Astronomy & Astrophysics manuscript no. aa_rev ©ESO 2021 September 14, 2021 Letter to the Editor The long period of 3He-rich solar energetic particles measured by Solar Orbiter on 2020 November 17–23 R. Bucˇík1, G. M. Mason2, R. Gómez-Herrero3, D. Lario4, L. Balmaceda4, 5, N. V. Nitta6, V. Kruparˇ4, 7, N. Dresing8, 9, G. C. Ho2, R. C. Allen2, F. Carcaboso3, J. Rodríguez-Pacheco3, F. Schuller10, A. Warmuth10, R. F. Wimmer-Schweingruber8, J. L. Freiherr von Forstner8, 11, G. B. Andrews2, L. Berger8, I. Cernuda3, F. Espinosa Lara3, W. J. Lees2, C. Martín8, 12, D. Pacheco8, M. Prieto3, S. Sánchez-Prieto3, C. E. Schlemm2, H. Seifert2, K. Tyagi2, 13, M. Maksimovic14, A. Vecchio14, 15, A. Kollhoff8, P. Kühl8, Z. G. Xu8, and S. Eldrum8 1 Southwest Research Institute, San Antonio, TX 78238, USA e-mail: radoslav.bucik@swri.org 2 Applied Physics Laboratory, Johns Hopkins University, Laurel, MD 20723, USA 3 Universidad de Alcalá, Space Research Group, 28805 Alcalá de Henares, Spain 4 Heliophysics Science Division, NASA Goddard Space Flight Center, Greenbelt, MD, USA 5 George Mason University, Fairfax, VA, USA 6 Lockheed Martin Advanced Technology Center, Palo Alto, CA 94304, USA 7 Goddard Planetary Heliophysics Institute, University of Maryland, Baltimore County, Baltimore, MD 21250, USA 8 Institut für Experimentelle und Angewandte Physik, Christian-Albrechts-Universität zu Kiel, Kiel, Germany 9 now at Department of Physics and Astronomy, University of Turku, FI-20014 Turku, Finland 10 Leibniz-Institut für Astrophysik Potsdam, D-14482 Potsdam, Germany 11 now at Paradox Cat GmbH, 80333 München, Germany 12 now at German Aerospace Center (DLR), Berlin, Germany, Dept. of Extrasolar Planets and Atmospheres 13 now at Univ. Colorado/LASP, Boulder, CO, USA 14 LESIA, Observatoire de Paris, Université PSL, CNRS, Sorbonne Université, Université de Paris, France 15 Radboud Radio Lab, Department of Astrophysics, Radboud University Nijmegen, The Netherlands Received ; accepted ABSTRACT We report observations of a relatively long period of 3He-rich solar energetic particles (SEPs) measured by Solar Orbiter. The period consists of several well-resolved ion injections. The high-resolution STEREO-A imaging observations reveal that the injections coin- cide with EUV jets/brightenings near the east limb, not far from the nominal magnetic connection of Solar Orbiter. The jets originated in two adjacent, large, and complex active regions as observed by the Solar Dynamics Observatory when the regions rotated to the Earth’s view. It appears that the sustained ion injections were related to the complex configuration of the sunspot group and the long period of 3He-rich SEPs to the longitudinal extent covered by the group during the analyzed time period. Key words. acceleration of particles – Sun: abundances – Sun: flares – Sun: particle emission 1. Introduction 3He-rich solar energetic particle (SEP) events show enormous enhancements of rare species such as the nuclide 3He and ultra- heavy elements by factors up to ∼104 above the nominal coro- nal abundances (e.g., Mason 2007; Reames 2021). The events are highly associated (>95%) with type III radio bursts (e.g., Reames & Stone 1986; Nitta et al. 2006), the emission generated by ∼10–100 keV outward streaming electrons. Solar sources of 3He-rich SEPs have been associated with EUV jets (Bucˇík 2020, and references therein), suggesting acceleration via magnetic reconnection involving field lines open to interplanetary space (Reames 2002). Progress in understanding 3He-rich SEPs has been hampered by the low intensities and short duration of these events. Solar Orbiter (Müller et al. 2020) will enable unprece- dented studies of small-size 3He-rich SEP events combining in- situ and remote-sensing observations close to the Sun. The first Solar Orbiter 3He-rich SEP events were measured during the spacecraft’s first perihelion pass from 0.52 to 0.96 au (Mason et al. 2020) in June–September 2020. Three events out of the five discrete events reported by Mason et al. (2020) have a 0.2–2 MeV/nucleon 3He/4He ratio above 10% with a maximum 3He/4He ratio of 0.61. In this paper, we report a relatively long period of 3He-rich SEPs, spanning almost 7 days in November 2020, observed by Solar Orbiter near 0.9 au. Such a long period may indicate a nearly continuous 3He-rich SEP injection into the interplanetary space (Mason 2007). 