Table Of Content

DraftversionFebruary25,2016 PreprinttypesetusingLATEXstyleemulateapjv.01/23/15 CHANDRAOBSERVATIONSOFCOMETSC/2012S1(ISON)ANDC/2011L4(PANSTARRS) BradfordSnios1,VasiliKharchenko1,CareyM.Lisse2,ScottJ.Wolk3,KonradDennerl4,andMichaelR.Combi5 1DepartmentofPhysics,UniversityofConnecticut,Storrs,CT06269,USA 2PlanetaryExplorationGroup,SpaceDepartment,JohnsHopkinsUniversityAppliedPhysicsLaboratory,Laurel,MD20723,USA 3ChandraX-RayObservatoryCenter,Harvard-SmithsonianCenterforAstrophysics,Cambridge,MA02138,USA 4Max-Planck-Institutfu¨rextraterrestrischePhysik,D-85748Garching,Germanyand 6 5DepartmentofClimateandSpaceSciencesandEngineering,UniversityofMichigan,AnnArbor,MI48109,USA 1 DraftversionFebruary25,2016 0 Abstract 2 WepresentourresultsontheChandraX-rayObservatoryAdvancedCCDImagingSpectrometer(ACIS)ob- b servationsofthebrightOortCloudcometsC/2012S1 (ISON)andC/2011L4(PanSTARRS).ISONwasob- e servedbetween2013October31–November06duringvariablespeedsolarwind(SW),andPanSTARRSwas F observedbetween 2013 April17–23during fast SW. ISON producedan extendedparabolic X-ray morphol- 4 ogyconsistentwithacollisionallythickcoma,whilePanSTARRSdemonstratedonlyadiffuseX-ray-emitting 2 region.Weconsidertheseemissionstobefromchargeexchange(CX)andmodeleachcomet’semissionspec- trum from first principles accordingly. Our model agrees with the observational spectra and also generates ] compositionratiosforheavy,highlychargedSWionsinteractingwiththecometaryatmosphere. Wecompare E ourderivedSWioncompositionstoobservationaldataandfindastrongagreementbetweenthem. Thesere- H sultsfurtherdemonstratetheutilityofCXemissionsasaremotediagnosticstoolofbothastrophysicalplasma . interactionandSWcomposition. Inaddition,weobservepotentialsoftX-rayemissionsviaACISaround0.2 h keV from both comets that are correlated in intensity to the hard X-ray emissions between 0.4–1.0keV. We p - fitourCXmodeltotheseemissions,butourlackofauniquesolutionatlowenergiesmakesitimpossibleto o concludeif theyare cometaryCX in origin. We lastly discussprobableemissionmechanismsourcesforthe r softX-raysandexplorenewopportunitiesthesefindingspresentinunderstandingcometaryemissionprocesses t s viaChandra. a Subject headings:comets: individual(Comet S1/ISON, Comet L4/PanSTARRS) – solar wind – techniques: [ spectroscopic–X-rays:general 2 v 2 1. INTRODUCTION theinnerplanetsandwasonlysparselyobserved.Itdid,how- 2 Cometary X-ray emissions, originally discovered by ever,demonstrateafantasticallyrichoutpouringofdustyma- 6 Lisseetal. (1996) and now observed in over 30 comets, are terialin2013March–Aprilasitpassedthroughperihelion,as 6 a well-studied phenomenon. It has been shown thatthe ma- seenbySTEREO(Raouafietal.2015). PanSTARRSisanun- 0 jority of these emissions are caused by solar wind Charge usuallydust-richcomet,withadust-to-gasmassratiogreater . than4 (Yangetal. 2014). Bycontrast, ISONwas seentobe 1 Exchange (CX) interactions between highly charged, heavy a dust-poor comet with a dust-to-gas mass ratio less than 1 0 solar wind (SW) ions (∼0.1% of all solar wind ions) and (Meechetal.2013).SincetheChandraobservationsforthese 6 neutral gas ejected from the comet nucleus into the coma twocometshavenotpreviouslybeenanalyzed,wedecideto 1 (Cravens1997;Krasnopolsky1997;Kharchenkoetal. 2003; examinetheir emissions for detailed analysis and interpreta- : Lisseetal. 2004; Bodewitsetal. 2007; Dennerl 2010). A v tionofthecometaryX-rayemissionspectraviamodeling. simplified theoretical description of interaction between the i To model CX X-ray emissions for these comets, we X SW plasma and cometary atmosphere shows that the emis- decide to expand upon previous modeling techniques sion originatespredominantlyfrom the sunward hemisphere r (Kharchenko&Dalgarno 2000, 2001; Krasnopolskyetal. a of the neutral coma and creates a projected paraboloid of 2002;Bodewitsetal.2007). Thoughthesemodelsarerobust, emissionwiththecometatitsfocalpoint(Ha¨berlietal.1997; most only incorporate the primary emission lines, generally Wegmannetal.2004). 10–20lines,outofthepossible700+linesthatmaybegener- Tworecentlydiscoveredcometsthatwereobservedbythe atedinanaverageCXinteractionwithacometaryatmosphere Chandra X-ray Observatory are comets C/2012 S1 (ISON) (Kharchenkoetal.2003).Thisistypicallyperformedbecause and C/2011 L4 (PanSTARRS), both of Oort cloud origin. each emission line is treated as a free modeling fit parame- ISON was a comet first detected at ∼10 AU from the Sun ter, and increasing the total number of parameters will sig- and was well studied during its first close inner system per- nificantly reduceconfidencein any results due to chi-square ihelion passage. We observed the comet at moderate activ- ity (Qgas ≈ 1028 mol s−1). ISON had also begun suffering testing. However,properconsiderationofstate selective CX crosssectionsofhighlychargedSWionswillreducemodelfit from a series of fragmentation events near the end of our parametersasallemissionlinesperionwillbesetatfixratios observations that markedly ramped up its outgassing activ- determinedbytheircrosssectionsandphotonemissionyields ity(Combietal.2014b). Ourobservationswerealsotakenat (Bodewitsetal.2007).Suchamodelwouldthereforeonlybe a time of variableSW speeds, as indicatedby the Advanced dependent on the heavy SW ion composition, reducing the CompositionExplorer(ACE). model from 700+ parameters down to 10-20. We therefore PanSTARRS, although a naked eye object from Earth in chooseto developa CX modelfromfirst principlesthatwill mid-March, did not have such a favorable close passage by 2 Sniosetal. TABLE1 ChandraCometObservationParameters Texp rc ∆ Lat⊙ Long⊙ QH O vp 2 Comet Obs. Date Prop. Num. (ks) (AU) (AU) (deg) (deg) (1028 mols−1) (kms−1) PanSTARRS 2013Apr17–23 14108442 45 1.10 1.44 84.16 150.5 5a 377∗ ISON 2013Oct31–Nov6 15100583 36 1.18 0.95 1.130 115.0 2b 313 Note. —Observationparameters arelistedasfollows: Chandraobservationdate, observationproposalnumber,exposuretimeTexp,comet-Sundistance rc,comet-Earthdistance∆,HeliosphericLatitudeLat⊙andLongitudeLong⊙,H2OproductionrateQH2O,andsolarwindprotonvelocityvpfromtheACE- SWEPAMonlinedataarchive. DuetothelargedifferenceinheliosphericlatitudebetweenACEandPanSTARRS,itisunlikelytheyexperiencedsimilarSW speeds.Morelikely,PanSTARRSencounteredfastSWduetoitshighaltitude.WedenoteouruncertaintyintheobservedSWspeedvaluewithanasterisk. aCombietal.(2014a) bCombietal.(2014b) Raw PanSTARRS Spectrum 0.8 Raw ISON Spectrum Nominal S3 Background Spectrum Nominal S3 Background Spectrum 0.4 0.6 1 1 0.4 -V0.2 -V e e 1 k 1 k0.2 -c -c e 0 e s s s s 0 nt nt u u o Background-Corrected PanSTARRS Spectrum o Background-Corrected ISON Spectrum C C 0.2 0.2 0 0 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 Energy [keV] Energy [keV] Fig.1.—SummedtotalspectraofX-rayphotoncountsforcometsPanSTARRSandISONextractedfromChandraobservations. Bothspectrautilizenominal S3chipbackgroundemissionsforbackgroundcorrectioncalculations. include all possible lines arising in radiative cascading pro- in drift-scan mode where no active guidance is enabled and cesses of excited SW ions with proper cross sections. This Chandra’spointingwasonlyupdatedtore-centerbeforethe should simplify input parameters of cometary X-ray model- cometmovedoffthechip. ing through limiting of input variables to SW ion composi- SW protonvelocitieswere extractedfromACE, a satellite tion while also improving its physical accuracy through the locatedattheL1Lagrangianpointthatcontinuouslyrecords increaseofemissionlines. Ourmodelmayalsobeutilizedas SW conditions. SW speed at each comet was calculated aremotediagnostictoolforsolarwindcomposition. throughtimeofflightcorrectionsbetweenACEandthecomet Inthisarticle,weanalyzetheChandraX-rayObservatory observations,andtheresultingSWvelocitieswerefoundtobe observations of comets C/2012 S1 (ISON) and C/2011 L4 consistentwithslowSW.However,wenotethelargediscrep- (PanSTARRS).Eachcomethaduniqueconditions,eitherso- ancyin heliosphericlatitude between PanSTARRS and ACE larorcometary,thatmayimpactCXemissions,andthesedif- duringourobservations. SincePanSTARRSwasathighlati- ferentconditionsshouldalsoprovideanexcellenttestforour tude,weinferthatitwasbombardedwithfastSW(Geissetal. model. We describe details regardingthe observations, data 1995; Schwadron&Cravens 2000). ISON was observed at extraction, and our modeling techniques in Section 2. Our similarheliosphericlatitudetoACE,soSWconditionsshould results are presented in Section 3. We discuss our findings besimilar betweenthetwo. In addition,solarX-rayactivity inSection4. Last, we providea summaryofourfindingsin detectedbytheGOESX-raysatelliteindicatesthatseveralM- Section5. classsolarflareeventsoccurredduringISON’sobservations, while solar activity was average for PanSTARRS. See Table 1foradditionaldetailsontheobservationparametersforboth 2. OBSERVATIONSANDANALYSIS comets. 2.1. ChandraObservations Since all observationswere performedin drift-scanmode, For both comets selected, the Chandra observations were we first convert all images to object-centered coordinates performed using the Advanced CCD Imaging Spectrometer throughuseofthesso freezeroutinefoundintheChandraIn- (ACIS). The comet was centered on the S3 chip as it offers teractiveAnalysisofObservations(CIAO)softwarepackage themostsensitivelow-energyresponseintheACISarray,and (Fruscioneetal.2006). Wethengenerateourresultingspec- ACIS was set to very faint mode for all observations to as- traviaCIAO’sspecextractroutineandarecombinedwiththe sistinfilteringoutbadX-rayevents,suchascosmicX-rays, combine spectraroutine. AllstepsareperformedwithCIAO fromthesourceevents. Eachobservationwasalsoperformed ChandraObservationsofC/2012S1andC/2011L4 3 1000 wecangenerateasyntheticbackgroundspectrumthatcanbe ACIS utilized with our data. This technique is beneficial for the HRC lowstatisticaluncertaintyitintroducestoourresultingspec- trum. Inaddition,blank-skycorrectionisoftenpreferredfor extendedsources,suchascomets,ason-chipbackgroundre- 2m] gionsmaybecontaminatedbythesource.Despitethesebene- c 100 a [ fits,thehighvariabilityoftheX-raybackgroundmayresultin e Ar blank-skycorrectionintroducingrandomuncertaintyintoour e calculationsthatwouldnotexistfromtheon-chipmethod. v cti Inouranalysisofbothtechniques,wefindthatthespectral e Eff uncertainty introduced via on-chip correction only becomes 10 significantatenergiesgreaterthan2keVforISONand1keV forPanSTARRS. Sincewe are focusedonanalyzingthe CX emissions up to 1 keV from each comet, we choose the on- chipcorrectionmethodtoavoidintroducingadditionaluncer- 0.1 1 10 tainty dueto the X-raybackgroundvariability. We uniquely Energy [keV] selectthebackgroundareaforeachobservationtoavoidcon- Fig.2.