History and development of coronal mass ejections as a key player in solar terrestrial relationship
- First Online:
- 1.5k Downloads
Coronal mass ejections (CMEs) are relatively a recently discovered phenomenon—in 1971, some 15 years into the Space Era. It took another two decades to realize that CMEs are the most important players in solar terrestrial relationship as the root cause of severe weather in Earth’s space environment. CMEs are now counted among the major natural hazards because they cause large solar energetic particle (SEP) events and major geomagnetic storms, both of which pose danger to humans and their technology in space and ground. Geomagnetic storms discovered in the 1700s, solar flares discovered in the 1800s, and SEP events discovered in the 1900s are all now found to be closely related to CMEs via various physical processes occurring at various locations in and around CMEs, when they interact with the ambient medium. This article identifies a number of key developments that preceded the discovery of white-light CMEs suggesting that CMEs were waiting to be discovered. The last two decades witnessed an explosion of CME research following the launch of the Solar and Heliospheric Observatory mission in 1995, resulting in the establishment of a full picture of CMEs.
Keywordscoronal mass ejections Flares Ground level enhancement events Solar energetic particle events
In the history of solar terrestrial relationships, CMEs are a very recent phenomenon, discovered only in the beginning of the 1970s (Hansen et al. 1971; Tousey 1973). However, some elements of CMEs were already known over the preceding decades. The purpose of this article is to provide an account of the discovery of various sub-structures of CMEs and the pre-eminent role played by space weather phenomena in finally obtaining the full picture of shock-driving CMEs, especially in visible light. Historical accounts on CMEs have been published a few times before (Kahler 1992; Gosling 1997; Howard 2006; Alexander et al. 2006; Webb and Howard 2012).
The wealth of knowledge accumulated on CMEs came from a series of coronagraphs: The first space borne coronagraph on board the seventh Orbiting Solar Observatory (OSO-7) mission (Tousey 1973), the Apollo Telescope Mount (ATM) Coronagraph on board Skylab (MacQueen et al. 1974), the Solwind coronagraph on board the Air Force satellite P78-1 (Michels et al. 1980), the Coronagraph/Polarimeter (CP) on board the Solar Maximum Mission (SMM) satellite (House et al. 1980), the Large Angle and Spectrometric Coronagraph (LASCO) on board the Solar and Heliospheric Observatory (SOHO) (Brueckner et al. 1995), the solar mass ejection imager (SMEI), a heliospheric imager on board the Coriolis spacecraft to observe interplanetary disturbances (Eyles et al. 2003), and the Sun Earth Connection Coronal and Heliospheric Investigation (SECCHI) on board the Solar Terrestrial Relations Observatory (STEREO) that includes the inner coronagraph COR1, the outer coronagraph COR2, the two heliospheric imagers covering the entire space (Howard et al. 2008). Finally, the ground-based Mauna Loa K Coronameter (Fisher et al. 1981a) has been providing information on the corona close to the Sun. Publications based on observations from these coronagraphs will be used in highlighting key results on CMEs. We also use numerous other observations made from remote-sensing and in situ observations that uncovered a number of phenomena associated with CMEs and hence contributed significantly to the development of the full picture of CMEs known today. Excellent reviews on CMEs and the related phenomena already exist (Schwenn 2006; Chen 2011; Akasofu 2011; Webb and Howard 2012), so I focus only on the key milestones that led to the current understanding of CMEs.
Section “Early days of solar terrestrial relationship” summarizes historical milestones that connected solar events to geomagnetic storms and cosmic ray modulation. The main space weather effect discussed—is the geomagnetic storm. Space weather implications of solar energetic particles became clear only in the Space Age. Section “Advances made during the Space Era” summarizes activities in the Space Era preceding the launch of SOHO. Section “CMEs in the SOHO Era” summarizes the exponential growth of knowledge on CMEs, aided by the continuous and uniform dataset obtained by SOHO. Results from the STEREO mission is also mentioned in this section. Finally, conclusions are presented in Section “Summary and conclusions.”
Early days of solar terrestrial relationship
Herschel (1801) reported of five prolonged periods of few sunspots to be correlated with high wheat prices in England. He inferred that a fewer sunspots indicated less heat and light from the Sun so the wheat production was low resulting in high wheat prices. Herschel was ridiculed for this report, but now we know that when there are more sunspots, the Sun emits more radiation because of the brighter faculae appearing around sunspots. This discovery can be thought of as an early recognition of the periodic nature of sunspot activity, recognized half a century later by Schwabe in 1843. Coining of the term “geomagnetic storm” by Humboldt in the early 1800s synthesized the knowledge on geomagnetic disturbances observed since the mid 1600s (see Howard 2006). Sabine (1852) found the synchronous variation of sunspot number and geomagnetic activity. In the backdrop of this long-term connection, Carrington and Hodgson independently discovered a sudden brightening on the Sun on September 1, 1859 and a huge geomagnetic storm was observed about 17 h later (Carrington 1859; Hodgson 1859). It took almost another half a century, back and forth, before connecting a geomagnetic event to a solar event (see discussion in Cliver 1994; Moldwin 2008), in particular, the ejection of particulate matter as a stream (FitzGerald 1892; Chapman 1918) or a plasma cloud (Lindemann 1919) was thought to be responsible for the geomagnetic event.