2. Observations The 3He-rich SEPs reported in this paper were measured by the Suprathermal Ion Spectrograph (SIS) of the Energetic Particle Detector (EPD) suite (Rodríguez-Pacheco et al. 2020) aboard Solar Orbiter. SIS is a time-of-flight mass spectrometer that mea- sures elemental composition from H through ultra-heavy nu- clei in the kinetic energy range of ∼0.1–10 MeV/nucleon. SIS has two telescopes, one pointing 30◦ (sunward) and the other Article number, page 1 of 7 ar X iv :2 10 9. 05 57 0v 1 [a str o- ph .SR ] 12 Se p 2 02 1 A&A proofs: manuscript no. aa_rev Earth Sun A SO 1,2,3 4 5 Fig. 1. Ecliptic plane projection of the Solar Orbiter (SO), STEREO- A (A), and the Earth in 2020 November 17–23. Two overlapping rings mark SO at the beginning and end of the examined period. The arrows indicate the longitude of the solar source associated with ion injections. Parker spiral for 340 km/s solar wind speed connecting to Solar Orbiter is shown. The dashed line corresponds to 1 au orbit. 160◦ (anti-sunward) to the west of the spacecraft-Sun line. We also use energetic electron measurements made by the Elec- tron Proton Telescope (EPT) of EPD, which covers energies (20–400 keV) between two other instruments of the EPD suite, STEP, and HET. The first year of operations, and details of the data products, provided by EPD, can be found in Wimmer- Schweingruber et al. (2021). Solar sources of 3He-rich SEPs were examined using high- resolution EUV images from the SECCHI/EUVI instrument (Howard et al. 2008) on the STEREO-A (STA). The EUVI pro- vides full-disk images of the Sun with 3′′ spatial and 5-minute nominal temporal resolution in four wavelength channels (304, 171, 195, and 284 Å). We use the 195 Å images that have the highest temporal resolution (5.0 and 2.5 minutes) in the exam- ined period. The Extreme-Ultraviolet Imager (EUI; Rochus et al. 2020) on Solar Orbiter provides images only with limited spa- tial and temporal resolution during the aforementioned period. We note that until November 2021, Solar Orbiter is in the cruise phase when remote-sensing instruments are only occasionally switched on for calibration. Further, we inspect radio spectro- grams for the presence of the associated type III radio bursts. The radio data are provided by the Solar Orbiter Radio and Plasma Waves (RPW; Maksimovic et al. 2020) and the STEREO-A Waves (Bougeret et al. 2008) instruments with a frequency range (<16 MHz) covering emission generated from about ∼2 R to 1 au. We also make use of full-disk line-of-sight magnetograms obtained from Helioseismic and Magnetic Imager (HMI; Scher- rer et al. 2012) onboard Solar Dynamics Observatory (SDO). The location of Solar Orbiter and STEREO-A during the in- vestigated period is shown in Fig. 1. Solar Orbiter traveled from 0.93 to 0.91 au; STEREO-A remained at 0.96 au. Both spacecraft were near the ecliptic plane, Solar Orbiter at −6◦ and STEREO- A at 7◦ of heliographic latitude. The angular separation between Solar Orbiter and STEREO-A was 180◦. SDO is in orbit around the Earth. Figure 2 displays Solar Orbiter EPT and SIS measurements in 2020 November 17–23. Figure 2a presents 30-minute electron intensities at different energy bins between 41.8 and 105.7 keV. 101 102 103 104 #/ (cm 2 s sr M eV ) 41.8 51.6 65.1 82.6 105.7 10-3 10-2 10-1 100 101 102 #/ (cm 2 s sr M eV /n uc ) H 3He 4He O Fe 10 M as s (A MU ) 17 18 19 20 21 22 23 24 2020 