—AcomparisonofeffectiveareafunctionsfortheACISandHRC tamination from other on-chip astronomical objects as they instruments, asdocumentedintheChandrahandbook. ISONobservations variedsignificantlyinlocationbetweeneachobservationdue were performed with both instruments as HRC has a higher sensitivity to totheclosecomet-detectordistanceandincreasingcometve- softX-rayemissionsthanACIS. locity. See Fig 1 for the resulting background spectra and thebackground-correctedsourcespectra.Wenotethatthere- v4.7. ThecumulativecometaryandbackgroundX-rayemis- sulting spectra possess largeerrorbars relativeto the scatter sionspectraareshowninFigure1. spreadofthedata,possiblyindicatinganunknownsourceof Inadditiontotheseobservations,ISONwasalsoobserved systematic error. Allsoftware toolsanddata reductiontech- withtheHighResolutionCamera(HRC)onChandrabetween niqueswerethereforetestedseparatelyforsuchanissue,and theACISvisits,alsoindrift-scanmode.HRCobservationsof nosourcesofsystematicerrorwerefoundinouranalysis. acomethadneverbeenperformedinconjunctionwithACIS observations before. Such observations were proposed for ISONduetoHRC’sincreasedsensitivitytosoftX-rayemis- 2.2.2. CXModelingofCometaryX-RayEmissions sionsoverACIS,seeFigure2,asitmakesthetwoinstruments As discussed in Section 1, a primary goal of our work is complementarytooneanother. Theresultingimagesandour todevelopaCX modelfromfirstprinciplesthatcanprovide discussionoftheirimplicationsareinSection4.1. more accurate diagnostic of the SW plasma interacting with cometarygas. 2.2. SpectrumAnalysis We begin by expanding upon the CX model outlined 2.2.1. BackgroundCorrection in Kharchenko&Dalgarno (2000), Kharchenko&Dalgarno (2001), Krasnopolskyetal. (2002), and Bodewitsetal. Giventhe lowcountrateandextendednatureofcometary (2007). The emitted intensity I of the photon flux in- X-rayemissions, bothproperbackgroundcorrectionandex- duced by CX collisions is defined as the total emission re- posuremapcalculationarecrucial.Sincethespecextractrou- sulting from the interaction between k species of cometary tine in CIAO correctly handles any differences in exposure atoms/moleculesandSWionsl,wherelisdependentonboth created from reprojectingan eventto the referenceframe of the elementand its charge, within the cometaryatmosphere. the comet, we focusour investigationon two possible back- Wedefineitasanintegraloverthelineofsightdistancesand groundcorrectiontechniquesthatwe couldemploy: on-chip thesolidviewingangleΩ , andblank-skycorrections. s On-chipcorrectionisthemostfrequentlyutilizedtechnique incometaryanalysisandisperformedbyisolatingaregionof I(~ω )= n nσ |v~ −v~|P(j)(~ω )dsdΩ , (1) purebackgroundonthesamechipthatobservedthesource.A j Z k l k,l k l k,l j s X backgroundspectrum,assumedtobeconstantovertheentire k,l areaofthe chip, isextractedfromthisregionandsubtracted where n is the cometary particle density, n is the SW ion k l fromtherawsourcespectrumtocreatethesourcespectrum. densityat the comet, σ is the chargetransfercross section k,l SuchamethodisvaluableforX-rayanalysisduetothehigh forcollisionsbetweenkneutralsandlions,v isthecometary k variabilityofX-raybackgroundovertimeasitensuresnear- particle velocity, v is the local SW velocity, and P(j) is the identicalbackgroundtothatfoundintherawsourcespectrum. l k,l photonyieldforemissionswiththeenergy~ω inthecollision However,observationsofextendedsourcesleavelittleareaon j chipforaproperbackgroundregiontobedefined.Inthecase betweenkandlspecies.ThetotalyieldofallX-rayphotonsis ofourcometobservations,thesourceoccupies70–90%ofthe normalizedtounity,where P(j)(~ω ) = 1,pereachunique k,l j S3chip,whichreducesthepossiblebackgroundstatisticswe Pj kandlinordertobevalidonapercollisionalbasis. can gather and can increase the uncertainty in the resulting For our equation’s parameters, both n and v are found sourcespectrum. k k fromobservationaldataonthecomets(Combietal.2014a,b). Thealternativemethodisblank-skybackgroundcorrection, whichisperformedbymatchingthecoordinatesofourcomet The physical parameters P(k,jl) and σk,l are obtained from observation to a similar blank-sky image where only back- previous lab and theoretical research (Dijkkampetal. 1985; ground emissions were observed. By scaling the blank-sky Janev&Winter 1985; Johnson&Soff 1985; Kelly 1987; observationto theexposuremapofouroriginalobservation, Suraudetal. 1991; Cann&Thakkar 1992; Janev 1995; 4 Sniosetal. Wieseetal.1996;Kharchenkoetal.2003;Koutroumpaetal. TABLE2 2006,2009). Inregardstonl andvl,wesetthemequaltoav- SWCompositionRatioInputsandResults eragevaluestakenfrompreviousanalysesoftheSWplasma (Bochsler2007). Avg. ISON PS Avg. ISON PS Ion Ratioa Ratio Ratio Ion Ratioa Ratio Ratio We initially find that our modeled spectral intensity does C6+ 0.318 0.318 0.310 Mg10+ 0.098 0.098 0.078 notaccuratelyfittheobservationaldataduetothehighvari- C5+ 0.210 0.240 0.240 Mg9+ 0.052 — — ance in SW conditions as a function of both time and solar N7+ 0.006 — — Si10+ 0.021 — — longitude, causing inaccurate modeled values for nl and vl. N6+ 0.058 — — S11+ 0.005 — — Since we lack any direct observations of these values at the N5+ 0.065 — — S10+ 0.016 — — comet, we therefore allow these parameters to vary within O8+ 0.070 0.100 0.040 S9+ 0.019 — — physical limits until the best agreement between our rela- O7+ 0.200 0.200 0.100 Fe13+ 0.002 — — tivemodeledintensityandtheobservationalintensityoverthe O6+ 0.730 0.700 0.860 Fe12+ 0.007 — — 0.3–1.0keVenergyrangeisfound. Ne9+ 0.004 0.020 0.004 Fe11+ 0.023 — — Ne8+ 0.084 0.068 0.084 Fe10+ 0.031 — — 2.2.3. CXModelComposition Note. — Model-calculated SW ion ratios for comets ISON and PanSTARRS(PS)incomparisontoaverageslowSWratios. Allratiosare The CX emission spectrum in our model is computed for normalizedwithrespecttothetotalSWoxygen. Allcalculated values are twoindependentmajorgroupsofheavySWions: foundtohaveanaverageuncertaintyof±15%.Valuesleftblankarebecause theobservationalspectrumdoesnotpossesstheresolutionrequiredtoaccu- ratelycalculatethoseratios,andsothemodeltreatsthemasconstants. 1. (Group A): key heavy ions for “3/4 keV” en- aBochsler(2007) ergy interval (C5+, C6+, N5+, N6+, N7+, O6+, O7+, O8+, Ne8+,and Ne9+). This group includes the CX emission spectra generated from collisions be- TABLE3 tween cometary neutrals (primarily H O) and H- SWCompositionComparisontoACE 2 like, He-like, and Li-like heavy SW ions. The CX Source C6+/C5+ O7+/O6+ O8+/O6+ spectra of these ions are reasonable constrained by ISON 1.35 0.28 0.14 lab and theoretical researches (Dijkkampetal. 1985; ACE 1.18+0.80 0.25+0.12 0.09+0.19 −0.48 −0.08 −0.06 Kelly 1987; Suraudetal. 1991; Wieseetal. 1996; PanSTARRS 1.29 0.12 0.05 Kharchenkoetal.2003;Koutroumpaetal.2006,2009; ACE 1.09+0.62 0.22+0.13 0.08+0.14 −0.39 −0.08 −0.05 Chutjianetal.2012). Note. —Acomparisonbetweenthemodel-calculatedSWionratiosand theaveragevaluesobservedbyACE.Allcalculatedvaluesarefoundtohave 2. (GroupB): heavier excited ions(Mgq+, Sq+, Siq+, and anaverageuncertainty of±20%andagreetotheobservational datawithin Feq+) that primarily contribute to the soft X-ray spec- uncertainty. We note that agreement between ACE and PanSTARRS is inconclusivegiventhesignificantdifferenceinheliosphericlatitudebetween tra(below0.4keV).Thecrosssectionsandrelativein- thetwo. tensity of different emission lines of the CX cascad- ing spectra for these ions are less known than for the ISON, which has 32 dof, we find χ2 = 1.1 over the 0.35– ionsfromGroupAbutarewellestimated(Hareletal. R 1.00keVrange.ForPanSTARRS,whichhas30dof,χ2 =1.2 1998;Simcicetal.2010). Theenergypositionofspec- R forthe0.35–1.00keVrange.Allspectraaretruncatedat0.35 trallinesarewelldefined(Kramidaetal.2014). keVduetothecarbonK-shellabsorptionedgedetectorcon- ThespectraofCX cascadingphotonsforGroupsAandB taminationpresentfrom ACIS atenergiesbelowthis thresh- arecomputedindependentlyandthenunifiedintoasynthetic old. These results provide a more complete and physically spectrum that represents the most probable emissions up to accuratepictureof the CX processin cometaryatmospheres 1 keV. Ion elemental and charge composition for all groups thanfoundinthepreviousgenerationofmodels. aretreatedasvariableparametersthatareinitiallysettoaver- Inadditiontoaccuratelymodelingthecometaryemissions, ageSWcompositionratios(Bochsler2007;Leprietal.2013). we compare our SW compositionsresults to contemporane- TheSWcompositionisthenvarieduntiltheχ2valueismini- ous composition ratios provided by ACE. A comparison of mized. DuetoChandra’slowsensitivitybelow0.35keVand ourmodelresultsto ACE,shownin Table 3, demonstratean the lack of accurate calibration near the carbon K-shell line agreementwithin uncertaintyforall observations. These re- at0.284keV,wefindthatvaryingseveralSWionstypesthat sults providean additional, and crucial, confirmationforthe predominantlyemitin thisregionproducesnochangeto χ2. physicalaccuracyofourmodelingtechnique. Italsoleadsus As a result, these SW ion types are left constant as average to considerusing our analysisand modelingof cometaryX- SWcompositionratios. TheinitialSWratiosandourresult- ray spectra in the future as a remote diagnostic tool for SW ingratiosforbothcometobservationsareshowninTable2. composition. 3. RESULTS 4. DISCUSSION Usingthemodeloutlinedintheprevioussection,wecreate 4.1. X-RayEmissionMorphology atheoreticalCXspectrumforthetwocometsobserved. Our WhenanalyzingcometaryX-rayimages,itisimportantto theoreticalspectra are comparedto the averagebackground- rememberthattheoverallemissionmorphologyisdetermined correctedobservationalspectra, and the results are shownin by SW interactionwith the cometaryatmosphereas the ma- Figure3. jorityoftheemittedintensityisduetoCX.Inthecollisionally Wealsocalculatethereducedχ2,alsoknownasχ2,bydi- thickcaseforanactivecomet,weexpectaparaboloidwiththe R vidingχ2 with the degreesof freedom(dof)foreach comet. cometatthefocuswherethemagnitudeofthesemimajoraxis Both comet observationsare binned with a minimum of six isdependentontheatmosphericdensity(Ha¨berlietal.1997; counts per spectral bin for proper gaussian statistics. For Wegmannetal.2004;Lisseetal.2005;Wegmann&Dennerl ChandraObservationsofC/2012S1andC/2011L4 5 0.001 0.002 PanSTARRS Observation ISON Observation -1V SWCX Model -1V SWCX Model ke ke0.0015 -1c -1c e e s s -2m 0.0005 -2m 0.001 c c s s n n oto oto0.0005 h h P P 0 0 10 10 2χ 0 2χ 0 ∆ ∆ -10 -10 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Energy [keV] Energy [keV] Fig.3.—ComparisonbetweenourCXmodelandtheaveragebackground-correctedobservationsforcometsPanSTARRSandISONandtheχ2residuals(∆χ2) ofthedata-modelcomparison. Eachobservationalspectrumhasbeengroupedwithaminimumofsixcountsperbinforproperstatistics. OurmodelvariesSW compositionratiosuntilabest-fitisfound.TheresultingSWcompositionratiosforeachmodelaredetailedinTable2. 