West Limb Research Articles

Streamers and adjacent regions observed by UVCS/SOHO: A comparison between different phases of solar activity

UVCS/SOHO observations of the $\ion{O}{vi}$ resonance doublet and $\ion{H}{i}$ Ly α line intensities and profiles, together with measurements of the visible linearly polarized radiance, have been performed during two MEDOC campaigns in 1997 and 2000, i.e. near solar minimum and approaching the solar maximum phase, respectively. During both observational runs mid-latitude coronal regions in the West limb of the Sun have been scanned over a range of heliocentric distance from 1.39 to 4.1 $R_{\odot}$, to study the plasma properties of streamers and adjacent regions, such as ion kinetic temperature, electron density and outflow velocity, paying particular attention to comparing plasma conditions deduced for different ions in coronal structures observed on different days and during different phases of solar activity. Besides confirming some previous findings on significant differences between open and closed field-line structures at solar minimum, our results provide some evidence for differences in kinetic temperature among mid-latitude solar minimum streamers observed on different days from about 2 $R_{\odot}$ outwards, as well as in their dynamical conditions at heliocentric distances greater than 3.6 $R_{\odot}$. For observations carried out in 2000, conversely, the mid-latitude coronal streamers and their surroundings are about 3 times, and more than one order of magnitude brighter, respectively, than their solar minimum counterparts and exhibit very similar kinetic and dynamical conditions. The kinetic temperatures in adjacent regions are higher than in streamers (by about a factor of 2) only within 2 $R_{\odot}$, while at greater heights such differences vanish, making it difficult to discriminate between open and closed structures. This is opposite to the behaviour detected at solar minimum, when adjacent regions appear to be characterized by kinetic temperatures progressively higher and higher than in streamers with increasing height, from 2 $R_{\odot}$ outwards. Therefore, a clear characterization of open and closed configurations near the solar maximum might be quite difficult, probably due both to the intrisically more complex magnetic configuration of the corona in this phase of the solar activity and the line-of-sight contamination effects that in a highly structured solar corona may strongly mix background and foreground plasma with different properties. The transition from the solar minimum to maximum also seems characterized by a global increase in the electron density inside streamers of about a factor of 4 at 1.7 $R_{\odot}$ and then it progressively decreases with height.

Multi-wavelength study of a strong impulsive solar limb flare on 2002 August 3

We made a detailed study of the impulsive solar flare of GOES class X1.0 which occurred near the west limb on 2002 August 3, peak time 19:07 UT. There is particularly good data coverage of this event, with simultaneous observations in EUV, soft and hard X-rays available. We used TRACE 171 Å images to study the morphology and evolution of this event. Soft X-ray spectra in the wavelength range 3.34–6.05 Å measured by the RESIK Bragg crystal spectrometer on CORONAS-F were used for determination of the evolution of the flare plasma temperature. Data from the RHESSI instrument were used to investigate properties of the higher-temperature plasma during the flare.

RATAN-600 Observations of Unusual Inversion of Polarization in Sunspot Associated Microwave Sources

We have discovered unusual inversion of circular polarization in both components of an active region. We have observed the region for six consecutive days in May 96 using the RATAN-600 radio telescope with high polarization accuracy up to 0.5 %. The observations were taken in the range of 1.7 to 30 cm. At short wavelengths the circular polarization scans revealed a bipolar structure, expected since the magnetic field is bipolar. However, for all six days and while the region was moving from the center of the disk to the west limb, the circular polarization dropped practically to zero at 14.5 cm and at longer wavelengths the sense of the polarization was reversed.

An unusually fast interplanetary coronal mass ejection observed by Ulysses at 5 AU on 15 November 2003

On 15 November 2003, at ∼2030 UT, Ulysses/SWOOPS observed the onset of a large, unusually fast interplanetary coronal mass ejection (ICME) in the solar wind at 5.2 AU. Ulysses measured the peak solar wind flow speed associated with this event, vmax = 993 km s−1, in the turbulent sheath region preceding the ICME. This is the fastest solar wind speed recorded by Ulysses/SWOOPS at this distance since November 1992. On 7 November 2003 at 1554 UT, SOHO/LASCO observed the solar counterpart of this ICME behind the west limb of the Sun; the CME had a plane‐of‐sky speed of 1995 km s−1. We believe this CME originated in AR 0486, the same active region responsible for the extremely fast ICMEs that impacted the Earth on 29 and 30 October 2003. The ICME took at least 4.3 days to propagate past Ulysses and had a radial width of ∼2.1 AU. The ICME was a magnetic cloud, with a depressed proton beta of ∼0.01 and a smooth rotation of the magnetic field polar angle from 43° to −62°. In addition, Ulysses/SWOOPS also observed bidirectional electron streaming throughout the ICME, indicating that even at this distance, closed field lines threaded the entire event. In the days preceding the ICME, Ulysses observed disturbed solar wind conditions, including six shocks from 6 November to 15 November and several small ICMEs, indicating a high degree of solar activity.

