ebook img

Terrestrial Planets Formation under Migration: the Systems near 4:2:1 Mean Motion Resonance PDF

8.1 MB·
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Terrestrial Planets Formation under Migration: the Systems near 4:2:1 Mean Motion Resonance

Mon.Not.R.Astron.Soc.000,1–??(2015) Printed12January2017 (MNLATEXstylefilev2.2) Terrestrial Planets Formation under Migration: the Systems near 4:2:1 Mean Motion Resonance Zhao Sun1,2, Jianghui Ji1⋆, Su Wang1, Sheng Jin1 7 1 1CASKeyLaboratoryofPlanetarySciences,PurpleMountainObservatory,ChineseAcademyofSciences,Nanjing210008,China 0 2UniversityofChineseAcademyofSciences,Beijing100049,China 2 n a Accepted.Received;inoriginalform J 1 1 ABSTRACT Inthiswork,weextensivelyinvestigatetheformationofnear4:2:1meanmotionresonances ] (MMRs)configurationbyperformingtwosetsofN-bodysimulations.Wemodeltheeccen- P tricitydamping,gasdrag,typeI and typeIIplanetarymigrationofplanetesimals,planetary E embryosand giant planetsin the first sets. For the simulations of giantplanets with type II . migration,the massiveterrestrialplanets,with amass uptoseveralEarthmasses, arelikely h p producedinthesystems.Wefurthershowthatbyshepherdingand/orscatteringmechanisms - throughJovianplanet’stypeIImigration,theterrestrialplanetsandgiantplanetsinthesys- o tems can be evolved into a near 4:2:1 MMRs. Moreover, the models are applicable to the r t formationofKepler-238and302systems.Inthesecondset, westudythe4:2:1MMRsfor- s mation in the terrestrial planetary systems, where the planets undergotype I migration and a eccentricitydamping.ByconsideringtypeImigration,∼17.1%ofthesimulationsindicate [ thatterrestrialplanetsareevolvedinto4:2:1MMRs.However,thisprobabilityshoulddepend 1 ontheinitialconditionsofplanets.Hence,weconcludethatbothtypeIandtypeIImigration v canplayacrucialroleinclose-interrestrialplanetformation. 4 9 Keywords: planetarysystems–methods:numerical–planetsandsatellites:formation. 9 2 0 . 1 1 INTRODUCTION 2014; Zhangetal. 2014),such as the4:2:1 MMRs. For example, 0 Kepler-238(Roweetal.2014)andKepler-302(Roweetal.2014) 7 The number of exoplanets has substantially increased in re- systems harbor close-in giant planets in near mean motion reso- 1 cent years. At the time of writing, over 3500 exoplanets (see : nance (MMR) configuration with other planets. Therefore, these v www.exoplanet.eu) are discovered, mostly through radial veloc- exciting observations motivate us to explore and understand the i ityand transiting surveys. In particular, asof Nov. 10, 2015, Ke- X formationandevolutionoftheplanetarysystems,especiallytoin- plermissionreleasedapproximately1030confirmedplanets,along vestigatetheshort-periodterrestrialplanetformationandthecon- r withover4696transitingplanetarycandidates(Fressinetal.2013). a figurationformationofthe4:2:1MMRs. Thismayimplythattheplanetsappeartobeverycommon,orbiting otherstarsbeyondourownsolarsystem. Severalmodelshavebeenproposed toexplaintheformation Theobservations show thatclose-in(short-period) terrestrial of close-in terrestrial planets (Raymondetal. 2008), and the for- planetsaretypicallycommonintheexoplanetarysystems.Roughly mationscenariosincludeinsituaccretion,orbitalmigrationarising speaking, one third to half of solar-type (FGK) stars can host fromplanetaryembryos(typeImigration(Goldreich&Tremaine at least one planet with a mass less than 10 M⊕ and an or- 1980))andgasgiantplanets(typeIImigration(Lin&Papaloizou bital period ranging from 50 to 100 days (Howardetal. 