[PPT] - S e a r c h f o r U l t r a H i g h E n e r g PowerPoint Presentation

SLIDE 1

S e a r c h f

r

U l t r a H i g h E n e r g y p h

t
n

s a t t h e P i e r r e A u g e r O b s e r v a t

r

y & c

n

t r i b u t i

n

t

A

u g e r P r i me

In the Auger group. Thesis Supervisor : Corinne Berat

SLIDE 2

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

O u t l i n e

U

l t r a H i g h E n e r g y C

s

m i c R a y s & E x t e n s i v e A i r S h

w

e r s

T

h e P i e r r e A u g e r O b s e r v a t

r

y & A u g e r P r i m e

S

c i n t i l l a t

r

S u r f a c e D e t e c t

r

s – C

n

s t r u c t i

n

& V a l i d a t i

n
S

e a r c h f

r

U l t r a

H

i g h E n e r g y p h

t
n

p r i m a r i e s – M

t

i v a t i

n

s & M u l t i

V

a r i a t e A n a l y s i s

2

SLIDE 3

U l t r a

H

i g h E n e r g y C

s

m i c R a y s U l t r a

H

i g h E n e r g y C

s

m i c R a y s & & E x t e n s i v e A i r S h

w

e r s E x t e n s i v e A i r S h

w

e r s

SLIDE 4

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

U l t r a H i g h E n e r g y C

s

m i c R a y s

S

p e c t r u m

UHECRs & EAS U l t r a H i g h E n e r g y C

s

m i c R a y s ( U H E C R s ) :

u

l t r a h i g h e n e r g i e s : E > 1

1 8

e V

v

e r y l i m i t e d fm u x : < 1 . k m

2

. y e a r

1
f

e a t u r e s i n t h e s p e c t r u m

n

u c l e u s f r

m

H t

F

e & n e u t r a l s ( n / γ / ν ) 4

SLIDE 5

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

A c c e l e r a t i

n

& P r

p

a g a t i

n

UHECRs & EAS

CRs ɣ ν

5

SLIDE 6

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

M u l t i

M

e s s e n g e r A s t r

p

h y s i c s

UHECRs & EAS Multi-Messenger era of astrophysics :

combine cosmic rays, gamma, neutrino and gravitational wave
bservations
interplay between all these messengers

Complementarity between the

bservations
Gamma-rays :

+: straight line

: UHE horizon < 10 Mpc
Neutrinos :

+: straight line, no interaction

: isotropic diffuse background
Cosmic-Rays :

+: direct accelerator probe

: deflected in magn. field

CRs ɣ ν

The 3 fluxes are linked! The 3 fluxes are linked! 6

SLIDE 7

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

E x t e n s i v e A i r S h

w

e r

UHECRs & EAS UHECRs interact with Earth’s atmosphere : generates an Extensive Air Shower (EAS) 3 main components :

muonic component (~4%)
hadronic component (~1%)
electromagnetic component (95%)

At ground : ~5.1010 particles (estimation for a 1019eV p-shower) 7

SLIDE 8

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

E x t e n s i v e A i r S h

w

e r

UHECRs & EAS UHECRs interact with Earth’s atmosphere : generates an Extensive Air Shower (EAS) 3 main components :

muonic component
hadronic component
electromagnetic component

Advantages of the EAS :

large footprint (up to 15 km)
multiple observations possibles
UHE hadronic physics laboratory...

Disadvantages of the EAS :

indirect information on primary’s

energy

inhomogeneous calorimeter
...model dependent

8

SLIDE 9

T h e P i e r r e A u g e r O b s e r v a t

r

y T h e P i e r r e A u g e r O b s e r v a t

r

y & & A u g e r P r i m e A u g e r P r i m e

SLIDE 10

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

T h e P i e r r e A u g e r O b s e r v a t

r

y

PAO & AugerPrime The Pierre Auger Observatory (PAO) :

Officially completed in 2008
started taking data in 2004
400 scientists from 18 countries
Location : pampa near Malargüe, Argentina
Altitude (mean) : 1400 m above sea-level
Surface (SD) : 3000 km2

Fluorescence Detector (FD) : Surface Detector (SD) :

1660 Water

Cherenkov Detector

Triangular spacing
f 1.5 km
Duty-Cycle : 100%
24 telescopes

located in 4 buildings.

