In the article, New Findings about Auroras Confirm Importance of Ionosphere in Space Weather (Eos, December 22,1998), the author states that imaging riometers respond mainly to keV electrons. It is more correct to say that they respond mainly to electrons of 10s to 100s keV. The absorption frequently maximizes at about 90 km altitude, and an electron giving maximum production at 90 km must enter the atmosphere with energy about 50 keV. The evidence comes from observations by rocket and by incoherent-scatter radar, together with direct comparisons between radiowave absorption in the auroral zone and electron spectra measured at geosynchronous orbit.
Yamagishi et al.  are quoted as finding that absorption events are far more common in winter than in summer. I have not been able to find this statement in the Yamagishi paper; moreover, conjugate-point studies show that near the maximum of the absorption zone the events are almost invariably conjugate that is, they are seen in both hemispheres. They do, however, tend to be more intense in the winter hemisphere. Yamagishi et al.s Table 1 gives the numbers of events that are more intense in one hemisphere or the other, and those results are not inconsistent with the earlier work on this point [Hargreaves and Cowley, 1967; Hargreaves, 1969]. Osepian and Smirnova  argue that the seasonal difference may be explained by seasonal changes in the effective recombination coefficient because of variations in the properties of the neutral air.