Isotopic Abundances in the Solar Corona as Inferred from
advertisement
Isotopic Abundances in the Solar Corona as Inferred from ACE Measurements of Solar Energetic Particles R. A. Leske*, R. A. Mewaldf, C. M. S. Cohen*, E. R. Christian^ A. C. Cummings*, P. L. Slocum**, E. C. Stone*, T. T. von Rosenvinge1^ and M. E. Wiedenbeck** ""California Institute of Technology, Pasadena, CA 91125 USA ^NASA/Goddard Space Flight Center, Greenbelt, MD 20771 USA ** Jet Propulsion Laboratory, Pasadena, CA 91109 USA Abstract. The isotopic composition of solar energetic particles (SEPs) has been measured using the Solar Isotope Spectrometer on the Advanced Composition Explorer. The measurements include up to 12 isotope abundance ratios for ten elements from C through Ni at energies of tens of MeV/nucleon in 18 large SEP events that have occurred since November 1997. These measurements clearly establish that SEP isotopic composition can vary widely (by factors of > 3) from event to event, presumably due to mass fractionation processes during particle acceleration and/or transport. Elemental and isotopic abundance ratios are strongly correlated, suggesting that elemental and isotopic fractionation relative to the coronal source are largely governed by the same processes. Using empirical correlations to correct for the fractionation and obtain the coronal isotopic composition yields preliminary abundance values in good agreement with those found in the solar wind, with comparable accuracy. INTRODUCTION up to Si [see, e.g. 7, and references therein]. To obtain adequate statistical accuracy, the earlier measurements sometimes required sums over several SEP events [8, 9, 10] and the resulting values usually agreed with terrestrial abundances but with rather large uncertainties. Isolated differences were noted for some gradual events [6, 7], and significant enrichments of 22Ne were found in 3 He-rich periods [11]. In recent studies using ACE data, enhancements by up to a factor of ~2 were reported in the 6 November 1997 SEP event for many heavy/light isotope abundance ratios from 13C/12C to 60Ni/58Ni [12]. Using the Solar Isotope Spectrometer (SIS) on ACE, the 22Ne/20Ne ratio was observed to vary by a factor of > 3 from event to event at energies of 24-72 MeV/nucleon [13], and similar or greater variability was found for Ne in other events at lower energies [14]. In the present work, we extend the previous ACE/SIS studies and present isotopic abundance measurements for C, O, Ne, Mg, Si, S, Ar, Ca, Fe, and Ni in as many as 18 individual SEP events. The isotopic composition is highly variable, but using the abundance correlations between different species we empirically correct for the variation and obtain preliminary coronal isotopic abundances from SEPs. Observations in a subset of the events over a more limited energy interval are also reported in [15, 16]. A primary goal of the Advanced Composition Explorer (ACE) mission is to better establish the elemental and isotopic composition of the Sun. Solar energetic particles (SEPs) provide a sample of solar material that may be used for such studies, but particle acceleration and transport processes can affect the arriving composition. Two distinct categories of SEP events, impulsive and gradual, are generally recognized [1]. Particles in gradual events are thought to originate as solar wind or coronal material accelerated by large shocks driven by coronal mass ejections. Elements in gradual SEP events have been measured for many years [e.g. 2]. Their abundances have been found to be highly variable from event to event but correlated with the ionic charge to mass ratio, Q/M [3]. When corrected for this fractionation [3, 4] or averaged over many events [5], SEP abundances can be used to determine the coronal elemental composition more accurately than is possible from spectroscopic measurements for some elements such as noble gases. In principle, the coronal isotopic composition can be similarly obtained from SEPs [6, 7], which has not been possible spectroscopically for more than a few isotopes. Unlike the case for SEP elemental measurements, before the launch of ACE there were only a few SEP heavy isotope measurements, and these included only elements CP598, Solar and Galactic Composition, edited by R. F. Wimmer-Schweingruber © 2001 American Institute of Physics 0-7354-0042-3/017$ 18.00 127 OBSERVATIONS AND ANALYSIS OBSERVATIONS OBSERVATIONSAND AND ANALYSIS ANALYSIS Using the dE/dx versus residual energy technique in a Using dE/dx versus technique in aa Using the dE/dx versusresidual residual energy technique in inpair ofthe silicon solid-state detectorenergy telescopes, the SIS pair of silicon solid-state detector telescopes, the SIS inpair of silicon detector telescopes, theand SIStotal instrument allowssolid-state the nuclear charge, Z, mass, M, strument allows nuclear charge, mass, M, total strument allowsthe theto nuclear charge,Z, Z, mass, M, and and total kinetic energy, E, be determined for particles with enkinetic toto bebedetermined particles with enkineticenergy, determinedfor for[17]. particles withstudy, energies ofenergy, 10 E, toE, 100 MeV/nucleon For this ergies ofof~1010 toto~events 100 [17]. For study, ergies 100MeV/nucleon MeV/nucleon [17]. Forthis thishigh study, we selected SEP with sufficient fluxes of enwe selected SEP events with sufficient fluxes of high we selected SEP events with sufficient fluxes of high enresergy heavy ions (E 15 MeV/nucleon, where massenwhere resergy ions ergyheavy heavy ions(E MeV/nucleon, where mass mass •>1515MeV/nucleon, olution is best) to(Eobtain