Overlooking the

atmosphere above the array

Duty-Cycle : 14%

10

SLIDE 11

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

T h e P i e r r e A u g e r O b s e r v a t

r

y

PAO & AugerPrime The Pierre Auger Observatory (PAO) :

Officially completed in 2008
started taking data in 2004
400 scientists from 18 countries
Location : pampa near Malargüe, Argentina
Altitude (mean) : 1400 m above sea-level
Surface (SD) : 3000 km2

Fluorescence Detector (FD) : Surface Detector (SD) :

1660 Water

Cherenkov Detector

Triangular spacing
f 1.5 km
Duty-Cycle : 100%
24 telescopes

located in 4 buildings.

Overlooking the

atmosphere above the array

Duty-Cycle : 14%

Water Cherenkov Detector (WCD) : Secondary particles (highly relativistic) going through the detector produce Cherenkov light. Sensitive to e±, γ, μ Collect timing and signal to reconstruct the showers. 11

SLIDE 12

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

T h e P i e r r e A u g e r O b s e r v a t

r

y

PAO & AugerPrime The Pierre Auger Observatory (PAO) :

Officially completed in 2008
started taking data in 2004
400 scientists from 18 countries
Location : pampa near Malargüe, Argentina
Altitude (mean) : 1400 m above sea-level
Surface (SD) : 3000 km2

Fluorescence Detector (FD) : Surface Detector (SD) :

1660 Water

Cherenkov Detector

Triangular spacing
f 1.5 km
Duty-Cycle : 100%
24 telescopes

located in 4 buildings.

Overlooking the

atmosphere above the array

Duty-Cycle : 14%

Fluorescence Telescope : An EAS excite the nitrogen molecules in the atmosphere → Fluorescence Emission Telescopes collect this UV light Direct measurement

f the shower

energy 12

SLIDE 13

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

T h e P i e r r e A u g e r O b s e r v a t

r

y

PAO & AugerPrime The Pierre Auger Observatory (PAO) :

Officially completed in 2008
started taking data in 2004
400 scientists from 18 countries
Location : pampa near Malargüe, Argentina
Altitude (mean) : 1400 m above sea-level
Surface (SD) : 3000 km2

Fluorescence Detector (FD) : Surface Detector (SD) :

1660 Water

Cherenkov Detector

Triangular spacing
f 1.5 km
Duty-Cycle : 100%
24 telescopes

located in 4 buildings.

Overlooking the

atmosphere above the array

Duty-Cycle : 14%

13

SLIDE 14

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

H y b r i d D e t e c t i

n

& O p e n q u e s t i

n

s

PAO & AugerPrime Hybrid Detection : Able to use the FD to calibrate the SD’s energy reconstruction Other complementary detection possible…

Flux suppression around 1020eV : What is causing it ? What particles are making up the UHECRs flux ?

14

SLIDE 15

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

A u g e r P r i m e U p g r a d e

PAO & AugerPrime

Science Motivation :

Probe the flux suppression
Look at the flux composition at the highest

energies

UHE hadronic physics

AugerPrime :

Add Scintillator Surface Detectors : on top of

the WCD, to disentangle muon/EM components

Upgraded electronics
Radio Upgrade : add antennas on top of WCD

to detect the showers radio-emissions 15

SLIDE 16

S c i n t i l l a t

r

S u r f a c e D e t e c t

r

s S c i n t i l l a t

r

S u r f a c e D e t e c t

r

s

C
n

s t r u c t i

n

& V a l i d a t i

n

C

n

s t r u c t i

n

& V a l i d a t i

n

SLIDE 17

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

S c i n t i l l a t

r

S u r f a c e D e t e c t

r

( S S D )

SSDs & Validation Objectives :

add an independent and different measurement of the EAS’ components, at the same

place as the WCD

reliable detector, with low maintenance

Detector’s signal :

the SSD’s signal will be dominated by the shower’s electrons, while the WCD’s signal

is dominated by the photons and muons.

particle → scintillator → light → fibers → PMT → signal
calibrated with atmospheric muons

17

SLIDE 18

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

SSDs & Validation

S S D – C

n

s t r u c t i

n

& T e s t s S e t u p

1200 SSDs built in 6 countries : 90 SSDs built at LPSC The scintillator boards are used as external triggers :

a particle (muon) goes through both

scintillators (up and down)

read-out electronics : “trigger”
record the signal from all PMTs inside

a user-defined time window (800 ns before trigger and 300 ns after trigger)

Important involvement by the SDI, electronics and administrative departments 18

SLIDE 19

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

S S D T e s t s – D a t a A n a l y s i s

SSDs & Validation

For each triggered event (50k per run) :

scan the SSD trace
look for a MIP and/or SPE peak
draw the distribution and fit
use timing information on the external trigger to select vertical events → VMIP