2005). In comparison, the X-ray emission structure is de- and the hard X-ray ACIS observations, most notably seen termined by the distribution of gas in the coma for the col- on the October 31 and November 03 visits. These intensity lisionally thin case. For such a situation, we expect to see fluctuations correlate with increases in SW speed as docu- regions of enhanced X-ray emission in regions of higher mentedbyACE,wherethemaximumSWspeedwasrecorded cometaryparticledensity,suchasthosefoundalongjetstruc- November03.SuchanassociationbetweenSWandcometary tures(Lisseetal.2013). emissionintensityispredictedbyourCXmodelasSWspeed We present the ACIS observation images of comet fluctuationsindicatefluctuationsinSWionfreeze-intemper- PanSTARRSinFigure4. Allimageshavebeencorrectedfor atures(Bochsler 2007). Suchtemperaturechangeswill shift differencesin exposuretime and are normalizedto the same theSWchargestatedistribution,producingavaryingaverage linear scale. These images show a constant intensity in X- cometary emission energy based on our normalized photon rayemissionsin allthe observations,aswe expectgiventhe emissionyieldfunction P(j)(~ω ). Aswe see asimilar shift- constantcometarydust/gasemissionratesandSWconditions k,l j ing of the average cometary emission energy, we therefore observed at the time of our observations. We also find that assertthatCXemissionsarethedominantcometaryemission theoverallmorphologyishighlynon-uniform,andsowecon- mechanism in the soft X-ray region, a fact that will become cludethatPanSTARRSwascollisionallythinduringitsobser- importantinourdiscussioninSection4.4. vations.Thisislikelyaresultofthehighdustdensitypresent inthecometaryatmosphereasdustparticlesaresignificantly less efficient in CX X-ray production than molecular gas 4.2. PanSTARRSSpectralAnalysis (Djuric´etal.2005;Wolketal.2009;Lisseetal.2013). Prior to the observation of comet PanSTARRS, there was TheISONobservationswereuniqueasitwasthefirsttime much speculation if its high dust-to-gas ratio would sig- HRCobservationsofacometwereperformedinconjunction nificantly affect its X-ray CX emission intensity as it is with ACIS observations. The extracted ACIS and HRC im- morefavorabletoproduceAugerelectronsinsteadofX-rays agesforthe threeISON observationsare shownin Figure5. when undergoingCX with dust particles (Djuric´etal. 2005; Eachimageset,eitherfromACISorHRC,hasbeencorrected Wolketal. 2009;Lisseetal. 2013). We thereforemake sure for differencesin exposuretime and has been normalizedto tonoteanyX-rayspectralirregularitieswithinourresultsand, thesamelinearscale. TheACISimagesdemonstratetheex- ifso,theirpossibilityofbeingduetodustparticles. pected paraboloid morphology of a collisionally thick case, Utilizing our CX model, we are able to successfully char- whiletheHRCobservationsdepictamorenon-uniformemis- acterize PanSTARRS’ spectrum without making any adjust- sion typical of a collisionally thin case. Given that HRC is mentstoourCX scenario. We findauniquesolutionforthe moresensitivetosoftX-raysthanACIS,thisresultmayindi- emissionspectrumthatfitswelltotheobservationsupto1.0 cate that the soft X-ray emissions are due to CX emissions keV. Above 1.0 keV, the uncertainty in the observations be- from lighter SW ions with smaller cross sections, such as comes too great to distinguish between noise and emission He2+, than the SW ions that emit hard X-rays, such as C6+ peaks. Analysis of SW composition through the use of our and O8+. The reduction in cross section will allow deeper modelshowsalowerthanaverageamountofhighlycharged penetrationintothecometaryatmosphere,anditmaybesub- ions,suchasO8+andNe9+,withanincreaseintheirloweren- stantialenoughtogeneratedifferencein these imagesets. It ergyvariants,likeO6+ andNe8+. Thisresultagreeswithour isalsopossiblethatthe softX-rayemissionsfromISON are previous assessment that PanSTARRS was observed at fast, fromadifferentemissionmechanismthatwouldnotproduce polar SW. Beyond this irregularity, PanSTARRS’ spectrum thesamemorphology,suchasscatteringorfluorescence. possesses no additional traits that would classify it different TheISON imagesetsalso demonstratesignificantfluctua- fromanyothercometX-rayspectrum. tioninthecometaryemissionintensityovertimeanda“see- Although we cannot infer from our spectral analysis how saw” in intensity between the soft X-ray HRC observations PanSTARRS’ largedustquantitiesmay have impactedother emissions mechanisms present within the cometary spectra, 6 Sniosetal. ACIS 100,000 km ACIS 100,000 km ACIS 100,000 km 0.3 Apr 17 Apr 21 Apr 23 0.25 0.2 0.15 0.1 Fig.4.—Chandra/ACIS-SobservationsofcometPanSTARRS.Theimagesarebinnedtoincludeall0.3–1.1keVphotonevents,exposurecorrected,onthe samelinearscale,andsmoothedwitha5×5pixelGaussianfilter.OurresultsshowfluctuationsinX-rayemissionintensity,andtheoverallmorphologyishighly non-uniform. ACIS 100,000 km ACIS 100,000 km ACIS 100,000 km 0.8 Oct 31 Nov 03 Nov 06 0.7 0.6 0.5 0.4 0.3 0.2 0.1 HRC 100,000 km HRC 100,000 km HRC 100,000 km 2.1 Oct 31 Nov 03 Nov 06 2.0 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 Fig.5.—Chandra/ACIS-SandHRC-IobservationsofcometISON.Eachsetofimages(eitherACISorHRC)areexposurecorrected,shownonalinearscale, andsmoothedwitha5×5pixelGaussianfilter. ACISimagesarealsobinnedtoincludeall0.3–2.0keVphotonevents. Theimagesshowa“see-saw”effect betweenthesoftX-rayHRCobservationsandhardX-rayACISobservationswhereanincreaseofintensityinHRCcorrelatestoadecreaseinACIS,andvice versa.ThisresultalsocorrelatestofluctuationsinSWspeedbetweenOctober31andNovember03asseenviaACE. our results indicate that it had little to no observable impact byGOES,werealsoobservedduringISON’sChandravisits. onthecomet’sCXX-rayemissions. Assuch,anydifferences ThesehighlyvolatileSWconditionsmaysignificantlyimpact presentaremorelikelyattributedtotheSWfluxdensityand ISON’semissionspectra. ionizationstateatthetimeofobservation. UsingtheACISobservationsandapplyingthesamemethod asdoneforPanSTARRS,weareabletomodelISON’sspec- 4.3. ISONSpectralAnalysis trumasCXbelow1keVandextractSWcompositionratios. Despite being one of the brightestcomets in recent years, ISON’s ratios confirm the above average SW speed with an the Chandra observations we analyzed were taken slightly overabundanceofhighlychargedSWions,likeO8+andNe9+, priortoISON’sdrasticincreaseingasproductionratestarting that produces a distinct plateau in the spectrum from 0.75– on 2013 November 13. Fluctuations in SW speeds, as con- 1.00keV.TheO7+ ratioistwicethatseenfromPanSTARRS, firmed by ACE, and severalM-class solar flares, as reported bestvisualizedviatheemissionpeakat0.6keV. ChandraObservationsofC/2012S1andC/2011L4 7 0.7 0.284keV,weplottheresultingemissionspectraforeachof Oct 31 the three ISON visits in Figure 6. Our plots show a soft X- 0.6 NNoovv 0036 rayregionat0.2keVthatthedetectorissensitivetoafterour corrections,evenshowingfluctuationsthatagreewiththesoft 0.5 X-rayemissionfluctuationsdetectedbyHRC(seeFigure5). -1V TheoverallshapeofthisfeatureislikelyduetotheACISef- e -1c k0.4 fective area function abruptly decaying toward zero at 0.18 e keV and is not due to any specific emission line. Also, the s nts 0.3 fluctuations between the visits exceed the spectral intensity ou uncertainty in this region, which is 0.08 counts s−1 keV−1, C 0.2 andsowebelievethesefeaturestobephysical. Based on our HRC results from Section 4.1, we assume 0.1 CXemissionstobethemostlikelycauseofthisfeature. We thereforeextendourCXmodeldowntothissoftX-rayregion 0 andplottheresultswiththeobservationaldata.Seethedotted 0 0.2 0.4 0.6 0.8 1 Energy [keV] linesinFigure7forourpredictedCXmodelforeachobser- Fig.6.—ACISspectralintensityforeachobservationofcometISON.All vation.CalculationofauniquesolutionofSWratiosrequired detector contamination fromthecarbonK-shellat0.284keV hasbeenre- to produce such intensities is not possible due to the abun- movedfromthespectra.Theobservedspectralfeatureat0.2keVhasthethe danceofover200uniquelinesfromover15differentSWion samefluctuationinintensitythatisobservedfromtheHRCobservationsand typesthatfallwithinACIS’resolutionofthissoftX-rayfea- exceedstheaveragespectralintensityuncertainty inthisregion, andsowe concludethatthisfeatureisphysical.Possibleoriginsofthisspectralfeature ture. We thereforechoosetoleaveourmodelataverageSW arediscussedinSection4.4. abundancesinthisregion. Wenotethatfixingtheseparame- tersproducesnodifferenceinthespectralfitoverthe0.3–1.0 In addition to these results, ISON exhibits some possible keVenergyrange. peak structures in its emission spectrum at energies above OurresultsshowthatouraverageCXmodelisnotcapable 1 keV: one peak at 1.35 keV and another at 1.85 keV, as ofproducingthenecessaryintensitytomatchtheobservations seen in Figure 1. Such peaks have been previously seen foreithertheISONemissionsorthePanSTARRSemissions, in Chandra’s observations of Comet 153P (Ikeya-Zhang) whichare notshown. Furthermore,theSW abundancesthat (Ewingetal.2013),anothercometviewedduringvolatileSW ourmodelwoulddemandtomatchthesefeaturesinintensity conditions.Ourmodelispresentlyunabletoaccuratelycalcu- far exceed their physical boundaries, with most abundances latetheoreticalCX emissionsinthisenergyrangeduetothe requiring an increase by an order of magnitude. Although lackofinformationaboutthepresenceofsuchhighlycharged such exotic SW compositions are not impossible given its ions in the solar wind plasma, but we may comment on the constantlyfluctuatingnature,theconsistentpresenceofthese possibleemissioncandidates. soft X-ray features during both fast and slow SW indicate Comparison to atomic emission line tables from NIST theseshouldbegeneratedunderaverageSWconditions. indicate that the most probable ions for each emission is AlthoughourcurrentCXmodeldoesnotagreewiththesoft Mg XI (1s21S–1s2p1,3P) for 1.35 keV and either Si XIII X-rayintensitiesdetected,weonlyconsiderasingleelectron (1s21S–1s2p1,3P) or Mg XII (1s2S–4p2P) for 1.85 keV captureeventperincomingSWion.Sequentialcaptureevents (Kramidaetal. 2014). However, it is unclear if these peaks mayoccurforanionifthecometaryatmosphereiscollision- couldbearesultfromCXastheseexoticcandidateshavenot allythick,increasingtheamountofsoftX-raysemittedfrom beendetectedviainsituobservationsofSWioncomposition the system as the ion charge state decreases (O8+→ O7+→ (vonSteigeretal.2000;Leprietal.2013). Furthermore,the- O6+→ ... ). We therefore modify our CX model to include oreticalmodelsdescribingthechargeabundanceofheavySW these sequential capture events per ion as it may solve our ionspredictanextremelylowprobabilityoffindingtheseions softX-rayintensitydeficit. because of the inability to reach such high freezing-in tem- Forouranalysis,wecalculateanupperlimitontheincrease peraturesinregularSWandcoronalmassejections(Bochsler tosoftX-rayCXemissionsfromsequentialcaptureeventsby 2007). Ontheotherhand,thesespectrallinesareclearlypre- assumingallSWionsareneutralizedthroughinteractionwith sented in the spectra of the solar X-ray flaresas well as in a thecometaryatmosphere. OurresultsarepresentedinFigure regularX-rayemission fromthe Sun (McKenzieetal. 1985; 7, andtheyshow thatthe additionalCX eventsarenotsuffi- Dereetal.1997;Landietal.2013).Itthereforemaybepossi- cienttoequaltheobservedsoftX-rayintensities.Wefindthat blethesepeaksareduetoadifferentmechanismwhoseemis- theupperlimitof CX emissionsonlyincreasesthe totalsoft sions are increased by solar flare activity, such as scattering X-rayintensity∼50%,whichisnotenoughtoaccountforthe of solar X-rays (Krasnopolsky 1997; Sniosetal. 