statistically meaningful isotope olution is best) to obtain statistically meaningful isotope olution is best) to obtain statistically meaningful isotope abundances. Time profiles of the 18 selected events are abundances. Time profiles 18 events are abundances. Time1. profiles ofthe theintensities 18 selected selected events are shown in Figure Their of peak vary by more shown in Figure 1. Their peak intensities vary by more shown in Figure 1. Their peak intensities vary by more than three orders of magnitude at these energies, with thanthree threeorders orders ofmagnitude magnitude these energies, with than atatthese energies, with the events in theofyear 2000 (bottom panel) containing theevents eventsininthe theyear year2000 2000 (bottom (bottom panel) panel) containing containing the most of the smallest events considered for this study as most of thesmallest smallestevents eventsconsidered considered for this this study study as as most wellofasthe the two largest SEP events infor this solar cycle. wellasasthe thetwo twolargest largest SEP SEP events events in in this this solar solar cycle. cycle. well The very high counting rates and correspondingly high The veryhigh high countingrates rates andcorrespondingly correspondingly high The ratevery chancecounting coincidences and at the peaks of thesehigh two rateofof ofchance chancecoincidences coincidences at the peaks peaks of of these these two two rate at the largest events severely degraded mass resolution in SIS largestevents eventsseverely severelydegraded degradedmass mass resolution resolution in in SIS SIS largest (although the resolution is still adequate to separate ele(althoughthe theresolution resolutionisisstill stilladequate adequateto to separate separate eleele(although ments), and therefore we we restricted restricted theisotopic isotopic analysis ments),and andtherefore therefore the analysis ments), we restricted the isotopic analysis to the decay phases of of these these two two events. events. thedecay decayphases phases totothe of these two events. 315 319 323 327 331 335 FIGURE 1. Time FIGURE Time profiles profiles of of the the 18 18 SEP SEPevents eventsexamined examined FIGURE 1. hourly-averaged Time profiles intensities of the 18 of SEP events examined here, using MeV/nucleon here, hourly-averaged intensities of21-64 21-64 MeV/nucleon here, using hourly-averaged of indicate 21-64 MeV/nucleon oxygen from SIS Shaded time oxygen from SIS on ACE. ACE. intensities Shaded bars bars indicate timeperiods periods oxygen from on ACE. Shaded barsrepresent indicate these time periods used for for the isotope analysis; symbols used the SIS isotope analysis; symbols represent theseperiods periods used for the3–5. isotope analysis; symbols represent these periods in Figures Figures 3-5. in in Figures 3–5. Isotopes of of elements Isotopes elements up up through through Ni Ni are aremeasured measuredby by Isotopes of elements up through Ni are measured SIS with a mass resolution which varies with for SIS with a mass resolution which varies withZZand andE; E;by for SIS a mass which varies with Z andranges E; for thewith species and energies here typically the species andresolution energies studied studied hereitit typically ranges the species and energies studied here it typically ranges from ~0.15 to ~0.3 amu. Details of the analysis required from 0.15 to 0.3 amu. Details of the analysis required from 0.15isotope to 0.3 amu. Details ofand the examples analysis required to obtain obtain isotope abundance ratios of to abundance ratios and examples ofmass mass tohistograms obtain isotope abundance ratios and examples of histograms are given elsewhere [12, 13], but in most are given elsewhere [12, 13], but inmass most histograms are given elsewheremakes [12, 13], but in most cases the the good good mass the cases mass resolution resolution makes thedetermination determination of abundances straightforward. cases the good mass resolution makes the determination of abundances straightforward. Obtaining coronal abundances from these data is not of abundances straightforward. Obtaining coronal abundances from these data is not so simple, however, due to the fact that SEP Obtaining coronal abundances thesethe data is isonot so simple, however, due to the from fact that the SEP isotopic abundances may vary significantly from eventisoto sotopic simple, however, due to the fact that the SEP abundances may vary22significantly from event to 20 eventabundances [13], as shown for the significantly Ne/ Ne ratio in Figure 2. topic may vary from event to 20 Ne ratio in Figure 2. event [13], as itshown forthat the2222 Ne/ 20 Ne Surprisingly, appears the composition event [13], as shown for the Ne/ ratio invariability Figure 2. Surprisingly, it appears that the composition variability Surprisingly, it appears that the composition variability 128 was greatly reduced and nearly absent in the 1999–2000 was greatly reduced andtonearly nearly absent inthe the1999-2000 1999–2000 was greatly and absent time framereduced compared that seen in in 1997–1998. The retime frame compared to that seen in 1997–1998. Thefirst re- 9 2 time frame to thatfitting seen in 1997-1998. The reobtained from a constant to the duced χ2 compared 2 obtained from fitting a constant to the first 9 duced χ duced constant firstit 9is a points%inobtained Figure 2from is 83,fitting while afor the lastto9 the points points inmore Figure 83,while while forthe thelast last 9points pointsititisisasiga points Figure 22 isis 83, for muchin reasonable 1.3, indicating a9statistically much more reasonable 1.3, indicating a