1. The PAMELA experiment: looking for antiparticles in Cosmic Rays
Firenze, 25 February 2009
The PAMELA experiment: looking for antiparticles in Cosmic Rays
Oscar Adriani
University of Florence and INFN Florence

2.  Oscar Adriani  Florence, February 25th, 2009 
Outline
The PAMELA experiment: short review
Results on cosmic-ray antimatter abundance:
Antiprotons
Positrons
Other results:
Cosmic-ray galactic light nuclei (primaries & secondaries)
Solar physics
Terrestrial physics
Conclusions
 Oscar Adriani  Florence, February 25th, 2009 

3. Tha PAMELA collaboration
Italy
Bari
Florence
Frascati
Trieste
Naples
Rome
CNR, Florence
Moscow
St. Petersburg
Russia
Germany
Siegen
Sweden
KTH, Stockholm
 Oscar Adriani  Florence, February 25th, 2009 

4. Why CR antimatter? Evaporation of primordial black holes
Anti-nucleosyntesis
First historical measurements of p-bar/p ratio
WIMP dark-matter annihilation in the galactic halo
Background:
CR interaction with ISM
CR + ISM  p-bar + …
 Oscar Adriani  Florence, February 25th, 2009 

5. Cosmic-ray Antimatter from Dark Matter annihilation
Annihilation of relic Weakly Interacting Massive Particles (WIMPs) gravitationally
confined in the galactic halo
 Distortion of antiproton and positron spectra from purely secondary production
A plausible dark matter candidate is
neutralino (c), the lightest SUSY
Particle (LSP).
Most likely processes:
cc  qq  hadrons anti-p, e+,…
cc  W+W-,Z0Z0,…  e+,…
 positron peak Ee+~Mc/2
 positron continuum Ee+~Mc/20
Another possible candidate is the lightest Kaluza-Klein Particle (LKP): B(1)
Fermionic final states no longer suppressed:
B(1)B(1)  e+e-
direct dicay  positron peak Ee+ ~ MB(1)
Halo
You are here
Milky Way
p-bar, e+ c c
 Oscar Adriani  Florence, February 25th, 2009 

6. Charge-dependent solar modulation
Solar polarity reversal 1999/2000
Asaoka Y. Et al. 2002 +
CR antimatter
Experimental scenario before PAMELA
Antiprotons
Positrons
Positron excess? ? ?
___ Moskalenko & Strong 1998
CR + ISM  p-bar + …
Propagation dominated by nuclear interactions
Kinematical threshold: Eth~5.6 for the reaction
CR + ISM  p± + x  m± + x  e± + x
CR + ISM  p0 + x  gg  e±
Propagation dominated by energy losses
(inverse Compton & synchrotron radiation)
Local origin 90% from <2kpc)
 Oscar Adriani  Florence, February 25th, 2009 

7.  Oscar Adriani  Florence, February 25th, 2009 
PAMELA detectors
Main requirements  high-sensitivity antiparticle identification and precise momentum measure
Time-Of-Flight
plastic scintillators + PMT:
Trigger
Albedo rejection;
Mass identification up to 1 GeV;
- Charge identification from dE/dX.
Electromagnetic calorimeter
W/Si sampling (16.3 X0, 0.6 λI)
Discrimination e+ / p, anti-p / e-
(shower topology)
Direct E measurement for e-
Neutron detector
High-energy e/h discrimination
Overview of the apparatus
GF: 21.5 cm2 sr Mass: 470 kg
Size: 130x70x70 cm3
Power Budget: 360W
Spectrometer
microstrip silicon tracking system permanent magnet
It provides:
- Magnetic rigidity  R = pc/Ze
Charge sign
Charge value from dE/dx
 Oscar Adriani  Florence, February 25th, 2009 

8. Principle of operation
with protons of known momentum
 MDR~1TV
Cross-check in flight with protons (alignment) and electrons (energy from calorimeter)
Iterative c2 minimization as a function of track state-vector components a
Magnetic deflection
|η| = 1/R
R = pc/Ze  magnetic rigidity
sR/R = sh/h
Maximum Detectable Rigidity (MDR)
R=MDR  sR/R=1
MDR = 1/sh
Track reconstruction
 Oscar Adriani  Florence, February 25th, 2009 