Nov (UT) 0 5 10 15 20 1/ Io n Sp ee d (ho ur/ AU ) 10-1 100 101 En er gy (M eV /nu c) En er gy (M eV /nu c) En er gy (M eV /nu c) En er gy (M eV /nu c) En er gy (M eV /nu c) Date 2020 Nov (a) (b) (c) (d) 2020/322 323 324 325 326 327 328 329 Fe Ca SSi Mg Ne ON C 4He 3He 1 23 4 5 Fig. 2. (a) 30-minute electron intensities from EPT (41.8–105.7 keV) sunward pointing sensor. Dashed vertical lines mark type III radio bursts listed in Table 1. (b) SIS 1-hr H, 3He, 4He, O, and Fe intensities at 0.23– 0.32 MeV/nucleon. (c) SIS mass spectrogram at 0.4–10 MeV/nucleon. (d) SIS 1/speed vs. arrival times of 2–70 AMU ions. Sloped dashed lines approximately mark the ion injections. SIS measurements are from both telescopes averaged together. The gap around the noon of November 19 was caused by EPD being shut down for software maintenance. Figure 2b shows hourly averages of the 0.23–0.32 MeV/nucleon H, 3He, 4He, O and Fe intensities as measured by both tele- scopes of SIS. Three major increases are seen at the intensity time profiles that start near the end of November 17, near the end of November 18, and around midday of November 21. It is clearly seen that the first two increases are 3He- and Fe- rich. The mass spectrogram in Fig. 2s¸hows almost continuous 3He presence from the middle of November 17 through the end of November 23, i.e., ∼6.5 days. The inverse ion-velocity time spectrogram in Fig. 2d shows at least three ion injections con- tributing to the first increase on November 17–18; one injection contributing to the increase on November 18–20 and one ion in- jection for the increase on November 21. These injections can be identified based on their characteristic triangular pattern in the inverse speed plots. Figure 3 shows the fluence energy spec- tra for selected ion species in injection #3 where enhancements in all ion species were observed without the inconvenient data gaps as in the case of injection #4. The 3He, O, and Fe show rollovers toward low energies as it has been previously reported in many 3He-rich SEP events (e.g., Mason et al. 2000). The ion fluences for injections #1, #2, and #3 are integrated in swoosh boxes bounded by slanted lines 1 & 2, 2 & 3, and 3 & 4, respec- tively. For injection #4 the swoosh box is between the slanted line 4 and November 20 19:26 UT and for injection #5 between slanted line 5 and November 22 08:24 UT. Table 1 lists the characteristics of the 3He-rich period. Col- umn 1 indicates the injection number, column 2 the ion injection time at the Sun (as a day of year and date), estimated by extrap- Article number, page 2 of 7 R. Bucˇík et al.: The long period of 3He-rich solar energetic particles Table 1. Characteristics of the 3He-rich period. See text for more details. Ion injection Type III STA EUVI event Separation Elec. injection 3He/4Hec Fe/Oc time (UT) start (UT) Typea Location angleb (◦) time (UT) 1 322.42 Nov-17 10:05 09:49 [41] B E90S22 20 09:20 0.61±0.08 2.00±0.37 2 322.62 Nov-17 14:53 15:28 [20] B E90S18 20 ... 0.22±0.03 0.63±0.06 3 322.74 Nov-17 17:46 18:20 [12] J E90S18 20 18:20 0.90±0.03 0.91±0.01 18:24 [16] . . . E90S18 20 . . . . . . . . . 4 323.54 Nov-18 12:58 13:08 [00] B E85S23 25 13:10 0.56±0.01 1.35±0.01 14:09 [01] B E85S23 25 14:07 . . . . . . 5 325.87 Nov-20 20:53 19:34 [26] B E48S19 62 19:30 0.32±0.03 0.76±0.03 20:33 [25] J E52S17 58 . . . . . . . . . 