Exploring of the ICMEs

1. Introduction

Coronal mass ejections (CMEs) are most dynamic global phenomena in solar atmosphere, and they are of great important because his influents on the interplanetary space and Earth magnetosphere. Traveling trough interplanetary space ejection dramatically changes stage of the solar wind and on arrival to the Earth, strong magnetic field frozen in CMEs creates the geomagnetic storms and depression in the cosmic rays (CRs) intensity. Consequences of these phenomena contribute to creation a new field of astrophysical-geophysical researches with a joint name Space Weather, current one of the most attractive part of the solar physics. Successful forecasting of these large disturbances which demands much better knowledge of their causing/driving processes would decrease negative performance on the space-borne as well as ground-based technological systems and decrease dangers for human life and health. Interplanetary counterpart of the coronal mass ejections is usually referred as ICME.

The attenuation of the CR flux due to passing of the ICME, nowadays referred as a Forbush decrease (FD), is dependent on the dynamic and topology of the CME, as well as the strength and orientation of the CME magnetic field and pre-shock conditions of interplanetary magnetic field (IMF). All these parameters are rather difficult to measure and therefore, the explanation of the FD mechanisms still lacking many details. Vice versa, in terms of understanding the internal magnetic topology of the CMEs in the interplanetary medium, modulations of the CRs should provide valuable information which cannot be obtained by any other type of in situ measurments. Thus, gaining better knowledge of the influence of CMEs on the modulations of the CRs, i.e. we improve prediction of the geomagnetic and radiation storms (hours before the arrival of CMEs at 1 AU). On this way, this project consolidating examinations of the physical processes in the solar atmosphere, interplanetary space and the Earth atmosphere. We will in particular make comprehensive use of the data harvested from the ground based Space Environmental Viewing and Analysis Network (SEVAN) particles detectors; as well as WIND, Solar and Heliospheric Observatory (SoHO), Solar TErrestrial RElations Observatory (STEREO) and Solar Dynamic Observatory (SDO) spacecrafts, combined with various other space-borne and ground-based data.

2. CME overview

Coronal mass ejections (CMEs) are the most violent activity signatures from our Sun. They appear as large-scale outward moving clouds of magnetized plasma expelled from the Sun to interplanetary space with velocities from only few kms-1 to nearly 3000 kms-1 (Gopalswamy, 2006). CMEs originate from active, filament or complexes regions and carry up to 1016 g of corona material. They are associated with a number of phenomena such as flares, solar energetic particles, prominence eruption, waves, shocks, radio bursts etc. CMEs have been observed for over four decade with a wide variety of instruments and are most often observed in photospheric white light scattered by coronal electrons (Thomson scattering; e.g., Gosling et al., 1974). A comprehensive books, reports and current theoretical overviews of CME are recently published (see e.g. Kunow et al., 2007; Alexander, Richardson, and Zurbuchen, 2006; Schwenn et al., 2006; Hudson, Bougeret, and Burkepile, 2006; Gopalswamy et al., 2006; Forbes et al., 2006). From this reports and quantitative studies we can see that our current knowledge of the physical characteristics and dynamics of CMEs are still limited and not well understood.

3. CME morphology

CMEs observed in white light and EUV are highly structured. The basic CME morphology has its roots in the pre-eruption closed magnetic field configuration (Low, 1996, and references therein). The basic magnetic pre-eruption structure of the solar atmosphere consists closed field regions, which are typically comprised of active or quiescent regions occasionally with filament which delineate opposite polarity patches. The filament is usually found in a coronal “void” enclosed by the closed field structure that is surmounted by the coronal streamer (cf., Engvold, 1987, and references therein). White-light observations of pre-eruption CMEs structure show a wealth of different morphologies. Many CMEs, especially the ones originating from active regions with filament, frequently show a three-part structure, traditionally called the bright frontal rim, the dark cavity, and the prominence (cf., Hundhausen, 1987). However, even in prominence related CMEs, the three-part appearance depends on the location of the underlying prominence (Cremades and Bothmer, 2004). If the CME moving faster than local Alfvenic speed, it can be expected to drive a shock ahead of it. Some structured CMEs are not three-part but two-part, being composed only of a bright frontal rim and a cavity, with no prominence in them. Jets and narrow CMEs (faint linear structures; widths <15˚) with no resemblance to the three-part structure have also been observed (Howard et al., 1985; Gilbert et al., 2001;Wang and Sheeley, 2002b; Yashiro et al., 2003). Therefore, morphologies of the CMEs ranging from amorphous blobs to simple narrow jet-like features, up to highly structured and complicated entities. The three-part structure is seen in only about 30% of CMEs (Gopalswamy et al., 2006), yet this is viewed as the “standard CME” configuration in observational and theoretical studies. In the three-part structure CMEs frontal structure is coronal material, the cavity also is coronal, but may have higher magnetic fields and lower density, while the bright core is the eruptive prominence. The circular pattern observed within the cavities of many CMEs suggests the existence of helical magnetic features (Chen et al., 1997; Dere et al., 1999; Wood et al., 1999), commonly known as flux ropes – used extensively by CME modelers.