Mass emission from the sun: prominences and flare plasma
We now know that the prominences form the core of most CMEs. After the demonstration by Janssen and Lockyer in 1868 that prominences can be observed outside of eclipses using spectroscopes, the knowledge on prominences grew rapidly. By 1871, Secchi classified prominences into active and quiescent prominences. By 1892, it was a common knowledge that prominences erupted with speeds exceeding 100 s of km/s (from Tandberg-Hanssen 1995). The disk counterpart of eruptive prominences was termed as disparition brusque (DB) by d’Azambuja and d’Azambuja (1948).
Energetic particles from the sun
While the Sun–Earth relationship from the perspective of geomagnetic activity was under discussion for almost a century, energetic particles from the Sun were discovered only in 1942 (Forbush 1946). A large flare occurred near the center of the solar disk at the time of the solar cosmic ray event in 1946 on July 25, allowing Forbush to conclude on the flare origin of the particles. Before discovering the so-called solar cosmic rays, Forbush had already found the worldwide reduction in the intensity of cosmic rays (Forbush decrease) at the time of geomagnetic storms (Forbush 1937). In hindsight, it is easy to see that material ejections from the Sun is the cause of geomagnetic storms and Forbush deceases. Solar cosmic ray events observed on the ground are now known as ground level enhancement (GLE). Solar energetic particle events are significant source of space weather (Feynman and Gabriel 2000; Shea and Smart 2012).
Radio bursts and the connection to mass emission and shocks
Payne-Scott et al. (1947) reported the observations of several solar radio bursts that showed a drift from higher to lower frequencies, which they interpreted as a physical agency moving away from the Sun with a speed of ~500–700 km/s similar to eruptive prominences. They also noted the association of the radio burst with a solar flare. These bursts were later classified as type II radio bursts (Wild and McCready 1950).
Around this time, detailed theory of shock waves in infinitely conduction media had been worked out (de Hoffmann and Teller 1950; Helfer 1953). Helfer (1953) even considered the possibility of shocks in solar prominences. Gold (1955) proposed that the sudden commencement (SSC) of the storm is caused by interplanetary shocks driven by gas emitted during solar flares. Type II bursts were also suggested to be due to MHD shocks (Uchida 1960). Morrison (1956) attributed cosmic ray modulation and Forbush decrease to large ionized gas clouds with tangled magnetic field ejected from the Sun. Parker (1957) worked out details on such magnetized gas clouds and showed that they can be ejected with speeds as high as the Alfvén speed of the ambient medium. Parker used the term “magnetic cloud” to describe large-scale structures with tangled magnetic fields, although this term is now used for flux rope structures with enhanced magnetic field, smooth rotation of the azimuthal component, and low temperature (Burlaga et al. 1981).
Boischot (1957) discovered moving type IV radio bursts, which are magnetic structures carrying non-thermal electrons, and associated with solar eruptions. The radio source motion as fast as 1000 km/s or faster was reported (Boischot 1958). This can be regarded as the direct observation of magnetized material ejection from the Sun, apart from the eruptive prominences.
Advances made during the Space Era
Interplanetary origin of forbush decrease
Direct detection of an interplanetary shock
On 1962 October 7, Mariner two was located between the Sun and Earth, slightly behind Earth and above the ecliptic. A shock was detected at 15:42 UT and was followed by an SSC ~4.7 h later (Sonnet et al. 1964). The SSC was not followed by a storm probably because just a flank of the shock impacted Earth. The shock detection and the associated SSC confirmed the suggestion by Gold (1955).
Early picture of a solar eruption
Energetic particles, shocks, and plasma clouds
With the blast wave model of shocks from the Sun (Parker 1961), there was an increased focus on particle acceleration by shocks (Weddell 1965; Hudson 1965; Jokipii 1966). Bryant et al. (1962) reported the detection of a large increase in the intensity of 2–15 MeV protons by Pioneer 12 two days after the 1961 September 28 flare. The particle increase began just before the sudden commencement of a magnetic storm and hence was called an ‘energetic storm particle’ event. Bryant et al. (1962) interpreted the event as protons from the Sun trapped within the plasma cloud that caused the magnetic storm. Examining similar events detected by the cosmic ray detectors on board Pioneer 6 and 7, Rao et al. (1967) concluded that the protons must have been accelerated by interplanetary shocks.
A K-coronal transient from ground-based observations
Malitson et al. (1973) were also able to track an IP shock—from the Sun to 1 AU using a type II radio burst observed by the IMP-6 mission. Watanabe et al. (1973) used the interplanetary scintillation (IPS) technique to identify transient interplanetary disturbances associated with solar flares. The transients were identified as an enhancement in the solar wind speed over a very short duration (see also Rickett 1975). Gosling et al. (1973) found that the ejecta behind interplanetary shocks detected by Vela 3 had unusually low proton temperature, high speed, low density, and helium enrichment. These authors suggested that the low temperature must have resulted from the magnetic bottle configuration in the ejecta and the adiabatic cooling of the plasma in the bottle. Gosling et al.’s (1973) observations provided direct confirmation of the Gold (1962) picture shown in Fig. 3.