21:00 [52] B E48S19 62 . . . . . . . . . Notes. (a) B: brightening; J: jet (b) between Solar Orbiter magnetic footpoint longitude on the Sun and the longitude of the EUVI event (c) 0.2– 2.0 MeV/nucleon Fig. 3. Fluence spectra for selected species in injection #3. olation of dispersive signature of individual ions in the inverted velocity-time spectrogram indicated by the inclined dashed red lines in Fig. 2d. The uncertainty in the injection time estimated by this method is ±45 min (Mason et al. 2000; Wang et al. 2016). Column 3 gives the associated type III radio burst start times as observed by STEREO-A/Waves at 16 MHz. Note, the RPW showed an enhanced level of interference at higher frequencies. Multiple type III bursts appear to contribute to injections #3, #4, and #5 (see Fig. 4). To compare with the estimated ion injection times, the square brackets show minutes of the type III bursts start times after subtraction of the light travel time (∼8 min). It is unclear if the type III burst at 19:34 UT is associated with in- jection #5; it occurs too early to be within ±45 min error of ion estimated release time. We note that injection #5 is the weak- est of all the injections and the magnetic connection to the site was interrupted early in the event, around ∼08:00 on November 21 (Fig. 2d), as indicated by the abrupt drop in ion counts at all energies. Therefore, the estimated injection time is only tenta- tive, and the association with the type III at 19:34 UT could not be ruled out. Columns 4 and 5 provide the type and location of the associated parent solar eruption, respectively, as seen in the EUVI on STEREO-A, where J indicates a clear EUV jet moving away from the parent active region and B indicates just a bright- ening seen in the EUV images without apparent outward move- ment. We cannot identify the type of EUVI event in 5-minute resolution images for the 2nd type III burst, corresponding to in- jection #3, that occurred only 4 minutes after the previous type III burst. Column 6 indicates a separation angle between Solar Orbiter magnetic footpoint longitude on the Sun and the longi- tude of the EUVI event. The magnetic footpoint of Solar Orbiter, based on simple Parker spiral approximation and assuming solar wind speed of 340 km/s, was ∼W70 which corresponds to E110 from STEREO-A view. The value of 340 km/s is the median so- lar wind speed measured by SWEPAM (McComas et al. 1998) on ACE nine days earlier (November 8–12), which corresponds to the solar rotation between the L1 and Solar Orbiter separated by 122◦. The Solar Orbiter Solar Wind Analyser (SWA; Owen et al. 2020) data were not available for the examined period. Col- umn 7 shows the electron injection time, estimated from the in- verted velocity-time spectrogram (not shown) of 1-min averaged EPD electron data. Columns 8 and 9 provide 3He/4He and Fe/O ratios at 0.2–2.0 MeV/nucleon. Figure 4 shows Solar Orbiter and STEREO-A radio spectro- grams where we have indicated the type III radio bursts asso- ciated with the ion injections. The presence of high frequencies at Waves in all the bursts suggests that the source was not be- hind the east limb as seen from STEREO-A. The 2nd type III bursts in injections #4 and #5 were weak at Solar Orbiter. Dur- ing ion injection #3, two small dispersive electron events were detected by EPD with solar injections on November 18 10:25, and 11:45 UT. The later one appears as a small peak in the EPT intensity-time profile at ∼12:00 UT on November 18 (Fig. 2a). Figure 4d shows a type III burst associated with the electron in- jection at 11:45 UT. The type III burst associated with the elec- tron injection at 10:25 UT is clearly observed only by Solar Or- biter (see Fig. 4d for low-frequency part, ∼0.05 MHz, between 11:30 and ∼13:00 UT). During ion injection #4, another small dispersive electron event was measured with solar injection on November 19 06:00 UT (see Fig. 2a for the peak at ∼07:00 UT on November 19). Figure 4e shows the associated type III radio burst. The type III bursts related to these electron events were accompanied by EUV jets (see Fig. A.3). To identify solar sources, we inspect full-disk solar images for EUV brightenings as seen by STEREO-A that temporally coincide with the type III radio burst associated with the ion in- jection. Figure 5 shows the EUV activity around the times of the type III radio burst for injection #1 (top row), #2 (middle row), and #3 (bottom