2010; 1986)), dynamical instabilities in systems of multiple gas gi- Mayoretal.2011).Thefrequencyofshort-periodterrestrialplan- ant planets, tidal circularization of eccentric terrestrial planets, ets is at least as high around M stars as around FGK stars, or and photo-evaporation of close-in giant planets. For the systems evenhigher(Howardetal.2012;Bonfilsetal.2013;Fressinetal. that are composed of short-period terrestrial planets, their pro- 2013). These terrestrial planets are usually found in multi-planet toplanetary disks are suggestive of very massive (Raymondetal. systems on compact but non-resonant orbits (Udryetal. 2007; 2008; Hansen&Murray 2012, 2013; Chiang&Laughlin 2013; Lovisetal. 2011; Lissaueretal. 2011). However, from the statis- Raymond&Cossou 2014), thus the planets can accrete in situ tical results of Kepler release data, a great many of planet pairs from a large number of planetesimals and planetary embryos in inthe systems areobserved tobe in near-resonant configurations thedisk. Theformationandfinalconfiguration of close-interres- (Zhangetal. 2010; Leeetal. 2013; Mart´ıetal. 2013; Wang&Ji trial planets are closely related to the strength of type-I migra- tion,andthesuppressionoftypeImigrationisrequiredincaseof in-situsuper-Earthsformation(Ogihara&Ida2009;Ogiharaetal. ⋆ E-mail:[email protected] 2015).Theobservedsystemsofhotsuper-Earthsmostlycancon- (cid:13)c 2015RAS 2 Z. Sun et al. tain 20 - 40 M⊕ in mass within a fraction of an AU of the host nally, we summarize the outcomes and give a brief discussion in star(Batalhaetal.2013).Star-planettidalinteractionsmayplaya Section4. roleincircularizingplanetsinhighly-eccentricorbitsandtherefore reducetheirsemi-majoraxisintheevolution(Ford&Rasio2006; Fabrycky&Tremaine2007;Beauge´&Nesvorny´ 2012;Dong&Ji 2013).However,Raymondetal.(2008)suggestedthattidaleffect canproducehotsuper-Earths,butonlyforthoserelativelymassive 2 MODELS planets (& 5M⊕) with very small perihelion distances (. 0.025 Following the empirical minimum-mass solar nebula (MMSN, AU),andeventhentheinwardmovementinorbitaldistanceisonly Hayashi 1981) model, the surface density of solid disk at stellar 0.1-0.15AUatmost,thereforetidesarenotstrongenoughtomove manyoftheKeplerplanetstothenowadaysobservedseparations, distanceaisdescribedas andadditionaldissipativeprocessesareatplay(Leeetal.2013). Alternatively, terrestrial planet may form inside a migrat- Σd =10fdγice a −3/2 gcm−2, (1) ing giant planet (Raymondetal. 2006). Gas-giants can shepherd 1AU planetesimals and embryos interior to their orbits as they mi- (cid:16) (cid:17) grate inward, which can further collide and merge into Earth- wherefdandγicearethesolidandthevolatileenhancementfactor, likeplanets (Zhouetal. 