Search

[1] Coronal mass ejections (CMEs) are major solar events that are known to cause large geomagnetic storms (Dst < −100 nT). Isolated geomagnetic storms typically have a main phase of 3–12 hours and a recovery phase of around 1 day. However, there are some storms with main and recovery phases exceeding ∼3 days. We trace the origin of these long-lived geomagnetic storms (LLGMS) to frontside halo CMEs. We studied 37 LLGMS events with Dst < −100 nT and the associated CMEs which occurred during 1998–2002. It is found that LLGMS events are caused by (1) successive CMEs, accounting for ∼64.9% (24 of 37); (2) single CMEs, accounting for ∼21.6% (8 of 37); and (3) high-speed streams (HSS) in corotating interaction regions (CIRs) with no related CME, accounting for ∼13.5% (5 of 37). The long duration of the LLGMS events was found to be due to successive CMEs and HSS events; the high intensity of the LLGMS events was related to the interaction of CMEs with other CMEs and HSS events. We find that the duration of LLGMS is well correlated to the number of participating CMEs (correlation coefficient r = 0.78). We also find that the intensity of LLGMS has a good correlation with the degree of interaction (the number of CMEs interacting with a HSS event or with themselves) (r = 0.67). The role of preconditioning in LLGMS events, where the Dst development occurred in multiple steps in the main and recovery phases, has been investigated. It is found that preconditioning does not affect the main phase of the LLGMS events, while it plays an important role during the recovery phase of the LLGMS events.

[1] It has been shown that Fe charge states greater than and equal to 16 are often observed in the presence of interplanetary coronal mass ejections (ICMEs) at 1 AU. These elevated charge states, representative of hot source regions on the Sun, provide an identifier of ICMEs independent of expansion processes and heliocentric distance. Using this new identifier, we present a comparative study of Fe charge state distributions as a function of latitude and time. We compare Fe charge state data from the ACE spacecraft in the ecliptic plane with that from the Ulysses spacecraft along its orbit from low to high latitudes. We discuss the frequencies of high Fe charge state distributions through a range of latitudes on the approach toward solar maximum. Close to solar maximum, we find almost an order of magnitude fewer hot ICMEs at higher latitudes than in the ecliptic. We offer an explanation for the existence of high Fe charge states in hot ICMEs as a result of magnetic connectivity to flaring regions.

[1] Using magnetic field, plasma, and energetic particle data from Wind and ACE, we analyze interplanetary features associated with the first strongly geoeffective interval in the rising phase of solar cycle 23, which affected Earth on 1–4 May 1998. As shown by Skoug et al. [1999], the configuration consisted of a compound stream made up of an interplanetary coronal mass ejection (ICME) containing a magnetic cloud and being trailed by a hot, faster flow. In addition, we find that the front boundary of the ICME is a rotational discontinuity and the leading edge of the fast stream has a zero normal magnetic field component and is followed by a magnetic field which is strongly enhanced (by a factor of ∼4) and whose fluctuations lie in a plane approximately parallel to the leading edge. Energetic particle and composition observations confirm that the field lines of the magnetic cloud were connected to at least two different flare sites in the same active region. We infer a lower limit for the size of the solar footprint of the connected flux tube of 0.02 Rs2, i.e., ∼1010 km2. A dramatic weakening of the halo electron distribution occurred during 3 May at the time when other experimenters have documented the presence of prominence material. We hypothesize that the solar wind halo population was scattered by enhanced frequency of Coulomb collisions in the dense and very cold plasma. We discuss our energetic particle observations in terms of local acceleration at interplanetary shocks and field discontinuities as well as in terms of acceleration in flares and CME-driven shocks. We also compare, in a specific formulation, the power and energy input of the May 1998 configuration to the magnetosphere with other much studied geoeffective events. We find that the power input far exceeded that in all previous geoeffective events in our sample and attribute this to the fact that the 1–4 May 1998 event consisted of a compound stream structure with an unprecedented power input during a ∼3-hour burst of high-speed flow on 4 May. A statistical survey using the OMNI database for the 6-year period 1995–2000 confirms these inferences and indicates further approximate saturation levels for energy and power input of 10 J m−2 and 0.3 mW m−2, only exceeded in exceptional events such as May 1998 and July 2000. The solar energetic particle event at the leading edge of the fast stream might be the only advance warning the Earth would receive of the approach of a configuration of such a concentrated geoeffective potential.