statistically sigmuch more reasonable 1.3, indicating a statistically significant change in the variability. Tracking the variability nificant change inthe thevariability. variability. Tracking the variability nificant in Tracking the variability of the change composition in future events may help to deterthe composition composition in future events may help deterofmine the in future events may help totodeterwhether this is merely a statistical aberration or an whether this is merely a statistical aberration oranan mine whether this is merely a statistical aberration or unexplained feature of the solar cycle. unexplained feature feature of ofthe thesolar solarcycle. cycle. unexplained 0.3 ; * O CN CN CN 0.1 0.03 1998 1999 Year 2000 2001 22 Ne/20 Ne ratio measured by ACE/SIS FIGURE 2. The The SEP 22 20 20 FIGURE 2. Ne ratio bybyACE/SIS FIGURE 2.MeV/nucleon TheSEP SEP 22Ne/ Ne/ Neversus ratiomeasured measured ACE/SIS at E 15 plotted the date of the event. at E > 15 MeV/nucleon plotted at E 15 MeV/nucleon plottedversus versusthe thedate dateofofthe theevent. event. In obtainingcoronal coronal abundances from the highly variIn abundances the highly In obtaining obtaining coronal abundancesfrom from theguided highlyvarivariable SEP isotope measurements, we are the able SEP isotope measurements, we are guided bybyby the able SEP isotope measurements, we are guided the experience gained in coping with a similar situation enexperience gained in with situation enexperience gained in coping coping withaasimilar similar situation encounteredfor forelemental elemental abundances. The variations countered abundances. The variations ofofof countered for elemental abundances. The variations heavyion ionelemental elementalabundances abundancesin inindividual individual gradual heavy gradual heavy ion elemental abundances in individual gradual events relative to coronal values have been found to scale events relative to values have been found events relativewell tocoronal coronal valueslaw have been foundto toscale scaleto reasonably as a power in the ionic charge reasonably well as a power law in the ionic charge to reasonably well as [3], a power in the power ionic charge to massratio, ratio,Q/M QM witha different alaw different index mass [3], with power lawlaw index mass ratio, Q M [3], with a different power law index eachSEP SEPevent. event.If IfthetheQ/M QM ratio indeed releforforeach is is indeed thethe relefor each SEP event. If the Qthen Mratio ratiosame is indeed the mechrelevant organizing parameter, physical vant organizing parameter, then thethe same physical mechvant organizing parameter, then the same physical mechanism responsible elemental fractionation, whatanism responsible forfor thethe elemental fractionation, whatanism responsible for the elemental fractionation, whateverit itis,is,should shouldalso alsoproduce produce variations isotopic ever variations in in thethe isotopic ever it is, should also variations inisotopes theisotopes isotopic abundances, since Q/M forfor twotwo of of abundances, since Qproduce Mwill willdiffer differ abundances, since Q M will differ for two isotopes of the number. This im-imthesame sameelement elementthrough throughthethemass mass number. This the same element through the mass number. This implies correlation between pliesthere thereshould shouldbebea predictable a predictable correlation between plies there should be a predictable between the ofofvarious species, in correlation particular between theabundances abundances various species, in particular between the abundances of various species, Following in Following particular elemental and abundances. [6],between if we elemental andisotopic isotopic abundances. [6], if we elemental and isotopic abundances. Following ifabunwe base the power law fractionation index on the abunbase the power law fractionation index on [6], the base the power law fractionation index on the abundance ratio of any two reference species, such as Fe/O, dance ratio of any two reference species, such as Fe/O, dance ratio ofingeneral any twoterms, reference such as Fe/O, Na/Mg, or,or,in remembering Na/Mg, general terms,Ri/R2, R 1species, /R2(and , (and remembering lny or, in lnx general terms, R /R Na/Mg, , (and remembering ln y lnx 1 2 that x = y ) it readily follows that we would expect that ln x y lnx y ) it readily follows that we would expect that x y ) it readily follows that we would expect the enhancement or depletion of the SEP abundance ratio the enhancement or depletion of the SEP abundance ratio the enhancement or depletion ofXthe SEP for b bof element to to be: forisotopes isotopesa aand and of element X be:abundance ratio for isotopes a and b of element X to be: ln ba /(G/M) M(G/M) ] (RI/R R1 •2JSEP R2 SEP ln lnQQMMln bR1aQQMM R2 (Ri/R RR1 2RR)c2SEP corona R1 1R2 2corona a 6 X) b SE P ( a X/ a XXb XX SEP SEP ( ^/ a X Xj b Xcorona a X b X corona corona __________ Rl R2 R 1 R 2 (D (1) using the fact that Q should be the same for two isotopes(1) using the fact that Q should be the same for two isotopes of the same element. using fact that Q should be the same for two isotopes of thethe same element. evaluate the expected enhancement from equaof To the same element. To evaluate the expected enhancement from equationTo(1), the ionic state enhancement Q must be known the evaluate thecharge expected fromforequation (1), the ionic charge state Q must be known for the reference species. Ionic charges are not be often measured tion (1), the ionic charge state Q must known for the reference species. Ionic charges are not and oftenthemeasured at SIS energies of