2005). Mandelletal. (2007) showed ma- respectively. γice is 4.2 exterior to the snow line or 1 interior to snowline.Theprofileofgasdensityisgivenas terials that have been shepherded interior to the migrating gi- ant planet by moving MMRs can accrete into close-in terres- trial planets. In addition, Izidoroetal. (2014) indicated that fast- a −3/2 migratingsuper-Earthsonlyweaklyperturbtheplanetesimalsdisk Σg =2.4×103fgfdep 1AU gcm−2, (2) and planetary embryos, whereas slowly migrating super-Earths (cid:16) (cid:17) shepherd rocky material interior totheir orbits inresonances and wherefgandfdeparethegasenhancementfactorandgasdepletion push toward the star. Moreover, the orbital migration and planet- factor, respectively. fdep=exp(−t/τdep), where t isthe timeand planet scattering play a vital role in producing short-period ter- thetimescaleτdep isaboutfewmillionyears(Haischetal.2001). restrial planets (Brunini&Cionco 2005; Terquem&Papaloizou Herein we adopt τdep = 106 yr. The inner edge of the gas disk locatesat0.1AU. 2007; Raymondetal. 2008; Cossouetal. 2014). According to As they are considered small bodies in our simulations, the core-accretion model, the planetary embryos in the terrestrial planet formation region, and the solid cores of giant planets, are planetesimalswillbeaffectedbytheaerodynamicaldragofthegas around them (Adachietal. 1976; Tanaka&Ida 1999). The force bothformedwithin∼ 1 Myrfromkilometer-sizedplanetesimals (Safronov1969;Wetherill1980).Subsequently, themassivesolid thataplanetesimalwithamassmwillsufferfromaerodynamical corescanfurtheraccretegasfromtheprotoplanetarydisktoform dragiswrittenas gas-giants(Kokubo&Ida2002;Ida&Lin2004)atMyrtimescale, beforethediskdisperses (Haischetal. 2001).Inthelatestageof 1 planetformation,afterthegasdiskclears,thegiantplanetsceases Faero =− CDπS2ρg|U|U, (3) 2 migrating, the numerous planetesimals and planetary embryos in thediskbecometurbulentduetodynamical stirringbygas-giants whereU=Vk−Vgistherelativevelocitybetweentheplanetesi- overhundredsofMyrsorevenlonger.Consequently,frequentor- mal’sKeplerianmotionVkandthegasmotionVg.CDisthedrag bital crossings and giant impacts are likely to occur, which may coefficient. If an object has a large Reynold number, CD can be eventually yield short-period Earth-like planets (Chambers 2001; takenas0.5. S andρg aretheplanetesimal’sradiusandgasden- Raymondetal.2004;Zhang&Ji2009;Jietal.2011). sity,respectively. Foraplanetaryembryoembeddinginthegaseousdisk,their Raymondetal. (2006) investigated terrestrial planets for- mutualinteractionswillleadtoaneccentricitydampingoftheem- mation under type II migration and in the model they in- bryowithatimescaleτ writtenas(Cresswell&Nelson2006) cluded a giant planet and gas drag. Based on Raymond et al’s damp model,Mandelletal.(2007)furtherconsideredanadditionalnon- tmheig3ra:2tinagndgi2a:n1tMplManRets.cInonofiugrueraatriloienrfsotrumdya,tiowne(hWavaengin&veJsiti2g0a1te4d) τdamp1= ee˙ = 0Q.7e8 Mm∗ ΣMa∗2 hr 4Ω−1 inthesystemobservedbyKeplermission,andsimplyconsidered (cid:16) (cid:17) (cid:18) (cid:19)(cid:18) g (cid:19)(cid:18) (cid:19) 1 r 3 the planets with their nominal masses that are over a few Earth × 1+ e , (4) 4 h masses without any growth (Wangetal. 2012; Wang&Ji 2014). (cid:20) (cid:16) (cid:17) (cid:21) Inthepresentwork,weaimtoexploreterrestrialplanetformation whereh, r, Ω,andearediskscaleheight,distancefromcentral especiallyinthesystemwith4:2:1configurationunder migration star,Keplerianangularvelocity,andtheeccentricityoftheembryo, by performing N-body simulations. In short, we have carried out respectively.Q = 0.1isanormalizedfactor inassociationwith e twokindsofsimulations,whereinthefirstmodel weincludethe hydrodynamicalsimulationresults.Moreover,theangularmomen- presenceofgiantplanetandasacomparison,inthesecondmodel