The first coronal transient from space: OSO-7
Soon after the report on the detection of the OSO-7 coronal transient, Eddy (1974) identified a coronal transient in the eclipse observations of the 1860 July 18 eclipse. Cliver (1989) pointed out that what was thought of as a comet during the 1893 April 16 eclipse is actually a CME event. Webb and Cliver (1995) searched a hundred years of eclipse observations looking for CMEs. There have also been other CME detections during modern eclipses (Rušin 2000; Hanaoka et al. 2014).
The Skylab ATM Coronagraph that had operations overlapping with OSO-7 (May 1973–Feb 1974) detected 110 coronal transients in white light over a period of 227 days and led to a number of key results. The transients also became known as coronal mass ejections around this time. Gosling et al. (1975) was the first to identify a Skylab CME with an interplanetary (IP) shock: the 1973 September 7 CME from the southwest quadrant of the Sun (S18W46) resulted in the shock observed by Pioneer 9 located at 0.98 AU directly above the source region on the Sun on September 9. The shock and the driving ejecta at Pioneer 9 had speeds of 722 and 600 km/s, respectively. The CME speed near the Sun was ≥960 km/s, indicating a clear deceleration in the IP medium. The CME mass was also estimated to be ~2.4 × 1016 g. Another key result was the significant correlation between the occurrence rate of CMEs and the sunspot number (Hildner et al. 1976). Kahler et al. (1978) studied 16 Skylab CMEs and found a good correlation between the CME speed and SEP intensity. These authors concluded that “energetic protons are accelerated in the shock front just ahead of the expanding loop structures observed as mass ejections.” This result formed the basis for the establishment of shock acceleration as the leading mechanism for SEPs (Cliver et al. 1982; Cane et al. 1988; Gosling 1993; Reames 1995). Mouschovias and Poland (1978) suggested that some CMEs can be thought of as a “twisted rope of magnetic-field lines expanding and broadening in the background coronal plasma and magnetic field” and modeled the 1973 August 10 Skylab transient as a rope. Based on the polarization measurements of the same CME, Crifo et al. (1983) suggested that the transient must be a three-dimensional bubble-like structure. Fisher and Munro (1984) introduced the ice-cream cone model of CMEs to explain a Mauna Loa Mark III CME observed, which is consistent with Crifo et al. (1983).
The Solwind CMEs
When the Solwind coronagraph on board the P78-1 mission became operational, (February 1979 to September 1985), it observed ~1700 CMEs. The Solwind observations confirmed many of the Skylab results and led to a number of new discoveries. Halo CMEs were identified as CMEs directed toward Earth appearing to surround the occulting disk in sky-plane projection (Howard et al. 1982). Solwind data also indicated rich morphological features in CMEs (Howard et al. 1985). Sheeley et al. (1980) reported a high-latitude CME and correctly inferred that such CMEs must be associated with polar crown filaments observed in large numbers during solar maxima. High-altitude CMEs represent the process by which the old magnetic flux is removed from polar regions, paving the way for the appearance of the new magnetic cycle according to the hypothesis by Low (1997).
The recognition that a halo CME is a regular CME propagating in the Earthward or anti-Earthward direction (Howard et al. 1982) is an important step in understanding how CMEs impact Earth. Howard et al. (1985) studied about 1000 Solwind CMEs (1979–1981) and found that there was no obvious correlation between CME rate and the sunspot number contradicting Hildner et al. (1976) results from Skylab. However, when the full set of data was used in combination with data from other missions, the CME rate was found to track the solar activity cycle in both amplitude and phase (Webb 1991; Webb and Howard 1994). Kahler et al.’s (1978) Skylab result that energetic particles are due to CME-driven shocks was also confirmed with a larger sample (Kahler et al. 1983). Cliver et al. (1983) associated the 1979 August 21 GLE with a Solwind CME with a sky-plane speed of ~700 km/s (real speed is likely to be >1000 km/s, when projection effects are taken into account).
The solar maximum mission CMEs
The Coronagraph/Polarimeter (CP) on board SMM observed CMEs in 1980 and 1984–1989, adding another 1200 CMEs to the CME data base. SMM was interrupted in 1980 when the satellite’s attitude control system failed. In 1984, the crew of space shuttle Challenger (STS-41-C) captured, repaired, and redeployed the SMM mission, which continued operations until 1989.
Other heliospheric observations
In addition to natural radio sources, scintillations of satellite signals have also been used to study shocks. Woo and Armstrong (1981) used the radio signals from Voyager 1, 2 to detect a fast (~3500 km/s) shock at a distance of 13.1 Rs from the east limb of the Sun.
Using the IPS technique, Gapper et al. (1982) identified solar disturbances to heliocentric distances beyond Earth’s orbit. Propagation of solar disturbances into the interplanetary medium was also inferred from HELIOS zodiacal photometer data as a short-term increase in polarized light. Richter et al. (1982) reported on eight clouds and estimated the size along the line of sight to be ~0.06 AU (see also Jackson 1985; Webb and Jackson 1990). Neutron monitors (Bieber and Evenson 1997) and muon detectors (Munakata et al. 2005) in the worldwide network remain an important source of information on Forbush decrease and GLEs. Low-frequency radio receivers also provide important information on CME-driven shocks in the inner heliosphere because the emission frequency indicates a heliocentric distance of the shock (Cane et al. 1987; Bougeret 1997).
