Investigation of High-Frequency Perturbations of the Surface Geomagnetic Field

Abstract

This dissertation presents a comprehensive analysis of high-frequency transient-large-amplitude (TLA) geomagnetic perturbations in ground magnetometer data. TLA events are large (≥ 6 nT/s), rapid (< 60 seconds) magnetic field changes, or dB/dt. This dissertation characterizes TLA signatures and investigates their relation to other space weather events in order to gain insight into the small-scale magnetosphere-ionosphere processes that cause them and may also give rise to GIC. In the initial discovery study, TLA events at five stations of the Magnetometer Array for Cusp and Cleft Studies (MACCS) throughout 2015 were identified. The events were characterized based on amplitude and frequency of occurrence, diurnal trend, and relation to geomagnetic storms, auroral substorms and nighttime geomagnetic disturbance events (GMD), also referred to as nighttime magnetic perturbation events (MPE). We show that TLA events occurred most often at local magnetic nighttime and while TLA events were observed at all five MACCS stations, a majority of individual events (74%) were observed at only one station, inferring a localized spatial scale smaller than ~580 km. The main driver for TLA events in 2015 was not sudden commencements (SC) or sudden impulses (SI) that are the most rapid global-scale space weather events. Rather, TLA events showed stronger association to smaller-scale processes like substorms and nighttime GMDs. The timescales and amplitudes of TLA dB/dt intervals are similar to noisy magnetic signatures caused by external interferences or internal instrumental defects, making detection of only TLA events a tedious and time-consuming manual process. An automated high-frequency magnetic disturbance classifier was developed to identify second-timescale, high-frequency dB/dt intervals in ground magnetic field data and discriminate between noise-type or geophysical TLA events. The full process utilizes insights gained from a statistical analysis of both types of events to implement constraints as well as a machine learning support vector machine to make the final classification of TLA or noise-type dB/dt. This method is a useful capability both for data quality control and the continued investigation of small-scale surface geomagnetic perturbations. Finally, the automated high-frequency disturbance classifier was used to gather a large database of TLA events for all latitude ranges and throughout Solar Cycle 24 from 2009 to 2019. Characteristics of the expanded TLA database show results consistent with the initial study and with nighttime GMDs. TLA event occurrence peaked in the declining phase of the solar cycle and trended similarly with the number of substorm onsets per day. Nearly all of the most extreme GMD events had associated TLA intervals in the same location and while GMDs have an effective radius of ~275 km, TLA events exhibited even more localized spatial scale. From an analysis of a TLA-related GMD event, we show that these events are associated with dipolarization fronts on the nightside at geosynchronous orbit and fast plasma flows toward Earth, and are closely temporally related to poleward boundary intensifications (PBI) and auroral streamers in the ionosphere. The highly localized behavior and connection to the most extreme GMD events suggests that TLA intervals are a ground manifestation of rapid and complex ionospheric current structures coupled to the magnetosphere that can drive GICs. The analysis of high-frequency TLA geomagnetic perturbation events in this dissertation gives new perspective of small-scale magnetosphere-ionosphere (M-I) phenomena that can lead to GIC.