[P-EM14] Heliosphere and Interplanetary Space

[PEM14-07] Identification of solar wind regimes using the non-hierarchical cluster analysis

*Shinichi Watari¹ (1.National Institute of Information and Communications Technology)

Keywords:non-hierarchical cluster analysis, k-means method, solar wind, identification of solar wind regimes

By identifying solar wind regimes, we can get information on their solar sources and situations. Experience is necessary for us to identify the solar wind regimes. We need an objective identification method for this. Richardson and Cane (2004) identified solar wind regimes using heavy ion charge states. However, heavy ion charge states are not always observed by satellites. A method using basic solar wind data (speed, density, temperature, density, and magnitude of magnetic field) is suitable as proposed by Xu and Borovsky (2015).

Here we report identification of the solar wind regimes using one of the non-hierarchical cluster analysis, the k-means method. Data are classified into given number of clusters optimizing the distances of data from the cluster centers by the k-means method. We used speed, temperature, density, and magnitude of magnetic field in the one-hour solar wind data from the OMNI data of the NASA/NSSDC. The data between 1998 and 2019 were applied for our analysis because almost continuous data are available during this period. We also used three parameters, entropy, Alfven speed, and ratio of expected temperature from speed to observed temperature, which were used by Xu and Borovsky (2015).