9. Principle of operation
Z measurement
Bethe Bloch
ionization energy-loss of heavy (M>>me) charged particles p d
3He
4He Li Be
B,C
track average e±
(saturation)
1st plane
 Oscar Adriani  Florence, February 25th, 2009 

10. Principle of operation
Particle low energy
Identify albedo (up-ward going particles b < 0 )
 NB! They mimic antimatter!
Velocity measurement
 Oscar Adriani  Florence, February 25th, 2009 

11. Principle of operation
Interaction topology
e/h separation
Energy measurement of electrons and positrons
(~full shower containment)
electron (17GV)
hadron (19GV)
Electron/hadron separation
+ NEUTRONS!!
 Oscar Adriani  Florence, February 25th, 2009 

12. The Resurs DK-1 spacecraft
Multi-spectral remote sensing of earth’s surface
near-real-time high-quality images
Built by the Space factory TsSKB Progress in Samara (Russia)
Operational orbit parameters:
inclination ~70o
altitude ~ km (elliptical)
Active life >3 years
Data transmitted via Very high-speed Radio Link (VRL)
Mass: 6.7 tons
Height: 7.4 m
Solar array area: 36 m2
Characteristic of the satellite
PAMELA mounted inside a pressurized container
moved from parking to data-taking position few times/year
 Oscar Adriani  Florence, February 25th, 2009 

13. PAMELA design performance
Maximum detectable rigidity (MDR)
Magnetic curvature & trigger
spillover
shower
containment
energy range particles in 3 years
Antiprotons 80 MeV ÷190 GeV O(104)
Positrons 50 MeV ÷ 270 GeV O(105)
Electrons up to 400 GeV O(106)
Protons up to 700 GeV O(108)
Electrons+positrons up to 2 TeV (from calorimeter)
Light Nuclei up to 200 GeV/n He/Be/C: O(107/4/5)
Anti-Nuclei search sensitivity of 3x10-8 in anti-He/He
Unprecedented statistics and new energy range for cosmic ray physics
(e.g. contemporary antiproton and positron maximum energy ~ 40 GeV)
Simultaneous measurements of many species
 Oscar Adriani  Florence, February 25th, 2009 

14.  Oscar Adriani  Florence, February 25th, 2009 
PAMELA milestones
Launch from Baikonur  June 15th 2006, 0800 UTC.
‘First light’  June 21st 2006, 0300 UTC.
• Detectors operated as expected after launch
• Different trigger and hardware configurations evaluated
 PAMELA in continuous data-taking mode since
commissioning phase ended on July 11th 2006
Main antenna in NTsOMZ
Trigger rate* ~25Hz
Fraction of live time* ~ 75%
Event size (compressed mode) ~ 5kB
25 Hz x 5 kB/ev  ~ 10 GB/day
(*outside radiation belts)
Till December 2008:
~800 days of data taking
~16 TByte of raw data downlinked
~16•108 triggers recorded and analysed
(Data from May till now under analysis)
 Oscar Adriani  Florence, February 25th, 2009 

15. Antiprotons

16. High-energy antiproton analysis
Analyzed data July 2006 – February 2008 (~500 days)
Collected triggers ~108
Identified ~ protons and ~ antiprotons between 1.5 and 100 GeV ( 100 p-bar above 20GeV )
Antiproton/proton identification:
rigidity (R)  SPE
|Z|=1 (dE/dx vs R)  SPE&ToF
b vs R consistent with Mp  ToF
p-bar/p separation (charge sign)  SPE
p-bar/e- (and p/e+ ) separation  CALO
Dominant background  spillover protons:
finite deflection resolution of the SPE  wrong assignment of high energy
proton spectrum harder than positron  p/p-bar increase for increasing energy
 Required strong SPE selection
 Oscar Adriani  Florence, February 25th, 2009 

