions in bins of Nch
• Fsim(i) = j j fsim,j(i) j j = 1
• 2 = i (Fsim(i)-Fexp(i))2/2(i)
• It has already been shown that:– KG can separate three mass groups:
• Light (H), Intermediate (He+C), Heavy (Si+Fe)
– QGSJet II gives a good description of the measured N/Nch distributions in the whole Nch range
– Light and Heavy mass groups are needed to describe the distributions measured in all Nch intervals 28
![Page 29: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/29.jpg)
• Analyzing the N/Nch distributions in the kth Nch interval we obtain the abundances j(k) of different mass groups.
• Nj(k) = j(k) Nexp(k)
• From Nj(k) we can calculate the flux in the kth Nch interval.
• By a full simulation we convert Nch into primary energy. We calculate the differential energy fluxes of single mass group
• Results heavily depend on the QGSjet II interaction model 29
From j to mass groups energy spectra
![Page 30: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/30.jpg)
30
Test spectra of the three mass groups reconstructed with the 2 analysis. The results are not yet corrected for migration effects.No spectral breaks are introduced by the algorithm.
Intermediate true
Intermediate rec
![Page 31: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/31.jpg)
Number of events Ni in the cells of the two dimensional spectra are given by:
Where pn is the conditional probability of measuring an event in the Log N, log Nch cell if the shower was induced by a particle of type nand primary energy E
Unfolding
![Page 32: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/32.jpg)
Conclusions• KASCADE-Grande operates in the 1016-1018 eV energy
range
• <1° arrival direction, <10 m core position, ~15% Nch , >20% N resolutions
• All particle Energy Spectrum– Agreement with KASCADE & EAS-TOP results at the threshold– Agreement between different reconstruction approaches– No single power law– Structures at the threshold and ~1017 eV
•Toward mass groups energy spectra• Resolve three mass groups• Mass group spectra will be soon presented
32
![Page 33: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/33.jpg)
33
![Page 34: The KASCADE-Grande experiment](https://reader036.fdocuments.net/reader036/viewer/2022081516/56814914550346895db64c19/html5/thumbnails/34.jpg)
• A feature like the one claimed by the GAMMA collaboration would have been detected. Even if broaden by KASCADE-Grande resolution and EAS development in atmosphere
34