Fig. 1 SOHO/LASCO C3 imagse show a different morphology of the CMEs a)three part structured CME; b) two part CME; c) flux rope CME and d) narrow, jet-like CME

Based on the speed spectrum of the CMEs, there are CMEs which rise gradually with slow speed (< 100 kms-1) over a period of severa hours and reach a terminal speed between 300 – 400 kms-1, which is equivalent to the speed of the slow solar wind at the 20 solar radii (Srivastava et al. 2000). Thise CMEs are generally associated with eruptive prominence. Futhermore, at the other end of the speed spectrum are fast events (speed > 600 kms-1) which undergo maximum acceleration in the lover corona and are generally associated with flares (Zang et al. 2004).

4. CME kinematics

Recognition of the processes that govern the eruption of the CMEs is the subject of many analyses, and many various models exist to explain these (REF). The common stand is that magnetic reconnection is responsible for the destabilization of the pre-eruption CMEs structure on the Sun, which than erupt trough the interplanetary medium into the solar wind to form the CMEs. During the CMEs propagation trought the interplanetary medium some of its parameters may greatly change. The variation of the CMEs propagation parameters could significantly influence the CMEs geoeffectiveness. For example variation of the propagation direction would determine whether a CME could arrive on Earth, or alterations of CME’s velocity may change the CME’s arrival time. Thus, the analise of the kinematic evolution of the CME in the low corona and the interplanetary medium is an important topic in space weather study. It has been reported by many authors, that the CMEs undergo deflection towards the equator during its propagation in the corona (REF).  The causes for this deflection are still debated. Scientices suggested that two factors would cause this deflection of the CME: (i) the initial magnetic polarity of the CME structure relative to the background magnetic field can influences its trajectory (MacQueen, Hundhausen, and Conover, 1986); (ii) the fast solar wind flow from polar CHsthat encompassed the CMEs’ expansion at higher latitude (Cremades, Bothmer, and Tripathi, 2006). Another important factor that would influence the CME’s space weather effect is the kinematic evolution of the CME, especially when the CME propagates in the low corona where the major acceleration of CMEs take place (Zhang et al., 2004, Maričić et al., 2004). Earlier results showed that majority CMEs usually undergo a multiphased kinematic evolution, including:

(i)  an slow rising motion (eruption is preceded by a swelling of the pre-eruptive CME structure, enhanced internal motions, and a slow rise with velocities on the order of 10 kms−1, this phase may last for hours (e.g., Low, 1982; Rompolt, 1990; Vršnak,1998));

(ii)  an acceleration phase (acceleration phase usually starts when the height of the structure becomes comparable with the footpoint half-separation (Vršnak, Ruždjak, and Rompolt, 1991; Chen and Krall, 2003), and is often (but not necessarily) characterized by an exponential-like development (Shanmugaraju et al., 2003). The acceleration maximum is attained within a distance of several solar radii (Chen and Krall, 2003));

(iii)  finally a post-acceleration phase (CME showing approximately constant velocity) [Zhang et al., 2001, 2004).

The slow rising motion and acceleration phases occur mainly in the inner corona (< 2.0 Solar radii of higher), while the post-acceleration phase is largely in the outer corona (>2.0 Solar radii). However, not all CMEs necessarily display a full three phase evolution. The CME lifting can be so impulsive that in extreme cases, the acceleration may be so large that the CME achieves a velocity on the order of few 1000 kms−1 (e.g., Vršnak, 2001; Zhang et al., 2001; Cliver et al., 2004). On the other hand, there are very gradual events, whose acceleration lasts for several hours, never exceeding 100 ms−2 (Zhang, 2005; Maričić et al., 2007; Zhang and Dere, 2006). A class of gradual CMEs characterized by a very weak but long duration acceleration has also been reported by Sheeley et al., 1999 and  Srivastava et al., 1999 and 2000. The fastest events are usually related to an impulsive acceleration, but not necessarily, since occasionally fast events show a gradual acceleration (Vršnak et al., 2005). Apparently, impulsive events are usually launched from active regions, whereas the events launched from quiet regions are usually more gradual (Andrews and Howard, 2001). Based on the different acceleration properties, Zhang et al. (2004) suggested that there were three classes of CMEs: (i) impulsive acceleration CMEs; (ii) intermediate acceleration CMEs; (iii) gradual acceleration CMEs.

