Classification of Solar Wind With Machine Learning

Abstract

We present a four‐category classification algorithm for the solar wind, based on Gaussian Process. The four categories are the ones previously adopted in Xu and Borovsky (2015): ejecta, coronal hole origin plasma, streamer belt origin plasma, and sector reversal origin plasma. The algorithm is trained and tested on a labeled portion of the OMNI data set. It uses seven inputs: the solar wind speed Vsw, the temperature standard deviation σT, the sunspot number R, the F10.7 index, the Alfven speed vA, the proton specific entropy Sp, and the proton temperature Tp compared to a velocity‐dependent expected temperature. The output of the Gaussian Process classifier is a four‐element vector containing the probabilities that an event (one reading from the hourly averaged OMNI database) belongs to each category. The probabilistic nature of the prediction allows for a more informative and flexible interpretation of the results, for instance, being able to classify events as “undecided.” The new method has a median accuracy larger than 90% for all categories, even using a small set of data for training. The Receiver Operating Characteristic curve and the reliability diagram also demonstrate the excellent quality of this new method. Finally, we use the algorithm to classify a large portion of the OMNI data set, and we present for the first time transition probabilities between different solar wind categories. Such probabilities represent the “climatological” statistics that determine the solar wind baseline.

Time‐Integral Correlations of Multiple Variables With the Relativistic‐Electron Flux at Geosynchronous Orbit: The Strong Roles of Substorm‐Injected Electrons and the Ion Plasma Sheet

“…Borovsky, ), while almost all other solar wind variables are strongly affected by compression or rarefaction. The ratio T p / v sw is often used in classification algorithms for the different types of solar wind plasma originating from different types of regions on the solar surface (e.g., Borovsky & Steinberg, ; Camporeale et al, ; Elliott et al, ; Neugebauer et al, ; Neugebauer et al, ; Reisenfeld et al, ; Xu & Borovsky, ).…”

Proton-proton collisional age to order solar wind types

“…The Xu & Borovsky (2015) scheme places three separating planes in the three-dimensional space spanned by Alfvén speed v A , specific proton entropy S p , and the ratio T rel = T p /T exp between the observed proton temperature T p and an expected proton temperature (in eV, 1 ) T exp = v p /258 3.113 depending on the solar wind proton speed v p in km/s. An improved version (Camporeale et al (2017) of the categorization scheme trains a Gaussian process based on the same handselected plasma data as in Xu & Borovsky (2015). For convenience, Xu & Borovsky (2015) also provided an expression for the decision boundaries based on proton speed, proton density, proton temperature, and magnetic field strength.…”

“…From the charge-state information alone, stream interaction regions cannot be uniquely identified, and their solar source regions cannot be unambiguously determined without additional or context information. 2) Proton plasma properties provide an alternative to determine the solar wind type (Xu & Borovsky 2015;Camporeale et al 2017). A clear advantage of this approach is that the required observables are available from more spacecraft.…”

On the Origins of the Intercorrelations Between Solar Wind Variables

“…In the panels of Figure 2 this give rise to a red-green-red temporal pattern, which produces the same intercorrelations of solar wind variables as does the red-green-blue-green-red temporal pattern. A third common switching pattern of the plasma types of the solar wind is between sector-reversal-region plasma and ejecta (Camporeale et al, 2017), with ejecta (Crooker et al, 1998;Foullon et al, 2011;Wang et al, 2000) and sector-reversal-region plasma (Suess et al, 2009;Susino et al, 2008) Figure 3) means that the velocity fluctuations and magnetic field fluctuations are uncorrelated. Note that the coronal-hole-origin plasma (red curve) is the most Alfvenic (median = 0.89), followed by streamer-belt-origin plasma (green curve, median = 0.76), then followed by ejecta (blue curve, median = 0.69), and finally sector-reversal-region plasma is the least Alfvenic (purple curve, median = 0.62).…”