Shock sheath and flux rope
An important refinement of the Gold picture in Fig. 3B was made by Burlaga et al. (1981) who identified the region between the shock and the magnetic cloud as the turbulent hot plasma compressed by the shock and consisting of highly variable magnetic field. Burlaga et al. (1981) used plasma and magnetic field data from five spacecraft (HELIOS-1,2, IMP-8, Voyager-1,2) to derive the 3-D structure of the driver consisting of high magnetic field, low temperature and density, and rotation of the field vectors indicating a loop. They called this structure a magnetic cloud consisting of helical magnetic field, subsequently known as flux ropes (Burlaga and Klein 1980; Klein and Burlaga 1981). The helical fields inferred from interplanetary observation are consistent with the magnetic ropes at the Sun as suggested by Mouschovias and Poland (1978), although the connection was not made at the time. These observations are consistent with the magnetic bottle inferred earlier (Fig. 3), except that the magnetic field is twisted. Goldstein (1983) modeled the magnetic structure of magnetic clouds as a flux rope based on the twisted cylindrical fields considered by Lundquist (1951) in conducting fluids.
Magnetic clouds and southward magnetic field component
The critical connection between CMEs and geomagnetic storms was made by Wilson (1987), who found that the southward component of the magnetic field in magnetic clouds is responsible for the storms. Wilson (1987) performed a superposed epoch analysis of 19 magnetic clouds and found that the Dst index simultaneously decreased to a large negative value at the onset of a large, sustained southward magnetic field in the magnetic cloud; the storm recovery started when the magnetic field turned northward. The results were consistent with the well-known connection between geomagnetic disturbances and the southward component of the external (interplanetary) magnetic field (Dungey 1961; Fairfield and Cahill Jr 1966). Zhang and Burlaga (1988) extended this study to various types of clouds. Southward interplanetary field can also occur in the shock sheaths (Gonzalez and Tsurutani 1987; Gosling and McComas 1987), which have been shown to be equally important in causing geomagnetic storms (Tsurutani et al. 1988).
CMEs as primary space weather driver
By late 1980s, it became clear that the southward component of the magnetic field in the interplanetary CMEs (ICMEs) and/or the shock sheath is directly responsible for intense geomagnetic storms (Wilson 1987; Gonzalez and Tsurutani 1987). It also became clear that CME-driven shocks are responsible for the large SEP events, although flares can accelerate particles that can be distinguished by the higher Fe/O ration (Reames 1990). However, flares continued to be considered as the primary source of shocks and CMEs (Maxwell and Dryer 1982). The term “flare-generated shock” continued to be used giving secondary importance to the CME (see e.g., Dryer 1975; Woo and Armstrong 1981). On the basis of a broad spectrum of evidences, Kahler (1992) concluded that CMEs play a primary role in accelerating SEPs and producing IP shocks. Gosling (1993), pointed out that the emphasis on flares is misplaced as far as geomagnetic storms and large SEP events are concerned and he referred to this as solar flare myth. Gosling’s (1993) conclusions represented the culmination of research on CMEs in the 1970s and 1980s bringing them to the forefront as the main players in affecting Earth’s space environment.
CMEs in the SOHO era
SOHO was launched in 1995 and has been observing the corona up to a distance of 32 Rs over the past 20 years. SOHO’s uniform and continuous data set has provided clarity to many of the results obtained from the past coronagraphs, and has led to many new discoveries. An order of magnitude more CMEs have been observed than the combined set of all CMEs observed in the pre-SOHO era (http://www.cdaw.gsfc.nasa.gov/CME_list, Yashiro et al. 2004; Gopalswamy et al. 2009a). Thousands of papers have been published using SOHO data, so it is impossible to list all the results. We highlight some key results that are relevant to space weather.
CME rate and correlation with the sunspot number
The fact that the CME rate is correlated with sunspot number (SSN) reported from pre-SOHO coronagraphs (Webb and Howard 1994) was readily confirmed with SOHO CMEs for cycle 23 (Gopalswamy et al. 2003a). When the correlation was considered separately for the rise, maximum, and declining phases, it was found that the maximum phase had the weakest correlation (Gopalswamy et al. 2010a). This was attributed to the fact that the number of CMEs associated with quiescent filament eruptions (i.e., non-spot) from the high and low latitudes increases during the maximum phase (Gopalswamy et al. 2003b, 2003c). This might be the reason that Howard et al. (1985) did not find good correlation between the CME rate and sunspot number for the period 1979–1981, which is the maximum phase of cycle 21.
Flux rope CMEs
Halo CMEs are so called because the excess brightness in these events appear to surround the occulting disk of the observing coronagraph as projected on the sky plane. Halo CMEs were discovered in Solwind images (Howard et al. 1982) but remained a novelty because only a handful were observed over the entire lifetime of Solwind. The SMM/CP did not observe many halos because most halos are observed at distances beyond ~6 Rs (outside SMM/CP FOV) and also SMM observed in quadrants which meant a lower cadence for the whole corona. SOHO has observed several hundred halo CMEs in the past 20 years, indicating that about 4 % of all CMEs are halos (Gopalswamy et al. 2010b). Frontside halos (those originating on the Earth-facing side of the Sun) are important for space weather because they are expected to directly impact Earth as magnetic clouds (Webb et al. 2000; St. Cyr 2005; Gopalswamy et al. 2007, 2010b). Halo CMEs are similar to normal CMEs, except that they are faster and wider and hence more energetic. For example, there were 696 fast (≥900 km/s) and wide (≥60°) CMEs were observed during January 1996 to March 2015. Over the same period, 662 halos were observed, of which 319 were fast. Some halos with speed <900 km/s may be faster in reality because of projection effects. With SOHO images alone, one does not know the intrinsic width of halo CMEs. When SOHO was in quadrature with one of the two STEREO spacecraft, it was possible to show that the STEREO limb CMEs were halo CMEs in SOHO and that they were wide (typical width >60°) and the speeds were in the range (510–2200 km/s) (Gopalswamy et al. 2013b). The fraction of halo CMEs in a CME population has been found to be a good indicator of the average CME energy in the population (Gopalswamy et al. 2010a).