Ionic tens ofcharges MeV/nucleon, meareference species. are not often measured at SIS energies of tens of MeV/nucleon, and the meaat SIS energies of tens of MeV/nucleon, and the mea- surements that exist surements that existshow showconsiderable considerablevariability variabilityfor forelelsurements that exist show considerable variability for elements such as Fe [see, e.g. 18, and references therein]. ements such as Fe [see, e.g. 18, and references therein]. ements such as Fe [see, e.g. 18, and references therein]. Although measurements Although measurementsofofQQatatlower lowerenergies energiesmight might Although measurements of Q at lower energies might bebe used [3], they may not apply to SIS data since sevused [3], they may not apply to SIS data since sevbe used [3], they may not apply to SIS data since several events have been clearly shown to exhibit energyeral events have been clearly shown to exhibit energyeral events have been clearly shown to exhibit energydependent charge states forfor heavy dependent charge states heavy elements [19,20, 20,21]. 21]. dependent charge states for heavyelements elements[19, [19, 20, 21]. addition, the relevant value of Q in equation (1) is In In addition, the relevant value of Q in equation (1) In addition, the relevant value of Q in equation (1) isis that which the particles have when the elemental and isothat which the particles have when the elemental and isothat which the particles have when the elemental and isotopic fractionation takes place, which may bequite quitedifdiftopic fractionation takes place, topic fractionation takes place,which whichmay maybe be quite different from the value at 1 AU if fractionation happens ferent from the value at 1 AU if fractionation happens ferent from the value at 1 AU if fractionation happens early and example, stripping occurs acceleration early and if,if, for example, stripping early and if,for for example, strippingoccurs occursinin inacceleration acceleration through corona [22, 23]. Another complicationnot not through thethe corona [22, through the corona [22,23]. 23].Another Anothercomplication complication not addressed the fact that ingradual gradualSEP SEP addressed in in equation (1)(1) is is the addressed inequation equation (1) is thefact factthat thatinin gradual SEP events abundances elements with lowfirst firstionizaionizaevents the abundances of elementswith withlow low first ionizaevents thethe abundances ofof elements tion potential (FIP), such as Fe, are generally enhanced potential (FIP), such Fe,are aregenerally generallyenhanced enhanced tiontion potential (FIP), such asasFe, over those with high FIP, such anamount amountwhich which over those with high FIP, such asO,O, O,byby byanan amount which over those with high FIP, such asas also varies from event to event [24, 25]. In spite ofthe the varies from event event[24, [24,25]. 25].InInspite spiteof of the alsoalso varies from event to toevent above considerations, a reasonable correlation has been above considerations, a reasonable has been above considerations, a reasonable correlation has been shown between isotopic abundances Fe/Oratio ratio shown between isotopic abundancesand andthe theFe/O Fe/O ratio shown between isotopic abundances [12,13,15], but uncertainties in the value Q(Fe) at SIS [12, 13, 15], but uncertainties of Q(Fe) at SIS [12, 13, 15], but uncertainties in the value of Q(Fe) at SIS energies make ititdifficult to directly compare the correlacorrelaenergies make difficult to directly energies make it difficult to directly compare the correlation with withpredictions. predictions. tiontion with predictions. 0.3 0.3 SEP 26 Mg/ 24 Mg / Solar System 22 20 26 24 20 26 24 FIGURE 4. 4. The The 22 Ne/ Ne versus Mg/ Mg isotopic ra-ra22Ne/ FIGURE Mg/ Mg isotopic raFIGURE 4. The Ne/20Ne Neversus versus 26 Mg/24 Mg isotopic tios in each of the SEP events shown in Figure normalized tios in in each 1,1,1, normalized tototo tios each of of the theSEP SEPevents eventsshown shownininFigure Figure normalized standard solar solar system systemvalues values [26]. The diagonal line shows the standard line shows the standard solar system values[26]. [26].The Thediagonal diagonal line shows the correlation expected expected usingequation equation (1). correlation correlation expectedusing using equation(1). (1). The predicted predicted correlationsare are verysensitive sensitivetotoQ,Q,and and The The predictedcorrelations correlations arevery very sensitive to Q, and small change change of of only only 1% 1%ininthe theQ(Na)/Q(Mg) Q(Na)/Q(Mg)ratio ratio aa small a small change of only 1% in the Q(Na)/Q(Mg) ratio changes the the expected expectedslope slopeconsiderably, considerably,asasalso alsoshown shown changes changes the expectedequilibrium slope considerably, as also shown in Figure Figure 3. 