17. Antiproton identification
Preliminary!!
-1  Z  +1
p (+ e+) p
e- (+ p-bar)
proton-consistency cuts
(dE/dx vs R and b vs R)
“spillover” p
p-bar
electron-rejection cuts based on calorimeter-pattern topology
5 GV
1 GV
( For |Z|=1, deflection=1/p )
 Oscar Adriani  Florence, February 25th, 2009 

18. Proton-spillover background
MDR = 1/sh
(evaluated event-by-event by the fitting routine)
Minimal track requirements
MDR > 850 GV
Strong track requirements:
strict constraints on c2 (~75% efficiency)
rejected tracks with low-resolution clusters along the trajectory
- faulty strips (high noise)
- d-rays (high signal and multiplicity)
 Oscar Adriani  Florence, February 25th, 2009 

19. Proton-spillover background
MDR = 1/sh
(evaluated event-by-event by the fitting routine)
p-bar
“spillover” p p
R < MDR/10
10 GV
50 GV
MDR depends on:
number and distribution of fitted points along the trajectory
spatial resolution of the single position measurements
magnetic field intensity along the trajectory
 Oscar Adriani  Florence, February 25th, 2009 

20. Antiproton-to-proton ratio
*preliminary*
PRL 102, (2009)
(Petter Hofverberg’s PhD Thesis)
 Oscar Adriani  Florence, February 25th, 2009 

21. Positrons

22. High-energy positron analysis
Analyzed data July 2006 – February 2008 (~500 days)
Collected triggers ~108
Identified ~ electrons and ~ positrons between 1.5 and 100 GeV (180 positrons above 20GeV )
Electron/positron identification:
rigidity (R)  SPE
|Z|=1 (dE/dx=MIP)  SPE&ToF
b=1  ToF
e-/e+ separation (charge sign)  SPE
e+/p (and e-/p-bar) separation  CALO
Dominant background  interacting protons:
fluctuations in hadronic shower development  p0 gg might mimic pure em showers
proton spectrum harder than positron  p/e+ increase for increasing energy
 Required strong CALO selection S1 S2
CALO S4
CARD
CAS
CAT
TOF
SPE S3 ND
 Oscar Adriani  Florence, February 25th, 2009 

23. Positron identification with CALO
Identification based on:
Shower topology (lateral and longitudinal profile, shower starting point)
Total detected energy (energy-rigidity match)
Analysis key points:
Tuning/check of selection criteria with:
test-beam data
simulation
flight data  dE/dx from SPE & neutron yield from ND
Selection of pure proton sample from flight data (“pre-sampler” method):
Background-suppression method
Background-estimation method
51 GV positron
80 GV proton
Final results make NON USE of test-beam and/or simulation calibrations.
The measurement is based only on flight data
with the background-estimation method
 Oscar Adriani  Florence, February 25th, 2009 

24. Positron identification
p (non-int)
p (int)
NB!
p-bar (int) e-
p-bar (non-int)
Z=-1
Z=+1
Rigidity: GV
Fraction of charge released along the calorimeter track
LEFT
HIT
RIGHT
strips
planes
0.6 RM
Total negative  3067
Total positive 
Selected electrons  2032
Selected positrons  322
 Oscar Adriani  Florence, February 25th, 2009  24

25. Positron identification
Energy-momentum match
Deflection in Tracker (MIP/GV)
Energy measured in Calo/ e-
( e+ )
 e
 h
p-bar p
 Oscar Adriani  Florence, February 25th, 2009 

26. Positron identification
p-bar e-
Z=-1
Z=+1
Rigidity: GV
(e+)
p (non-int)
p (int)
NB!
p-bar (int) e-
p-bar (non-int)
Z=-1
Z=+1
Rigidity: GV
Fraction of charge released along the calorimeter track +
Constraints on:
Energy-momentum match
Total negative  3067
Total positive 
Selected electrons  2032
Selected positrons  322
 Oscar Adriani  Florence, February 25th, 2009  26

27. Positron identification
51 GV positron
Shower starting-point
Longitudinal profile
80 GV proton
 Oscar Adriani  Florence, February 25th, 2009 