CME propagation: deflection, interaction, and rotation
A systematic equatorward deflection of CMEs observed during 1973–1974 by ~2° was reported in the pre-SOHO era (Hildner 1977; MacQueen et al. 1986). SOHO images abundantly confirmed this, but helped to quantify the extent of CME deflection by the enhanced magnetic field in coronal holes. Gopalswamy and Thompson (2000) found that both the prominence and the CME showed the deflection, suggesting that the CME deflected as a whole, including sub-structures. Furthermore, the deflection was over ~30°, much larger than what was reported in the pre-SOHO era. Deflections were also observed in the longitudinal direction due to the magnetic field of equatorial coronal holes. In fact the deflection could be in any direction, depending on the relative position of the coronal hole and the eruption region (Cremades et al. 2006; Gopalswamy et al. 2001). For space weather effects, deflections relative to the Sun–Earth line are important (Gopalswamy et al. 2009b). Deflections toward the Sun–Earth line can make a CME more geoeffective, while a deflection in the opposite direction can make a CME miss Earth. Coronal-hole deflection can also result in the lack of alignment between the CME and shock (Wood et al. 2012) if the shock is weakly driven. The physical reason for the deflection has been shown to be due to the global magnetic pattern surrounding the eruption region (Filippov et al. 2001; Sterling et al. 2011; Gui et al. 2011; Panasenco et al. 2013; Kay et al. 2013; Möstl et al. 2015).
CME–CME interaction in the SOHO coronagraph field of view was reported by Gopalswamy et al. (2001). The interaction caused a change in trajectory of the preceding CME and a broadband enhancement of the associated type II burst occurred suggesting additional electron acceleration. Higher SEP intensity results when there is CME–CME interaction (Gopalswamy et al. 2004) and the arrival of CMEs at 1 AU gets delayed when they interact with preceding CMEs (Manoharan et al. 2004; Temmer et al. 2012). CME interaction can also result in a single shock at Earth even though there are multiple CMEs are ejected at the Sun (Gopalswamy et al. 2012a).
CME Rotation is somewhat controversial. It has been known for a long time that the orientation of interplanetary flux ropes are generally aligned with the polarity inversion line (or filament) at the Sun (Marubashi 1997; Bothmer and Schwenn 1998; Yurchyshyn et al. 2001; Marubashi et al. 2009). However, several authors have interpreted CME observations to indicate rotation during coronal and interplanetary propagation. Yurchyshyn et al. (2009) fitted ellipses to the outlines of halo and partial halo CMEs and compared them with the axis of the associated post-eruption arcades. They reported that CMEs appear to rotate by about 10° for most of the events with about 30–50° for some events (see also Isavnin et al. 2014). Vourlidas et al. (2011) reported an event with a rotation rate of 60° per day. Recently, Marubashi et al. (2015) analyzed a set of more than 50 well observed CME-ICME pairs and found strong support to the idea that an erupted flux rope has its main axis parallel to the polarity inversion line and remains so as it propagates through the interplanetary space. This is also consistent with models in which the flux rope is formed due to reconnection (Kusano et al. 2012). A clear definition of rotation is needed in three dimensions, for example, with respect to the radial direction from the source region.
Although CMEs are responsible for space weather effects, they can be still closely related to flares: the eruption process has been understood to be common to both the mass ejection and flare heating (see e.g., Gosling 1990). SOHO observations support the idea that the flare and CME are two manifestations of the same energy release (e.g., Low 1994; Harrison 1995), although the CME kinetic energy has been found to be the single largest component of the released energy (Emslie et al. 2012). The connection between flares and CMEs has been recognized in a number of ways: coupled evolution indicated by the similarity of CME speed profiles to that of flare soft X-ray emission (Zhang et al. 2001), significant correlation between CME kinetic energy and soft X-ray fluence (Burkepile et al. 2004; Gopalswamy et al. 2009a, b), flare origin of enhanced charge state in ICMEs (Reinard 2008; Gopalswamy et al. 2013a), CME nose located radially above the flare site (Yashiro et al. 2008; Veronig et al. 2010), and the correlation between reconnection flux and the azimuthal flux of IP flux ropes (Longcope and Beveridge 2007). It is well known that CMEs do occur with almost no surface signatures (Wagner 1984; Robbrecht et al. 2009), suggesting that all the released energy must be going into mass motion. However, flare energy, measured as the thermal energy content of even extremely weak post-eruption arcades from the polar CMEs, seems to have quantitative relationship with CME kinetic energy (Gopalswamy et al. 2015b). The flare–CME relationship is somewhat muddled when CMEs erupt from complex magnetic regions involving active regions and filaments (Harrison et al. 2012).