2014). At factorsofthreetosixbetweenthemodelandtheobservations. present, we cannot conclude what is the primary source of Furthermore,westressthattheactualrateofsequentialcap- these exoticemissionsdetectedfromISON asfurtheranaly- ture events present in these cometary systems is lower than sisofISON’sspectrumwitharevisedCXmodel,andpossibly thisupperlimit,sotheactualemissionintensitieswillreside alsoascatteringemissionmodel,isrequired. between our model and the upper limit. We therefore find itunlikelythatsequentialCX eventscouldaccountforthese 4.4. PotentialSoftX-RayEmissionsfromACIS softX-rayfeatures. While examining the ACIS spectra from ISON, we found AsweareconfidentthatbothourCXmodel’sresultingSW a peak-like feature located at 0.2 keV. This feature was also compositionandphotonyieldemissionratesareaccurate,we foundinthePanSTARRSobservations,butwithalowerrela- therefore consider two possible explanations for the soft X- tivespectralintensity.Afterremovingtheportionofthespec- raydiscrepancy: tra caused by the carbon K-shell detector contamination at 8 Sniosetal. 0.7 0.7 Oct 31 Obs Nov 03 Obs CX Model CX Model 0.6 0.6 Model Upper Model Upper -1V0.5 Limit -1V0.5 Limit e e k k 1 0.4 1 0.4 -c -c e e s s0.3 s s0.3 nt nt u u o0.2 o0.2 C C 0.1 0.1 0 0 40 40 2χ 0 2χ 0 ∆ ∆ -40 -40 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Energy [keV] Energy [keV] 0.7 Nov 06 Obs CX Model 0.6 Model Upper -1V0.5 Limit e k 1 0.4 -c e s s0.3 nt ou0.2 C 0.1 0 40 2χ 0 ∆ -40 0 0.2 0.4 0.6 0.8 1 Energy [keV] Fig.7.—ACISspectralintensity foreachobservation ofcometISON(solidlines)anditsrespective modeledCXusingaverageSWcompositions (dotted coloredlines). AlldetectorcontaminationfromthecarbonK-shellat0.284keVhasbeenremovedfromthespectra. Despitetheexcellentagreementabove0.4 keV,ourmodelfailstomatchtheshapeandintensityofthesoftX-rayspectralfeature.WealsocalculateanupperlimittosoftX-rayCXemissionsbyaccounting forsequentialCXeventsandassumingthatallSWionsareneutralizedthroughinteractionwiththecometaryatmosphere(dottedblacklines),andourresults showtheseadditionstobeinsufficienttoequaltheobservedsoftX-rayintensities. WethereforebelievethesesoftX-rayfeaturestobeCXfromanunaccounted SWion(suchasHe2+),detectorcontamination,oracombinationoftheseoptions. 1. Since the CX modeldoes notmatch the observational cometaryorigins,orifthisfeaturehasmanifesteditself intensities, it is possible we lack the SW ion type re- overtime,indicatingadetectorissue. quired to produce this feature. He2+ CX emissions, currently not included in our analysis, would be de- The required analysis for each of these possibilities is be- tectable in this soft X-ray region due to the low res- yondthe scope of thisarticle, butwe believethatany future olution of ACIS, and its high abundance may provide workonthesesoftX-rayfeaturesfromACISshouldprovide the required order-of-magnitude increase in intensity athoroughanalysisofeachpossibleexplanationtodetermine (Kharchenko&Dalgarno 2001; Bodewitsetal. 2004). thecauseoftheseuniquefindings. Such SW ions would also have deeper penetration in thecometaryatmosphere,whichmightalsoexplainthe 5. CONCLUSIONS collisionally thin appearance of the HRC morphology Insummary,wehaveusedChandrato studytwo verydif- discussedinSection4.1.Futureiterationsofourmodel ferentOortCloudcomets,thegas-richC/2012S1ISONand should include this ion and compare the modified re- thedust-richC/2011L4PanSTARRS.Bothcometswereob- sultstothesoftX-rayemissionsfromACIS. served within 1 AU heliocentric distance of the Sun, when they were active. The observed X-ray morphologies were 2. ThesoftX-rayfeaturemaybearesultofpreviouslyun- dramaticallydifferent,however,withISONdisplayinganex- documenteddetectorcontaminationordegradationthat tended,well-developedX-raycomaandPanSTARRSproduc- sharplydecaysbelow0.2keV,producingapeakinob- ing an unformed X-ray haze. The two comets also experi- served spectrum. Examination of similar comets ob- encedmarkedlydifferentSWconditions,withISONimpact- servedatdifferentstagesofACIS’lifetimewouldshow inganexcitedwind,whilePanSTARRStraveledthroughfast ifsuchasoftX-rayfeatureisalwayspresent,indicating SW. ChandraObservationsofC/2012S1andC/2011L4 9 WedevelopedanupdatedCXemissionmodelthatincludes andMg10+.WithfurtherdevelopmentofCXX-raymodeling, large amountsof ion spectral lines induced in CX collisions suchanapplicationwouldbepossibleforanyCXemissions, and simplifies input variables while improving the physical not just those from comets. Our model also simplifies the accuracyin comparisonto previousmodels. Ourmodelwas variable inputs and provides an additional information on used to analyze Chandra observations of comets ISON and SW composition. We therefore intend to use such a model PanSTARRS, andwe foundstrongemissionsinducedin CX for all future CX analyses of cometary and planetary X-ray collisionsofSWionsnormallypresentwithincometaryemis- emissionsaswellasforinvestigationsofCXX-raysinduced sions(C5+, C6+, N5+, N6+, N7+, O6+, O7+, O8+, Ne8+, Ne9+) ininteractionbetweentheSWplasmaandinterstellargas. from both comets. Analysis of ISON spectra shows higher Inadditiontoourmodelingresults,wefoundthepossibil- concentrationsofO8+ andNe9+thanPanSTARRS,indicating ityofsoftX-rayemissionsaround0.2keVdetectedfromboth higherSWionfreeze-intemperatureduringitsobservations. comets ISON and PanSTARRS via ACIS. These soft X-ray AnalysisofISON’sspectrumalsoshowshigh-energyspec- features fluctuate similarly to those observed from the HRC tral features above 1 keV. To clarify the physical origin of observationsandexceedtheaveragespectralintensityuncer- these cometary“hard”X-rays, we intend in the futureto in- tainty,leadingustobelievethesefeaturestobecometaryCX cludeCXemissionsfromexoticSWions,suchasMg11+and in origin. We extended our CX model to this soft X-ray re- Si13+,whichwillextendourmodelbeyond1keV.Analysisof gion to compare, only to find our results lower in intensity high-energyspectralfeatureswillallowustopredictthetotal thanthe observationsbyan orderofmagnitude. We also re- ratioofexoticSW ionsandtodiscusswhetherthosequanti- vise ourmodelto includesequentialCX captureeventsas it tiescouldbeobservableusingcurrenttools. Wewillalsoin- will incease soft X-ray intensities, but we find that even the cludeemissioncontributionsfromscatteringandfluorescence inclusionofmorecaptureeventsisnotsufficienttomatchthe ofenergeticsolarX-rays,especiallyduringsolarX-rayflares observed intensities. Based on our confidence in the model events, and X-ray emissions of non-thermal energetic elec- from its previous results, we believe this discrepancy to be tronsduetoelectronicimpactorbremsstrahlungmechanisms. aresultofeitheralackofSWiontypesthatproducesignifi- Accurateinvestigationsofthespectralmorphology,whichare cantemissionsinthesoftX-rayregion(suchasHe+),detector differentforeachmechanism,willalsobeperformed.Sucha contaminationordegradation,oracombinationofthesepos- discussionwouldestablishahierarchyofpotentialcontribut- sibilities. Investigations of these soft X-ray features should ingmechanismsincometaryX-rayspectraabove1keVand carefully explore each explanation as confirmation of these provideinsightontheoriginoftheobservedhighenergyspec- featuresasphysicalemissionswouldopennewopportunities tralfeatures. inunderstandingcometaryemissionprocessesviaChandra. Beyond successfully analyzing comets ISON and PanSTARRS, we also demonstrated our model’s poten- WewouldliketoacknowledgeNationalOceanicandAtmo- tialuseasaSW ioncompositionanalyzer. Ourcomposition spheric Administration for their Advanced Composition Ex- resultsagreewellwithotherSWcompositiontoolsavailable, plorer data. The work of B.S. and V.K. on this project has such as ACE, while also calculating unique composition beensupportedbyNASAgrantNNX10AB88G.Theworkof ratiosnotavailablethroughtheseothertools,likeNe8+,Ne9+, C.L.hasbeensupportedbyChandragrantGO4-15001A. REFERENCES Bochsler,P.2007,A&ARev.,14,1 Koutroumpa,D.,Lallement,R.,Kharchenko,V.,etal.2006,A&A,460,289 Bodewits,D.,Christian,D.J.,Torney,M.,etal.2007,A&A,469,1183 Koutroumpa,D.,Lallement,R.,Raymond,J.C.,&Kharchenko,V.2009, Bodewits,D.,McCullough,R.W.,Tielens,A.G.G.M.,&Hoekstra,R. ApJ,696,1517 2004,PhysicaScripta,70,C17 Kramida,A.,Ralchenko,Y.,Reader,J.,&NISTASDTeam.2014,NIST Cann,N.M.&Thakkar,A.J.1992,Phys.Rev.A,46,5397 AtomicSpectraDatabase(ver.5.2) Chutjian,A.,Simcic,J.,Madzunkov,S.M.,etal.2012,JPCS,388,012042 Krasnopolsky,V.A.1997,Icarus,128,368 Combi,M.R.,Bertaux,J.-L.,Que´merais,E.,etal.2014a,ApJ,147,126 Krasnopolsky,V.A.,Christian,D.J.,Kharchenko,V.,etal.2002,Icarus, Combi,M.R.,Fougere,N.,Ma¨kinen,J.T.T.,etal.2014b,ApJ,788,L7 160,437 Cravens,T.E.1997,Geophys.Res.Lett.,24,105 Landi,E.,Young,P.R.,Dere,K.P.,Zanna,G.D.,&Mason,H.E.2013, Dennerl,K.2010,SpaceSci.Rev.,157,57 ApJ,763,86 Dere,K.P.,Landi,E.,Mason,H.E.,Fossi,B.C.M.,&Young,P.R.1997, Lepri,S.T.,Landi,E.,&Zurbuchen,T.H.2013,ApJ,768,94 A&AS,125,149 Lisse,C.,Christian,D.,Wolk,S.,etal.2013,Icarus,222,752 Dijkkamp,D.,Gordeev,Y.S.,Brazuk,A.,Drentje,A.G.,&deHeer,F.J. Lisse,C.M.,Christian,D.J.,Dennerl,K.,etal.2005,ApJ,635,1329 1985,J.Phys.B,18,737 Lisse,C.M.,Cravens,T.E.,&Dennerl,K.2004,inCometsII,ed. Djuric´,N.,Lozano,J.A.,Smith,S.J.,&Chutjian,A.2005,ApJ,635,718 M.Festou,H.U.Keller,&H.A.Weaver(Tucson:TheUniversityof Ewing,I.,Christian,D.J.,Bodewits,D.,etal.2013,ApJ,763,66 ArizonaPress),631 Fruscione,A.,McDowell,J.C.,Allen,G.E.,etal.2006,Proc.SPIE,6270, Lisse,C.M.,Dennerl,K.,Englhauser,J.,etal.1996,Science,274,205 62701V McKenzie,D.L.,Landecker,P.B.,Feldman,U.,&Doschek,G.A.1985, Geiss,J.,Gloeckler,G.,&Steiger,R.V.1995,SpaceSci.Rev.,72,49 ApJ,289,849 Ha¨berli,R.M.,Gombosi,T.I.,Zeeuw,D.L.D.,Combi,M.R.,&Powell, Meech,K.J.,Yang,B.,Kleyna,J.,etal.2013,ApJL,776,L20 K.G.1997,Science,276,939 Raouafi,N.-E.,Lisse,C.M.,Stenborg,G.,Jones,G.H.,&Schmidt,C.A. Harel,C.,Joiun,H.,&Pons,B.1998,At.Data.Nucl.DataTables,68,279 2015,J.Geophys.Res.Space,120,5329 Janev,R.K.1995,AtomicandMolecularProcessesinFusionEdgePlasmas Schwadron,N.A.&Cravens,T.E.2000,ApJ,544,558 (NewYork:Plenum) Simcic,J.,Schultz,D.R.,Mawhorter,R.J.,etal.2010,Phys.Rev.A,81, Janev,R.K.&Winter,H.1985,Phys.Rep.,117,265 062715 Johnson,W.R.&Soff,G.1985,At.DataNucl.DataTab.,33,405 Snios,B.,Lewkow,N.,&Kharchenko,V.2014,A&A,568,A80 Kelly,R.L.1987,J.Phys.Chem.Ref.Data,16,649 Suraud,M.G.,Hoekstra,R.,deHeer,F.J.,Bonnet,J.J.,&Morgenstern,R. Kharchenko,V.&Dalgarno,A.2000,J.Geophys.Res,105,18351 1991,J.Phys.B,24,2543 Kharchenko,V.&Dalgarno,A.2001,ApJ,554,L99 vonSteiger,R.,Schwadron,N.A.,Fisk,L.A.,etal.2000,J.Geophys.Res, Kharchenko,V.,Rigazio,M.,Dalgarno,A.,&Krasnopolsky,V.A.2003, 105,27217 ApJ,585,L73 Wegmann,R.&Dennerl,K.2005,A&A,430,L33 10 Sniosetal. Wegmann,R.,Dennerl,K.,&Lisse,C.M.2004,A&A,428,647 Wolk,S.J.,Lisse,C.M.,Bodewits,D.,Christian,D.J.,&Dennerl,K.2009, Wiese,W.L.,Fuhr,J.R.,&Deters,T.M.1996,AtomicTransition ApJ,694,1293 ProbabilitiesofCarbon,Nitrogen,andOxygen:ACriticalData Yang,B.,Keane,J.,Meech,K.,Owen,T.,&Wainscoat,R.2014,ApJL, Compilation(Washington,DC:Am.Chem.Soc.) 784,L23

ebook img

Chandra Observations of Comets C/2012 S1 (ISON) and C/2011 L4 (PanSTARRS) PDF

0.41 MB·
by  Bradford Snios
#journals #arxiv
Save to my drive
Quick download
Download
Upgrade Premium
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Chandra Observations of Comets C/2012 S1 (ISON) and C/2011 L4 (PanSTARRS)

