Space climate implications from substorm frequency

Abstract

AbstractThe solar wind impacting the Earth varies over a wide range of time scales, driving a corresponding range of geomagnetic activity. Past work has strongly indicated that the rate of merging on the frontside magnetosphere is the most important predictor for magnetospheric activity, especially over a few hours. However, the magnetosphere exhibits variations on other time scales, including UT, seasonal, and solar cycle variations. Much of this geomagnetic variation cannot be reasonably attributed to changes in the solar wind driving—that is, it is not created by the original Russell‐McPherron effect or any generalization thereof. In this paper we examine the solar cycle, seasonal, and diurnal effects based upon the frequency of substorm onsets, using a data set of 53,000 substorm onsets. These were identified through the SuperMAG collaboration and span three decades with continuous coverage. Solar cycle variations include a profound minimum in 2009 (448 substorms) and peak in 2003 (3727). The magnitude of this variation (a factor of 8.3) is not explained through variations in estimators of the frontside merging rate (such as dΦMP/dt), even when the more detailed probability distribution functions are examined. Instead, v, or better, n1/2v2 seems to be implicated in the dramatic difference between active and quiet years, even beyond the role of velocity in modulating merging. Moreover, we find that although most substorms are preceded by flux loading (78.5% are above the mean and 83.8% above median solar wind driving), a high solar wind v is almost as important (68.3% above mean, 74.8% above median). This and other evidence suggest that either v or n1/2v2 (but probably not p) plays a strong secondary role in substorm onset. As for the seasonal and diurnal effects, the elliptical nature of the Earth's orbit, which is closest to the Sun in January, leads to a larger solar wind driving (measured by Bs, vBs, or dΦMP/dt) in November, as is confirmed by 22 years of solar wind observations. However, substorms peak in October and March and have a UT dependence best explained by whether a conducting path established by solar illumination exists in at least one hemisphere in the region where substorm onsets typically occur.