Four Corona Mass Ejections and their Associated Surface Activity Observed on 26 October 2003

Four coronal mass ejections (CMEs) occurs successively from the solar disk near the west limb on October 26, 2003. They, together with the associated activities of the solar surface, were observed by various instruments both in space and on ground, such as the Large Angle and Spectrometric Coronagraph Experiment (LASCO), the Extreme Ultraviolet Imaging Telescope, and the Michelson Doppler Imager on board the Solar and Heliospheric Observatory, as well as the Huairou Solar Observing Station and the Big Bear Solar Observatory. These four events start with a filament eruption that manifests a two-ribbon flare in a spotless region, destroyed a helmet streamer, and give rise to an X1.2 flare in the active region AR0484. We notice that these eruptions occur either in active region, or in quiescent region, or in the region without any precursors. The time profiles of the CME (filament) heights show that the main acceleration takes place within one solar radius .To search for other articles by the author(s) go to: http://adsabs.harvard.edu/abstract_service.html

SINCLINAL OURO FINO REVISITADO, QUADRILÁTERO FERRÍFERO, MINAS GERAIS: UMA HIPÓTESE SOBRE A SUA ORIGEM E EVOLUÇÃO

The Ouro Fino Syncline is a regional structure with a controversial origin. Two structural models represent the synthesis to its tectonic evolution. On the one hand, according to Ladeira & Viveiros (1984), the Ouro Fino syncline is the result of the first folding event which affected the Minas Supergroup. These folds present E-W trending axes and associated penetrative cleavage. The tectonic transport vector is northward. On the other hand, Fonseca (1990), interprets Ouro Fino syncline's origin as a result of compressive forces directed to westward. The axis has N-S trending and westward vergence according to Dorr (1969). This second point of view is the most accepted model. However the present investigation of the Ouro Fino syncline allowed lo characterize four families of structures and to present a third hypothesis for its origin and evolution. The first family is characterized by tight and reclined folds with moderate ESE plunging fold axis. The other structures of this phase are the axial plane schistosity and the mineralization parallel to intersection lineation. The mineral lineation is characterized as a "b" type fabric. The iron mineralization is closely related to the strain parameters of the first deformation phase. The fold closure occurs in the southern region of the Ouro Fino Syncline defined by structural relationships between bedding and schistosity and by subsidiary "S" and "Z" folds in the west and east limbs respectively. These relations allow to redefine the Ouro Fino structure as a reclined antiformic synclinorial fold with southward vergence and axis plunging to ESE. In this case the paleoproterozoic Minas Supergroup units occupy the fold nucleus being enveloped by the archean Rio das Velhas Supergroup metapelitic sequence. The second deformation phase is characterized by tight to open folds with sub-horizontal N-S trending axes genetically related to the Córrego das Flechas fault. The notable fold for this phase is the footwall Córrego das Flechas synform with westward vergence. The other families are represented by crenulations and E-Wand N-S axial trending gentle folds.

Sinclinal Ouro Fino revisitado, Quadrilátero Ferrífero, Minas Gerais : uma hipótese sobre a sua origem e evolução.