DraftversionFebruary25,2016 PreprinttypesetusingLATEXstyleemulateapjv.01/23/15 CHANDRAOBSERVATIONSOFCOMETSC/2012S1(ISON)ANDC/2011L4(PANSTARRS) BradfordSnios1,VasiliKharchenko1,CareyM.Lisse2,ScottJ.Wolk3,KonradDennerl4,andMichaelR.Combi5 1DepartmentofPhysics,UniversityofConnecticut,Storrs,CT06269,USA 2PlanetaryExplorationGroup,SpaceDepartment,JohnsHopkinsUniversityAppliedPhysicsLaboratory,Laurel,MD20723,USA 3ChandraX-RayObservatoryCenter,Harvard-SmithsonianCenterforAstrophysics,Cambridge,MA02138,USA 4Max-Planck-Institutfu¨rextraterrestrischePhysik,D-85748Garching,Germanyand 6 5DepartmentofClimateandSpaceSciencesandEngineering,UniversityofMichigan,AnnArbor,MI48109,USA 1 DraftversionFebruary25,2016 0 Abstract 2 WepresentourresultsontheChandraX-rayObservatoryAdvancedCCDImagingSpectrometer(ACIS)ob- b servationsofthebrightOortCloudcometsC/2012S1 (ISON)andC/2011L4(PanSTARRS).ISONwasob- e servedbetween2013October31–November06duringvariablespeedsolarwind(SW),andPanSTARRSwas F observedbetween 2013 April17–23during fast SW. ISON producedan extendedparabolic X-ray morphol- 4 ogyconsistentwithacollisionallythickcoma,whilePanSTARRSdemonstratedonlyadiffuseX-ray-emitting 2 region.Weconsidertheseemissionstobefromchargeexchange(CX)andmodeleachcomet’semissionspec- trum from first principles accordingly. Our model agrees with the observational spectra and also generates ] compositionratiosforheavy,highlychargedSWionsinteractingwiththecometaryatmosphere. Wecompare E ourderivedSWioncompositionstoobservationaldataandfindastrongagreementbetweenthem. Thesere- H sultsfurtherdemonstratetheutilityofCXemissionsasaremotediagnosticstoolofbothastrophysicalplasma . interactionandSWcomposition. Inaddition,weobservepotentialsoftX-rayemissionsviaACISaround0.2 h keV from both comets that are correlated in intensity to the hard X-ray emissions between 0.4–1.0keV. We p - fitourCXmodeltotheseemissions,butourlackofauniquesolutionatlowenergiesmakesitimpossibleto o concludeif theyare cometaryCX in origin. We lastly discussprobableemissionmechanismsourcesforthe r softX-raysandexplorenewopportunitiesthesefindingspresentinunderstandingcometaryemissionprocesses t s viaChandra. a Subject headings:comets: individual(Comet S1/ISON, Comet L4/PanSTARRS) – solar wind – techniques: [ spectroscopic–X-rays:general 2 v 2 1. INTRODUCTION theinnerplanetsandwasonlysparselyobserved.Itdid,how- 2 Cometary X-ray emissions, originally discovered by ever,demonstrateafantasticallyrichoutpouringofdustyma- 6 Lisseetal. (1996) and now observed in over 30 comets, are terialin2013March–Aprilasitpassedthroughperihelion,as 6 a well-studied phenomenon. It has been shown thatthe ma- seenbySTEREO(Raouafietal.2015). PanSTARRSisanun- 0 jority of these emissions are caused by solar wind Charge usuallydust-richcomet,withadust-to-gasmassratiogreater . than4 (Yangetal. 2014). Bycontrast, ISONwas seentobe 1 Exchange (CX) interactions between highly charged, heavy a dust-poor comet with a dust-to-gas mass ratio less than 1 0 solar wind (SW) ions (∼0.1% of all solar wind ions) and (Meechetal.2013).SincetheChandraobservationsforthese 6 neutral gas ejected from the comet nucleus into the coma twocometshavenotpreviouslybeenanalyzed,wedecideto 1 (Cravens1997;Krasnopolsky1997;Kharchenkoetal. 2003; examinetheir emissions for detailed analysis and interpreta- : Lisseetal. 2004; Bodewitsetal. 2007; Dennerl 2010). A v tionofthecometaryX-rayemissionspectraviamodeling. simplified theoretical description of interaction between the i To model CX X-ray emissions for these comets, we X SW plasma and cometary atmosphere shows that the emis- decide to expand upon previous modeling techniques sion originatespredominantlyfrom the sunward hemisphere r (Kharchenko&Dalgarno 2000, 2001; Krasnopolskyetal. a of the neutral coma and creates a projected paraboloid of 2002;Bodewitsetal.2007). Thoughthesemodelsarerobust, emissionwiththecometatitsfocalpoint(Ha¨berlietal.1997; most only incorporate the primary emission lines, generally Wegmannetal.2004). 10–20lines,outofthepossible700+linesthatmaybegener- Tworecentlydiscoveredcometsthatwereobservedbythe atedinanaverageCXinteractionwithacometaryatmosphere Chandra X-ray Observatory are comets C/2012 S1 (ISON) (Kharchenkoetal.2003).Thisistypicallyperformedbecause and C/2011 L4 (PanSTARRS), both of Oort cloud origin. each emission line is treated as a free modeling fit parame- ISON was a comet first detected at ∼10 AU from the Sun ter, and increasing the total number of parameters will sig- and was well studied during its first close inner system per- nificantly reduceconfidencein any results due to chi-square ihelion passage. We observed the comet at moderate activ- ity (Qgas ≈ 1028 mol s−1). ISON had also begun suffering testing. However,properconsiderationofstate selective CX crosssectionsofhighlychargedSWionswillreducemodelfit from a series of fragmentation events near the end of our parametersasallemissionlinesperionwillbesetatfixratios observations that markedly ramped up its outgassing activ- determinedbytheircrosssectionsandphotonemissionyields ity(Combietal.2014b). Ourobservationswerealsotakenat (Bodewitsetal.2007).Suchamodelwouldthereforeonlybe a time of variableSW speeds, as indicatedby the Advanced dependent on the heavy SW ion composition, reducing the CompositionExplorer(ACE). model from 700+ parameters down to 10-20. We therefore PanSTARRS, although a naked eye object from Earth in chooseto developa CX modelfromfirst principlesthatwill mid-March, did not have such a favorable close passage by 2 Sniosetal. TABLE1 ChandraCometObservationParameters Texp rc ∆ Lat⊙ Long⊙ QH O vp 2 Comet Obs. Date Prop. Num. (ks) (AU) (AU) (deg) (deg) (1028 mols−1) (kms−1) PanSTARRS 2013Apr17–23 14108442 45 1.10 1.44 84.16 150.5 5a 377∗ ISON 2013Oct31–Nov6 15100583 36 1.18 0.95 1.130 115.0 2b 313 Note. —Observationparameters arelistedasfollows: Chandraobservationdate, observationproposalnumber,exposuretimeTexp,comet-Sundistance rc,comet-Earthdistance∆,HeliosphericLatitudeLat⊙andLongitudeLong⊙,H2OproductionrateQH2O,andsolarwindprotonvelocityvpfromtheACE- SWEPAMonlinedataarchive. DuetothelargedifferenceinheliosphericlatitudebetweenACEandPanSTARRS,itisunlikelytheyexperiencedsimilarSW speeds.Morelikely,PanSTARRSencounteredfastSWduetoitshighaltitude.WedenoteouruncertaintyintheobservedSWspeedvaluewithanasterisk. aCombietal.(2014a) bCombietal.(2014b) Raw PanSTARRS Spectrum 0.8 Raw ISON Spectrum Nominal S3 Background Spectrum Nominal S3 Background Spectrum 0.4 0.6 1 1 0.4 -V0.2 -V e e 1 k 1 k0.2 -c -c e 0 e s s s s 0 nt nt u u o Background-Corrected PanSTARRS Spectrum o Background-Corrected ISON Spectrum C C 0.2 0.2 0 0 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 Energy [keV] Energy [keV] Fig.1.—SummedtotalspectraofX-rayphotoncountsforcometsPanSTARRSandISONextractedfromChandraobservations. Bothspectrautilizenominal S3chipbackgroundemissionsforbackgroundcorrectioncalculations. include all possible lines arising in radiative cascading pro- in drift-scan mode where no active guidance is enabled and cesses of excited SW ions with proper cross sections. This Chandra’spointingwasonlyupdatedtore-centerbeforethe should simplify input parameters of cometary X-ray model- cometmovedoffthechip. ing through limiting of input variables to SW ion composi- SW protonvelocitieswere extractedfromACE, a satellite tion while also improving its physical accuracy through the locatedattheL1Lagrangianpointthatcontinuouslyrecords increaseofemissionlines. Ourmodelmayalsobeutilizedas SW conditions. SW speed at each comet was calculated aremotediagnostictoolforsolarwindcomposition. throughtimeofflightcorrectionsbetweenACEandthecomet Inthisarticle,weanalyzetheChandraX-rayObservatory observations,andtheresultingSWvelocitieswerefoundtobe observations of comets C/2012 S1 (ISON) and C/2011 L4 consistentwithslowSW.However,wenotethelargediscrep- (PanSTARRS).Eachcomethaduniqueconditions,eitherso- ancyin heliosphericlatitude between PanSTARRS and ACE larorcometary,thatmayimpactCXemissions,andthesedif- duringourobservations. SincePanSTARRSwasathighlati- ferentconditionsshouldalsoprovideanexcellenttestforour tude,weinferthatitwasbombardedwithfastSW(Geissetal. model. We describe details regardingthe observations, data 1995; Schwadron&Cravens 2000). ISON was observed at extraction, and our modeling techniques in Section 2. Our similarheliosphericlatitudetoACE,soSWconditionsshould results are presented in Section 3. We discuss our findings besimilar betweenthetwo. In addition,solarX-rayactivity inSection4. Last, we providea summaryofourfindingsin detectedbytheGOESX-raysatelliteindicatesthatseveralM- Section5. classsolarflareeventsoccurredduringISON’sobservations, while solar activity was average for PanSTARRS. See Table 1foradditionaldetailsontheobservationparametersforboth 2. OBSERVATIONSANDANALYSIS comets. 2.1. ChandraObservations Since all observationswere performedin drift-scanmode, For both comets selected, the Chandra observations were we first convert all images to object-centered coordinates performed using the Advanced CCD Imaging Spectrometer throughuseofthesso freezeroutinefoundintheChandraIn- (ACIS). The comet was centered on the S3 chip as it offers teractiveAnalysisofObservations(CIAO)softwarepackage themostsensitivelow-energyresponseintheACISarray,and (Fruscioneetal.2006). Wethengenerateourresultingspec- ACIS was set to very faint mode for all observations to as- traviaCIAO’sspecextractroutineandarecombinedwiththe sistinfilteringoutbadX-rayevents,suchascosmicX-rays, combine spectraroutine. AllstepsareperformedwithCIAO fromthesourceevents. Eachobservationwasalsoperformed ChandraObservationsofC/2012S1andC/2011L4 3 1000 wecangenerateasyntheticbackgroundspectrumthatcanbe ACIS utilized with our data. This technique is beneficial for the HRC lowstatisticaluncertaintyitintroducestoourresultingspec- trum. Inaddition,blank-skycorrectionisoftenpreferredfor extendedsources,suchascomets,ason-chipbackgroundre- 2m] gionsmaybecontaminatedbythesource.Despitethesebene- c 100 a [ fits,thehighvariabilityoftheX-raybackgroundmayresultin e Ar blank-skycorrectionintroducingrandomuncertaintyintoour e calculationsthatwouldnotexistfromtheon-chipmethod. v cti Inouranalysisofbothtechniques,wefindthatthespectral e Eff uncertainty introduced via on-chip correction only becomes 10 significantatenergiesgreaterthan2keVforISONand1keV forPanSTARRS. Sincewe are focusedonanalyzingthe CX emissions up to 1 keV from each comet, we choose the on- chipcorrectionmethodtoavoidintroducingadditionaluncer- 0.1 1 10 tainty dueto the X-raybackgroundvariability. We uniquely Energy [keV] selectthebackgroundareaforeachobservationtoavoidcon- Fig.2.