3.Detailed Detailedequilibrium calculations[28] [28]show show in calculations in Figure 3. Detailed equilibrium calculations [28] show 9 for that that at at aa constant constanttemperature temperatureQ(Na)/Q(Mg) Q(Na)/2(Mg)isis>00.9 for 0 9 for that at a constant temperature Q(Na)/Q(Mg) is all all temperatures temperatures from from 0.5–10 0.5-10MK, MK,that thatis,is,the theexpected expected all temperatures 0.5–10than MK, that is, line the expected correlation should be the solid correlation shouldfrom besteeper steeper than the solid linein inFigFigcorrelation should be steeper than the solid line inthe Figure ure 33 and and more more discrepant discrepantwith withthe thedata. data.However, However, the ure 3 and more discrepant with the data. However, assumption assumption of ofaaconstant constanttemperature temperaturefor forallallelements elementsisthe is assumption of a constant temperature for all elements probably InInfact, isisoften probably unwarranted. unwarranted. fact,Mg Mgininparticular particular oftenis probably unwarranted. Intypical fact, Mg in particular is often found ofofhigher temperatures found to to have have aa mean meanQQ typical higher temperatures found to have a mean Q typical of higher temperatures than most other elements [29, 30], which would result than most other elements [29, 30], which would result in more consistent than most value other elements [29, 30], which wouldwith result in aa lower lower value of of Q(Na)/Q(Mg) Q(Na)/Q(Mg) more consistent with the observed correlation. In any case, the relatively small in a lower value of Q(Na)/Q(Mg) more consistent with the observed correlation. In any case, the relatively small amount of scatter in Figure 3 suggests that at the time the observed correlation. In any case, the relatively small amount of scatter in Figure 3 suggests that at the time of Q(Na)/Q(Mg) byby more amount of scatter in Figure 3 does suggests that the time of fractionation, fractionation, Q(Na)/Q(Mg) doesnot notdiffer differat more than aa few from This isisconsiderof fractionation, does differ by more than few percent percentQ(Na)/Q(Mg) from event eventtotoevent. event.not This considerably the ininindividual than fewthan percent from event tovariability event. This isindividual considerably aless less than theevent-to-event event-to-event variability charge states observed at 1 AU [31] and may aa charge at 1 AU [31] and may provide ably lessstates than observed the event-to-event variability inprovide individual clue as to how and when mass fractionation takes place. clue as states to howobserved and whenatmass takes place. a charge 1 AUfractionation [31] and may provide If the scatter ininmass Figure 33isisindeed If as thetoresidual residual scatter Figure indeeddue duetotovarivariclue how and when fractionation takes place. ability in the value ififQQisisthe ability the Q(Na)/Q(Mg) Q(Na)/Q(Mg) valueand thesame If theinresidual scatter in Figure 3and is indeed due tosame varifor of then correfor 22 isotopes isotopes of the thesame sameelement, element, thenifa abetter ability in the Q(Na)/Q(Mg) value and Qbetter is thecorresame lation should possible using an ratio the lation should be be possible using anisotope isotope ratioasascorrethe for 2 isotopes ofThis the issame element, then a better reference value. illustrated by the correlation bereference value. This is illustrated byisotope the correlation belation should be possible using an ratio as the 22 20 26 24 tween 22Ne/ 20Ne and 26Mg/ 24Mg ratios in Figure 4 tween the thevalue. Ne/ This Ne and Mg/ Mg ratios in Figure be4 reference is illustrated by the correlation which, for most of the events with24the better-determined 22 20 26 which, for most of the events with the better-determined tween andwithMg/ Mg ratios in Figure 4 values,the agreesNe/ veryNe well the expected correlation. values, for agrees very well withwith the expected correlation. which, most of the events the better-determined At least no systematic deviation from the expected trend At leastagrees no systematic deviation from the expected trend values, very well with the correlation. is evident. Although the outliers areexpected small events with is evident. Although the outliers are small events with At least no systematic deviation from the expected trend large uncertainties and most are not seriously discrepant large uncertainties and most are not seriously discrepant is evident. Although the outliers events with statistically, it is interesting to note are thatsmall they tend to lie statistically, it is interesting to note that they tend to lie near auncertainties value of unity onmost one of axes, asdiscrepant if only large and arethe nottwo seriously near a value of unity on one ofwhile the two axes, as if only one isotope ratio fractionated the other is unafstatistically, it is isinteresting to note that they tend to lie one isotope ratio is fractionated while the other is unaf- 0.3 0.03 0.1 22 0.03 Na/Mg Ratio 20 26 24 FIGURE 3.3. The The 22Ne/ Ne/20Ne Ne (left) (left) and and 26Mg/ Mg/24Mg Mg (right) (right) FIGURE 22 20 26 24 isotopic ratios versus the Na/Mg elemental abundance ratio isotopic ratios versus the Na/Mg elemental abundance ratio FIGURE 3. The Ne/ Ne (left) and Mg/ Mg (right) each the SEPthe events of Figure Figure 1. Both Both isotope ratios inineach ofof the SEP events of 1. isotope ratios isotopic ratios versus Na/Mg elemental abundance ratio havebeen been normalized their respective “solarisotope system”ratios value normalized tototheir respective "solar system" value in have each of the SEP events of Figure 1. Both [26].Diagonal Diagonallines linesshow showthe thecorrelations correlations expected expected from from equaequa[26]. have been normalized to their respective “solar system” value tion(1), (1),assuming assumingthe thecharge charge states states of of Na Na and and Mg Mg as as indicated. indicated. tion [26]. Diagonal lines show the correlations expected from equation (1), assuming the charge states of Na and Mg as indicated. Withthe theappropriate appropriate choice choice of of reference reference species, With species, such such as Na/Mg, we do find that the isotopic and elemental as Na/Mg, we do find that the isotopic and elemental With the appropriate choice of reference species, such abundances tend to to be correlated correlated approximately approximately as as exexas