28. Positron identification
p-bar e-
Z=-1
Z=+1
Rigidity: GV e+ p e-
Rigidity: GV
Z=-1
Z=+1
Fraction of charge released along the calorimeter track +
Constraints on:
Energy-momentum match
Shower starting-point
Longitudinal profile
Lateral profile
Total negative  3067
Total positive 
Selected electrons  2032
Selected positrons  322
BK-suppression method
 Oscar Adriani  Florence, February 25th, 2009  28

29. The “pre-sampler” method
Selection of a pure sample of protons from flight data
CALORIMETER: 22 W planes: 16.3 X0
2 W planes: ≈1.5 X0
20 W planes: ≈15 X0
 Oscar Adriani  Florence, February 25th, 2009  29

30. Proton background evaluation
Rigidity: GV e-
Fraction of charge released along the calorimeter track (left, hit, right) +
Constraints on:
p (pre-sampler)
Energy-momentum match
Shower starting-point
Total negative  3067
Total positive 
Selected electrons  2032
Selected positrons  322 e+ p
 Oscar Adriani  Florence, February 25th, 2009  30

31.  Oscar Adriani  Florence, February 25th, 2009 
Positron fraction
astro-ph
Accepted by Nature
 Oscar Adriani  Florence, February 25th, 2009 

32. Do we have any antimatter excess in CRs?
 Oscar Adriani  Florence, February 25th, 2009 

33. Antiproton-to-proton ratio Secondary Production Models
CR + ISM  p-bar + …
(Moskalenko et al. 2006) GALPROP code
Plain diffusion model
Solar modulation: drift model ( A<0, a=15o )
(Donato et al. 2001)
Diffusion model with convection and reacceleration
Solar modulation: spherical model (f=500MV )
 Uncertainty band related to propagation parameters
 Additional uncertainty of ~25% due to production cs should be considered !!
(Ptuskin et al. 2006) GALPROP code
Plain diffusion model
Solar modulation: spherical model ( f=550MV )
No evidence for any antiproton excess
 Oscar Adriani  Florence, February 25th, 2009 

34. Positron fraction Secondary Production Models
CR + ISM  p± + …  m± + …  e± + …
CR + ISM  p0 + …  gg  e±
(Moskalenko & Strong 1998)
GALPROP code
Plain diffusion model
Interstellar spectra
 Oscar Adriani  Florence, February 25th, 2009 

35. Charge dependent solar modulation+ +
A > 0
Positive particles
A < 0
Drift model
(Clem & Evenson 2007)
 Oscar Adriani  Florence, February 25th, 2009 

36. Positron fraction Secondary Production Models
CR + ISM  p± + …  m± + …  e± + …
CR + ISM  p0 + …  gg  e±
The positron fraction depends on the primary (+secondary)
electron spectrum
(Moskalenko & Strong 1998)
GALPROP code
Plain diffusion model
Interstellar spectra
Soft electron spectrum (g = 3.54)
Hard electron spectrum (g = 3.34)
(Delahaye et al. 2008)
Plain diffusion model
Solar modulation: spherical model (f=600MV)
Uncertainty band related to e- spectral index
(ge = 3.44±0.1  MIN÷MAX)
Additional uncertainty due to propagation parameters should be considered (factor
Moskalenko & strong 1998  electron gamma = 2.1 below 10GeV  2.4 above
 Oscar Adriani  Florence, February 25th, 2009 

37. Positron fraction Secondary Production Models
Moskalenko & strong 1998  electron gamma = 2.1 below 10GeV  2.4 above
Preferred by Pamela
electron data!
Quite robust evidence for a positron excess
 Oscar Adriani  Florence, February 25th, 2009 

38. Primary positron sources
Dark Matter
e+ yield depend on the dominant decay channel
LSPs seem disfavored due to suppression of e+e- final states
low yield (relative to p-bar)
soft spectrum from cascade decays
LKPs seem favored because can annihilate directly in e+e-
high yield (relative to p-bar)
hard spectrum with pronounced MLKP (>300 GeV)
Boost factor required to have a sizable e+ signal
NB: constraints from p-bar data!!
LKP -- M= 300 GeV
(Hooper & Profumo 2007)
 Oscar Adriani  Florence, February 25th, 2009 