row). We do not see clear jets for injections #1 and #2 in EUVI images. The EUV images of the solar source Article number, page 3 of 7 A&A proofs: manuscript no. aa_rev Fig. 4. (a–f) Solar Orbiter/RPW and STEREO-A/Waves radio spectrograms. (a–d) and (f) correspond to ion injections #1 – #4 and #5, respectively. (e) corresponds to an electron event. The vertical dashed lines mark the start times of type III radio bursts associated with the ion injections. for injections #4 and #5 are shown in Appendix A. We note that SDO was not well located to observe EUV activity related to the origin of these ion injections (Fig. 1). However, for injection #5 the SDO Atmospheric Imaging Assembly (AIA; Lemen et al. 2012) observed the EUV jets, from region ∼16◦ behind the east limb, that temporally match all three type III bursts. On November 17–18, the EUI on Solar Orbiter provides im- ages only with 1-hr cadence. On November 19–20, there are also higher cadence data, but they either cover only short pe- riods or they have a low spatial resolution. In any case, periods of jets/brightening on November 19 and 20 were missed by EUI. Figure 6 shows SDO HMI magnetograms on November 24 00:00 UT (Left) and November 28 00:00 UT (Right). The EUV activity observed by STEREO-A on November 17-20 likely originated in two adjacent, large active regions (ARs) 12785 and 12786 that appeared near the east limb, as viewed from Earth early on November 23. The latitude of the jets/brightenings as indicated in Table 1 and as seen in Fig. 5 and Fig. A.1– A.3 matches well with the latitudes of these two ARs. It is particularly well seen in Fig. A.2, where the constellations of bright areas are similar to the positions of these two ARs. Thus, the brightenings in injection #5 can be clearly associ- ated with AR 12785, while the jet occurred between these ARs. The AR 12786 shows a complex βγ magnetic class1 and sunspot area of 1000 millionths of the solar hemisphere (MH) on November 25–26. The AR 12785 has a simple β magnetic class and sunspot area of 140 MH (November 23–24). This in- formation is provided by the USAF/NOAA Solar Region Sum- mary (ftp.swpc.noaa.gov/pub/warehouse/2020/SRS). The mag- 1 βγ denotes a bipolar sunspot group with no clearly marked line sep- arating spots of opposite polarity; β indicates a bipolar sunspot group -600 -400 -200 17-Nov-2020 09:45:30 injection #1 17-Nov-2020 09:50:30 17-Nov-2020 09:55:30 -600 -400 -200 17-Nov-2020 15:25:30 injection #2 17-Nov-2020 15:30:30 17-Nov-2020 15:35:30 -1100 -900 -700 -600 -400 -200 17-Nov-2020 18:20:30 injection #3 -1100 -900 -700 17-Nov-2020 18:25:30 -1100 -900 -700 17-Nov-2020 18:30:30 X (arcsec) Y (ar cs ec ) Fig. 5. STEREO-A 195 Å EUV running difference images correspond- ing to injection #1 (top row), #2 (middle row), and #3 (bottom row). The arrow marks the solar source. The heliographic longitude-latitude grid has a 15◦ spacing. netic complexity of the ARs decreased after they crossed the central meridian (as observed by SDO) on November 29–30. Article number, page 4 of 7 R. Bucˇík et al.: The long period of 3He-rich solar energetic particles -1000 -500 0 500 1000 X (arcsec) -1000 -500 0 500 1000 Y (ar cs ec ) 24-Nov-2020 00:00:24 12786 12785 -500 0 500 1000 28-Nov-2020 00:00:23 X (arcsec) 12786 12785 12787 12789 Fig. 6. SDO HMI line-of-sight magnetograms (scaled to ±100 G). The numbers mark the NOAA active regions of interest. The heliographic longitude-latitude grid has a 15◦ spacing. These ARs were seen by STEREO-A in EUV for the first time on November 19 (they were not reported in the previous rota- tion), and therefore we do not know what their properties were on November 17. As STEREO-A does not have a magnetograph, the magnetic class and area of the ARs were unknown when the examined activity was occurring. Marked are also two small ARs 12787 and 12789 (Fig. 6 Right) that could be located close to the Solar Orbiter nominal magnetic footpoint longitude. If these regions were in the hidden hemisphere, we cannot confirm/rule out that there was some simultaneous activity occurring in them as well. However, it is improbable that these regions dominated the observed long period of 3He-rich SEPs as all type III radio bursts temporally coincide with jets/brightening in AR 12786 and 12785. 