Detection of of shocks using white-light imaging
CME relevance for space weather
All the subsets of CMEs have average speeds above that of the general population. CMEs producing purely metric type II bursts are of lowest speed (611 km/s), which is sufficient to drive shocks near the Sun (within about 1 Rs from the surface). CMEs accelerating SEPs with ground level enhancement are the fastest (~2000 km/s). SOHO CMEs were identified with all GLE events of cycle 23 and found to be the fastest subset driving shocks throughout the inner heliosphere (Kahler et al. 2003; Gopalswamy et al. 2005b; Cliver 2006). CMEs associated with decameter-hectometric (DH) type II busts and SEP events are similar because the same shock is responsible for both the phenomena. CMEs associated with magnetic clouds (MCs), ejecta (EJ), IP shocks (S), and geomagnetic storms (GM) represent plasma impact on Earth while the others represent particle acceleration by CME-driven shocks. All subsets except for those associated with MCs have speeds close to 1000 km/s or higher. MCs are associated with CMEs from close to the disk center, so their true speed may be higher. Thus we see that only about 1000–2000 CMEs (or <10 %) are important for space weather.
The two additional views provided by the twin STEREO spacecraft helped understand the three-dimensional structure of CMEs (Howard et al. 2008). For example, the same CME observed in two views confirmed that halo CMEs are regular CMEs but viewed head-on (Gopalswamy et al. 2010a). STEREO instruments helped us observe CMEs much closer to the Sun (see e.g., Gopalswamy et al. 2009c; Patsourakos et al. 2010) and all the way to 1 AU (Harrison and Davies 2014 and references therein) because of the appropriate instrument suite includes the inner coronagraph COR1 and the heliospheric imagers (HIs). CMEs were tracked in three dimensions (Mierla et al. 2010) all the way to 1-AU using the HI images (see e.g., Möstl et al. 2010; DeForest et al. 2013; Liu et al. 2014). DeForest et al. (2013) exploited the availability of quantitative photometry from STEREO/SECCHI to distinguish between the mass picked up from the wind and the mass that is part of the original CME. HI observations (from SMEI and STEREO) have helped improve the prediction of arrival times of CMEs (Webb et al. 2009) and provide advance warning of storm-causing CMEs. In particular, the arrival of the CME-driven shock can be viewed in HI images, normally inferred from solar wind plots (Möstl et al. 2010). It was also possible to obtain the height of GLE release directly from STEREO/COR1 images (Gopalswamy et al. 2013c; Thakur et al. 2014), confirming earlier results from Kahler (1994). Observations close to the Sun also helped quantify the initial acceleration of CMEs (Bein et al. 2011), which was also possible using the Mauna Loa K Coronameter images (Fisher 1984; St Cyr et al. 1999; Gopalswamy et al. 2012b). Multiple views also helped recognize many SEP events that are observed all around the Sun (see e.g., Dresing et al. 2012; Lario et al. 2013) providing opportunities to test the CME origin and heliospheric propagation of energetic particles.
Summary and conclusions
Coronal mass ejections are a novelty because they were discovered only in 1971, some 15 years into the Space Era. It took another 30 years and half a dozen space missions to obtain the complete picture of CMEs. The discovery of CMEs can be traced to the attempts by scientists to understand the cause of geomagnetic disturbances, Forbush decreases, and solar energetic particles. Simply because flares were discovered first, it was natural to identify flares as the cause of the interplanetary disturbances. Yet, people realized early on that it was matter rather than electromagnetic radiation that caused large geomagnetic and interplanetary disturbances. A shock was proposed to explain the sudden commencement of storm and energetic storm particles. A shock was also needed to explain type II radio bursts. The driver was identified to be a magnetized plasma structure, although initially thought to be just a plasma cloud. The last thing to be identified in the corona is the white-light shock.
Space weather effects can be directly attributed to CME sub-structures. While the outermost structure (shock) accelerates particles and causes sudden commencement, the interior structures (sheath and flux rope) cause geomagnetic storms if they possess a strong southward component of the magnetic field. It has become clear that CMEs are a natural hazard because they produce space weather effects that pose danger to human technology in space and ground: from telegraph systems to GPS navigation; from power grids on the ground to payload in Mars orbit. We still have a long way to go in predicting when a CME will occur on the Sun, and when it will arrive at a given destination in the heliosphere and what the magnetic structure would be. We still need to figure out the relative contribution of flares and CMEs to the observed energetic particles from solar eruptions.
I thank R. Selvakumaran for checking the references. I also thank V. Bothmer, J. Chen, C. E. DeForest, R. A. Harrison, Y. Liu, B. C. Low, K. Marubashi, M. Mierla, C. Moestl, V. N. Obridko, D. V. Reames, A. Reinard, O. C. St. Cyr, A. Sterling, N. Thakur, and D. F. Webb for helpful comments. This work was supported by NASA’s LWS and TR&T programs. This work benefitted greatly from the open data policy of NASA.
The author declares that he has no competing interests.