Furthermore, the relative kinematics between different parts of the CME is still poorly known. Studies addressing this aspect of CME dynamics are rare, since it is generally difficult to identify and measure the overlying frontal rim during the initiation of a CME. Such analyses offer an exceptionally important input for the modelling of CME dynamics (Schmahl and Hildner, 1977; Fisher, Garcia, and Seagraves, 1981; Illing and Hundhausen, 1985; Low and Hundhausen, 1980; Maričić et al., 2004; Plunkett et al., 1997; Dere et al., 1999; Wood et al., 1999; Srivastava et al., 2000; Plunkett et al., 2000; Forbes, 2000).Works on CMEs often treated the prominence eruption as a secondary process during the CME because prominence may not have enough energy to drive CMEs (Hundhausen 1999; Smith, Hildner, & Quin 1992). However, there are alternative viewpoints suggesting a primary role for the prominences: Filippov (1998) has shown that CMEs can be caused by the eruption of inverse-polarity prominences. A comparison of the height, speed, and acceleration time profiles of different components of CME provides an insight into the initiation process and the dynamics of CME until late phases of eruption.

Fig. 2 Coronal mass ejection of 16 February 2003. Top: Sequence of a) EIT FeXII 195 Å running difference image; b) MK-IV; c) LASCO-C2 and d) LASCO-C3 image. In all images the cross marks the measured position of the frontal rim. Below: Kinematics: a) complete distance-time b) the acceleration phase data and the smoothed R(t) curve; c) the velocity time-profile, v(t), d) the acceleration time-profile, a(t). (put the better kinematics measurement with STEREOs)

At larger distances from the Sun, a weak deceleration/acceleration is sometimes measured (Vršnak et al., 2004a; Yashiro et al., 2004). It is expected that, moving trough IP space CMEs interacts with the solar wind and interplanetary magnetic field. The interaction of CMEs and the solar wind plasma is under favourable conditions “visualized” by type II radio burst, which are excited at the IP shocks that are driven by CMEs (Gopalswamy et al. 2001a).  Studies of the CMEs velocity trough the corona show that slow CMEs are accelerated towards the speed of the solar wind and fast CMEs deccelerated. It has been proposed that deceleration/acceleration is presumably caused by effect of the aerodynamic drag on the CME in the solar wind (Gopalswamy et al., 2001; Vršnak et al. 2004a, Byrne et al. 2010).

However, the quantification of the CME expansion trough interplanetary magnetic field and solar wind is currently lacking, mostly because of the limits of the coronagraph observations from the single viewpoints.  The projected against the plane of sky, two dimensional nature of this observations introduces uncertainties in kinematical and morphological analyses, and therefore the 3D structure and evolution of CMEs are still difficult to resolve (Howard et al., 1985; Hundhausen, 1993; Webb and Howard, 1994; Sheeley et al., 1999; St. Cyr et al., 2000).