abundances Na/Mg, wetend do findbethat the isotopic and elemental pected from equation (1), as shown in Figure 3. Both Na pected from equation (1), as shown in Figure 3. Both Na abundances tend to be correlated approximately as exandMg Mgare arelow-FIP low-FIP elements elements so so this this ratio ratio is unaffected and is unaffected pected from equation (1), as shown Figureout 3. by Both Na byvariable variable FIP fractionation. fractionation. As in pointed FIP As pointed outunaffected by Cohen Cohen andby Mg are low-FIP elements so this ratio is et al. [27], both elements are theoretically expected to et al. [27], both elements are theoretically expected by variable FIP fractionation. As apointed out by Cohento have 2 electrons attached over broad range of coronal ~2 electrons attached over a broad range of coronal 23 et have al. [27], both elements are theoretically expected to temperatures [28], and since 23 Na is neutron-rich with temperatures [28], and since Na is neutron-rich with 24 have 2 electrons attached over a broad range of coronal respect to Mg, there is a significant difference in Q M. respect to 24Mg, is a significant difference in Q/M. 23 temperatures [28],inthere and since neutron-rich with The solid lines Figure 3 showNa theiscorrelations expected The solid lines in Figure 3 show the correlations expected 24 Mg, there is a significant difference in Q M. respect to from equation (1), assuming Q(Na)=9 and Q(Mg)=10. from (1), assuming Q(Na)=9 and Q(Mg)=10. The solidequation lines in Figure 3 show correlations expected While this very simple modelthe provides a good first orWhile this very simple model provides a good first orderequation fit to the (1), data,assuming both isotope correlations appear to be from Q(Na)=9 and Q(Mg)=10. der fit to the data, both isotope correlations appear to be shallower thansimple expected. While this very model provides a good first or- shallower than expected. der fit to the data, both isotope correlations appear to be shallower than expected. near a value of unity on one of the two axes, as if only one isotope ratio is fractionated while the other is unaf- 129 10 F 0.5 0.1 9R CL5 0.1 0.5 0.1 0.5 Mg Ratio FIGURE 5. Eleven SEP isotope abundance ratios (normalized to standard abundances [26]) plotted versus the 26 Mg/24 Mg ratio. 24 Symbols 5. indicate the SEP SEP isotope events shown in Figure The diagonaltolines showabundances the correlations using equation FIGURE Eleven abundance ratios1.(normalized standard [26])expected plotted versus the 26Mg/(1). Mg ratio. Symbols indicate the SEP events shown in Figure 1. The diagonal lines show the correlations expected using equation (1). fected. Since these outlying events are among the small- The isotopic composition of Ne is interesting since it differs in various solar system materials. Therefore, of The isotopic composition of Ne is interesting since it the 2 isotopic ratios in Figure 4, we chose the 26 Mg/24 Mg differs in various solar system materials. Therefore, of ratio as our abundance standard so that we can solve the 2 isotopic ratios in Figure 4, we chose the 26Mg/24Mg for the SEP Ne composition in this study. The SEP for the SEP Ne for composition in thisforstudy. Thefrom SEP abundance values 11 isotope ratios elements abundance 11 isotope for elements from C to Ni arevalues shownfor plotted versusratios this reference ratio in C to Ni5.are shown plotted versusslopes this reference ratio Figure The different expected arise from thein different5.relative mass number ratios, and the seem Figure The different expected slopes arisedata from the to follow relative these expected trends for certain such as different mass number ratios, andspecies the data seem Ne,follow Mg, Si, andexpected Ca. For many thecertain heavy elements fromas to these trendsoffor species such S and the elements more limited Ne, Mg,above, Si, andthe Ca.agreement For many between of the heavy from data the expectations is not as clear.the Themore agreement S andand above, the agreement between limited M from mightand break with increasing in Qagreement data the down expectations is not asdistance clear. The Mg if break the actual on Qdistance M is not a simple might downdependence with increasing in Q/M from 13 C, 12 of the 15 data power law as we assumed. For Mg if the actual dependence on Q/M is not a simple 13 points fall expectedFor correlation, power lawabove as wetheassumed. C, 12 including of the 15most data 26 24 Mg/ correlation, Mg and other ratios show of those which points fallforabove thethe expected including most little or no The24reason forother this isratios not atshow all of those forfractionation. which the 26Mg/ Mg and clear,or butnoiffractionation. this preliminary holds up, itis suggests little Theresult reason for this not at all that 13but C routinely is enhancedresult or 12 Cholds is depleted in SEP clear, if this preliminary up, it suggests 13 relative to terrestrial abundances. 