The Ouro Fino Syncline is a regional structure with a controversial origin. Two structural models represent the synthesis to its tectonic evolution. On the one hand, according to Ladeira & Viveiros (1984), the Ouro Fino syncline is the result of the first folding event which affected the Minas Supergroup. These folds present E-W trending axes and associated penetrative cleavage. The tectonic transport vector is northward. On the other hand, Fonseca (1990), interprets Ouro Fino syncline’s origin as a result of compressive forces directed to westward. The axis has N-S trending and westward vergence according to Dorr (1969). This second point of view is the most accepted model. However the present investigation of the Ouro Fino syncline allowed to characterize four families of structures and to present a third hypothesis for its origin and evolution. The first family is characterized by tight and reclined folds with moderate ESE plunging fold axis. The other structures of this phase are the axial plane schistosity and the mineral lineation parallel to intersection lineation. The mineral lineation is characterized as a “b” type fabric. The iron mineralization is closely related to the strain parameters of the first deformation phase. The fold closure occurs in the southern region of the Ouro Fino Syncline defined by structural relationships between bedding and schistosity and by subsidiary “S” and “Z” folds in the west and east limbs respectively. These relations allow to redefine the Ouro Fino structure as a reclined antiformic synclinorial fold with southward vergence and axis plunging to ESE. In this case the paleoproterozoic Minas Supergroup units occupy the fold nucleus being enveloped by the archean Rio das Velhas Supergroup metapelitic sequence. The second deformation phase is characterized by tight to open folds with sub-horizontal N-S trending axes genetically related to the Corrego das Flechas fault. The notable fold for this phase is the footwall Corrego das Flechas synform with westward vergence. The other families are represented by crenulations and E-W and N-S axial trending gentle folds.

Observations of 1000 km s-1Doppler shifts in 107 K solar flare supra-arcade

An X1.5 flare on the west limb of the Sun on 21 April 2002 developed a large supra-arcade about 30 min after flare onset. The growth of the supra-arcade can be followed in both TRACE 195 A images and SUMER spectra. Its growth seems to be associated with dark (in TRACE images), sunward moving channels that descend onto the arcade from above. SUMER recorded Doppler shifts of 800–1000 km s−1 in Fe xxi 1354 A from positions where this sunward flow interacts with the arcade tops. We describe the observations, focusing on the relationship of the high Fe xxi line shifts to the sunward moving dark flows.

SUMER spectral observations of post-flare supra-arcade inflows

On 21 April 2002 a large eruptive flare on the west limb of the Sun developed a bright, very dynamic, post-flare arcade. In TRACE 195 A images, a series of dark, sunward moving flows were seen against the bright extreme ultraviolet (EUV) arcade. SUMER obtained a series of spectra of the dark EUV flows in the lines C ii, Fe xii, and Fe xxi at a fixed position above the limb. These spectra give spatially resolved line-of-sight velocities and emission measures for the arcade plasma over a temperature range 2×104 to 107 K. The flows are dark in all SUMER lines. The UV continuum longward (∼ 1350 A) and shortward (∼ 675 A) of the hydrogen Lyman limit is used to determine whether the dark 195 A inflows are due to regions of low plasma density (plasma voids) or cold absorbing material. There is some evidence of absorption near the front of one of the inflows; however, along most of the dark channels there is no change in continuum ratio and we therefore conclude, as originally suggested by McKenzie and Hudson (1999), that they are plasma voids.

Elemental Abundances for the 1996 Streamer Belt

The purpose of this paper is to explore the observed physical characteristics of the solar corona streamer belt from 1996 June 1 to August 5. The UV spectral data was collected by the Ultraviolet Coronagraph Spectrometer (UVCS) instrument on board the Solar and Heliospheric Observatory (SOHO) spacecraft. From this data the abundances of oxygen, silicon, and magnesium were determined on an almost daily basis during this time period at both the west and east limbs. The streamer belt was composed of both active and quiescent streamers, which allows for the comparison of these two categories of magnetic field structures. The depletion of these three elemental abundances in the core of quiescent streamers was found, as in previous observations. The variance in abundance with solar rotation was investigated as a consequence of the long time frame considered here. The first ionization potential (FIP) effect was present in the data, and the danger of exploring this phenomena by the traditional FIP bias was also covered. A comparison with in situ elemental abundance data from the Charge, Element, and Isotope Analysis System (CELIAS) instrument on SOHO provides evidence suggesting that active-region streamers and the outer ‘‘ leg ’’ structural component of quiescent streamers are definite contributors to the slow solar wind. Subject headings: solar wind — Sun: abundances — Sun: corona

Energetic Particles and Coronal Mass Ejections: a Case Study From ace and Ulysses