Efforts to infer internal structure of the CMEs were made with two-dimensional images recorded by the Large Angle Spectroscopic Coronagraphs (LASCO; Brueckner et al., 1995) aboard Solar and Heliospheric Observatory (SOHO Delaboudiniere et al., 1995; Brueckner et al., 1995), situated at the first Lagrangian point (L1). These efforts were based on a comparison of observations with in situ and on-disk data or assumed geometry of the CMEs (Hundhausen, 1993; Sheeley et al., 1999; St. Cyr et al., 2000). In such a coronagraph observations, which image CMEs in projection against the plane of sky, observational CMEs parameters such as the angular width, the structure, propagation direction, quantities such as radial distance, velocities and accelerations are biased by projection effects (see, e.g., Burkepile et al., 2004; Schwenn et al., 2005). The extent of projection mainly depends on the location of the CME source region on the solar surface. Thouse, to determine the  “true” (i.e., de-projected) CMEs parameters we need to know the location of their source regions on the Sun. Recently, observations from the Solar Terrestrial Relations Observatory (STEREO; Kaiser et al., 2008), launched by NASA in 2006, provide us with unprecedented opportunities to investigate the changing dybamic and 3D morphology of the CMEs. STEREO consists of two near-identical spacecrafts, positioned ahead (A) and behind (B) the Earth on its orbit around the Sun. The spacecrafts drift away from the Sun – Earth line at a rate of ±22 ° per year. Thus, STEREO observes the Sun from two different vantage points (different from that of LASCO coronagraphs)), provides a unique twin perspective of the Sun and inner heliosphere, and enables the implementation of a variety of methods for studying CMEs in 3D. Many CME studies have used tie-pointing techniques with the COR1 and COR2 coronagraphs of the Sun-Earth Connection Coronal and Heliospheric Investigation (SECCHI; REF) aboard STEREO. The additional use of SECCHIs Heliospheric Imagers (HI1/2; REF) allows a study of CMEs out to distances of 1 astronomical unit. Applying differente methods and tehnique to every image observed by SOHO, STEREO, recently launched (February 2010) Solar Dynamic Observatory (SDO; O’Dwyer et al., 2010) missions as well as ground based the Mark-IV K-coronameter of the Mauna Loa Solar Observatory (MLSO; Bruno et al., 1984), the CME propagation direction and the source region can be determined, which enables us to investigate the changing dynamics and morphology of the CMEs and remove, at least partly, projection effects from the kinematics. This is of great importance, since in particular fast and ICMEs may cause strong geomagnetic disturbances (Zhang et al., 2003; Michałek, Gopalswamy, and Yashiro, 2003; Webb et al., 2006). Summarising, we note that with reliable estimations of true CME speeds and propagation directions, as well as  the CME evolution in the solar wind, we are poised to increase the accuracy of forecasting transit times and achieve unprecedented success in space weather prediction.

5. CMEs geospehre consecvence and cosmic rays

The relationship between geomagnetic disturbances and solar activity was recognized in the 19th century (Sabine, 1852; Carrington, 1860). Currently we know that, coronal holes (CHs) and CMEs are essentially the two solar phenomena responsible for most disturbances in the interplanetary space. CHs contain open magnetic field lines that carry high speed solar wind streams, while CMEs originate from closed field regions on the Sun. Furthermore, the recurrent geomagnetic storms due to CHs are generally weaker than the non-recurrent geomagnetic storms caused by the CMEs. If CME is super-Alfvenic it can create MHD shocks, which accelerate the electrons, protons and ions of the solar wind to very high energies. The accelerated heavier ions and protons are known as solar energetic particles (SEPs). Furthermore, on the way to the Earth CMEs also modulate the intensity of cosmic rays (CRs), introducing anisotropy and changing energy spectra of the previously isotropic population of protons and stripped nuclei accelerated in the numerous galactic sources. Changes in the rather stable flux of CR are detected by space-born spectrometers (rigidities up to ~1GV) and by world-wide networks of particle detectors (rigidities up to ~100GV) located at different latitudes, longitudes and altitudes.

Therefore, three primary consequences of the CMEs are geomagnetic storms (depending on the CMEs dynamic and morphology), SEPs (depending on the energy of CMEs) and modulation of the CRs (depending on the size, kinematics and magnetic field strength of CMEs).

Geomagnetic storms originate when the shock and CMEs magnetic field entering the Earth magnetosphere and reconnect. Since the geomagnetic field is directed northward, the CME magnetic field needs to have southward component (-Bz) for reconnection (Valtonen, 2007 and references therein). After reconnection, southward directed field are transfer to the night side, where they reconnect once more, driving the plasma form the tail region deep toward the Earth. Some of this energetic particles cause aurora. During geomagnetic storms, particles in magnetosphere can be accelerated to levels that they become hazardous to astronauts and space borne technological systems, as well as on the ground currents can also be induced, which may disrupt a pipelines, railroads and electric power transformers.

