Substorm cycle dependence of various types of aurora

Abstract

We combine 4861 identifications of substorm onsets from satellite global imagers (Polar UVI and IMAGE FUV) with DMSP particle data to investigate the substorm cycle dependence of the diffuse aurora (electrons and ions), monoenergetic aurora (produced by quasi‐static electric fields), and broadband aurora (associated with Alfvénic acceleration). Although all types of aurora increase at substorm onset, broadband aurora shows a particular association with substorms and, especially, substorm onset. While diffuse electron and monoenergetic auroral precipitating power rises by 79% and 90%, respectively, following an onset, wave aurora rises by 182%. In the first 10–15 min following onset, the power associated with Alfvénic acceleration is comparable to monoenergetic acceleration (also called “inverted V” events). In general, this is not the case before onset, or indeed, during recovery. The rise time of the electron diffuse aurora following onset is much slower, about 50 min, and thus presumably extends into recovery. Discrete acceleration, which rises over just a few minutes, is already deep into decline, while diffuse auroral power is still rising. Most of the sharp increase in Alfvénic precipitating power at onset is associated with a 132% jump in the characteristic energy associated with broadband acceleration. Previous work has associated PiB signatures (which are compressional) in the magnetotail with dipolarization. More recently, FAST observations near the auroral zone have been used to show that there are nearly simultaneous sharp increases in Alfvén waves, apparently created by mode coupling. The present work strongly supports the idea that the Alfvénic aurora has a particular connection to substorm onsets, with much of the auroral power over a substorm cycle concentrated shortly after onset. Finally, a superposed epoch analysis of commonly used coupling functions indicates a drop in the mean solar wind driving starting 20 min before substorm onset. However, the distribution of increases and decreases in solar wind driving is roughly symmetric about a peak at zero (no change) but with (rare) large drops in solar wind driving (“northward turnings”) outweighing large increases. Comparison with the distribution of random changes in solar wind driving leads us to suggest that probably only a minority of substorms are externally driven.