Hadron background rejection for Very for Very Hadron background - - PowerPoint PPT Presentation

▶
hadron background rejection for very for very hadron
SMART_READER_LITE
LIVE PREVIEW

Hadron background rejection for Very for Very Hadron background - - PowerPoint PPT Presentation

Dec 01, 2022 284 likes •610 views

Hadron background rejection for Very for Very Hadron background rejection High Energy gamma ray astronomy High Energy gamma ray astronomy with ARGO- -YBJ YBJ with ARGO Ivan DE MITRI Ivan DE MITRI Dipartimento di di Fisica Fisica

slide-1
SLIDE 1

Hadron background rejection Hadron background rejection for Very for Very High Energy gamma ray astronomy High Energy gamma ray astronomy with ARGO with ARGO-

  • YBJ

YBJ

Ivan DE MITRI Ivan DE MITRI

Dipartimento Dipartimento di di Fisica Fisica – – Università Università di di Lecce Lecce and and Istituto Istituto Nazionale Nazionale di di Fisica Fisica Nucleare Nucleare Lecce Lecce, ITALY , ITALY

On behalf of the ARGO On behalf of the ARGO-

  • YBJ Collaboration

YBJ Collaboration

HE γ-ray workshop, Bari, June 2004

slide-2
SLIDE 2

2 2

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Physics Physics goals goals

γ-Ray Astronomy:

Search for point-like galactic and extra-galactic sources at few hundreds GeV energy threshold

Diffuse γ-Rays

from the Galactic plane and SuperNova Remnants

Gamma Ray Burst physics (full GeV / TeV energy range) Cosmic ray physics:

  • anti-p / p ratio at TeV energy
  • spectrum and composition around “knee” (Eth ∼ 10 TeV)

Sun and Heliosphere physics (Eth ∼ 10 GeV)

through the observation of Extensive Air Showers produced in the atmosphere by γ’s and primary nuclei

slide-3
SLIDE 3

3 3

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

The ARGO The ARGO-

  • YBJ

YBJ experiment experiment

  • Collaboration between:

Collaboration between:

  • Istituto

Istituto Nazionale Nazionale di di Fisica Fisica Nucleare Nucleare (INFN) (INFN) – – Italy Italy

  • Chinese Academy of Science (CAS)

Chinese Academy of Science (CAS)

  • Site:

Site: Cosmic Ray Observatory @ Cosmic Ray Observatory @ Yangbajing Yangbajing (Tibet), 4300 m a.s.l. (Tibet), 4300 m a.s.l.

slide-4
SLIDE 4

4 4

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Astrophysical Radiation Ground-based Observatory @ YangBaJing

slide-5
SLIDE 5

5 5

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

High Altitude Cosmic Ray Laboratory @ YangBaJing

(Site Coordinates: longitude 90° 31’ 50” E, latitude 30° 06’ 38” N)

slide-6
SLIDE 6

6 6

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

ARGO ARGO-

  • YBJ layout

YBJ layout

Detector layout

Layer (∼92% active surface) of Resistive Plate Chambers (RPC), covering a large area + sampling guard ring + 0.5 cm lead converter

time resolution ~1 ns space resolution = strip 10 Pads (56 x 62 cm2) for each RPC 8 Strips (6.5 x 62 cm2) for each Pad

1 CLUSTER = 12 RPC 78 m 111 m 99 m 74 m

BIG PAD

ADC

RPC

Read-out

  • f the charge

induced on “Big Pads”

(∼43 m2)

slide-7
SLIDE 7

7 7

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Main detector features and performance Main detector features and performance

▼

  • good pointing accuracy (≤1°)
  • detailed space-time image of the shower front
  • capability of small shower detection (⇒ low E threshold)
  • large aperture (→2π) and high “duty-cycle” (→100%)

⇒ continuous monitoring of the sky (-10°<δ <70°)

Active element: Resistive Plate Chamber ⇒ time resolution ∼1 ns Time information from Pad (56 x 62 cm2) Space information from Strip (6.5 x 62 cm2) Full coverage and large area (∼ 10,000 m2) High altitude (4300 m a.s.l.)

slide-8
SLIDE 8

8 8

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Experiment Hall

slide-9
SLIDE 9

9 9

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

slide-10
SLIDE 10

10 10

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

P a d

  • X

100 105 110 115 120 125 130 135 140 Pad-Y 45 50 55 60 65 70 75 80

Pads Entries 512 Pads Entries 512

pady:padx - 16 Clusters (Data: Feb 2003) Run = 2840 - Ev = 11

100 105 110 115 120 125 130 135 140 45 50 55 60 65 70 75 80 20 40 60 80 100 120 140 160 180 200