3. Discussion and conclusion The relatively long period of 3He-rich SEPs observed by So- lar Orbiter is related to the recurrent activity (brightening and jets) in a large and complex group of sunspots in two adjacent ARs. Recurrent 3He-rich SEP events have been found to origi- nate from active regions at the boundary of low-latitude coronal holes (e.g., Wang et al. 2006; Bucˇík et al. 2014). There are only a few reports of 3He-rich SEPs associated with sunspot jets (Nitta et al. 2008; Bucˇík et al. 2018) and none report recurrent ion in- jection. The configuration with two large and complex nearby ARs may be favorable for the recurrent particle injections in the sense that there may be long-lived interaction between the negative polarity of one AR and the positive polarity of neigh- boring AR leading to the magnetic reconnection. Furthermore, these two ARs produce a longitudinally extended source (∼40◦) where spacecraft may be magnetically connected for a long pe- riod as the Sun rotates. We note that this extended region is rotat- ing away from Solar Orbiter such that the magnetic connection is presumably weakening with time. Kocharov et al. (2008) have studied extended periods of 3He- rich SEPs where most of them showed no dispersive onset. The authors have suggested that the temporal confinement of ions in the solar wind structures is an essential factor in the occurrence of such periods. Chen et al. (2015) have reported two relatively long, 4-day, periods of 3He-rich SEPs that were produced by recurring injections originating in plage regions from dispersed sunspots. While Chen et al. (2015) have identified two injections per period, we report at least five ion injections responsible for a long period. There might be other unresolved ion injections dur- ing the decay phase of the 1st and the 2nd ion intensity increases. The recurrent production of 3He-rich SEPs appears to occur in different magnetic environments that include plages, coronal holes, and sunspots and may be the result of a common process. Further studies may confirm whether complex and longitudinally extended sunspot groups are responsible for longer 3He-rich SEP periods compared to simple and small size sources. Acknowledgements. The Suprathermal Ion Spectrograph (SIS) is a European facility instrument funded by ESA. The SIS instrument was constructed by the JHU/Applied Physics Lab. and CAU Kiel. We thank the many individ- uals at ESA and within the Energetic Particle Detector team for their sup- port in its development. Post launch operation of SIS at APL is funded by NASA contract NNN06AA01C, and we thank NASA headquarters and the NASA/GSFC Solar Orbiter project office for their continuing support. The UAH team acknowledges the financial support by the Spanish Ministerio de Cien- cia, Innovación y Universidades FEDER/MCIU/AEI Projects ESP2017-88436- R and PID2019-104863RB- I00/AEI/10.13039/501100011033. The CAU Kiel team thanks the German Federal Ministry for Economic Affairs and Energy and the German Space Agency (Deutsches Zentrum für Luft- und Raumfahrt, e.V., (DLR)) for their unwavering support under grant numbers 50OT0901, 50OT1202, 50OT1702, and 50OT2002; and ESA for supporting the build of SIS under contract number SOL.ASTR.CON.00004, and the University of Kiel and the Land Schleswig-Holstein for their support of SIS. F. Carcaboso ac- knowledges the financial support by the Spanish MINECO-FPI-2016 predoc- toral grant with FSE. V. Kruparˇ acknowledges the support by NASA under grants 18-2HSWO2182-0010 and 19-HSR-192-0143. The work was partly sup- ported by NASA grant 80NSSC19K0079. LB acknowledges the support from the NASA program NNH17ZDA001N-LWS (Awards Nr. 80NSSC19K1261 and 80NSSC19K1235). References Bougeret, J. L., Goetz, K., Kaiser, M. L., et al. 2008, Space Sci. Rev., 136, 487 Bucˇík, R. 2020, Space Sci. Rev., 216, 24 Bucˇík, R., Innes, D. E., Mall, U., et al. 2014, Astrophys. J., 786, 71 Bucˇík, R., Wiedenbeck, M. E., Mason, G. M., et al. 2018, Astrophys. J. Lett., 869, L21 Chen, N.