HR: 1340h
AN: SM13A-0332    [Abstracts]
TI: Statistical Properties and Time Series Modeling of Dst-Index: Long-Term Space Weather Forecast
AU: * Tsubouchi, K
EM: tsubo@rish.kyoto-u.ac.jp
AF: Research Institute for Sustainable Humanosphere, Kyoto University, Uji, Kyoto, 611-0011 Japan
AU: Omura, Y
EM: omura@rish.kyoto-u.ac.jp
AF: Research Institute for Sustainable Humanosphere, Kyoto University, Uji, Kyoto, 611-0011 Japan
AU: Matsumoto, H
EM: matsumot@rish.kyoto-u.ac.jp
AF: Research Institute for Sustainable Humanosphere, Kyoto University, Uji, Kyoto, 611-0011 Japan
AB: Time series data set of Dst-index from 1957 to 2001 is analyzed for evaluating the occurrence frequency of intense geomagnetic storms (peak Dst < -100 nT). We first investigate the occurrence rate and intensity of storms by means of the peaks-over-threshold (POT) model; which is a method specialized in treating the distribution of very rare (or even never observed until now) events. The 'T-year return level' parameters of the storm strength (max |Dst|), which suggests the event occurring once in T years, are derived from this model. Next, we construct a numerical model of the long-term Dst sequence which is based on its historical statistics. Temporal variation of Dst is designed as a stochastic process, where the variance of one-hour difference Δ Dstⁱ ({Dst}ⁱ⁺¹-Dstⁱ) is assumed to be conditional for time i. Conventional time series modeling, such as GARCH, is applied to show the appearance of a storm-associated steep fall of Dst. The distribution of such an event is compared with the results from the POT model in order to evaluate its validity as a tool for a long-term space weather forecast.
DE: 3270 Time series analysis (1872, 4277, 4475)
DE: 4468 Probability distributions, heavy and fat-tailed (3265)
DE: 7900 SPACE WEATHER
DE: 7924 Forecasting (2722)
DE: 7954 Magnetic storms (2788)
SC: SPA-Magnetospheric Physics [SM]
MN: Fall Meeting 2005