More observations of the magnetosphere under conditions of low-density solar wind
As discussed in this space a few weeks back, in mid-May 1999 the density of the solar wind dropped well below 1 proton per cubic centimeter for a period of almost two days. The reduction of the ram pressure on the Earth’s magnetosphere caused it to balloon to several times its normal linear dimensions and become much more symmetrical. During the period of low solar wind density, the orbit-averaged fluxes of 1.5-3.5 MeV electrons observed by the SAMPEX PET instrument dropped significantly, then recovered as the solar wind density rose toward normal; then about a day later they dropped again to low levels, and remained there for months despite conditions in the interplanetary medium that might have been expected to energize the outer-zone relativistic electrons on several occasions during this period. These observations were reported by Dan Baker et al. at the Fall 1999 meeting of the AGU, in a special session dedicated to this event.
To learn more about the effects of periods of low solar-wind density on the outer-zone electrons, we looked at other SAMPEX observations during this period, and also during several other periods since 1996 (identified in the Proton Monitor dataset from the CELIAS/MTOF instrument aboard SOHO, courtesy of F. Ipavich) when solar wind density has dropped below 1 proton per cc for periods from a few hours to almost two days. First, we split the SAMPEX data set into trapped particles and particles in the “drift loss cone”, the subset of observations where particles drifting around the earth can be expected to be absorbed by the atmosphere on a timescale comparable to their drift periods; such particles must therefore have been scattered from the trapped population along the field lines into the loss cone fairly recently. The upper panel of the accompanying plot shows an orbit-by-orbit “spectrogram” of 1.5-3.5 MeV electrons in the loss cone, with color denoting intensity on a plot of L (distance from Earth in the equatorial plane, measured in Earth radii) vs. time for the month of May 1999. The red bars label the period when the solar wind density was below 1 proton per cc; the flux dropout during this period is easily visible, indicating that the plasma waves that usually scatter electrons into the loss cone became very quiet. In fact, most of the reduction in orbit-averaged fluxes due to this event (as reported earlier) was due to this emptying of the loss cone; the trapped electrons remained steady. We also looked at electron observations at other energies; the bottom panel shows a similar L vs. time “spectrogram” for electrons above 0.4 MeV in the loss cone. Note that while this energy channel shows the dropout during the period of low solar wind density, it does not show a significant weakening of the radiation belt comparable to that seen in the higher-energy channel starting on about May 13. Thus the long-lasting flux dropout was a high-energy effect; we have also confirmed this with observations from POLAR and several geostationary and high-inclination satellites, as reported by Dan Baker and myself, et al., at the Fall AGU meeting. Finally, though there were almost a dozen other extended periods of low solar wind density since 1996, none of the others showed any particular effects, either prompt or delayed, on the electron fluxes. Thus this event, with its prompt dropout at all energies and the possibly related long-lasting decline in high-energy fluxes a day or so later, is unique in our observations to date.
Contributed by Mark Looper, The Aerospace Corporation «Return to the Results page