Space-Time projection

Space view of an event with 512 hits detected by 16 Clusters (pixel ≡ Pad) Projected space-time view

  • f the same event:

Pad number Time (ns) P a d

  • X

100 105 110 115 120 125 130 135 140

Pad-Y

45 50 55 60 65 70 75 80

Time (ns)

20 40 60 80 100 120 140 160 180 200

3D view of the same event

slide-11
SLIDE 11

11 11

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

P a d

  • X

100 105 110 115 120 125 130 135 140 P a d

  • Y

45 50 55 60 65 70 75 80 Pads Entries 797 Pads Entries 797

pady:padx - 16 Clusters (Data: Feb 2003) Run = 537 - Ev = 970

100 105 110 115 120 125 130 135 140 45 50 55 60 65 70 75 80 20 40 60 80 100 120 140 160 180 200

Space-Time projection

View of an event with ~ 800 hits detected by 16 Clusters

Pad number Time (ns)

slide-12
SLIDE 12

12 12

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

100 105 110 115 120 125 130 135 140 45 50 55 60 65 70 75 80 20 40 60 80 100 120 140 160 180 200 100 105 110 115 120 125 130 135 140 45 50 55 60 65 70 75 80

Pads Entries 444 Mean x 110.2 Mean y 60.09 RMS x 10.41 RMS y 9.59 Pads Entries 444 Mean x 110.2 Mean y 60.09 RMS x 10.41 RMS y 9.59

ARGO-YBJ (16 Clusters) - Run: 545, Event:2558

A shower giving 444 hits

  • n 16 Clusters.

The shower core is well contained into the detector area.

Time (ns) Pad number

slide-13
SLIDE 13

13 13

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Distribution of reconstructed zenith angle Distribution of reconstructed zenith angle

Zenith (degree) 10 20 30 40 50 60 70 80 90 500 1000 1500 2000 2500 3000

Sample of 122,000 events on 16 Clusters (Trigger: N_pad ≥ 20)

slide-14
SLIDE 14

14 14

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Even Even-

  • odd angle difference
  • dd angle difference vs

vs pad pad multiplicity multiplicity

(run on 6 Clusters) (run on 6 Clusters)

Very preliminary Very preliminary σθ(°)

multiplicity

slide-15
SLIDE 15

15 15

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Real Event taken with a fraction of the detector

slide-16
SLIDE 16

16 16

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Real Event taken with a fraction of the detector

slide-17
SLIDE 17

17 17

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Simulated Photon Event

slide-18
SLIDE 18

18 18

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Gamma/hadron Gamma/hadron discrimination discrimination

Photon Shower Proton Shower

The photon signal is statistically identified by looking for an excess, coming from a given direction, over the isotropic background due to charged cosmic rays (H, He, Li, .. nuclei) In addition to this tool the study of the shower

1) space (and time) patterns 2) muon content

can be useful to have higher discrimination power and then a larger sensitivity

slide-19
SLIDE 19

19 19

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

The multiscale approach The multiscale approach

m m l1 = m/2 l1 = m/2 1st step m m ln = m/2n ln = m/2n nth step m lN = 1 lN = 1 Nth step 2N = m m

= ) , , ( l y x φ

tot

N l y x l y x p ) , , ( ) , , ( φ =

content of the cell at position (x,y) as seen at scale length l

Ntot = total map content

slide-20
SLIDE 20

20 20

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

The multifractal analysis (MFA) The multifractal analysis (MFA)

Structures displaying self-similar properties are called fractals. They can be quantitatively described by their fractal dimension. To fully characterize self-similar distributions an infinite number of fractal dimensions is required. Multifractals can be analyzed with the box-counting method.

q y x q

l y x p l Z

∑

=

} , {

) , , ( ) (

The MFA moment of order q at length scale l is defined by:

) ( 1

~ ) (

q l q

l l Z

τ

  → →

When scaling is observed

log Zq(l) log l

The dependence of the MFA scaling exponent τ(q) on the order q, gives the main information on the image.

slide-21
SLIDE 21

21 21

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

) ( 1 } {

~ ) , ( ) , ( ) (

q l q x q

l l l x p l x p l W

β

 →  + − =

→

∑

The DWA moment of order q at length scale l is directly related to the coefficients of the DW transform of φ(x). It is defined by:

+ -

The discrete wavelet analysis (DWA) The discrete wavelet analysis (DWA)

In the 2-D case, three Haar mother wavelets can be used:

