Forecasting the properties of the solar wind using simple pattern recognition

• Published in: Space Weather 2017-03, Vol.15 (3), p.526-540
• Main Authors: Riley, Pete, Ben‐Nun, Michal, Linker, Jon A., Owens, M. J., Horbury, T. S.
• Format: Article
• Language: English
• Subjects: Algorithms; Charged particles; Coronal mass ejection; Euclidean geometry; forecasting; Interplanetary magnetic field; Interplanetary magnetic fields; Magnetic fields; Magnetic properties; Mathematical models; Parameters; Pattern recognition; prediction; Predictions; Solar corona; Solar magnetic field; Solar wind; Space weather; Space weather forecasting; Weather forecasting
• ISSN: 1542-7390; 1539-4964; 1542-7390
• DOI: 10.1002/2016SW001589

An accurate forecast of the solar wind plasma and magnetic field properties is a crucial capability for space weather prediction. However, thus far, it has been limited to the large‐scale properties of the solar wind plasma or the arrival time of a coronal mass ejection from the Sun. As yet there are no reliable forecasts for the north‐south interplanetary magnetic field component, Bn (or, equivalently, Bz). In this study, we develop a technique for predicting the magnetic and plasma state of the solar wind Δt hours into the future (where Δt can range from 6 h to several weeks) based on a simple pattern recognition algorithm. At some time, t, the algorithm takes the previous Δt hours and compares it with a sliding window of Δt hours running back all the way through the data. For each window, a Euclidean distance is computed. These are ranked, and the top 50 are used as starting point realizations from which to make ensemble forecasts of the next Δt hours. We find that this approach works remarkably well for most solar wind parameters such as v, np, Tp, and even Br and Bt, but only modestly better than our baseline model for Bn. We discuss why this is so and suggest how more sophisticated techniques might be applied to improve the prediction scheme. Key Points Solar wind parameters can be predicted with lead times of up to several weeks IMF Bn, arguably the most important space weather parameter, is also the least predictable Dynamic time warping may lead to improvements in the forecasting abilities of these algorithms