-H., Bucˇík, R., Innes, D. E., & Mason, G. M. 2015, Astron. Astrophys., 580, A16 Article number, page 5 of 7 A&A proofs: manuscript no. aa_rev Howard, R. A., Moses, J. D., Vourlidas, A., et al. 2008, Space Sci. Rev., 136, 67 Kocharov, L., Laivola, J., Mason, G. M., Didkovsky, L., & Judge, D. L. 2008, Astrophys. J. Suppl. Ser., 176, 497 Lemen, J. R., Title, A. M., Akin, D. J., et al. 2012, Sol. Phys., 275, 17 Maksimovic, M., Bale, S. D., Chust, T., et al. 2020, A&A, 642, A12 Mason, G. M. 2007, Space Sci. Rev., 130, 231 Mason, G. M., Dwyer, J. R., & Mazur, J. E. 2000, Astrophys. J. Lett., 545, L157 Mason, G. M., Ho, G. C., Allen, R., et al. 2020, A&A, in press McComas, D. J., Bame, S. J., Barker, P., et al. 1998, Space Sci. Rev., 86, 563 Müller, D., St. Cyr, O. C., Zouganelis, I., et al. 2020, A&A, 642, A1 Nitta, N. V., Mason, G. M., Wiedenbeck, M. E., et al. 2008, Astrophys. J. Lett., 675, L125 Nitta, N. V., Reames, D. V., De Rosa, M. L., et al. 2006, Astrophys. J., 650, 438 Owen, C. J., Bruno, R., Livi, S., et al. 2020, A&A, 642, A16 Reames, D. V. 2002, Astrophys. J. Lett., 571, L63 Reames, D. V. 2021, Solar Energetic Particles, Vol. 978 (Springer, Cham.) Reames, D. V. & Stone, R. G. 1986, Astrophys. J., 308, 902 Rochus, P., Auchère, F., Berghmans, D., et al. 2020, A&A, 642, A8 Rodríguez-Pacheco, J., Wimmer-Schweingruber, R. F., Mason, G. M., et al. 2020, A&A, 642, A7 Scherrer, P. H., Schou, J., Bush, R. I., et al. 2012, Sol. Phys., 275, 207 Wang, L., Krucker, S., Mason, G. M., Lin, R. P., & Li, G. 2016, Astron. Astro- phys., 585, A119 Wang, Y. M., Pick, M., & Mason, G. M. 2006, Astrophys. J., 639, 495 Wimmer-Schweingruber, R. F., Janitzek, N. P., Pacheco, D., et al. 2021, A&A, submitted Article number, page 6 of 7 R. Bucˇík et al.: The long period of 3He-rich solar energetic particles -600 -400 -200 18-Nov-2020 13:05:30 18-Nov-2020 13:10:30 18-Nov-2020 13:15:30 -1100 -900 -700 -600 -400 -200 18-Nov-2020 14:05:30 -1100 -900 -700 18-Nov-2020 14:10:30 -1100 -900 -700 18-Nov-2020 14:18:10 X (arcsec) Y (ar cs ec ) Fig. A.1. Same as Fig. 5 but for injection #4. The top row corresponds to the 1st, and the bottom to the 2nd type III burst. -600 -400 -200 20-Nov-2020 19:33:00 20-Nov-2020 19:35:30 20-Nov-2020 19:38:00 -600 -400 -200 20-Nov-2020 20:30:30 20-Nov-2020 20:33:00 20-Nov-2020 20:35:30 -1100 -900 -700 -600 -400 -200 20-Nov-2020 20:58:00 -1100 -900 -700 20-Nov-2020 21:00:30 -1100 -900 -700 20-Nov-2020 21:03:00 X (arcsec) Y (ar cs ec ) Fig. A.2. Same as Fig. 5 but for injection #5. The top row corresponds to the 1st, middle to the 2nd, and bottom to the 3rd type III burst. Appendix A: EUV images of the solar sources The solar sources associated with injections #4, #5, and the elec- tron events/type III bursts that occurred during decay phase of the 1st and 2nd ion intensity increases are shown in Fig. A.1, Fig. A.2, and Fig. A.3, respectively. -600 -400 -200 18-Nov-2020 10:25:30 18-Nov-2020 10:30:30 18-Nov-2020 10:35:30 -600 -400 -200 18-Nov-2020 11:50:30 18-Nov-2020 11:55:30 18-Nov-2020 12:00:30 -1100 -900 -700 -600 -400 -200 19-Nov-2020 06:05:30 -1100 -900 -700 19-Nov-2020 06:10:30 -1100 -900 -700 19-Nov-2020 06:15:30 X (arcsec) Y (ar cs ec ) Fig. A.3. Same as Fig. 5 but for type III bursts that occurred during the decay phase of the 1st (top and middle panels) and the 2nd ion intensity increase (bottom panel). Article number, page 7 of 7