+ - + -

  • +

+ -

  • -

+ +

) ( 1 ) 1 (

) 1 (

~ ) (

q l q

l l W

β

 →  →

) ( 1 ) 2 (

) 2 (

~ ) (

q l q

l l W

β

 →  →

) ( 1 ) 3 (

) 3 (

~ ) (

q l q

l l W

β

 →  →

For isotropic cases

) ( ) ( ) ( ) (

) 3 ( ) 2 ( ) 1 (

q q q q β β β β ≡ = =

The dependence of the DWA scaling exponent β(q) on the order q, gives the main information on the image.

slide-22
SLIDE 22

22 22

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

The smallest pixel is taken at (2 ×2) pad ~ 1m2

slide-23
SLIDE 23

23 23

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Simulated data sample Simulated data sample

Gamma and proton induced showers have been simulated with CORSIKA + ARGOG with the following characteristics:

power spectrum between 10GeV and 300TeV with a spectral index γ = -2.5 and -2.7

for photons and protons respectively

azimuth between 0 and 15 degrees core at the detector center

Since the photons and hadrons of the same energy produce different pad multiplicities, the data sample has been divided into five multiplicity windows 3367 3367 18.0 18.0 5145 5145 11.3 11.3 1500 1500 -

  • 6000

6000 2770 2770 7.7 7.7 3397 3397 4.6 4.6 800 800 -

  • 1500

1500 1951 1951 4.9 4.9 2885 2885 2.9 2.9 500 500 -

  • 800

800 7601 7601 1.7 1.7 11902 11902 1.1 1.1 100 100 -

  • 500

500 4160 4160 0.8 0.8 6955 6955 0.5 0.5 50 50 – – 100 100 N Np

p

< <E Ep

p> (

> (TeV TeV) ) N Nγ

γ

< <E Eγ

γ>

> ( (TeV TeV) ) N Npad

pad

slide-24
SLIDE 24

24 24

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Single Event Analysis Single Event Analysis

Compute the MFA and DWA moments as a function of the scale length for different values of the order q. Fit these curves and get the scaling exponents τ and β

Zq vs the scale length l Wq vs the scale length l Example for a 7.9TeV photon initiated shower……

slide-25
SLIDE 25

25 25

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Study of the scaling exponents Study of the scaling exponents

q = 4, 6, 8

The linear dependence of τ and β on q allows the analysis just for three sample values.

slide-26
SLIDE 26

26 26

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Scaling exponents distributions Scaling exponents distributions

As expected from the previous graphs, the values of τ and β for different q’s give similar but different distributions. The scaling exponents are very good candidates to be the input values for an Artificial Neural Network able to discriminate between photon and hadron induced showers.

slide-27
SLIDE 27

27 27

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Artificial Neural Network Artificial Neural Network

Eight parameters have been identified and used as input for an (8,4,1) ANN. Nhit τ(q=4) τ(q=6) τ(q=8) β(q=4) β(q=6) β(q=8) <x3 >/ <y3>

slide-28
SLIDE 28

28 28

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

ANN results ANN results

Results have been obtained by using, in each multiplicity window, 400 events (200 γ + 200 p ). Fluctuations here are essentially due to this limited statistics.

slide-29
SLIDE 29

29 29

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Artificial Neural Network training Artificial Neural Network training

Different ANN’s (with the same topology) have been trained in the different multiplicity windows. The number of training epochs has been optimized in order to maximize the efficiencies and minimize the processing times.

εγ 1−εp

slide-30
SLIDE 30

30 30

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

ANN results ANN results

Reduced time interval needed to identify sources Larger equivalent effective area Sensitivity to smaller fluxes

days Q TCrab 120 ) 1 (

5

= =

σ

days Q TCrab 30 ) 2 (

5

= =

σ

h

Q ε ε γ − ≡ 1

h

N N S

γ

≡

h

ε ε γ − × 1

h

Q ε ε γ − ≡ 1

slide-31
SLIDE 31

31 31

  • I. De Mitri
  • I. De Mitri

Hadron background Hadron background rejection rejection for for VHE gamma VHE gamma ray ray astronomy astronomy with with ARGO ARGO-

  • YBJ

YBJ

Conclusions Conclusions

Multiscale image analysis has been showed to provide an Multiscale image analysis has been showed to provide an efficient tool for gamma/hadron discrimination efficient tool for gamma/hadron discrimination Results are encouraging and allow to nearly double the Results are encouraging and allow to nearly double the detector sensitivity. detector sensitivity. The best response is obtained in the The best response is obtained in the few few TeV TeV range. range. The study is now being extended to all event categories The study is now being extended to all event categories The measurement of the The measurement of the muon muon content of the shower content of the shower allows hadron background rejection at higher energies allows hadron background rejection at higher energies

… see next talk