Minimum Ionising Particle (MIP) :

Minimum energy deposited by a through-going relativistic particle (i.e muon)

Single Photo-Electron (SPE) peak :

signal picked-up by the PMT for a single photoelectron inside the detector

19

SLIDE 20

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

S S D T e s t s – D a t a A n a l y s i s

SSDs & Validation

For each triggered event (50k per run) :

scan the SSD trace
look for a MIP and/or SPE peak
draw the distribution and fit
use timing information on the external trigger to select vertical events → VMIP

Minimum Ionising Particle (MIP) :

Minimum energy deposited by a through-going relativistic particle (i.e muon)

Single Photo-Electron (SPE) peak :

signal picked-up by the PMT for a single photoelectron inside the detector

Test results : stable so far

20

SLIDE 21

S e a r c h f

r

p h

t
n

p r i m a r i e s S e a r c h f

r

p h

t
n

p r i m a r i e s

P

r i m a r y D e p e n d e n t V a r i a b l e s P r i m a r y D e p e n d e n t V a r i a b l e s & & M u l t i

V

a r i a t e A n a l y s i s M u l t i

V

a r i a t e A n a l y s i s

SLIDE 22

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

Wh y l

k

f

r

p h

t
n

s h

w

e r s ?

Photon Search Why are photon primaries interesting? :

point back to source directly
UHE photons follow-up searches
an excess of photons could independently

confirm the GZK effect

constrain astrophysical scenarios

Greisen Zatsepin Kuzmin effect : Interaction between UHECRs and Cosmic Microwave Background photons. Above an energy threshold : Emin ~ 1019eV 22

SLIDE 23

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

P h

t
n
i

n d u c e d s h

w

e r s

Photon Search

Develop deeper into the

atmosphere

On average, more fluctuations of

the shower maximum

Less muons on ground

23

SLIDE 24

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

P h

t
n

A n a l y s i s i n A u g e r

Photon Search Two official analyses in Auger collaboration :

Hybrid (SD+FD) analysis : more observables
SD only analysis : more data

Aim of my analysis : Use features inside the WCD signal to estimate the muonic content of the showers and thus differentiate between photon/proton primaries Improve the discrimination power of the analysis SD only analysis = 100% duty cycle Trace = WCD signal

24

SLIDE 25

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

M u

n

i c i t y M e t h

d

– s t

l

e v e l v a r i a b l e

Photon Search Muon peak Decay-tail Deconvolution filter ck → c’k Same peak after deconvolution Same integral:

Sk

Core of the Muonicity Method : perform a linear deconvolution to remove the exponential decay-tail of a muon signal while integrating the total muon-signal.

Simulated WCD trace

25

SLIDE 26

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

M u

n

i c i t y M e t h

d

– s t

l

e v e l v a r i a b l e

Photon Search

Calculate Sk for each peak
Sum Sk for each station : Speaks

For each station j deconvolute the VEM trace, identify the peaks :

Core of the Muonicity Method : perform a linear deconvolution to remove the exponential decay-tail of a muon signal while integrating the total muon-signal. 26

SLIDE 27

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

Photon Search

M u

n

i c i t y M e t h

d

– e v t

l

e v e l v a r i a b l e

Value calculated on trace, for proton and photon showers : Value predicted by proton-trained Machine Learning model :

→ → the model is supposed to better reconstruct proton- (closer to ) the model is supposed to better reconstruct proton- (closer to )

Machine Learning Model :

Parameters : Distance, Zenith, SD_Energy
Predict values from parameters

Training with proton simulations Training with proton simulations Predict a value from parameters Predict a value from parameters

27

SLIDE 28

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

M u

n

i c i t y M e t h

d

– P r e l i m i n a r y R e s u l t s

Photon Search Simplified pipeline of the Muonicity Method On its own and without further parameter tuning, the Muonicity variable has some separation power Merit Factor η=0.91 → Can we extract more discrimination from the muonic component of the SD traces? 28

SLIDE 29

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

S m

t

h i n g / M u

n

i c i t y M e t h

d

Photon Search Simplified pipeline of the Muonicity Method Muonicity :

deconvolute muon peaks
sum peaks signal

→ Speaks

29

SLIDE 30

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

Simplified pipeline of the Smoothing Method Smoothing :

smoothen out the muon peaks
estimate the EM signal

→ fmu

S m

t

h i n g / M u

n

i c i t y M e t h

d

Photon Search 30

SLIDE 31

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

Simplified pipeline of the Smoothing Method Smoothing :

smoothen out the muon peaks
estimate the EM signal

→ fmu

S m

t

h i n g / M u

n

i c i t y M e t h

d

Photon Search 31 Smoothing variable’s separation power : Merit Factor η=1.50 → Let’s combine the 2 variables!