In 2001 the Ulysses spacecraft crossed the ecliptic plane near perihelion. The heliographic longitude with respect to the Earth was within ± 20° of the west solar limb while it was ± 15° of the ecliptic plane, which meant that coronal mass ejections seen off the solar west limb were likely to pass over Ulysses. On 10 May the largest > 38 keV electron intensity of the mission, since the Jovian encounter in 1992, was observed, which was accompanied by a fast perpendicular shock. This event was preceded by a fast coronal mass ejection some two and a half days earlier which is the probable source of the shock. However, both the ACE spacecraft and Ulysses observed, simultaneously, an intense, prompt electron event on 7 May from a solar flare associated with earlier coronal mass ejections also observed off the west limb; Ulysses was magnetically connected to a longitude well behind the west limb. ACE did not observe any (at the 0.1% level) energetic electrons which were associated with the 10 May event seen at Ulysses. We discuss in detail the energetic particles seen at the two spacecraft during 7–11 May, with the objective of understanding the origin of the intense electron event seen on 10 May and the manner in which particles escaping from the shock populate the inner heliosphere. The energy spectrum of the ions at both ACE and Ulysses exhibits a maximum at around 400 keV; this form of the spectrum was seen at the shock itself. It appears that the strong shock driven by the fast coronal mass ejection is able to populate a large fraction of the inner heliosphere with accelerated ions. The shock-accelerated electrons do not pervade the inner heliosphere in the same manner as the ions. We suggest that the electron acceleration was enhanced by the presence of multiple coronal mass ejections.

RHESSI observations of the coronal component ofsolar flare hard X-ray sources

Abstract In flares that occur behind the limb, the intense chromospheric (foot-point) part of the hard X-ray source is occulted, thus permitting good observations of the coronal component. Between 15 and 18 April 2002, RHESSI observed a series of small (GOES Class C) flares produced by the active region NOAA 9905 as it rotated behind the west limb. A preliminary analysis of the observed hard X-ray sources in the 17–18 April 2002 flares has confirmed that flare-associated sources of gradual 12–25 keV X-ray emission can exist in the corona at heights up to 27000 km.

Low‐latitude solar wind during the Fall 1998 SOHO‐Ulysses quadrature

The Fall 1998 Solar and Heliospheric Observatory (SOHO)‐Ulysses quadrature occurred when Ulysses was at 5.2 AU, 17.4°S of the equator, and off the west limb of the Sun. SOHO coronal observations, at heliocentric distances of a few solar radii, showed that the line through the solar center and Ulysses crossed, over the first days of observations, a dark, weakly emitting area and through the northern edge of a streamer complex during the second half of the quadrature campaign. Ulysses in situ observations showed this transition to correspond to a decrease from higher‐speed wind typical of coronal hole flow to low‐speed wind. Physical parameters of the low‐latitude coronal plasma sampled over the campaign are determined using constraints from what is the same plasma measured later in situ and simulating the intensities of the hydrogen Lyman‐α and OVI 1032 and 1037 Å lines, measured by the Ultraviolet Coronagraph Spectrometer on SOHO. It appears that low‐latitude wind from small coronal holes and polar wind have different characteristics in the corona, differences well known at interplanetary distances through in situ experiments. Small, low‐latitude coronal holes have a higher expansion factor than typical polar holes, and their plasma moves at a lower speed than plasma from polar holes, reaching, at 3.5 Rsun, only about one‐fifth of the terminal speed. Wind emanating from bright regions, above streamer complexes, is, at the altitudes we analyzed (i.e., 3.5 and 4.5 Rsun), about a factor 3 slower than the low‐latitude coronal hole wind, implying a shift to even higher altitudes of the region where plasma gets accelerated. We surmise that open field regions, interspersed amidst closed coronal loops/streamers, may be at least partially responsible for the well‐known slow wind speed variability. As in polar fast wind, OVI ions move faster than protons, over the range of altitudes we sampled, and are frozen‐in at temperatures of ≈1.3–1.5 106 K, depending on the site where the outflow originates. An oxygen abundance variation from a value of 8.55, in low‐latitude holes, to 8.73 in bright areas, has also been inferred.

Real-time forecasting of ICME shock arrivals at L1 during the "April Fool’s Day" epoch: 28 March – 21 April 2001

