The Cosmic Ray electron spectrum below 20 GeV Melissa - - PowerPoint PPT Presentation

The Cosmic Ray electron spectrum below 20 GeV

Melissa Pesce-Rollins

INFN–Pisa melissa.pesce.rollins@pi.infn.it

• n behalf of the Fermi LAT

collaboration RICAP May 13, 2009

Outline

◮ Monte Carlo simulations

◮ Instrument simulation ◮ International Geomagnetic Reference Field (IGRF)

◮ Analysis strategy ◮ Event selection ◮ Estimation of the background ◮ Preliminary results

◮ Comparison between Fermi-LAT, PAMELA and AMS ◮ Searching for east-west asymmetries ◮ Count rates of secondaries particles.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 2 / 25

Cosmic ray electrons and the Fermi LAT

Two different set of analyses, two different sources Fermi-LAT does not distinguish between e− and e+, we use the term electrons to refer to the sum of the two.

Diagnostic (DGN) on-board filter

◮ Unbiased sample of all events that trigger the LAT. ◮ Prescaled 1:250 on-board(due to bandwidth limitations). ◮ Designed for diagnostic purposes ◮ And study of filter efficiencies ◮ ∼ 20 Hz rate ◮ Excellent data source of cosmic ray electrons!

Gamma on-board filter

◮ Main on-board filter for γs ◮ High pass condition:

◮ Downlinks all events with onboard

energy greater than 20 GeV

◮ Excellent data source for cosmic ray

electrons above 20 GeV.

◮ Overlap region (20 GeV - 100 GeV) also used for comparison of two analysis approaches. ⇒ Large fluxes for CR electrons below few GeV compensates for DGN 1:250 prescale!

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 3 / 25

Monte-Carlo simulations

◮ Accurate detector model

◮ over 45000 volumes

◮ Physical interactions with

GEANT4

◮ Uses real LAT calibrations ◮ Monte Carlo is crucial for

◮ Event selection and LAT

performance

◮ Instrument Response

Functions

◮ Estimation of residual

hadron contamination

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 4 / 25

International Geomagnetic Reference Field (IGRF)

◮ We use the 10th generation version of IGRF. ◮ http://www.ngdc.noaa.gov/IAGA/vmod/igrf.html ◮ Fermi orbits at 565 km altitude with 25.6◦ inclination

◮ Fixes the geomagnetic latitude range and ◮ fixes the minimum energy at which we can measure the

primary CR electrons ∼ 5 GeV.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 5 / 25

Selection criteria

Fraction of electrons after minimal quality cuts ◮ Electron fraction has large variations over the

energy range of interest.

◮ The LAT does not distinguish electrons from

positrons, we take the sum of the two contributions.

◮ Therefore the term electrons refers to the sum

• f electrons and positrons.

◮ Event selection based on LAT’s

capability to discriminate between EM and hadronic showers in the subdetectors

◮ Similar approach to photon

analysis

◮ Make use of Classification

Trees for final boost in particle selection.

◮ Selection cuts need to be a

smooth function of energy

◮ Make sure not to introduce

any artificial features in the measured spectrum

◮ Could also be a function of

geomagnetic latitude.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 6 / 25

Geometry factor and rejection power

Geometry Factor

Energy (MeV)

3

10

4

10

5

10

Geometry Factor (m^2 sr)

0.005 0.01 0.015 Preliminary Results

Hadron Rejection Power

Energy (MeV)

3

10

4

10 Hadron rejection power 1 10

2

10

3

10

4

10

5

10

Preliminary Results quality quality + acd quality + acd + HEEProb quality + acd + HEEProb + cal quality + acd + HEEProb + cal + CT

◮ Geometry Factor: GF(E) = FOV · Aeff (E) ◮ The geometry factor is small due to the 1:250 prescale of the

DGN filter.

◮ Hadron rejection power at various stages of the event

selection.

◮ Final boost in rejection power from the CT analysis is clear.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 7 / 25

Orbital averaged hadron contamination

Energy (MeV)

3

10

4

10

Orbital Ave contamination

0.05 0.1 0.15 Preliminary Results

◮ Average contamination below 20% throughout the energy

range.

◮ Varies with energy. ◮ Varies for different orbital positions. ◮ Optimization of the selection cuts as a function of

geomagnetic latitude still needs to be performed.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 8 / 25

Monte Carlo validation with flight data

Preliminary results

◮ An ACD variable at an

intermediate stage of the analysis.

◮ Total energy deposited in the

ACD tiles per event.

◮ Analysis variables carefully

checked over the energy range with flight data.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 9 / 25

Monte Carlo validation with flight data

Preliminary results

◮ The low energy CT variable at

an intermediate stage of the analysis

◮ Continuous probability

variable

◮ Overall agreement is fair but we

plan to optimize the CT variable for better agreement.