—AcomparisonofeffectiveareafunctionsfortheACISandHRC tamination from other on-chip astronomical objects as they instruments, asdocumentedintheChandrahandbook. ISONobservations variedsignificantlyinlocationbetweeneachobservationdue were performed with both instruments as HRC has a higher sensitivity to totheclosecomet-detectordistanceandincreasingcometve- softX-rayemissionsthanACIS. locity. See Fig 1 for the resulting background spectra and thebackground-correctedsourcespectra.Wenotethatthere- v4.7. ThecumulativecometaryandbackgroundX-rayemis- sulting spectra possess largeerrorbars relativeto the scatter sionspectraareshowninFigure1. spreadofthedata,possiblyindicatinganunknownsourceof Inadditiontotheseobservations,ISONwasalsoobserved systematic error. Allsoftware toolsanddata reductiontech- withtheHighResolutionCamera(HRC)onChandrabetween niqueswerethereforetestedseparatelyforsuchanissue,and theACISvisits,alsoindrift-scanmode.HRCobservationsof nosourcesofsystematicerrorwerefoundinouranalysis. acomethadneverbeenperformedinconjunctionwithACIS observations before. Such observations were proposed for ISONduetoHRC’sincreasedsensitivitytosoftX-rayemis- 2.2.2. CXModelingofCometaryX-RayEmissions sionsoverACIS,seeFigure2,asitmakesthetwoinstruments As discussed in Section 1, a primary goal of our work is complementarytooneanother. Theresultingimagesandour todevelopaCX modelfromfirstprinciplesthatcanprovide discussionoftheirimplicationsareinSection4.1. more accurate diagnostic of the SW plasma interacting with cometarygas. 2.2. SpectrumAnalysis We begin by expanding upon the CX model outlined 2.2.1. BackgroundCorrection in Kharchenko&Dalgarno (2000), Kharchenko&Dalgarno (2001), Krasnopolskyetal. (2002), and Bodewitsetal. Giventhe lowcountrateandextendednatureofcometary (2007). The emitted intensity I of the photon flux in- X-rayemissions, bothproperbackgroundcorrectionandex- duced by CX collisions is defined as the total emission re- posuremapcalculationarecrucial.Sincethespecextractrou- sulting from the interaction between k species of cometary tine in CIAO correctly handles any differences in exposure atoms/moleculesandSWionsl,wherelisdependentonboth created from reprojectingan eventto the referenceframe of the elementand its charge, within the cometaryatmosphere. the comet, we focusour investigationon two possible back- Wedefineitasanintegraloverthelineofsightdistancesand groundcorrectiontechniquesthatwe couldemploy: on-chip thesolidviewingangleΩ , andblank-skycorrections. s On-chipcorrectionisthemostfrequentlyutilizedtechnique incometaryanalysisandisperformedbyisolatingaregionof I(~ω )= n nσ |v~ −v~|P(j)(~ω )dsdΩ , (1) purebackgroundonthesamechipthatobservedthesource.A j Z k l k,l k l k,l j s X backgroundspectrum,assumedtobeconstantovertheentire k,l areaofthe chip, isextractedfromthisregionandsubtracted where n is the cometary particle density, n is the SW ion k l fromtherawsourcespectrumtocreatethesourcespectrum. densityat the comet, σ is the chargetransfercross section k,l SuchamethodisvaluableforX-rayanalysisduetothehigh forcollisionsbetweenkneutralsandlions,v isthecometary k variabilityofX-raybackgroundovertimeasitensuresnear- particle velocity, v is the local SW velocity, and P(j) is the identicalbackgroundtothatfoundintherawsourcespectrum. l k,l photonyieldforemissionswiththeenergy~ω inthecollision However,observationsofextendedsourcesleavelittleareaon j chipforaproperbackgroundregiontobedefined.Inthecase betweenkandlspecies.ThetotalyieldofallX-rayphotonsis ofourcometobservations,thesourceoccupies70–90%ofthe normalizedtounity,where P(j)(~ω ) = 1,pereachunique k,l j S3chip,whichreducesthepossiblebackgroundstatisticswe Pj kandlinordertobevalidonapercollisionalbasis. can gather and can increase the uncertainty in the resulting For our equation’s parameters, both n and v are found sourcespectrum. k k fromobservationaldataonthecomets(Combietal.2014a,b). Thealternativemethodisblank-skybackgroundcorrection, whichisperformedbymatchingthecoordinatesofourcomet The physical parameters P(k,jl) and σk,l are obtained from observation to a similar blank-sky image where only back- previous lab and theoretical research (Dijkkampetal. 1985; ground emissions were observed. By scaling the blank-sky Janev&Winter 1985; Johnson&Soff 1985; Kelly 1987; observationto theexposuremapofouroriginalobservation, Suraudetal. 1991; Cann&Thakkar 1992; Janev 1995; 4 Sniosetal. Wieseetal.1996;Kharchenkoetal.2003;Koutroumpaetal. TABLE2 2006,2009). Inregardstonl andvl,wesetthemequaltoav- SWCompositionRatioInputsandResults eragevaluestakenfrompreviousanalysesoftheSWplasma (Bochsler2007). Avg. ISON PS Avg. ISON PS Ion Ratioa Ratio Ratio Ion Ratioa Ratio Ratio We initially find that our modeled spectral intensity does C6+ 0.318 0.318 0.310 Mg10+ 0.098 0.098 0.078 notaccuratelyfittheobservationaldataduetothehighvari- C5+ 0.210 0.240 0.240 Mg9+ 0.052 — — ance in SW conditions as a function of both time and solar N7+ 0.006 — — Si10+ 0.021 — — longitude, causing inaccurate modeled values for nl and vl. N6+ 0.058 — — S11+ 0.005 — — Since we lack any direct observations of these values at the N5+ 0.065 — — S10+ 0.016 — — comet, we therefore allow these parameters to vary within O8+ 0.070 0.100 0.040 S9+ 0.019 — — physical limits until the best agreement between our rela- O7+ 0.200 0.200 0.100 Fe13+ 0.002 — — tivemodeledintensityandtheobservationalintensityoverthe O6+ 0.730 0.700 0.860 Fe12+ 0.007 — — 0.3–1.0keVenergyrangeisfound. Ne9+ 0.004 0.020 0.004 Fe11+ 0.023 — — Ne8+ 0.084 0.068 0.084 Fe10+ 0.031 — — 2.2.3. CXModelComposition Note. — Model-calculated SW ion ratios for comets ISON and PanSTARRS(PS)incomparisontoaverageslowSWratios. Allratiosare The CX emission spectrum in our model is computed for normalizedwithrespecttothetotalSWoxygen. Allcalculated values are twoindependentmajorgroupsofheavySWions: foundtohaveanaverageuncertaintyof±15%.Valuesleftblankarebecause theobservationalspectrumdoesnotpossesstheresolutionrequiredtoaccu- ratelycalculatethoseratios,andsothemodeltreatsthemasconstants. 1. (Group A): key heavy ions for “3/4 keV” en- aBochsler(2007) ergy interval (C5+, C6+, N5+, N6+, N7+, O6+, O7+, O8+, Ne8+,and Ne9+). This group includes the CX emission spectra generated from collisions be- TABLE3 tween cometary neutrals (primarily H O) and H- SWCompositionComparisontoACE 2 like, He-like, and Li-like heavy SW ions. The CX Source C6+/C5+ O7+/O6+ O8+/O6+ spectra of these ions are reasonable constrained by ISON 1.35 0.28 0.14 lab and theoretical researches (Dijkkampetal. 1985; ACE 1.18+0.80 0.25+0.12 0.09+0.19 −0.48 −0.08 −0.06 Kelly 1987; Suraudetal. 1991; Wieseetal. 1996; PanSTARRS 1.29 0.12 0.05 Kharchenkoetal.2003;Koutroumpaetal.2006,2009; ACE 1.09+0.62 0.22+0.13 0.08+0.14 −0.39 −0.08 −0.05 Chutjianetal.2012). Note. —Acomparisonbetweenthemodel-calculatedSWionratiosand theaveragevaluesobservedbyACE.Allcalculatedvaluesarefoundtohave 2. (GroupB): heavier excited ions(Mgq+, Sq+, Siq+, and anaverageuncertainty of±20%andagreetotheobservational datawithin Feq+) that primarily contribute to the soft X-ray spec- uncertainty. We note that agreement between ACE and PanSTARRS is inconclusivegiventhesignificantdifferenceinheliosphericlatitudebetween tra(below0.4keV).Thecrosssectionsandrelativein- thetwo. tensity of different emission lines of the CX cascad- ing spectra for these ions are less known than for the ISON, which has 32 dof, we find χ2 = 1.1 over the 0.35– ionsfromGroupAbutarewellestimated(Hareletal. R 1.00keVrange.ForPanSTARRS,whichhas30dof,χ2 =1.2 1998;Simcicetal.2010). Theenergypositionofspec- R forthe0.35–1.00keVrange.Allspectraaretruncatedat0.35 trallinesarewelldefined(Kramidaetal.2014). keVduetothecarbonK-shellabsorptionedgedetectorcon- ThespectraofCX cascadingphotonsforGroupsAandB taminationpresentfrom ACIS atenergiesbelowthis thresh- arecomputedindependentlyandthenunifiedintoasynthetic old. These results provide a more complete and physically spectrum that represents the most probable emissions up to accuratepictureof the CX processin cometaryatmospheres 1 keV. Ion elemental and charge composition for all groups thanfoundinthepreviousgenerationofmodels. aretreatedasvariableparametersthatareinitiallysettoaver- Inadditiontoaccuratelymodelingthecometaryemissions, ageSWcompositionratios(Bochsler2007;Leprietal.2013). we compare our SW compositionsresults to contemporane- TheSWcompositionisthenvarieduntiltheχ2valueismini- ous composition ratios provided by ACE. A comparison of mized. DuetoChandra’slowsensitivitybelow0.35keVand ourmodelresultsto ACE,shownin Table 3, demonstratean the lack of accurate calibration near the carbon K-shell line agreementwithin uncertaintyforall observations. These re- at0.284keV,wefindthatvaryingseveralSWionstypesthat sults providean additional, and crucial, confirmationforthe predominantlyemitin thisregionproducesnochangeto χ2. physicalaccuracyofourmodelingtechnique. Italsoleadsus As a result, these SW ion types are left constant as average to considerusing our analysisand modelingof cometaryX- SWcompositionratios. TheinitialSWratiosandourresult- ray spectra in the future as a remote diagnostic tool for SW ingratiosforbothcometobservationsareshowninTable2. composition. 3. RESULTS 4. DISCUSSION Usingthemodeloutlinedintheprevioussection,wecreate 4.1. X-RayEmissionMorphology atheoreticalCXspectrumforthetwocometsobserved. Our WhenanalyzingcometaryX-rayimages,itisimportantto theoreticalspectra are comparedto the averagebackground- rememberthattheoverallemissionmorphologyisdetermined correctedobservationalspectra, and the results are shownin by SW interactionwith the cometaryatmosphereas the ma- Figure3. jorityoftheemittedintensityisduetoCX.Inthecollisionally Wealsocalculatethereducedχ2,alsoknownasχ2,bydi- thickcaseforanactivecomet,weexpectaparaboloidwiththe R vidingχ2 with the degreesof freedom(dof)foreach comet. cometatthefocuswherethemagnitudeofthesemimajoraxis Both comet observationsare binned with a minimum of six isdependentontheatmosphericdensity(Ha¨berlietal.1997; counts per spectral bin for proper gaussian statistics. For Wegmannetal.2004;Lisseetal.2005;Wegmann&Dennerl ChandraObservationsofC/2012S1andC/2011L4 5 0.001 0.002 PanSTARRS Observation ISON Observation -1V SWCX Model -1V SWCX Model ke ke0.0015 -1c -1c e e s s -2m 0.0005 -2m 0.001 c c s s n n oto oto0.0005 h h P P 0 0 10 10 2χ 0 2χ 0 ∆ ∆ -10 -10 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Energy [keV] Energy [keV] Fig.3.—ComparisonbetweenourCXmodelandtheaveragebackground-correctedobservationsforcometsPanSTARRSandISONandtheχ2residuals(∆χ2) ofthedata-modelcomparison. Eachobservationalspectrumhasbeengroupedwithaminimumofsixcountsperbinforproperstatistics. OurmodelvariesSW compositionratiosuntilabest-fitisfound.TheresultingSWcompositionratiosforeachmodelaredetailedinTable2. 