12 events In future that C routinely is enhanced or C is depleted inwork SEP we planrelative to extend the isotopeabundances. measurements to include events to terrestrial In future work 15 to see if it is similarly affected. weNplan to extend the isotope measurements to include 15 To correct for the QM-dependent mass fractionation, N to see if it is similarly affected. weTo solve equation for the coronal isotope ratios. For correct for the(1)Q/M-dependent mass fractionation, 24 Mg ratio as the reference Mg/the example, using the(1)26for we solve equation coronal isotope ratios. For 22 Ne/ 2620 Ne 24 ratio R1 /Rusing value from 2 , the the example, Mg/ coronal Mg ratio asobtained the reference any SEP event would be: 22 20 ratio Ri/R2, the Ne/ Ne coronal value obtained from any SEPevent would be: 22 Ne 26 Mg24Mg lnln20242226 22 Ne 2220 Ne 2626Mg2424Mg cor ln(20/22) 20 Ne 22 ( Mg/ Mg)SEP ln(24/26) / Ne\ cor / Ne\ SEP (2) ratio as our abundance standard so that we can solve (2) fected. these outlying eventsamount are among the smallest in Since our sample, even a small of contaminaest infrom our sample, a small of contamination impulsiveeven events mightamount significantly alter their tion from impulsive might 5significantly alterFetheir composition. Of theevents 6 outliers, exhibit modest enhancements, Of with 0.5. Most of themodest events Fe in our composition. theFe/O ~6 outliers, 5 exhibit en3 He [16, 32], study containwith significant enhancements of events hancements, Fe/O~0.5. Most of the in our 3 whichcontain could be due to residual materialoffrom study significant enhancements Heimpulsive [16, 32], flares could resident theto interplanetary medium which is later which beindue residual material from impulsive accelerated shock [33]. The ion cyclotron wave resflares residentbyinathe interplanetary medium which is later onances [34, or cascading Alvén waves [36] responaccelerated by35] a shock [33]. The ion cyclotron wave res3 He enrichment of impulsive events might sible for[34, the35] onances or cascading Alven waves [36] responalso for selectively otherofspecies withevents discrete valsible the 3He enhance enrichment impulsive might 3 M at higher harmonics of the He cyclotron freues of Q also selectively enhance other species with discrete val3 quency [37], in fact significant enrichments of both ues of Q/M at and higher harmonics of the He cyclotron fre22 Ne and 26 Mg have been reported in 3 He-rich periods quency [37], and in fact significant enrichments of both 22[11, 38]. 26 3 it were for theinresonance affect Ne and IfMg have possible been reported He-rich to periods a narrow enough frequency range, perhaps only of [11, 38]. If it were possible for the resonance to one affect these two species might be enhanced, resulting in a pata narrow enough frequency range, perhaps only one of tern such as appears to be in Figure 4. Itinwill be these two species might be present enhanced, resulting a patinteresting to see if this pattern persists as more events tern such as appears to be present in Figure 4. It will be accumulatetoduring cycle. interesting see if this thissolar pattern persists as more events accumulate during this solar cycle. RESULTS RESULTS 130 10°^ B 10 -1 o > o o 10 -2 U ACE/SIS weighted average I ACE/SIS SEP-derived corona 1 Solar wind - - Anders & Grevesse Abundance Ratio FIGURE 6. Deduced coronal source isotopic abundance ratio averages from SIS SEP measurements without correcting for fractionation (open boxes) and after correction (light grey boxes) as in equation (2). For comparison, standard solar system values (dashed lines; [26]) and measured solar wind values (dark grey boxes; [see 39, and references therein]) are shown. The 26Mg/24Mg ratio served as the reference value for the fractionation corrections for everything other than the 26Mg/24Mg ratio, for which Na/Mg at the ionic charge ratios considered in Figure 3 was used. Note that the effect of uncertainties in the selected reference value is easy to determine from this expression, the propagation of errors is straightforward, and the exponents are simple constants and do not depend on the measured values. Preliminary solar coronal isotopic abundances obtained following the example of equation (2) and averaging over all the SIS measurements are shown in Figure 6. For comparison, we have also calculated the weighted average without correcting for the fractionation, but including in the weighting the width of the parent population distribution added in quadrature to the statistical uncertainties. This may be a more representative value for cases such as S or Fe where the data may not follow the expected fractionation correlations in Figure 5, and with a large enough data set (if unbiased by selection effects) may even average out to the coronal value as seems to be the case for elemental abundances [5]. Also, the uncorrected average is the appropriate one to consider for assessing the average arriving solar particle composition at 1 AU. For example, the uncorrected average 22Ne/20Ne ratio we find here is consistent with the value of ~0.09 of the so-called SEP component detected in lunar soils [40]. Both our corrected and uncorrected SEP values are 131 also compared with the standard solar system values [26] and existing solar wind values [39] in Figure 6. Although the results are preliminary, it is encouraging that this early attempt to obtain coronal abundances from the fractionated SEPs seems to yield reasonable values. With the exception of 13C, all of the isotope abundances are within 2.5a of the Anders and Grevesse [26] "solar system" values. Both the corrected noble gas isotopes 22 Ne and 38 Ar appear a bit low compared to Anders and Grevesse, but for these two species Anders and Grevesse adopted the solar wind values as their standard without accounting for mass fractionation in the solar wind of perhaps several percent [41]. So far, Ni isotope abundances have not been reported from solar wind data, so the SEP value given here is the first determination of the 60 Ni/58Ni ratio in the corona. In many cases the uncertainties on the SEP-derived coronal isotope values are comparable to those obtained from solar wind measurements. Accumulating additional SEP events will help reduce the uncertainties for the heaviest species such as Ar to Ni where there are still only a few measurements, but for most of the others a better theoretical understanding of the mass fractionation process is required to make much further progress. Tracking some of the puzzles uncovered here through the declining part of the solar cycle, such as the decreasing variability in recent events (Figure 2), the possible fractionation of only some isotope ratios (Figure 4), and the apparent common enhancement of 13C (Figure 5) may help to shed light on the nature of the fractionation process. 