Sun modulation of the CRs is observed as: a) a daily variation; b) a 27- day variation; c) an anti-correlation with 11-year solar activity; d) a few days decreases due to passing of CME and d) flare generated CR increases that last only a few minutes. The attenuation of the CR intensity due to passing CME is referred as Forbush decrease (Fd). When the sun ejects a huge amount of the material and magnetic field in to interplanetary space, the CRs interact with this plasma/field and his intensity decrease. Mainly, attenuation takes less than a day to reach the minimum and recovery back to the previous intensity values for a several days after the CME. Most observed attenuations have two components i.e. two steps. First step is due to the shock wave created in the medium ahead of the fast CME material, while the second step is due to the CME body. Cane (2000) explains the continuation of the CME in the interplanetary space and indicates that according to position of the observer, CR decrease may be observed with both shock and CME, and also only with the shock or only with CME. Emphasize also, if the CME is connected with flare, before the FD, CR monitors usually detected a sharp increase do to solar protons generated by the flares, referred as ground-level enhancements (GLEs). Although, there is no clear relationship between flares and the CMEs (Zhang et al. 2001; Temmer et al., 2008; Temmer et al., 2009; Temmer et al., 2010). Ther is a only statistical tendency that CME with stronger accelerations are associated with more powerful flares (Maricic et al., 2007).  Furthermore, FD is dependent on the speed and size of the CME, the magnetic field strength and orientation of the CME and pre-shock conditions of interplanetary magnetic field (IMF) (Chilingarian and Bostanjyan,  2010). All these parameters are rather difficult to measure; therefore, the explanation of the FD mechanisms still lacking many details.

The size and magnetic field strength of CMEs are correlated with the CME modulation effects on the energy spectra and direction of CRs. At the same time the presence of strong and long-duration southward magnetic field (-Bz) in CMEs is the primary requirement for their geoeffectiveness. Thus, strong magnetic field frozen in CMEs is both modulation agent of CR and driver of geomagnetic storms. Although there is no one-to-one dependence between the variations of the CR and the strength of geomagnetic storm (Kudela and Brenkus, 2004) and there exist other drivers of storms and modulation agents of CRs, the large Bz value associated with approaching CMEs is best known diagnostics of geomagnetic storm strength. Appropriate observations of the variations of the primary and secondary cosmic rays can be a proxy of Bz value available long before CMEs reach the L1 libration point where Bz is measured directly (see e.g., Kudela and Storini, 2006).

The estimated arrival time of the CMEs at the Earth is an important input for forecasting the time of occurrence of the resulting geomagnetic storms. CMEs dynamic properties show that ICMEs arrive at 1 AU within 5 days of their departure from the Sun (Gopalswamy et al., 2000a). Based on the estimation of the CME acceleration an empirical model has been developed to predict arrival time of the CMEs at 1 AU (Gopalswamy et al., 2001b). Although this model cannot tell us about the magnetic field orientation that the CMEs impose when intercut with magnetosphere. It can provide a simple means of advance warning or the arrival of CMEs. Ground based particle detectors measure time series of secondary particles born in cascades originating in the atmosphere caused by primary ions and solar neutrons. The networks of particle detectors can predict upcoming geomagnetic and radiation storms hours before the arrival of CMEs at the Advanced Composition Explorer (ACE) and SOHO spacecrafts. The less than one hour lead time (the time it takes for the CME to travel from the spacecraft to the magnetosphere), provided by particle detectors located at ACE and SOHO at the libration point 1,5 million kilometres from the Earth, is too brief to take effective mitigating acti

Fig.3 left: the basic detecting unit of the middle to low latitude particle detectors network called SEVAN (Space Environmental Viewing and Analysis Network); right: posible locations of the SEVAN detectors.

ons to protect satellites and surface industries from the harm of major geomagnetic storms. Therefore, measurements of secondary fluxes can be used for probing CMEs, providing highly effective information on the key characteristics of these interplanetary disturbances. The Space Environmental Viewing and Analysis Network (SEVAN) particle detectors measure neutral and charged fluxes coming from different directions by particle detectors located at 5 different altitudes (see Chilingarian, et al., 2005; Gevorgyan, et al., 2005; Chilingarian et al., 2003a and 2003b). One of these monitors was installed in February 2009 in the Zagreb Observatory (Croatia). Each species has different most probable energy of primary parent proton/nuclei. Particle detectors operate at energy range from 7 to 200 GeV. Therefore, from SEVAN particle detectors data we can estimate the CR energy range affected by CME and reconstruct actual spectra of the cosmic rays incident terrestrial atmosphere.


The Fd magnitude measured at SEVAN particle detectors during the 23rd solar activity cycle ranges from about 1.5% to 20% in the secondary neutron flux, 1-15% in the charged low-energy particle flux and 0.6-6% in the >5 GeV muon flux (Bostanjyan and Chilingarian,  2009). The modulation strength of the CMEs is highly correlated with the speed and size of CMEs, but not with its density.

Summarizing, we note that quite a variety of models for CMEs – CRs relationship are available but these are not constrained enough by detailed observations, such as kinematics and morphology of CMEs in the interplanetary space, and detailed complementary observations of the evolution of the CRs modulation. To perform such studies is the core of our proposal.