- Bieber J, Evenson P (1997) Neutron monitors and space weather. ICRC 1997(1):341Google Scholar
- Boischot A (1957) Caractères d’un type d’émission hertzienne associé à certaines éruptions chromosphériques. Comptes Rendus 244:1326–1329Google Scholar
- Boischot A (1958) Étude du rayonnement radioélectrique solaire sur 169 MHz é l’aide d’un grand interféromètre à réseau. Ann d’Astrophys 21:273Google Scholar
- Bougeret J-L (1997) The Heliospheric Signatures of Coronal Mass Ejections. In: Wilson A (Ed) Correlated Phenomena at the Sun, in the Heliosphere and in Geospace. 31st ESLAB Symposium held 22–25 Sep 1997, at ESTEC, The Netherlands. European Space Agency, ESA SP-415. ESA Publications Division, Noordwijk, p 523Google Scholar
- Burlaga L, Klein L (1980) Magnetic clouds in the solar wind, NASA. Tech Memo 80668Google Scholar
- Chen PF (2011) Coronal mass ejections: models and their observational basis. LRSP 8:1Google Scholar
- d’Azambuja L and d’Azambuja M (1948) Ann Obs Paris Meudon 6, fasc. 7:1Google Scholar
- Dresing N, Gómez-Herrero R, Klassen A, Heber B, Kartavykh Y, Dröge W (2012) The Large Longitudinal Spread of Solar Energetic Particles During the 17 January 2010 Solar Event. Sol Phys 281:281–300Google Scholar
- Eddy JA (1974) A nineteenth-century coronal transient. A&A 34:325Google Scholar
- Eyles CJ, Simnett GM, Cooke MP, Jackson BV, Buffington A, Hick PP, Waltham NR, King JM, Anderson PA, Holladay PE (2003) The solar mass ejection imager (SMEI). Solar Phys 217:319–347Google Scholar
- Fisher RR, Munro RH (1984) Coronal transient geometry. I-The flare-associated event of 1981 March 25. Astrophys J 280:428–439Google Scholar
- Fitzgerald GF (1892) Sunspots and magnetic storms. Electrician 30:48Google Scholar
- Ginzburg VL, Zhelezniakov VV (1958) On the Possible Mechanisms of Sporadic Solar Radio Emission (Radiation in an Isotropic Plasma). Sov Astr—AJ. 2:653Google Scholar
- Gold T (1955) Discussion of Shock Waves and Rarefied Gases. In: van de Hulst JC, Burgers JM (eds) Gas Dynamics of Cosmic Clouds, North-Holland Publishing Co., Amsterdam, p103Google Scholar
- Goldstein H (1983) On the field configuration in magnetic clouds, In: JPL Solar Wind Five. pp 731–733Google Scholar
- Gopalswamy N (2004) A global picture of CMEs in the inner heliosphere. In: Poletto G, Suess ST (eds) The Sun and the Heliosphere as an Integrated System, Kluwer Acad, Boston, chapter 8, pp 201–251Google Scholar
- Gopalswamy N (2010) Coronal mass ejections: a summary of recent results. In: Proceedings of the 20th National Solar Physics Meeting, held 31 May–4 June, 2010, Papradno, Slovakia, pp 108–130Google Scholar
- Gopalswamy N, Thompson BJ (2000) Early life of coronal mass ejections. JASTP 62:1457–1469Google Scholar
- Gopalswamy N, Lara A, Yashiro S, Nunes S, Howard RA, (2003a) Coronal mass ejection activity during solar cycle 23. In: Wilson A (ed) Solar variability as an input to the Earth’s environment. International Solar Cycle Studies (ISCS) Symposium, 23–28 June 2003, Tatranská Lomnica, Slovak Republic, ESA SP-535. ESA Publications Division, Noordwijk, pp 403–414Google Scholar
- Gopalswamy N, Xie H, Yashiro S, Usoskin I (2005b) Coronal Mass Ejections and Ground Level Enhancements. ICRC 1:169Google Scholar
- Gopalswamy N, Akiyama S, Yashiro S, Mäkelä P (2010a) Coronal mass ejections from sunspot and non-sunspot regions. In: Hasan SS, Rutten RJ (eds) Magnetic Coupling Between the Interior and Atmosphere of the Sun. Astrophysics and Space Science Proceedings, Springer, Berlin, pp 289–307CrossRefGoogle Scholar
- Gopalswamy N, Yashiro S, Michalek G, Xie H, Mäkelä P, Vourlidas A, Howard RA (2010b) Catalog of Halo Coronal Mass Ejections from SOHO. Sun and Geosphere 5:7–16Google Scholar
- Gosling JT (1997) Coronal mass ejections: An overview. In: Crooker N, Joselyn JA, Feynman J (eds) Coronal Mass Ejections, Geophys. Monogr. Ser. 99. American Geophysical Union, Washington DC, pp 9–16Google Scholar
- Hansen RT, Garcia CJ, Grognard RJM, Sheridan KV (1971) A coronal disturbance observed simultaneously with a white-light corona-meter and the 80 MHz Culgoora radioheliograph. Proc Astron Soc Aust 2:57Google Scholar
- Harrison RA (1995) The nature of solar flares associated with coronal mass ejection. A and A 34:585Google Scholar
- Hildner E (1977) Mass ejections from the corona into the interplanetary space. In: Shea MA, Smart DF, Wu ST (eds) Study of Traveling Interplanetary Phenomena. D. Reidel, Hingham, pp 3–21Google Scholar