39. Primary positron sources
Astrophysical processes
Local pulsars are well-known sites of e+e- pair production:
 they can individually and/or coherently contribute to the e+e- galactic flux and explain the PAMELA e+ excess (both spectral feature and intensity)
No fine tuning required
if one or few nearby pulsars dominate, anisotropy could be detected in the angular distribution
possibility to discriminate between pulsar and DM origin of e+ excess
All pulsars (rate = 3.3 / 100 years)
(Hooper, Blasi, Seprico 2008)
~80 theoretical paper on Pamela data since our ArXiv publication!!!!!
 Oscar Adriani  Florence, February 25th, 2009 

40.  Oscar Adriani  Florence, February 25th, 2009 
PAMELA positron excess might be connected with ATIC electron+positron structures?
(Chang et al 2008)
 Oscar Adriani  Florence, February 25th, 2009 

41. Primary positron sources
PAMELA positron fraction alone insufficient to understand the origin of positron excess
Additional experimental data will be provided by PAMELA:
e+ higher energy (up to 300 GeV)
individual e- e+ spectra
anisotropy (…maybe)
high energy e++e- spectrum (up to 2 TV)
Complementary information from:
gamma rays
neutrinos
 Oscar Adriani  Florence, February 25th, 2009 

42. Galactic cosmic-ray origin & propagation

43.  Oscar Adriani  Florence, February 25th, 2009 
H and He spectra
(statistical errors only)
Preliminary!!
Proton of primary origin
Diffusive shock-wave acceleration in SNRs
Local spectrum:
injection spectrum  galactic propagation
Local primary spectral shape:
study of particle acceleration mechanism
Very high statistics over a wide energy range
 Precise measurement of spectral shape
 Possibility to study time variations and transient phenomena
 Oscar Adriani  Florence, February 25th, 2009 

44.  Oscar Adriani  Florence, February 25th, 2009 
Secondary nuclei
Preliminary!!
B nuclei of secondary origin:
CNO + ISM  B + …
Local secondary/primary ratio sensitive to average amount of traversed matter (lesc) from the source to the solar system
Local secondary abundance:
study of galactic CR propagation
(B/C used for tuning of propagation models)
 Oscar Adriani  Florence, February 25th, 2009 

45. Solar modulation Solar Energetic Particle events (SEPs)
Solar physics
Solar modulation
Solar Energetic Particle events (SEPs)

46.  Oscar Adriani  Florence, February 25th, 2009 
Solar modulation
Interstellar spectrum
Preliminary!!
(statistical errors only)
Increasing
GCR flux
July 2006
August 2007
February 2008
solar activity
Decreasing
sun-spot number
Ground neutron monitor
PAMELA
 Oscar Adriani  Florence, February 25th, 2009 

47. December 2006 CME/SEP events
SOHO/LASCO
SOHO/EIT
Coronal Mass Ejection (CME)
X-ray flares
Solar Energetic Particles (SEPs)
protons: 1÷100 MeV
alphas: 150÷500 MeV
 Oscar Adriani  Florence, February 25th, 2009 

48.  Oscar Adriani  Florence, February 25th, 2009 
December 13th event
from to
Increase of low energy component
 Oscar Adriani  Florence, February 25th, 2009 

49.  Oscar Adriani  Florence, February 25th, 2009 
December 13th event
from to
from :23:02 to :57:46
Increase of low energy component
 Oscar Adriani  Florence, February 25th, 2009 

50.  Oscar Adriani  Florence, February 25th, 2009 
December 13th event
from to
from :23:02 to :57:46
from :57:46 to :49:09
Increase of low energy component
 Oscar Adriani  Florence, February 25th, 2009 

51.  Oscar Adriani  Florence, February 25th, 2009 
December 13th event
from to
from :23:02 to :57:46
from :57:46 to :49:09
from :49:09 to :32:56
Decrease of high energy component
Increase of low energy component
Increase of low energy component
 Oscar Adriani  Florence, February 25th, 2009 