2005). In comparison, the X-ray emission structure is de- and the hard X-ray ACIS observations, most notably seen termined by the distribution of gas in the coma for the col- on the October 31 and November 03 visits. These intensity lisionally thin case. For such a situation, we expect to see fluctuations correlate with increases in SW speed as docu- regions of enhanced X-ray emission in regions of higher mentedbyACE,wherethemaximumSWspeedwasrecorded cometaryparticledensity,suchasthosefoundalongjetstruc- November03.SuchanassociationbetweenSWandcometary tures(Lisseetal.2013). emissionintensityispredictedbyourCXmodelasSWspeed We present the ACIS observation images of comet fluctuationsindicatefluctuationsinSWionfreeze-intemper- PanSTARRSinFigure4. Allimageshavebeencorrectedfor atures(Bochsler 2007). Suchtemperaturechangeswill shift differencesin exposuretime and are normalizedto the same theSWchargestatedistribution,producingavaryingaverage linear scale. These images show a constant intensity in X- cometary emission energy based on our normalized photon rayemissionsin allthe observations,aswe expectgiventhe emissionyieldfunction P(j)(~ω ). Aswe see asimilar shift- constantcometarydust/gasemissionratesandSWconditions k,l j ing of the average cometary emission energy, we therefore observed at the time of our observations. We also find that assertthatCXemissionsarethedominantcometaryemission theoverallmorphologyishighlynon-uniform,andsowecon- mechanism in the soft X-ray region, a fact that will become cludethatPanSTARRSwascollisionallythinduringitsobser- importantinourdiscussioninSection4.4. vations.Thisislikelyaresultofthehighdustdensitypresent inthecometaryatmosphereasdustparticlesaresignificantly less efficient in CX X-ray production than molecular gas 4.2. PanSTARRSSpectralAnalysis (Djuric´etal.2005;Wolketal.2009;Lisseetal.2013). Prior to the observation of comet PanSTARRS, there was TheISONobservationswereuniqueasitwasthefirsttime much speculation if its high dust-to-gas ratio would sig- HRCobservationsofacometwereperformedinconjunction nificantly affect its X-ray CX emission intensity as it is with ACIS observations. The extracted ACIS and HRC im- morefavorabletoproduceAugerelectronsinsteadofX-rays agesforthe threeISON observationsare shownin Figure5. when undergoingCX with dust particles (Djuric´etal. 2005; Eachimageset,eitherfromACISorHRC,hasbeencorrected Wolketal. 2009;Lisseetal. 2013). We thereforemake sure for differencesin exposuretime and has been normalizedto tonoteanyX-rayspectralirregularitieswithinourresultsand, thesamelinearscale. TheACISimagesdemonstratetheex- ifso,theirpossibilityofbeingduetodustparticles. pected paraboloid morphology of a collisionally thick case, Utilizing our CX model, we are able to successfully char- whiletheHRCobservationsdepictamorenon-uniformemis- acterize PanSTARRS’ spectrum without making any adjust- sion typical of a collisionally thin case. Given that HRC is mentstoourCX scenario. We findauniquesolutionforthe moresensitivetosoftX-raysthanACIS,thisresultmayindi- emissionspectrumthatfitswelltotheobservationsupto1.0 cate that the soft X-ray emissions are due to CX emissions keV. Above 1.0 keV, the uncertainty in the observations be- from lighter SW ions with smaller cross sections, such as comes too great to distinguish between noise and emission He2+, than the SW ions that emit hard X-rays, such as C6+ peaks. Analysis of SW composition through the use of our and O8+. The reduction in cross section will allow deeper modelshowsalowerthanaverageamountofhighlycharged penetrationintothecometaryatmosphere,anditmaybesub- ions,suchasO8+andNe9+,withanincreaseintheirloweren- stantialenoughtogeneratedifferencein these imagesets. It ergyvariants,likeO6+ andNe8+. Thisresultagreeswithour isalsopossiblethatthe softX-rayemissionsfromISON are previous assessment that PanSTARRS was observed at fast, fromadifferentemissionmechanismthatwouldnotproduce polar SW. Beyond this irregularity, PanSTARRS’ spectrum thesamemorphology,suchasscatteringorfluorescence. possesses no additional traits that would classify it different TheISON imagesetsalso demonstratesignificantfluctua- fromanyothercometX-rayspectrum. tioninthecometaryemissionintensityovertimeanda“see- Although we cannot infer from our spectral analysis how saw” in intensity between the soft X-ray HRC observations PanSTARRS’ largedustquantitiesmay have impactedother emissions mechanisms present within the cometary spectra, 6 Sniosetal. ACIS 100,000 km ACIS 100,000 km ACIS 100,000 km 0.3 Apr 17 Apr 21 Apr 23 0.25 0.2 0.15 0.1 Fig.4.—Chandra/ACIS-SobservationsofcometPanSTARRS.Theimagesarebinnedtoincludeall0.3–1.1keVphotonevents,exposurecorrected,onthe samelinearscale,andsmoothedwitha5×5pixelGaussianfilter.OurresultsshowfluctuationsinX-rayemissionintensity,andtheoverallmorphologyishighly non-uniform. ACIS 100,000 km ACIS 100,000 km ACIS 100,000 km 0.8 Oct 31 Nov 03 Nov 06 0.7 0.6 0.5 0.4 0.3 0.2 0.1 HRC 100,000 km HRC 100,000 km HRC 100,000 km 2.1 Oct 31 Nov 03 Nov 06 2.0 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 Fig.5.—Chandra/ACIS-SandHRC-IobservationsofcometISON.Eachsetofimages(eitherACISorHRC)areexposurecorrected,shownonalinearscale, andsmoothedwitha5×5pixelGaussianfilter. ACISimagesarealsobinnedtoincludeall0.3–2.0keVphotonevents. Theimagesshowa“see-saw”effect betweenthesoftX-rayHRCobservationsandhardX-rayACISobservationswhereanincreaseofintensityinHRCcorrelatestoadecreaseinACIS,andvice versa.ThisresultalsocorrelatestofluctuationsinSWspeedbetweenOctober31andNovember03asseenviaACE. our results indicate that it had little to no observable impact byGOES,werealsoobservedduringISON’sChandravisits. onthecomet’sCXX-rayemissions. Assuch,anydifferences ThesehighlyvolatileSWconditionsmaysignificantlyimpact presentaremorelikelyattributedtotheSWfluxdensityand ISON’semissionspectra. ionizationstateatthetimeofobservation. UsingtheACISobservationsandapplyingthesamemethod asdoneforPanSTARRS,weareabletomodelISON’sspec- 4.3. ISONSpectralAnalysis trumasCXbelow1keVandextractSWcompositionratios. Despite being one of the brightestcomets in recent years, ISON’s ratios confirm the above average SW speed with an the Chandra observations we analyzed were taken slightly overabundanceofhighlychargedSWions,likeO8+andNe9+, priortoISON’sdrasticincreaseingasproductionratestarting that produces a distinct plateau in the spectrum from 0.75– on 2013 November 13. Fluctuations in SW speeds, as con- 1.00keV.TheO7+ ratioistwicethatseenfromPanSTARRS, firmed by ACE, and severalM-class solar flares, as reported bestvisualizedviatheemissionpeakat0.6keV. ChandraObservationsofC/2012S1andC/2011L4 7 0.7 0.284keV,weplottheresultingemissionspectraforeachof Oct 31 the three ISON visits in Figure 6. Our plots show a soft X- 0.6 NNoovv 0036 rayregionat0.2keVthatthedetectorissensitivetoafterour corrections,evenshowingfluctuationsthatagreewiththesoft 0.5 X-rayemissionfluctuationsdetectedbyHRC(seeFigure5). -1V TheoverallshapeofthisfeatureislikelyduetotheACISef- e -1c k0.4 fective area function abruptly decaying toward zero at 0.18 e keV and is not due to any specific emission line. Also, the s nts 0.3 fluctuations between the visits exceed the spectral intensity ou uncertainty in this region, which is 0.08 counts s−1 keV−1, C 0.2 andsowebelievethesefeaturestobephysical. Based on our HRC results from Section 4.1, we assume 0.1 CXemissionstobethemostlikelycauseofthisfeature. We thereforeextendourCXmodeldowntothissoftX-rayregion 0 andplottheresultswiththeobservationaldata.Seethedotted 0 0.2 0.4 0.6 0.8 1 Energy [keV] linesinFigure7forourpredictedCXmodelforeachobser- Fig.6.—ACISspectralintensityforeachobservationofcometISON.All vation.CalculationofauniquesolutionofSWratiosrequired detector contamination fromthecarbonK-shellat0.284keV hasbeenre- to produce such intensities is not possible due to the abun- movedfromthespectra.Theobservedspectralfeatureat0.2keVhasthethe danceofover200uniquelinesfromover15differentSWion samefluctuationinintensitythatisobservedfromtheHRCobservationsand typesthatfallwithinACIS’resolutionofthissoftX-rayfea- exceedstheaveragespectralintensityuncertainty inthisregion, andsowe concludethatthisfeatureisphysical.Possibleoriginsofthisspectralfeature ture. We thereforechoosetoleaveourmodelataverageSW arediscussedinSection4.4. abundancesinthisregion. Wenotethatfixingtheseparame- tersproducesnodifferenceinthespectralfitoverthe0.3–1.0 In addition to these results, ISON exhibits some possible keVenergyrange. peak structures in its emission spectrum at energies above OurresultsshowthatouraverageCXmodelisnotcapable 1 keV: one peak at 1.35 keV and another at 1.85 keV, as ofproducingthenecessaryintensitytomatchtheobservations seen in Figure 1. Such peaks have been previously seen foreithertheISONemissionsorthePanSTARRSemissions, in Chandra’s observations of Comet 153P (Ikeya-Zhang) whichare notshown. Furthermore,theSW abundancesthat (Ewingetal.2013),anothercometviewedduringvolatileSW ourmodelwoulddemandtomatchthesefeaturesinintensity conditions.Ourmodelispresentlyunabletoaccuratelycalcu- far exceed their physical boundaries, with most abundances latetheoreticalCX emissionsinthisenergyrangeduetothe requiring an increase by an order of magnitude. Although lackofinformationaboutthepresenceofsuchhighlycharged such exotic SW compositions are not impossible given its ions in the solar wind plasma, but we may comment on the constantlyfluctuatingnature,theconsistentpresenceofthese possibleemissioncandidates. soft X-ray features during both fast and slow SW indicate Comparison to atomic emission line tables from NIST theseshouldbegeneratedunderaverageSWconditions. indicate that the most probable ions for each emission is AlthoughourcurrentCXmodeldoesnotagreewiththesoft Mg XI (1s21S–1s2p1,3P) for 1.35 keV and either Si XIII X-rayintensitiesdetected,weonlyconsiderasingleelectron (1s21S–1s2p1,3P) or Mg XII (1s2S–4p2P) for 1.85 keV captureeventperincomingSWion.Sequentialcaptureevents (Kramidaetal. 2014). However, it is unclear if these peaks mayoccurforanionifthecometaryatmosphereiscollision- couldbearesultfromCXastheseexoticcandidateshavenot allythick,increasingtheamountofsoftX-raysemittedfrom beendetectedviainsituobservationsofSWioncomposition the system as the ion charge state decreases (O8+→ O7+→ (vonSteigeretal.2000;Leprietal.2013). Furthermore,the- O6+→ ... ). We therefore modify our CX model to include oreticalmodelsdescribingthechargeabundanceofheavySW these sequential capture events per ion as it may solve our ionspredictanextremelylowprobabilityoffindingtheseions softX-rayintensitydeficit. because of the inability to reach such high freezing-in tem- Forouranalysis,wecalculateanupperlimitontheincrease peraturesinregularSWandcoronalmassejections(Bochsler tosoftX-rayCXemissionsfromsequentialcaptureeventsby 2007). Ontheotherhand,thesespectrallinesareclearlypre- assumingallSWionsareneutralizedthroughinteractionwith sented in the spectra of the solar X-ray flaresas well as in a thecometaryatmosphere. OurresultsarepresentedinFigure regularX-rayemission fromthe Sun (McKenzieetal. 1985; 7, andtheyshow thatthe additionalCX eventsarenotsuffi- Dereetal.1997;Landietal.2013).Itthereforemaybepossi- cienttoequaltheobservedsoftX-rayintensities.Wefindthat blethesepeaksareduetoadifferentmechanismwhoseemis- theupperlimitof CX emissionsonlyincreasesthe totalsoft sions are increased by solar flare activity, such as scattering X-rayintensity∼50%,whichisnotenoughtoaccountforthe of solar X-rays (Krasnopolsky 1997; Sniosetal. 