Probabilistic Forecasts of Storm Sudden Commencements From Interplanetary Shocks Using Machine Learning

Abstract

In this study we investigate the ability of several different machine learning models to provide probabilistic predictions as to whether interplanetary shocks observed upstream of the Earth at L1 will lead to immediate (Sudden Commencements, SCs) or longer lasting magnetospheric activity (Storm Sudden Commencements, SSCs). Four models are tested including linear (Logistic Regression), nonlinear (Naive Bayes and Gaussian Process), and ensemble (Random Forest) models and are shown to provide skillful and reliable forecasts of SCs with Brier Skill Scores (BSSs) of ∼0.3 and ROC scores >0.8. The most powerful predictive parameter is found to be the range in the interplanetary magnetic field. The models also produce skillful forecasts of SSCs, though with less reliability than was found for SCs. The BSSs and ROC scores returned are ∼0.21 and 0.82, respectively. The most important parameter for these predictions was found to be the minimum observed B Z. The simple parameterization of the shock was tested by including additional features related to magnetospheric indices and their changes during shock impact, resulting in moderate increases in reliability. Several parameters, such as velocity and density, may be able to be more accurately predicted at a longer lead time, for example, from heliospheric imagery. When the input was limited to the velocity and density the models were found to perform well at forecasting SSCs, though with lower reliability than previously (BSSs ∼ 0.16, ROC Scores ∼ 0.8), Finally, the models were tested with hypothetical extreme data beyond current observations, showing dramatically different extrapolations.

Timing of the solar wind propagation delay between L1 and Earth based on machine learning

“…6) when the input data is outside of the parameterspace used for training. This finding is in line with Smith et al (2020), who showed that ML based classification of sudden commencements can face misclassifications when applied outside of the trained parameter space.…”

“…In recent years there has been an ever-increasing number of studies in the field of space weather that have made use of ML algorithms. More specifically, these ML algorithms have been particularly successful for the purpose of prediction, including the prediction of CME arrival times based on images of the Sun (Liu et al, 2018), solar wind properties (Yang et al, 2018), geomagnetic indices (Zhelavskaya et al, 2019) and even predictive classification of (storm) sudden commencements from solar wind data (Smith et al, 2020). For an overview on ML applications for space weather purposes we recommend the review by Camporeale (2019).…”

The Impact of Sudden Commencements on Ground Magnetic Field Variability: Immediate and Delayed Consequences

“…Looking to forecasting other significant phenomena, approximately 75% of SCs are preceded by the observation of an interplanetary shock upstream of the Earth at L1 (Smith et al., 2020; Wang et al., 2006), providing a significant amount of warning and the opportunity to forecast the consequences of the shock. Excellent correlations have historically been observed between large geomagnetic storms and interplanetary shocks (Chao & Lepping, 1974; Gosling et al., 1991); while statistically between ∼45% and 60% of interplanetary shocks incident at the Earth being linked to geomagnetic storm activity in the days that follow (Echer & Gonzalez, 2004).…”

Interplanetary Shock Impact Angles Control Magnetospheric ULF Wave Activity: Wave Amplitude, Frequency, and Power Spectra

“…The most geoeffective IP shocks are associated with intense, step‐like variations in the interplanetary magnetic field (IMF) and solar wind parameters (Burguess, 1995). These rapidly increasing conditions abruptly enhance the ambient dayside solar wind ram pressure, which in turn leads to an immediate magnetic field response in the geospace (Kokubun, 1983; Smith et al, 1986) and on the ground (Araki, 1994; Chao & Lepping, 1974; Smith et al, 2020). Such magnetic perturbations are usually associated with spacecraft surface charging (Baker et al, 2017) and geomagnetically induced currents (Pilipenko et al, 2018; Pulkkinen et al, 2017), which can be highly detrimental to human assets in space and on the ground, respectively.…”