Extreme value statistics in the solar wind: An application to correlated Lévy processes

摘要: [1] The interplay between the solar wind and Earth's magnetosphere is a longstanding challenging problem. An estimate of energy influx into given by Akasofu e parameter. Extreme values this parameter are interest not only for magnetospheric response but also design satellites, space stations, considerations astronaut safety. For time series derived from ACE spacecraft measurements years 2000–2007, we find that its distribution extreme over windows about 18 h longer can be accurately described parametric models based on mathematical theory generalized value statistics. These predict significantly larger than observed to date expected during any 50 year period. While our findings seem suggest correlations and/or nonstationarities do play significant role statistics parameter, show contrary in fact true. To isolate effect finite observation periods, consider maximal changes compare it recently proposed fractional Levy motion-type model. However, motion does reliably capture extremal behavior series.