◮ Analysis variables carefully

checked over the energy range with flight data

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 10 / 25

Comparison with PAMELA and AMS

◮ Geomagnetic latitude interval: 0.0 ≤ λ ≤ 0.3 ◮ Good overall agreement with PAMELA and AMS. ◮ Some differences in the energy range ∼3 GeV and 10 GeV. ◮ Systematics not included! Estimated to be around 20%

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 11 / 25

Comparison with PAMELA and AMS

◮ Geomagnetic latitude interval: 0.3 ≤ λ ≤ 0.6 ◮ Cutoff position in rough agreement. ◮ Systematics not included! Estimated to be around 20% ◮ Large differences at lower energies.

◮ Different fraction of time spent in each lambda interval. ◮ Different acceptance.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 12 / 25

Fraction of time spent in each latitude band

Fraction of time Fermi and PAMELA spend in each latitude band

Fermi: ◮ 0.0 ≤ λ ≤ 0.3, 49% of total orbital time. ◮ 0.3 ≤ λ ≤ 0.6 spend 46% of total orbital time

◮ Roughly half of which is spent in 0.3

≤ λ ≤ 0.4 interval. PAMELA: ◮ 0.0 ≤ λ ≤ 0.1, 23% of total orbital time. ◮ 0.3 ≤ λ ≤ 0.6 roughly equal fraction of time spent in each lambda interval. ◮ Comparison at this stage intended to be a simple sanity check. ◮ Cutoff position is quite similar between the three experiments in both intervals. ◮ Fluxes below the cutoff are not isotropic: ◮ The three experiments have different acceptances (Fermi is accepting particles from ∼ 60 ◦). ◮ Larger fraction of time spent in lower geomagnetic latitude could explain the higher fluxes seen by Fermi in 0.3 ≤ λ ≤ 0.6 interval (larger concentration of secondaries located near the magnetic equator). ◮ PAMELA data correspond to August-September 2006; solar modulation?

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 13 / 25

Searching for East West asymmetries

Flux East west ratio

Geomagnetic latitude interval 0.1< λ <0.2 ◮ Magnetic rigidity is also a function of the particle’s charge, expect to see an east/west

effect on the magnetic cutoff.

◮ Clear separation between east and west intensities. ◮ Larger intensities coming from the east.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 14 / 25

Searching for East West asymmetries

Flux East west ratio

Geomagnetic latitude interval 0.4< λ <0.5 ◮ Cutoff energy decreases with increasing lambda as expected. ◮ Gradual decrease in the east west ratio at higher geomagnetic latitude.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 15 / 25

Searching for East West asymmetries

◮ Asymmetry peak value in energy decreases with increaing

lambda values, this is expected due to the rigidity cutoff.

◮ The amount of asymmetry decreases with increasing lambda

(maximum in the geomagnetic equator region).

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 16 / 25

AMS secondary electrons and positrons

◮ Geographical origin of electrons (a) and positrons (b)

measured by AMS 01.

◮ Secondary electrons and positrons are concentrated around

the geomagnetic equator.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 17 / 25

Distribution of secondary electrons by Fermi

Preliminary Results!

◮ Count rate of secondary electrons and positrons as measured

by Fermi.

◮ Flux distribution is still work in progress.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 18 / 25

Conclusions

◮ Event selection in an advanced stage ◮ Orbital averaged hadron contamination well below 20% ◮ Preliminary cosmic ray electron fluxes are roughly consistent

with past experiments.

◮ Still some discrepancy at high geomagnetic latitude ◮ Working to pin down the cause.

◮ Preliminary signs of east west asymmetries ◮ Trapped count rates show a concentration in the geomagnetic

equitorial region (as also observed by M.Aguillar et al., Phys. Rep.,366, 331 (2002))

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 19 / 25

Future work

◮ Study the systematic uncertainties

◮ Data/Monta Carlo simulations ◮ Background model

◮ Spectrum unfolding ◮ Geomagnetic latitude dependent event selections ◮ Update the background models with Fermi electron data.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 20 / 25

Extra Slides

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 21 / 25

Energy resolution

◮ Energy resolution integrated over all angles. ◮ Compatible with energy resolution of the LAT for photons.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 22 / 25

Energy dependent cuts

◮ At high energy the necessary rejection power increases with

energy.

◮ Most of the cuts do not depend on energy. ◮ Even when this is not the case a linear function of log energy is

enough.

◮ At low energy the dependancy of the cut values on energy

need to be more complecated to achieve the necessary rejection power over the entire range.

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 23 / 25

On orbit trigger rates

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 24 / 25

East/West effects

Illustration of east/west effect in primaries

B field N E W S Positively charged particles drift to west Negatively charged particles drift to the east

◮ The Lorentz force:

F = q( v/cx B)

◮ Direction of the velocity of the particle given

by the circle.

◮ If particle is positive it will drift towards the

east and we will see a larger flux coming from the west.

◮ Vice versa for negative particles. The right hand rule

• M. Pesce-Rollins (INFN)

RICAP May 13, 2009 25 / 25