Spectral Properties of Large Gradual Solar Energetic Particle Events Paper No. SH041 35th ICRC July 12-21, 2017 Busan, South Korea Mihir I. Desai Southwest Research Institute San Antonio, Texas G. M. Mason, M. A. Dayeh, R. W. Ebert, D. J. McComas, G. Li, C. M. S. Cohen, N. A. Schwadron, R. A. Mewaldt, & C. W. Smith

Outline CME associated SEP events Model predictions/expectations Spectral Properties Q/M-dependence Summary Implications for SPP and SolO Desai, M. I., et al. (2016a). The Astrophysical Journal, 816(2), 68. http://doi.org/10.3847/0004-637X/816/2/68 Desai, M. I., & Giacalone, J. (2016). Living Reviews in Solar Physics, 1–132. http://doi.org/10.1007/s41116-016-0002-5 Desai, M. I., et al. (2016b). The Astrophysical Journal, 828(2), 1–19. http://doi.org/10.3847/0004-637X/828/2/106

CME Shocks: SEPs and ESPs CMEs drive shocks in the corona and the interplanetary medium Accel. Near-SunSEPs Accel. Near-EarthESPs ESP SEP Reames.SSR, 1999

Q/M-dependent Spectral breaks Heavy ion spectral breaks show a Q/M-dependence Consistent with turbulence near and upstream of CME shock Mewaldt et al. (2008) Biggest Explosions (in tons of TNT equivalent) Conventional TNT 3.8x103 Atomic bomb (Hiroshima) 1.3x104 Hydrogen bomb 2x107 Tunguska event (1908) 1x107 Mount Pinatubo (1991) 7x107 Largest earthquake (Chile 1960) 3x109 Asteroid that wiped out dinosaurs 1014 Comet Shoemaker-Levy 9 on Jupiter 1014 Largest solar flare 1016 Top right: Flare/CME eruption on July 11, 1998 195 Angstrom 500,000 to 2,000,000 K. This sequence of extreme ultraviolet images shows brightening and expansion of coronal loops accompanied by a violent ejection of material away from the Sun's surface. A bubble of hot, ionized gas or plasma as well as cooler, dark filamentary material, erupts from the solar corona and travels through space at a high speed. The cross-shaped pattern at the peak of the loop brightening is an artifact of the instrument response rather than a feature on the Sun. Bottom left: Flare/CME eruption on July 11, 1998 1600 Angstrom 4,000 to 10,000 K. This sequence of ultraviolet images shows the the same CME/flare event as seen in the chromosphere. The dark region at the center of the activity is a sunspot. Image Credits: NASA/TRACE: http://trace.lmsal.com/Science/ScientificResults/trace_cdrom/html/mov_page.html

Latest Models Acceleration at Perpendicular shockslower turbulence weaker Q/M-dependence Acceleration at quasi-parallel shocks  enhanced turbulence stronger Q/M-dependence Using scattering in interplanetary medium, Li & Lee (2016); Zhao et al (2016) predict double power-laws with weaker Q/M-dependence Li et al. (2009) Biggest Explosions (in tons of TNT equivalent) Conventional TNT 3.8x103 Atomic bomb (Hiroshima) 1.3x104 Hydrogen bomb 2x107 Tunguska event (1908) 1x107 Mount Pinatubo (1991) 7x107 Largest earthquake (Chile 1960) 3x109 Asteroid that wiped out dinosaurs 1014 Comet Shoemaker-Levy 9 on Jupiter 1014 Largest solar flare 1016 Top right: Flare/CME eruption on July 11, 1998 195 Angstrom 500,000 to 2,000,000 K. This sequence of extreme ultraviolet images shows brightening and expansion of coronal loops accompanied by a violent ejection of material away from the Sun's surface. A bubble of hot, ionized gas or plasma as well as cooler, dark filamentary material, erupts from the solar corona and travels through space at a high speed. The cross-shaped pattern at the peak of the loop brightening is an artifact of the instrument response rather than a feature on the Sun. Bottom left: Flare/CME eruption on July 11, 1998 1600 Angstrom 4,000 to 10,000 K. This sequence of ultraviolet images shows the the same CME/flare event as seen in the chromosphere. The dark region at the center of the activity is a sunspot. Image Credits: NASA/TRACE: http://trace.lmsal.com/Science/ScientificResults/trace_cdrom/html/mov_page.html

Survey of 46 large SEPs Selected 46 SEP events from solar cycles 23 & 24 with no ESP or local shock accelerated component Mostly western events Identified SEP sampling time intervals, taking account of velocity dispersion effects Obtained ~0.1-500 MeV/nucleon H-Fe energy spectra from ACE/ULEIS, ACE/SIS, GOES, SAMPEX, SoHO/ERNE Fit the Band function (double-power law) to each species in each SEP event  obtain 2 power-law indices and a spectral break energy Analyze spectral properties, spectral breakpoints vs. Q/M-dependence