SLIDE 32

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

M u l t i

V

a r i a t e A n a l y s i s

Photon Search

As expected the 2 variables are correlated
But the separation power is still better by combining the two
Using a Support Vector Machine classifier with RBF kernels
At 50% signal efficiency, 99.6% background rejection

32

SLIDE 33

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

Wh a t I d i d s

f

a r

S

S D s a s s e m b l y :

–

Wr

t

e a p r

c

e d u r e t

s

t a n d a r d i z e t h e S S D s c

n

s t r u c t i

n

p r

c

e s s

–

H e l p e d d e s i g n a n d i n s t a l l t h e S S D s t e s t s e t u p

–

P e r f

r

m e d t h e S S D s v a l i d a t i

n

s

P

h

t
n
s

e a r c h :

–

I m p l e m e n t e d d i s c r i m i n a t i

n
b

s e r v a b l e s c a l c u l a t i

n

i n s i d e t h e A u g e r f r a m e w

r

k

–

D e s i g n e d :

a

fm e x i b l e m e t h

d

f

r

p h

t
n

/ p r

t
n

d i s c r i m i n a t i

n
a

p i p e l i n e f

r

M V A b a s e d

n

m a c h i n e l e a r n i n g

S

h i f t s , F

r

m a t i

n

s & O u t r e a c h :

–

P a r t i c i p a t e d t

F

D s h i f t s a n d b e t a

t

e s t e d t h e S D s h i f t s

–

O u t r e a c h t

s

c h

l

a r s

–

L a b e l R E I

–

S O S s c h

l

, I S A P P s c h

l

Conclusion 33

SLIDE 34

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

Wh a t ’ s n e x t ?

S

S D s t e s t s :

–

R e

r

u n t h e a n a l y s i s

n

t h e w h

l

e d a t a s e t

n

c e t h e p r

d

u c t i

n

i s fj n i s h e d

M

u

n

i c i t y & S m

t

h i n g m e t h

d

s :

–

O p t i m i z e t h e

b

s e r v a b l e s

–

R e

r

u n t h e p i p e l i n e w i t h t u n e d p a r a m e t e r s , u p d a t e d s i m u l a t i

n

s a n d h e a v i e r p r i m a r i e s .

–

T r y c h a n g i n g t h e M L i n p u t p a r a m e t e r s

–

A d d

t

h e r

b

s e r v a b l e s ?

A

p p l y t h e m e t h

d

t

d

a t a

D

e t e c t U H E p h

t
n

s

r

i m p r

v

e t h e l i m i t s

n

t h e fm u x

Outlook 34

SLIDE 35

T h a n k y

u

!

SLIDE 36

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

P r i m a r y d e p e n d e n c e

f

t h e s h

w

e r s

Back-up Differences in the “shape” of the showers :

First interaction further down for light CRs
Spread of the z(Nmax) tighter for heavier

CRs Differences in the composition of the showers :

Heavier primary → faster energy loss in the

atmosphere

Muon ratio α Aprimary

36

SLIDE 37

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

S m

t

h i n g M e t h

d

– s t

l

e v e l v a r i a b l e

Back-up Core of the Smoothing Method : extract the muonic component of the trace by performing a sliding-window averaging to remove spikes The smoothened trace is then compared to the original trace to obtain a st-level variable : fmu fmu here, has the same role as Speaks in the Muonicity Method 37

SLIDE 38

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

P h

t
n

A n a l y s i s i n A u g e r

Back-up

PRELIMINARY

Hybrid (SD+FD) analysis : more observables
SD only analysis : more data

Aim of my analysis : Use features inside the WCD signal to estimate the muonic content of the showers and thus differentiate between photon/proton primaries SD only analysis = 100% duty cycle Trace = WCD signal

38

SLIDE 39

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

A n i s

t

r

p

y m a p s

Back-up 39

SLIDE 40

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

C

n

s t r u c t i n g a n d v a l i d a t i

n

s t a t u s

Back-up 40

SLIDE 41

T h e s i s M

n

i t

r

i n g P r e s e n t a t i

n

J u l i e n S

u

c h a r d

R a d i

D

e t e c t i

n

Back-up 41

SLIDE 42

SLIDE 43

SLIDE 44

SLIDE 45

SLIDE 46

SLIDE 47