tumexchangebetweentheembryosandthegasdiskwillresultin we consider the configuration formation in the system with only orbitalmigrationofplanets.Whentheplanetsarelessmassive,they terrestrialplanets. will undergo type I migration, whereas they grow large enough, The paper is structured as follows. In Section 2, we de- theywillexperiencetypeIImigration(Ida&Lin2004). scribe the adopted models of planetary formation, including the ThetimescaleoftypeImigrationcanbeassessedusinglinear diskmodelandtheplanetarymigrationscenarios.InSection3we model and the net loss on embryo (Goldreich&Tremaine 1979; presentnumericalsetupandmajorresultsofourinvestigation.Fi- Ward1997;Tanakaetal.2002)isexpressedas (cid:13)c 2015RAS,MNRAS000,1–?? TerrestrialPlanetsFormationunderMigration 3 where r and V represent the position and velocity vectors of i i planetm andallvectorsaregiveninstellar-centriccoordinates. τmigI= |aa˙| = (2.7+11.1β) Mm∗ ΣMa∗2 i (cid:18) (cid:19)(cid:18) g (cid:19) h 2 1+( er )5 × 1.3h Ω−1, (5) (cid:18)a(cid:19) "1−(1e.1rh)4# wheree, r, h,andΩarethesamemeaningasgiveninEquation 4. Using the gas density profile in Equation (2), we can achieve 3 NUMERICALSIMULATIONSANDRESULTS β=−dlnΣ /dlna=1.5. g 3.1 Initialconditions Recent workshowed that type Imigration couldbe directed eitherinwardoroutwarddependingondifferentplanetaryandgas Inoursimulations,allbodiesareassumedtobeinitiallyincopla- diskproperties(Cossouetal.2013),butingeneral,outwardtypeI narandnear-circulartrajectoryorbitingthecentralstar,wherethe migrationcansimplyatworkwhenthemassoftheplanetscanbe argumentofpericenter,meananomaly,andlongitudeofascending severalM⊕(Cossouetal.2014;Legaetal.2014),whicharemuch nodearerandomlydistributedbetween0◦to360◦,respectively.In largerthantheembryosandplanetesimalsinourmodel,sowedo total, fiveruns were carried out for theinvestigation of planetary notconsideroutwardtypeImigrationinthiswork. formation. The planet will experience type II migration when a planet InsimulationS1-S5,theplanetarysystemiscomposedofthe grows to a massive one (M > Mcrit) in the viscous disk hoststar,oneortwogiantplanets,and2000equal-massplanetes- (Lin&Papaloizou1993).ThetimescaleoftypeIImigrationisde- imals. As shown in Table 1 and Equation (8), each planetesimal scribedas(Ida&Lin2008) performsaerodynamicaldrag,eccentricitydamping,TypeImigra- tionandthegravitationalperturbationsfromJupiterorSaturnand τmigII1≃5×105fg−1 otherplanetesimalsinthesystem.Thegiantplanetsaremodeledto sufferfromeccentricitydamping,TypeIImigration,andtheinter- × C2α −1 m a 1/2 M∗ −1/2, actionfromtheplanetsandsmallbodies.Inseveralruns,thegiant (cid:18)10−4(cid:19) (cid:18)MJ(cid:19)(cid:16)1AU(cid:17) (cid:18)M⊙(cid:19) planetsareassumed tobefixedabout theoriginal region without τmigII2= |aa˙| =0.7×105(cid:16)10α−3(cid:17)−1(cid:16)1AaU(cid:17)(cid:18)MM⊙∗(cid:19)−1/2, (6) metaigryraIntfiootrnhm.eaIsnteiotshniemufunoldlaletoirwotnihnse,g,wcwiercemuwmaiislnltlabynrcieeexflpoylfosrutehmetmheexaritsieztreerneecsaetcrihoaflcapgslieaa.nn-t whereαandC2areefficiencyfactorofangularmomentumtrans- planets.Thus,ineachsimulationwestartwithaswarmofplanetes- port and reduce factor. Herein C2α = 10−4. τmigII1 is fit tothe imals,togetherwithoneortwogiantplanetsinthesystem,orbiting case that the planet mass is comparable to theentire mass of the