52.  Oscar Adriani  Florence, February 25th, 2009 
December 13th event
from to
from :23:02 to :57:46
from :57:46 to :49:09
from :49:09 to :32:56
from :32:56 to :59:16
Increase of low energy component
 Oscar Adriani  Florence, February 25th, 2009 

53.  Oscar Adriani  Florence, February 25th, 2009 
December 13th event
from to
from :23:02 to :57:46
from :57:46 to :49:09
from :49:09 to :32:56
from :32:56 to :59:16
from :17:54 to :17:34
SEPs accelerated during CMEs
SEP spectral shape and time evolution
 study of particle acceleration mechanisms in CME
Additional information will be provided by:
electrons & positrons
nuclear and isotopic (3He, 4He) composition
Increase of low energy component
 Oscar Adriani  Florence, February 25th, 2009 

54. Interactions of CRs with the atmosphere
Terrestrial physics
Magnetosphere
Radiation belts & SAA
Interactions of CRs with the atmosphere

55. Analysis of a PAMELA orbit
SAA S1 S2 S3
Low energy particles stops inside the apparatus
 Counting rates: S1>S2>S3
Outer radiation belt
– 15/02/ :35:00 MWT
NP SP
EQ EQ
95 min S1 S2 S3
Outer rb = elettroni. Piu` vicina alla terra. Si attraversa al polo sud perche` il satellite e` piu` lontano. 600km a sud 350km nord.
La SSA e` un’estenzione della fascia interna. Causata dal fatto che in centro del dipolo magnetico e` spostato rispetto al cetro della terra.
Counting rates along the orbit. Comparison between different detectors (tof scintillators, neutron detector, anticounters) allows to check detector behavior in the space environment.
Conclusion: counting rates are consistent with satellite position along the orbit
Rate on S1, S2 S3:
- Duration of a orbit
- Environment: poles, equators and SSA
Inner radiation belt
(SAA)
Ratemeters
Independent from trigger
orbit 3751
orbit 3752
orbit 3753
 Oscar Adriani  Florence, February 25th, 2009 

56.  Oscar Adriani  Florence, February 25th, 2009 
Trigger rate
SAA
trapped
 Oscar Adriani  Florence, February 25th, 2009 

57. Trapped-particle spectrum
Preliminary!!
(statistical errors only)
Trapped
Preliminary!!
Galactic
B < 0.19 G
0.19 G ≤ B < 0.20 G
0.20 G ≤ B < 0.21 G
B > 0.30 G
0.22 G ≤ B < 0.23 G
0.21 G ≤ B < 0.22 G
Always: 10 GV < cutoff < 11
 Oscar Adriani  Florence, February 25th, 2009 

58. Secondary reentrant-albedo protons
Geomagnetic cutoff
(statistical errors only)
Magnetic poles
( galactic protons)
Secondary reentrant-albedo protons
Atmospheric neutrino contribution
Astronaut dose
Indirect measurement of cs in the atmosphere
Agile e Glast background estimation
Preliminary!!
0.4 to 0.5
1.0 to 1.5
1.5 to 2.0
2 to 4
> 14
10 to 14
7 to 10
4 to 7
Geomagnetic
cutoff (GV/c)
--- M. Honda, 2008
Magnetic equator
Up-ward going albedo excluded
SAA excluded
 Oscar Adriani  Florence, February 25th, 2009 

59.  Oscar Adriani  Florence, February 25th, 2009 
Conclusions
PAMELA is the first space experiment which is measuring the antiproton and positron cosmic-ray components to the high energies (>100GeV) with an unprecedented statistical precision
 search for evidence of DM candidates
 “direct” measurement of particle acceleration in astrophysical sources (pulsars?)
Furthermore:
PAMELA is providing measurements on low-mass elemental (and isotopical) spectra with an unprecedented statistical precision
 study of particle origin and propagation in the interstellar medium
PAMELA is able to measure the high energy tail of solar particles.
PAMELA is measuring composition and spectra of trapped and re-entrant albedo particles in the Earth magnetosphere
 Oscar Adriani  Florence, February 25th, 2009 