19. 20. 21. 22. 23. ACKNOWLEDGMENTS 24. This research was supported by NASA at the California Institute of Technology (under grant NAG5-6912), the Jet Propulsion Laboratory, and the Goddard Space Flight Center. 25. 26. REFERENCES 1. 2. 3. 4. 5. 6. 27. Reames, D. V., Revs. Geophys., 33, 585-589 (1995). Teegarden, B. J., von Rosenvinge, T. T., and McDonald, F. B., Astrophys. J., 180, 571-581 (1973). Breneman, H. H., and Stone, E. C., Astrophys, J, Lett,, 299, L57-L61 (1985). Garrard, T. L., and Stone, E. C., Proc, 23rd Internat, Cosmic Ray Conf. (Calgary), 3, 384-387 (1993). Reames, D. V., Adv. Space Res., 15, (7)41-(7)51 (1995). Mewaldt, R. A., and Stone, E. C., Astrophys. J., 337, 28. 29. 30. 31. 959-963 (1989). 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. Williams, D. L., Leske, R. A., Mewaldt, R. A., and Stone, E. C., Space Sci. Rev., 85, 379-386 (1998). Dietrich, W. K, and Simpson, J. A., Astrophys. J. Lett., 231, L91-L95 (1979). Dietrich, W. F., and Simpson, J. A., Astrophys. J. Lett., 245, L41-L44 (1981). Simpson, J. A., Wefel, J. P., and Zamow, R., Proc. 18th Internat. Cosmic Ray Conf. (Bangalore), 10, 322-325 (1983). Mason, G. M., Mazur, J. E., and Hamilton, D. C., Astrophys. J., 425, 843-848 (1994). Leske, R. A., et al., Geophys. Res. Lett., 26, 153-156 (1999). Leske, R. A., et al., Geophys. Res. Lett., 26, 2693-2696 (1999). Dwyer, J. R., Mason, G. M., Mazur, J. E., Gold, R. E., and Krimigis, S. M., Proc. 26th Internat. Cosmic Ray Conf. (Salt Lake City), 6, 147-150 (1999). Leske, R. A., et al., Proc. 26th Internat. Cosmic Ray Conf. (Salt Lake City), 6, 139-142 (1999). Leske, R. A., et al., "Measurements of the Heavy-Ion Elemental and Isotopic Composition in Large Solar Energetic Particle Events from ACE", in High Energy Solar Physics: Anticipating HESSI, edited by R. Ramaty and N. Mandzhavidze, ASP Conf. Ser. 206, Astronomical Society of the Pacific, San Francisco, 2000, pp. 118-123. Stone, E. C., et al., Space Sci. Rev., 86, 357^08 (1998). Leske, R. A., Mewaldt, R. A., Cummings, A. C., Stone, E. C., and von Rosenvinge, T. T., "The Ionic Charge State Composition at High Energies in Large Solar Energetic Particle Events in Solar Cycle 23", in Proc. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 132 Joint SOHO-ACE Workshop 2001, AIP Conf. Proc., AIP, New York, 2001, this volume. Mazur, J. E., Mason, G. M., Looper, M. D., Leske, R. A., and Mewaldt, R. A., Geophys. Res. Lett., 26, 173-176 (1999). Mobius, E., et al., Geophys. Res. Lett., 26, 145-148 (1999). Oetliker, M., et al., Astrophys. J., 477, 495-501 (1997). Reames, D. V., Ng, C. K., and Tylka, A. J., Geophys. Res. Lett., 26, 3585-3588 (1999). Barghouty, A. F, and Mewaldt, R. A., Astrophys. J. Lett., 520, L127-L130 (1999). Garrard, T. L., and Stone, E. C., Adv. Space Res., 14, (10)589-(10)598 (1994). Mewaldt, R. A., et al., "Variable Fractionation of Solar Energetic Particles According to First lonization Potential", in Acceleration and Transport of Energetic Particles Observed in the Heliosphere: ACE 2000 Symposium, edited by R. A. Mewaldt et al., AIP Conf. Proc. 528, AIP, New York, 2000, pp. 123-126. Anders, E., and Grevesse, N., Geochim. Cosmochim. Acta, 53, 197-214 (1989). Cohen, C. M. S., et al., Geophys. Res. Lett., 26,2697-2700 (1999). Arnaud, M., and Rothenflug, R., Astron. Astrophys. Suppl, 60, 425^57 (1985). Luhn, A., et al., Proc. 19th Internat. Cosmic Ray Conf. (LaJolla), 4, 241-244 (1985). Leske, R. A., Cummings, J. R., Mewaldt, R. A., Stone, E. C., and von Rosenvinge, T. T., Astrophys. J. Lett., 452, L149-L152 (1995). Mobius, E., et al., "Survey of Ionic Charge States of Solar Energetic Particle Events During the First Year of ACE", in Acceleration and Transport of Energetic Particles Observed in the Heliosphere: ACE 2000 Symposium, edited by R. A. Mewaldt et al., AIP Conf. Proc. 528, AIP, New York, 2000, pp. 131-134. Wiedenbeck, M. E., et al., "Enhanced Abundances of 3He in Large Solar Energetic Particle Events", in Acceleration and Transport of Energetic Particles Observed in the Heliosphere: ACE 2000 Symposium, edited by R. A. Mewaldt et al., AIP Conf. Proc. 528, AIP, New York, 2000, pp. 107-110. Mason, G. M., Mazur, J. E., and Dwyer, J. E., Astrophys. J. Lett., 525, L133-L136 (1999). Fisk, L. A., Astrophys. J., 224, 1048-1055 (1978). Temerin, M., and Roth, L, Astrophys. J. Lett., 391, L105-L108 (1992). Miller, J. A., Space Sci. Rev., 86, 79-105 (1998). Bochsler, P., and Kallenbach, R., Meteoritics, 29, 653-658 (1994). Slocum, P. L., et al., "Measurements of Heavy Elements and Isotopes in Small Solar Energetic Particle Events", in Proc. Joint SOHO-ACE Workshop 2001, AIP Conf. Proc., AIP, New York, 2001, this volume. Wimmer-Schweingruber, R. F, Bochsler, P., and Wurz, P., "Isotopes in the Solar Wind: New Results from ACE, SOHO, and WIND", in Solar Wind Nine, edited by S. R. Habbal et al., AIP Conf. Proc. 471, AIP, New York, 1999, pp. 147-152. Wieler, R., Space Sci. Rev., 85, 303-314 (1998). Kallenbach, R., et al., Space Sci. Rev., 85, 357-370 (1998).
