Pitch angle evolution of energetic electrons at geosynchronous orbit during disturbed times

1
Pitch angle evolution of energetic electrons at
geosynchronous orbit during disturbed times

• R. Friedel, Y. Chen, G. Reeves, T. Cayton
• ISR-1, Los Alamos National Laboratory, USA
• Yuri Shprits
• University of California, Los Angeles, USA

2
Contents

• Rationale
• Geosynchronous pitch angle distributions
• Instrumentation and Data
• Drift shell splitting example and explanation
• Mapping to constant L 6.5
• Assumptions
• Demonstration of method
• Quiet time test of method (10-13 December 2002)
• Application to small relativistic electron event
• August 2-5, 2002
• Theoretical predictions
• Summary/Conclusion

3
Rationale

• Energetic electron pitch angle distributions show
  clear local time variations due to the asymmetry
  of the Earths field Drift shell splitting
  L f (pitch angle)
• These geometric effects may mask the changes
  that may be due to in-situ acceleration or pitch
  angle scattering processes.
• We remove here the geometric effects by mapping
  the observed pitch angle distributions to a fixed
  3rd (L,F) adiabatic invariant preserving the 1st
  (µ) and 2nd (K,J) adiabatic invariants.

4
Geosynchronous Instrumentation and Data

• Data is presented from the relativistic electron
  channels of the LANL SOPA instrument 50 keV
  1.5 MeV.
• As there is no magnetometer on the LANL GEO
  spacecraft, the magnetic field direction is
  inferred using the MPA plasma measurements by
  deducing the symmetry axis of the pressure tensor
  (Thomsen et al, 1996).
• GEO spacecraft have a 10 sec spin period. SOPA
  data sampling is at 0.16 seconds. Data is
  collected in 32 azimuthal bins averaged over 10
  minutes.
• Pitch angle resolved GEO data is available for
  LANL-97a, 1991-080 and 1990-095 for most of Jul
  2002 Dec 2003.

5
Geosynchronous Orbit Drift shell splitting
example and explanation
6
Mapping to L6.5Assumptions
Drift shell splitting at geo orbit leads to
observations over L 6 7. We map observations
to a fixed L 6.5 at constant µ (1st) and K
(2nd) invariant using the following assumptions

• Phase space density gradients near GEO are flat
  or small (SCATHA Fennell and GEO/Polar Chen
  observations).
• Over this small range of L we can approximate
  our µ mapping using a dipolar
  approximation.
• The change in the mapping of K to pitch angle is
  over this range of L is negligible.

7
Mapping to L6.5Demonstration of method near
midnight
8
Mapping to L6.5Map at constant K or pitch
angle?

• We tested the change in the 2nd invariant K to
  pitch angle mapping at a variety of local times
  for L6 and L7, the maximum mapping needed in
  this study.
• Changes in pitch angle at constant K are lt 3 Deg,
  which below our 10 deg pitch angle resolution.

Near geosynchronous orbit we thus are safe to
map at constant pitch angle.
9
Pitch Angle MappingQuiet time test of Method -
Olson Pfitzer Model
10
Pitch angle MappingSmall rel. electron Event
August 2-5, 2002
11
Evolution of Pitch Angle Distributions
Modeling effects of Waves - UCLA

• Addition of EMIC waves leads to rapid loss of
  electrons at energies down to 0.5 MeV
• Higher pitch angles are affected for higher
  energies
• lt 60o 1 MeV
• lt 30o 400 keV

12
LANL GEO Pitch Angle observations at all
energies 50 keV 1.5 MeV

• Losses to lt60o for 1MeV
• Losses become less severe as energy decreases
• Observations are roughly consistent with EMIC
  theory and modeling

13
Summary / Conclusion

• Pitch angle mapping seems to work and can reveal
  the real PA distribution dynamics
• We show that method works in principle however
  many assumptions probably violated during very
  active periods
• For the week relativistic electron event of
  August 2-5, 2002
• the pitch angle distribution seems to show
  evidence of acceleration processes (-gt peaked at
  90o)
• The loss period at the end of the event is
  clearly associated with cold dense plasma and
  losses are due to precipitation -gt field aligned
  electrons vanish
• Association with EMIC waves Detailed evolution
  modeling of PA distributions shows roughly
  consistent behavior with data.