2014). At factorsofthreetosixbetweenthemodelandtheobservations. present, we cannot conclude what is the primary source of Furthermore,westressthattheactualrateofsequentialcap- these exoticemissionsdetectedfromISON asfurtheranaly- ture events present in these cometary systems is lower than sisofISON’sspectrumwitharevisedCXmodel,andpossibly thisupperlimit,sotheactualemissionintensitieswillreside alsoascatteringemissionmodel,isrequired. between our model and the upper limit. We therefore find itunlikelythatsequentialCX eventscouldaccountforthese 4.4. PotentialSoftX-RayEmissionsfromACIS softX-rayfeatures. While examining the ACIS spectra from ISON, we found AsweareconfidentthatbothourCXmodel’sresultingSW a peak-like feature located at 0.2 keV. This feature was also compositionandphotonyieldemissionratesareaccurate,we foundinthePanSTARRSobservations,butwithalowerrela- therefore consider two possible explanations for the soft X- tivespectralintensity.Afterremovingtheportionofthespec- raydiscrepancy: tra caused by the carbon K-shell detector contamination at 8 Sniosetal. 0.7 0.7 Oct 31 Obs Nov 03 Obs CX Model CX Model 0.6 0.6 Model Upper Model Upper -1V0.5 Limit -1V0.5 Limit e e k k 1 0.4 1 0.4 -c -c e e s s0.3 s s0.3 nt nt u u o0.2 o0.2 C C 0.1 0.1 0 0 40 40 2χ 0 2χ 0 ∆ ∆ -40 -40 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Energy [keV] Energy [keV] 0.7 Nov 06 Obs CX Model 0.6 Model Upper -1V0.5 Limit e k 1 0.4 -c e s s0.3 nt ou0.2 C 0.1 0 40 2χ 0 ∆ -40 0 0.2 0.4 0.6 0.8 1 Energy [keV] Fig.7.—ACISspectralintensity foreachobservation ofcometISON(solidlines)anditsrespective modeledCXusingaverageSWcompositions (dotted coloredlines). AlldetectorcontaminationfromthecarbonK-shellat0.284keVhasbeenremovedfromthespectra. Despitetheexcellentagreementabove0.4 keV,ourmodelfailstomatchtheshapeandintensityofthesoftX-rayspectralfeature.WealsocalculateanupperlimittosoftX-rayCXemissionsbyaccounting forsequentialCXeventsandassumingthatallSWionsareneutralizedthroughinteractionwiththecometaryatmosphere(dottedblacklines),andourresults showtheseadditionstobeinsufficienttoequaltheobservedsoftX-rayintensities. WethereforebelievethesesoftX-rayfeaturestobeCXfromanunaccounted SWion(suchasHe2+),detectorcontamination,oracombinationoftheseoptions. 1. Since the CX modeldoes notmatch the observational cometaryorigins,orifthisfeaturehasmanifesteditself intensities, it is possible we lack the SW ion type re- overtime,indicatingadetectorissue. quired to produce this feature. He2+ CX emissions, currently not included in our analysis, would be de- The required analysis for each of these possibilities is be- tectable in this soft X-ray region due to the low res- yondthe scope of thisarticle, butwe believethatany future olution of ACIS, and its high abundance may provide workonthesesoftX-rayfeaturesfromACISshouldprovide the required order-of-magnitude increase in intensity athoroughanalysisofeachpossibleexplanationtodetermine (Kharchenko&Dalgarno 2001; Bodewitsetal. 2004). thecauseoftheseuniquefindings. Such SW ions would also have deeper penetration in thecometaryatmosphere,whichmightalsoexplainthe 5. CONCLUSIONS collisionally thin appearance of the HRC morphology Insummary,wehaveusedChandrato studytwo verydif- discussedinSection4.1.Futureiterationsofourmodel ferentOortCloudcomets,thegas-richC/2012S1ISONand should include this ion and compare the modified re- thedust-richC/2011L4PanSTARRS.Bothcometswereob- sultstothesoftX-rayemissionsfromACIS. served within 1 AU heliocentric distance of the Sun, when they were active. The observed X-ray morphologies were 2. ThesoftX-rayfeaturemaybearesultofpreviouslyun- dramaticallydifferent,however,withISONdisplayinganex- documenteddetectorcontaminationordegradationthat tended,well-developedX-raycomaandPanSTARRSproduc- sharplydecaysbelow0.2keV,producingapeakinob- ing an unformed X-ray haze. The two comets also experi- served spectrum. Examination of similar comets ob- encedmarkedlydifferentSWconditions,withISONimpact- servedatdifferentstagesofACIS’lifetimewouldshow inganexcitedwind,whilePanSTARRStraveledthroughfast ifsuchasoftX-rayfeatureisalwayspresent,indicating SW. ChandraObservationsofC/2012S1andC/2011L4 9 WedevelopedanupdatedCXemissionmodelthatincludes andMg10+.WithfurtherdevelopmentofCXX-raymodeling, large amountsof ion spectral lines induced in CX collisions suchanapplicationwouldbepossibleforanyCXemissions, and simplifies input variables while improving the physical not just those from comets. Our model also simplifies the accuracyin comparisonto previousmodels. Ourmodelwas variable inputs and provides an additional information on used to analyze Chandra observations of comets ISON and SW composition. We therefore intend to use such a model PanSTARRS, andwe foundstrongemissionsinducedin CX for all future CX analyses of cometary and planetary X-ray collisionsofSWionsnormallypresentwithincometaryemis- emissionsaswellasforinvestigationsofCXX-raysinduced sions(C5+, C6+, N5+, N6+, N7+, O6+, O7+, O8+, Ne8+, Ne9+) ininteractionbetweentheSWplasmaandinterstellargas. from both comets. Analysis of ISON spectra shows higher Inadditiontoourmodelingresults,wefoundthepossibil- concentrationsofO8+ andNe9+thanPanSTARRS,indicating ityofsoftX-rayemissionsaround0.2keVdetectedfromboth higherSWionfreeze-intemperatureduringitsobservations. comets ISON and PanSTARRS via ACIS. These soft X-ray AnalysisofISON’sspectrumalsoshowshigh-energyspec- features fluctuate similarly to those observed from the HRC tral features above 1 keV. To clarify the physical origin of observationsandexceedtheaveragespectralintensityuncer- these cometary“hard”X-rays, we intend in the futureto in- tainty,leadingustobelievethesefeaturestobecometaryCX cludeCXemissionsfromexoticSWions,suchasMg11+and in origin. We extended our CX model to this soft X-ray re- Si13+,whichwillextendourmodelbeyond1keV.Analysisof gion to compare, only to find our results lower in intensity high-energyspectralfeatureswillallowustopredictthetotal thanthe observationsbyan orderofmagnitude. We also re- ratioofexoticSW ionsandtodiscusswhetherthosequanti- vise ourmodelto includesequentialCX captureeventsas it tiescouldbeobservableusingcurrenttools. Wewillalsoin- will incease soft X-ray intensities, but we find that even the cludeemissioncontributionsfromscatteringandfluorescence inclusionofmorecaptureeventsisnotsufficienttomatchthe ofenergeticsolarX-rays,especiallyduringsolarX-rayflares observed intensities. Based on our confidence in the model events, and X-ray emissions of non-thermal energetic elec- from its previous results, we believe this discrepancy to be tronsduetoelectronicimpactorbremsstrahlungmechanisms. aresultofeitheralackofSWiontypesthatproducesignifi- Accurateinvestigationsofthespectralmorphology,whichare cantemissionsinthesoftX-rayregion(suchasHe+),detector differentforeachmechanism,willalsobeperformed.Sucha contaminationordegradation,oracombinationofthesepos- discussionwouldestablishahierarchyofpotentialcontribut- sibilities. Investigations of these soft X-ray features should ingmechanismsincometaryX-rayspectraabove1keVand carefully explore each explanation as confirmation of these provideinsightontheoriginoftheobservedhighenergyspec- featuresasphysicalemissionswouldopennewopportunities tralfeatures. inunderstandingcometaryemissionprocessesviaChandra. Beyond successfully analyzing comets ISON and PanSTARRS, we also demonstrated our model’s poten- WewouldliketoacknowledgeNationalOceanicandAtmo- tialuseasaSW ioncompositionanalyzer. Ourcomposition spheric Administration for their Advanced Composition Ex- resultsagreewellwithotherSWcompositiontoolsavailable, plorer data. The work of B.S. and V.K. on this project has such as ACE, while also calculating unique composition beensupportedbyNASAgrantNNX10AB88G.Theworkof ratiosnotavailablethroughtheseothertools,likeNe8+,Ne9+, C.L.hasbeensupportedbyChandragrantGO4-15001A. REFERENCES Bochsler,P.2007,A&ARev.,14,1 Koutroumpa,D.,Lallement,R.,Kharchenko,V.,etal.2006,A&A,460,289 Bodewits,D.,Christian,D.J.,Torney,M.,etal.2007,A&A,469,1183 Koutroumpa,D.,Lallement,R.,Raymond,J.C.,&Kharchenko,V.2009, Bodewits,D.,McCullough,R.W.,Tielens,A.G.G.M.,&Hoekstra,R. ApJ,696,1517 2004,PhysicaScripta,70,C17 Kramida,A.,Ralchenko,Y.,Reader,J.,&NISTASDTeam.2014,NIST Cann,N.M.&Thakkar,A.J.1992,Phys.Rev.A,46,5397 AtomicSpectraDatabase(ver.5.2) Chutjian,A.,Simcic,J.,Madzunkov,S.M.,etal.2012,JPCS,388,012042 Krasnopolsky,V.A.1997,Icarus,128,368 Combi,M.R.,Bertaux,J.-L.,Que´merais,E.,etal.2014a,ApJ,147,126 Krasnopolsky,V.A.,Christian,D.J.,Kharchenko,V.,etal.2002,Icarus, Combi,M.R.,Fougere,N.,Ma¨kinen,J.T.T.,etal.2014b,ApJ,788,L7 160,437 Cravens,T.E.1997,Geophys.Res.Lett.,24,105 Landi,E.,Young,P.R.,Dere,K.P.,Zanna,G.D.,&Mason,H.E.2013, Dennerl,K.2010,SpaceSci.Rev.,157,57 ApJ,763,86 Dere,K.P.,Landi,E.,Mason,H.E.,Fossi,B.C.M.,&Young,P.R.1997, Lepri,S.T.,Landi,E.,&Zurbuchen,T.H.2013,ApJ,768,94 A&AS,125,149 Lisse,C.,Christian,D.,Wolk,S.,etal.2013,Icarus,222,752 Dijkkamp,D.,Gordeev,Y.S.,Brazuk,A.,Drentje,A.G.,&deHeer,F.J. Lisse,C.M.,Christian,D.J.,Dennerl,K.,etal.2005,ApJ,635,1329 1985,J.Phys.B,18,737 Lisse,C.M.,Cravens,T.E.,&Dennerl,K.2004,inCometsII,ed. Djuric´,N.,Lozano,J.A.,Smith,S.J.,&Chutjian,A.2005,ApJ,635,718 M.Festou,H.U.Keller,&H.A.Weaver(Tucson:TheUniversityof Ewing,I.,Christian,D.J.,Bodewits,D.,etal.2013,ApJ,763,66 ArizonaPress),631 Fruscione,A.,McDowell,J.C.,Allen,G.E.,etal.2006,Proc.SPIE,6270, Lisse,C.M.,Dennerl,K.,Englhauser,J.,etal.1996,Science,274,205 62701V McKenzie,D.L.,Landecker,P.B.,Feldman,U.,&Doschek,G.A.1985, Geiss,J.,Gloeckler,G.,&Steiger,R.V.1995,SpaceSci.Rev.,72,49 ApJ,289,849 Ha¨berli,R.M.,Gombosi,T.I.,Zeeuw,D.L.D.,Combi,M.R.,&Powell, Meech,K.J.,Yang,B.,Kleyna,J.,etal.2013,ApJL,776,L20 K.G.1997,Science,276,939 Raouafi,N.-E.,Lisse,C.M.,Stenborg,G.,Jones,G.H.,&Schmidt,C.A. Harel,C.,Joiun,H.,&Pons,B.1998,At.Data.Nucl.DataTables,68,279 2015,J.Geophys.Res.Space,120,5329 Janev,R.K.1995,AtomicandMolecularProcessesinFusionEdgePlasmas Schwadron,N.A.&Cravens,T.E.2000,ApJ,544,558 (NewYork:Plenum) Simcic,J.,Schultz,D.R.,Mawhorter,R.J.,etal.2010,Phys.Rev.A,81, Janev,R.K.&Winter,H.1985,Phys.Rep.,117,265 062715 Johnson,W.R.&Soff,G.1985,At.DataNucl.DataTab.,33,405 Snios,B.,Lewkow,N.,&Kharchenko,V.2014,A&A,568,A80 Kelly,R.L.1987,J.Phys.Chem.Ref.Data,16,649 Suraud,M.G.,Hoekstra,R.,deHeer,F.J.,Bonnet,J.J.,&Morgenstern,R. Kharchenko,V.&Dalgarno,A.2000,J.Geophys.Res,105,18351 1991,J.Phys.B,24,2543 Kharchenko,V.&Dalgarno,A.2001,ApJ,554,L99 vonSteiger,R.,Schwadron,N.A.,Fisk,L.A.,etal.2000,J.Geophys.Res, Kharchenko,V.,Rigazio,M.,Dalgarno,A.,&Krasnopolsky,V.A.2003, 105,27217 ApJ,585,L73 Wegmann,R.&Dennerl,K.2005,A&A,430,L33 10 Sniosetal. Wegmann,R.,Dennerl,K.,&Lisse,C.M.2004,A&A,428,647 Wolk,S.J.,Lisse,C.M.,Bodewits,D.,Christian,D.J.,&Dennerl,K.2009, Wiese,W.L.,Fuhr,J.R.,&Deters,T.M.1996,AtomicTransition ApJ,694,1293 ProbabilitiesofCarbon,Nitrogen,andOxygen:ACriticalData Yang,B.,Keane,J.,Meech,K.,Owen,T.,&Wainscoat,R.2014,ApJL, Compilation(Washington,DC:Am.Chem.Soc.) 784,L23

See more

The list of books you might like

Upgrade Premium
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.
DMCA & Copyright: Dear all, most of the website is community built, users are uploading hundred of books everyday, which makes really hard for us to identify copyrighted material, please contact us if you want any material removed.
Copyright @ PDFdrive.to, 2022 | We love our users