Abstract. The Sun was extremely active during the "April Fool’s Day" epoch of 2001. We chose this period between a solar flare on 28 March 2001 to a final shock arrival at Earth on 21 April 2001. The activity consisted of two presumed helmet-streamer blowouts, seven M-class flares, and nine X-class flares, the last of which was behind the west limb. We have been experimenting since February 1997 with real-time, end-to-end forecasting of interplanetary coronal mass ejection (ICME) shock arrival times. Since August 1998, these forecasts have been distributed in real-time by e-mail to a list of interested scientists and operational USAF and NOAA forecasters. They are made using three different solar wind models. We describe here the solar events observed during the April Fool’s 2001 epoch, along with the predicted and actual shock arrival times, and the ex post facto correction to the real-time coronal shock speed observations. It appears that the initial estimates of coronal shock speeds from Type II radio burst observations and coronal mass ejections were too high by as much as 30%. We conclude that a 3-dimensional coronal density model should be developed for application to observations of solar flares and their Type II radio burst observations.Key words. Interplanetary physics (flare and stream dynamics; interplanetary shocks) – Magnetosheric physics (storms and substorms)

UV line intensity and flow velocity distributions in two coronal mass ejections as deduced by UVCS-SOHO observations

The Ultraviolet Coronagraph Spectrometer (UVCS) instrument onboard the SOHO satellite observed two coronal mass ejections on November 2 and 3, 2000, related to the eruptions of a large filament structure in an active region close to the West limb of the Sun, and of a prominence near the South Pole, respectively. Intensity and profile of the O VI resonance doublet lines at 1032 and 1037 A and of Lyβ (1026 A) line, together with the intensity of some other minor ions, were observed using the O VI channel of UVCS. We analysed these spectroscopic observations in order to get information about the distributions of ionic densities and flow velocities in the solar coronal plasma ejected during these transient events. Emission in ions ranging from C II to O VI indicates a temperature range between 10 4 . 5 and 10 5 . 5 K. The morphology of the bright emission regions suggests the development of several strands of plasma irregularly distributed inside the CME structures, whose temporal evolution is significantly different from each other. The velocities determined for each bright element also give a complex picture of the plasma kinematics characterizing these coronal mass ejections.

Determining coronal heating of plasma loops through differential emission measure analysis

Insights into the nature of the heating mechanism in coronal plasmas can be gained through the analysis of differential emission measure curves localized along coronal loops. This technique is especially desirable since it does not require the isothermal approximation for potentially dynamic loop plasmas. Of particular interest are the parameters describing the heating rate per unit volume in the corona as a function of radial height and/or arc length along loop structures. Using simultaneous observations taken on 20 April 98 with the Solar and Heliospheric Observatory Coronal Diagnostics Spectrometer and the Yohkoh Soft X-ray Telescope, plasma temperature distributions were constructed in the relevant temperature domain (Log T = 5.5–7.5) along an isolated coronal loop on the west limb, ranging from one footpoint to the loop top. Subsequent analysis of the differential emission measure curves using abundance values from Fludra and Schmelz (1999) at each pixel combined with knowledge of the loop geometry helped to pin down the coronal heating mechanism using the method described by Priest et al. (1998).

Hard X-ray Microflares down to 3 keV

The excellent sensitivity, spectral and spatial resolution, and energy coverage down to 3 keV provided by the Reuven Ramaty High-Energy Solar Spectroscopic Imager mission (RHESSI) allows for the first time the detailed study of the locations and the spectra of solar microflares down to 3 keV. During a one-hour quiet interval (GOES soft X-ray level around B6) on 2 May, 1:40–2:40 UT, at least 7 microflares occurred with the largest peaking at A6 GOES level. The microflares are found to come from 4 different active regions including one behind the west limb. At 7′′ resolution, some events show elongated sources, while others are unresolved point sources. In the impulsive phase of the microflares, the spectra can generally be fitted better with a thermal model plus power law above ∼ 6–7 keV than with a thermal only. The decay phase sometimes can be fitted with a thermal only, but in some events, power-law emission is detected late in the event indicating particle acceleration after the thermal peak of the event. The behind-the-limb microflare shows thermal emissions only, suggesting that the non-thermal power law emission originates lower, in footpoints that are occulted. The power-law fits extend to below 7 keV with exponents between −5 and −8, and imply a total non-thermal electron energy content between 1026–1027 erg. Except for the fact that the power-law indices are steeper than what is generally found in regular flares, the investigated microflares show characteristics similar to large flares. Since the total energy in non-thermal electrons is very sensitive to the value of the power law and the energy cutoff, these observations will give us better estimates of the total energy input into the corona. (Note that color versions of figures are on the accompanying CD-ROM.)

