The top panel shows the calculated electron flux, colour coded as a function of time and L*. The dashed white line shows the location of the GOES satellite at geosynchronous orbit. The solid white line shows the location of the outer boundary of the geomagnetic field.

The second panel shows a comparison between the model and the observations by GOES (for integrated electron flux only).

The third panel shows the >10 MeV proton flux. GOES electron data are unreliable when the proton flux exceeds the red line.

The fourth panel shows the solar wind velocity in red and the z component of the interplanetary magnetic field in black. Fast solar wind usually ‘pumps up’ the radiation belts. When IMF Bz is negative energy is transferred from the solar wind into the geomagnetic field.

The fifth panel shows the Dst index, colour coded, which is a measure of geomagnetic storms, and the solar wind dynamic pressure. Strong pressure usually pushes the outer boundary of the magnetic field inwards.

The bottom panel shows the Kp index, colour coded, and the AE index. Kp is a measure of geomagnetic activity and AE is a measure of substorm electron injections.

Internal charging:  High energy electrons can penetrate the outer layers of a satellite and accumulate in insulators such as cables and dielectrics on circuit boards.  If the charge builds up faster than it can leak away it can cause an electrostatic discharge and break down the material permanently.  This has led to loss of service and in some cases total satellite loss.  The periods most at risk correspond to periods of high electron flux.


Surface charging:  Bursts of low energy electrons can accumulate on surfaces, particularly when the spacecraft is in the Earth’s shadow and there is no photoelectron emission.  Since only half the spacecraft can ever be in sunlight large electric fields can build up across the satellite and cause an electrostatic discharge.  Surface charging has led to loss of solar array strings.  The periods most at risk correspond to periods of substorm injections when the satellite is in eclipse.


Single event upsets: High energy ions can penetrate electronic components and deposit charge in them.  This can corrupt memory circuits.  High energy ions can also dislodge ions in the crystal structure causing increased noise and degradation of performance.  Periods most at risk are during solar energetic particle events when the number of high energy protons can increase a thousand fold.  This has led to repeated memory corruption over a period of a few days and up to 2% loss of solar array power.

The BAS radiation belt model is used to provide a forecast of the 2 MeV electron flux in the outer Van Allen radiation belt.  The model uses data from the ACE satellite and a forecast of geomagnetic activity to make the forecasts.  The model solves a diffusion equation that takes into account the transport of electrons across the magnetic field towards and away from the planet (radial diffusion), electron acceleration by wave-particle interactions, electron loss into the atmosphere by wave-particle interactions and collisions with atmospheric gasses.  Changes in the interplanetary magnetic field and the solar wind dynamic pressure are used to determine the outer boundary of the Earth’s magnetic field which affects radial transport.  The injection of low energy electrons during substorms is represented by changes in the electron flux at the low energy boundary and by scaling the wave power by geomagnetic activity.  Three types of wave-particle interactions are included in the model.


The flux of electrons can vary significantly around the geostationary arc.  The GOES satellites only provide a snapshot at one or two locations.  The model can be tailored to other locations.


Key features are

  • The forecasts cover the three main orbit types where most commercial satellites fly
  • The forecasts are based on a physical model
  • The forecasts include the physics of wave-particle interactions
  • The forecasts can be tailored to specific satellites at GEO or other orbits.

For more details on the model see

Glauert et al. J. Geophysical Res., [2014]:

Horne et al. Space Weather, [2013]: