G. Atkinson, HOW DOES MAGNETOSPHERIC CONVECTION RELATE TO THE EXPANSION ONSET OF SUBSTORMS, Journal of atmospheric and terrestrial physics, 55(8), 1993, pp. 1151-1157
The suddenness of substorm expansions suggests that they are instabili
ties in the magnetosphere. The convection during the growth phase of i
solated substorms is examined with the intent of determining condition
s leading to the instability. Properties of the large-scale two-celled
convection considered are the intensity of the convection, the size o
f the convection cells and the time rate of change of both of these. T
here is considerable variation in values of the intensity, its rate of
change, and the size of the convection cells just before expansion on
set. On the other hand, onsets frequently occur when the rate of incre
ase of the size of the convection cell is slowing down or has reached
zero. Since the time-scale of this 'slowing down' is an order of magni
tude greater than the time-scale of expansion, the slowing down is not
a result of the instability, but is one of the conditions leading up
to the onset of instability. A number of examples of this behaviour of
the growth of the size of the convection cells are discussed. The fol
lowing interpretation is proposed. The slowing down indicates that the
magnetospheric convection cells are approaching a steady-state size w
hich corresponds to the enhanced dayside merging rate that exists in g
rowth phases. However, the steady-state is unstable against contractio
ns in size and this is the cause of expansions. Triggered substorms an
d substorms without growth phases occur when a decrease in the dayside
merging rate requires a contraction of the cell size. Theories of sub
storms compatible with this behaviour include the ballooning instabili
ty theory and a model in which magnetospheric dipolar convection contr
ols merging in the tail.