- Howard RA (1987) A Critical Comparison of CME results from Solwind and SMM. In: Pizzo VJ, Holzer T, Sime DG (eds) Proc Sixth International Solar Wind Conference, 23–28 August, 1987 at Estes Park, Colorado. NCAR Technical Note NCAR/TN-306 + Proc, Volume 2. NCAR, Boulder, p 233Google Scholar
- Howard RA (2006) A Historical Perspective on Coronal Mass Ejections. In: Gopalswamy N, Mewaldt R, Torsti R (eds). Solar Eruptions and Energetic Particles. Geophysical Monograph Series, Vol 165. American Geophysical Union, Washington DC, pp 7–12Google Scholar
- Kahler SW, McGuire RE, Reames DV, von Rosenvinge TT, Jr Sheeley N R, Howard RA, Michels DJ, Koomen MJ (1983) The correlation of coronal mass ejections with energetic flare proton events. ICRC 4:6–9Google Scholar
- Kahler SW, Simnett GM, Reiner MJ (2003) Onsets of solar cycle 23 ground level events as probes of solar energetic particle injections at the sun. ICRC 2003:3415Google Scholar
- Low BC (1994) Coronal mass ejections and magnetic helicity. In: Hunt JJ (ed) Solar Dynamic Phenomena and Solar Wind Consequences, Proceedings of the Third SOHO Workshop held 26–29 September, 1994 in Estes Park, Colorado, ESA SP-373. European Space Agency, Noordwijk, p 123Google Scholar
- Marubashi K (1997) Interplanetary magnetic flux ropes and solar filaments. In: Crooker N, Joselyn JA, Feynman J (eds) Coronal Mass Ejections. Geophysical Monograph Series 99. American Geophysical Union, Washington DC, pp 147–156Google Scholar
- Newton HW (1936) Eruptive Dark Flocculi and their Possible Relation to Magnetic Disturbances. The Observatory 59:51–52Google Scholar
- Patsourakos S, Vourlidas A, Stenborg G (2010) The genesis of an impulsive coronal mass ejection observed at ultra-high cadence by AIA on SDO. A&A 724:L188–L193Google Scholar
- Reames DV (1990) Acceleration of energetic particles by shock waves from large solar flares. Astrophys J 358:L63–L67Google Scholar
- Richter I, Leinert C, Planck B (1982) Search for short term variations of zodiacal light and optical detection of interplanetary plasma clouds. A and A 110:115–120Google Scholar
- Rušin V (2000) Shape and Structure of the White-Light Corona over Solar Cycles (Invited review). In: Livingston W, Özgüç A (eds) Last Total Solar Eclipse of the Millennium, ASP Conference Series, vol 205. Astronomical Society of the Pacific, San Francisco, p 17Google Scholar
- Schmahl EJ (1972) A model of the evolution of the flux density and polarization in isolated moving type IV bursts. Proc Astron Soc Austr 2:95Google Scholar
- Schwenn R (2006) Space weather: the solar perspective. LSRP 3:2Google Scholar
- Smerd, S, Dulk G (1971) in Solar Magnetic Fields, Proc IAU symposium 43, (ed. R. Howard), D. Reidel, Drodrecht, p 616Google Scholar
- Tandberg-Hanssen E (1995) The nature of solar prominences. Astrophysics and Space Science Library, vol. 199. Kluwer Academic Publishers, DordrechtGoogle Scholar
- Temmer M, Vršnak B, Rollett T, Bein B, de Koning C, Liu Y, Bosman E, Davies J, Möstl C, Žic T, Veronig AM, Bothmer V, Harrison R, Nitta N, Bisi M, Flor O, Eastwood J, Odstrcil D, Forsyth R (2012) Characteristics of kinematics of a coronal mass ejection during the 2010 August 1 CME-CME interaction event. APJ 749:57CrossRefGoogle Scholar
- Tousey R (1973) The Solar Corona. In: Rycroft MJ, Runcorn SK (eds) Space Research XIII. AkademieVerlag, Berlin, p 713Google Scholar
- Uchida Y (1960) On the Exciters of Type II and Type III Solar Radio Bursts. PASJ 12:376Google Scholar
- Vourlidas A, Lynch BJ, Howard RA, Li Y (2013) How many CMEs have flux ropes? Deciphering the signatures of shocks, flux ropes, and prominences in coronagraph observations of CMEs. Sol Phys 284:179–201Google Scholar
- Wagner WJ (1984) Spontaneous coronal mass ejections. Bull American Astron Soc 16:536Google Scholar
- Webb DF, Howard TA (2012) Coronal mass ejections: observations. LSRP 9:3Google Scholar
- Webb DF, Howard TA, Fry CD, Kuchar TA, Mizuno DR, Johnston JC, Jackson BV (2009) Studying geoeffective interplanetary coronal mass ejections between the Sun and Earth: Space weather implications of Solar Mass Ejection Imager observations. Space Weather 7:S05002. doi:10.1029/2008SW000409
- Wild JP, McCready LL (1950) Observations of the Spectrum of High-Intensity Solar Radiation at Metre Wavelengths I The Apparatus and Spectral Types of Solar Burst Observed. Australian J Sci Res A 3:387Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.