Sampling Interval using 1/Ion speed Sampling time intervals take account of velocity dispersion effects Integrate over whole event

Spectral Fits Band Function yields a low and a high-energy power-law index and a break energy for each species Band et al., (1993); Tylka et al.(2001;2005); Mewaldt et al. (2005)

Scatterplot of Band-fit slopes γa<γb i.e., spectra steepen with increasing energy

H, Fe, O Low-energy slopes Spectral slopes are correlated All species have similar slopes during an event

H, Fe, O High-energy slopes Range: 2-7 Spectral slopes are correlated All species have similar slopes during an event

Types of Q/M-dependence Both axes on right normalized to H Modest Q/M-dependence (α~1) Average Large SEP Q-states (Klecker et al 2000)

Types of Q/M-dependence Weak Q/M-dependence (α~0.2)

Types of Q/M-dependence GLE event Strong Q/M-dependence (α~2.8) Strong scattering

Histogram of alpha EB/EX∝[Q/M]α α – power-law exponent of the Q/M-dependence of spectral break energy α is constrained within 0.2-2; Mean and median ~1 α >2 in 3 events; 2 are GLEs Consistent with Li et al. (2009); Schwadron et al. (2015)

Alpha vs. CME speed Larger α for faster CMEs 6 GLE-associated SEPs have larger α; 2 events α>2 At odds with models that predict double power-laws in GLEs are due to scatter-dominated transport (Li & Lee; 2015; Zhao et al.2016)

Alpha vs. ~0.2 Fe/O Weak correlation Larger α and higher Fe/O for “extreme” SEPs

Alpha vs. 3He/4He No overall trend 21 events are 3He-rich 5 of the 9 extreme SEPs are also 3He-rich SW Value

Observations/Results Conclusions/Implications Summary Observations/Results Conclusions/Implications SEP spectra are well-represented by double power-laws with species-independent spectral slopes that steepen above break energies Formation of double power-laws is consistent with CME shock acceleration (Li et al., 2009; Schwadron et al., 2015) Break energies exhibit Q/M-dependence that scales as EB/EX∝[Q/M]α Consistent with the equal diffusion coefficient or resonance condition (Li et al. 2009; Schwadron et al., 2015) α is constrained within 0.2-2, and α ≥1.4 in extreme SEPs Consistent with acceleration-dominated (Li et al. 2009) models rather than transport-dominated models (Li & Lee 2015; Zhao et al 2016) Many extreme events are Fe- and 3He-rich, and have large values of α (≥1.4) Enhanced wave power enable near-Sun CME shocks to accelerate flare-rich material more efficiently than the ambient coronal material (Desai et al. 2016a;b) Measurements of SEP energy spectra for multiple species over a broad energy range can remotely probe the geometry, turbulence, and acceleration conditions at near-Sun shocks

Probes of near-Sun CME shocks H<4 for most events Weak correlation with CME speed Higher compression ratios for extreme events

Probe near-Sun CME shocks H<4 for most events Weak correlation with CME speed Higher compression ratios for extreme events Extreme events in red Stronger wave power than Kolmogorov -5/3 index

Conclusions Interplanetary turbulence levels are not likely to be strong enough to produce the strong Q/M-dependence of spectral break energies that result in α ≥1.4 in GLE events Rather, the spectral breaks are consistent with Q/M-dependent acceleration at near-Sun CME shocks Future Outlook: Simultaneous measurements of SEP spectra (SPP/IS⊙IS and SolO/EPD), plasma parameters, and fields closer to the regions where the acceleration processes begin, will unravel the roles of variable seed populations, self-excited waves, turbulence, and shock geometry Data from these missions will provide the ground truth and definitive tests for acceleration and transport-dominated models of SEP spectral properties

Thank You

Alpha vs. γb-γa Events with α>1.4 have harder spectra at high and low-energies

O EB vs. γb-γa Two groups? Extreme SEPs: EB Increases as γb-γa1; flat spectra at high and low energy Second group: EB increases as γb-γa increases

H, O, Fe Break Energies Range: ~0.2-100 EH>EO>EFe Break energies are also generally correlated within events

Alpha vs. γb-γa Events with α>1.4 have harder spectra at high and low-energies

H-Fe Spectral Break Energies EB for H is generally larger than other species

Distribution of Break Energies In a given SEP event, spectral variability due to break energy Large spread in break energies

Histograms of all Spectral Slopes All species are shown Some overlap Low-energy slopes <2.5 Most high-energy slopes>2

H-Fe Low-energy slopes Low-energy spectral slopes are correlated within events

H-Fe High-energy slopes High-energy spectral slopes are correlated within events

α in large SEPs Estimated α for each event 33 events have α ranging between >0.2-3

Distribution of spectral slopes In a given event, the low- and high- energy spectral slopes are species-independent