Hard X-Radiation from a Fast Coronal Ejection

We have observed a high-speed coronal ejection in hard X-rays, detectable to an altitude of some 2 × 105 km in the Yohkoh 23-53 keV energy bands. Simultaneous imaging at 17 and 34 GHz from the Nobeyama radioheliograph shows complex moving features simultaneous with the ejection, including a compact source that we identify with the rapid X-ray source motion. The hard X-ray and microwave observations agree on ejection velocities in the vicinity of 1000 km s-1. The hard X-ray sources also corresponded in position angle to a bright coronal mass ejection (CME) detected about 15 minutes later and temporally to both fast-drift and slow-drift radio bursts in the decimeter-meter bands. Other components of coronal hard X-ray emission were also detected, including an extended long-duration event with a nonthermal spectrum. We suggest that a major eruptive flare occurred in NOAA Active Region 9415, approximately 26° beyond the west limb at the time of the event. Estimating a source density of 4 × 109 cm-3 from the compact source observed at 17 GHz, we find a total electron number (>20 keV) of approximately N20 ~ 1.3 × 1036 for the compact part of the source. We infer that these electrons were trapped in expanding loops forming a part of the CME and may have contributed substantial pressure within these loops.

Near‐Sun and near‐Earth manifestations of solar eruptions

We compare the near‐Sun and near‐Earth manifestations of solar eruptions that occurred during November 1994 to June 1998. We compared white‐light coronal mass ejections, metric type II radio bursts, and extreme ultraviolet wave transients (near the Sun) with interplanetary (IP) signatures such as decameter‐hectometric type II bursts, kilometric type II bursts, IP ejecta, and IP shocks. We did a two‐way correlation study to (1) look for counterparts of metric type II bursts that occurred close to the central meridian and (2) look for solar counterparts of IP shocks and IP ejecta. We used data from Wind and Solar and Heliospheric Observatory missions along with metric radio burst data from ground‐based solar observatories. Analysis shows that (1) most (93%) of the metric type II bursts did not have IP signatures, (2) most (80%) of the IP events (IP ejecta and shocks) did not have metric counterparts, and (3) a significant fraction (26%) of IP shocks were detected in situ without drivers. In all these cases the drivers (the coronal mass ejections) were ejected transverse to the Sun‐Earth line, suggesting that the shocks have a much larger extent than the drivers. Shocks originating from both limbs of the Sun arrived at Earth, contradicting earlier claims that shocks from the west limb do not reach Earth. These shocks also had good type II radio burst association. We provide an explanation for the observed relation between metric, decameter‐hectometric, and kilometric type II bursts based on the fast mode magnetosonic speed profile in the solar atmosphere.

Solar Origin of a Series of Well‐collimated Electron Events

We investigate the solar origin and propagation of a series of well-collimated energetic electron events on 1997 November 28, measured in situ by the Electron, Proton, and Alpha Monitor (EPAM) experiment on the Advanced Composition Explorer (ACE) spacecraft. EPAM measures electrons in the energy range from 40 to 300 keV over a wide range of look directions and with better than 1 minute time resolution. During the events in our study, the particles are strongly collimated along the magnetic field. As such, these near-relativistic (beta = 0.4-0.7) particles tend to be scatter free; their observed arrival at ACE provides a good estimate of the release time back to the Sun. EPAM results are extended to 20 keV using energetic electron data from the Three-Dimensional Plasma Analyzer (3DP) experiment on the Wind spacecraft. We combine these observations with fast imaging of the solar corona in the meter wave domain, provided by the Nancay radioheliograph, and dynamic spectral information from the WAVES experiment on the Wind spacecraft. Together, this complement of observations of solar energetic particle events provides insight into the onset times and sites of particle acceleration in the low corona. The imaging observations at metric radio frequencies show three series of similar events off the west limb of the Sun that extend into interplanetary type III bursts. The third event corresponds only to a strong in situ particle event. During this time period, the coronagraphic observations from the Large Angle and Spectrometric Coronagraph experiment on the Solar and Heliospheric Observatory reveal a dimming and some morphological changes in the vicinity of the injection site, suggesting the opening of the magnetic field. We find that for the strongest event, the onset times for the energetic particles measured at the spacecraft reveal a dispersion in the inferred release times back at the Sun versus energy. The onset times measured for electrons with energies above 60 keV are not compatible with the release time back at the Sun inferred from particles of lower energies. This dispersion can be successfully explained by the radio data: the radio observations reveal two successive energetic electron injections separated by a few minutes. The observations thus show that there is a later release of the particles detected in situ at higher energies and that the origin of the phenomenon is in the low corona.