A Real-Time Technique for Sudden Commencement Detection

Abstract:

Sudden Commencements (SCs) have been defined as rapid changes in Earth’s geomagnetic field in response to solar wind drivers such as shocks in the solar wind. SC’s can be further classified as those that are followed by geomagnetic storms, Storm Sudden Commencements (SSC’s), or those that are not followed by storms, Sudden Impulses (SI’s). In this work we concentrate on the general class of SC’s observed in ground-based magnetometer data that can sometimes provide forecasters with a useful indication of an impending geomagnetic storm. We have created a real-time, computationally inexpensive algorithm that assimilates magnetic field data across a wide range of geomagnetic longitudes to identify an SC. In addition to describing this technique, and potential future improvements, we use the International Association of Geomagnetism and Aeronomy’s (IAGA’s) non-real-time SC event lists as truth for a quantitative assessment of our algorithm’s performance. We compare the IAGA lists of SC’s to current Space Weather Prediction Center forecaster performance as well as the performance obtained with an automated legacy approach developed by Joselyn (1985). Given observations from ground stations well distributed in geomagnetic longitude, the results demonstrate skill scores comparable to forecasters. We argue that forecasters could use these results, in conjunction with additional information that this algorithm does not currently consider (i.e. data from L1), to improve on current warnings and notifications of SCs and geomagnetic storms.