thecentralstar.TheinitialpositionsforJupiterandSaturnaresetto gasdisk.Whiletheplanetmassislowerthanthemassofthegas be5.0AUand9.54AU,respectively.Originally,theinnerregionof disk,τmigII2istherightfit.Weadoptanempiricalformulaforthe thesystemconsistsof2000planetesimalsthateachownsamassof eccentricity damping from the gasdisk, (τdamp2)−1 = (e˙/e) = 5×10−3M⊕,therebyleadingtoanentiremassoftheplanetesimal −K|a˙/a|(Lee&Peale2002).Herein,wechooseK=10.Thecriti- diskof10M⊕.Hereinforallcases,theplanetesimalsareinitially calmassofplanetthatcanproduceagapisgivenas distributedintheregion[0.5,3.78]AU. S1:Inthisscenario,theplanetarysystemcomprisesthehost Mcrit ≃30(cid:16)10α−3(cid:17)(cid:16)1AaU(cid:17)1/2(cid:18)MM⊙∗(cid:19)M⊕. (7) ssdtuyamnr,aetmsheitchpaalltaetnvheoetelgusitiaimonnatl.ps,laannedtosunfefeJruspfitreorm-mtayspsepIlIanmeitg.rTahtieonmoovdeerltahse- In thiswork, besides mutual gravitational interactionamong S2: Theinitialsparameters for Jupiter and planetesimals are the objects in the system, we also consider the orbital migration, thesameasthosegiveninS1.Asacomparison,inthissimulation eccentricitydampingandaerodynamicaldragofplanetaryembryo themodeldoesnotaccountfortypeIImigrationforJupiter. (planetesimal).Theaccelerationoftheplanetorplanetaryembryo S3, S4 and S5: in this three runs, the initial system consists (planetesimal)withmassm isexpressedas i ofthehoststar,theplanetesimals,twogiantplanets–Jupiterand Saturn.InsimulationS3,bothJupiterandSaturndonotmigratein- ddtVi =−G(Mr∗+2 mi) rri + N Gmj |(rrj−−rri|)3 − rrj3 wanadrdSbaututrrnemaraeinalsltoawbleedotrobimtsi.gIrnatseiminuwlaatriodninS4thaenddySn5a,mbioctahlJeuvpoiltue-r i (cid:18) i(cid:19) Xj6=i (cid:20) j i j(cid:21) tion. + Fdamp1+Faero+FmigI(forplanetesimals/embryos) (8) WeintegrateEquation(8)usingahybridsymmetricalgorithm Fdamp2+FmigII (forgiantplanets) inMERCURYpackage(Chambers1999).However,wehavemod- (cid:26) ifiedcodestoincorporategasdragandorbitalmigrationscenarios where foroursimulation.Intheseruns,mutualinteractionsofallbodies are fully taken into account. Two bodies are considered to be in F =−2(Vi·ri)ri , collisionstagewhenevertheirdistanceislessthanthesumoftwo damp1,2 ri2τdamp1,2 physicalradii(Chambers1999).Iftwoobjectscollideeachother, theycanmergeandformasinglelargerbodywithoutanyfragmen- FmigI=−2τVmiigI, (9) taattiiomne.Isnteopuorfsi2mdualaytsioann,deaacBhurulinrsicshi-nStetogerratteodlefroarn2ce-1o0f01M0−y1r2w.Aiths usual, when the simulation ends up, the variations of energy and V FmigII=−2τmiigII, angularmomentaare10−3and10−11,respectively. (cid:13)c 2015RAS,MNRAS000,1–?? 4 Z. Sun et al. Table1.Theinitialconditionsoftheruns.ThemeaningofForceisgivenbyEquation8. Name Massofplanetesimal No. Jupiter Saturn Force M⊕ Planetesimals YesorNot YesorNot ongiantplanets S1 5×10−3 2000 Y N FmigII S2 5×10−3 2000 Y N Nomigration S3 5×10−3 2000 Y Y Nomigration S4 5×10−3 2000 Y Y FmigII S5 5×10−3 2000 Y Y FmigII Table2.Thestatisticsofthefinaldestinationoftheplanetesimalsineachruns. Name Time Percentageofmass PercentageofMass Saturn Force Myr (inside0.1AU) (beyond5AU) YesorNot ongiantplanets S1 5 79.5% 8% N FmigII S2 3.9 14.1% 0 N Nomigration S3 3.7 17.9% 0 Y Nomigration S4 5 50.5% 14.8% Y FmigII S5 5 36.9% 5.9% Y FmigII 3.2 Terrestrialplanetsformationwithgiantplanets somecaseswater-richearth-massplanetcanformoutsidethemi- gratinggiant,locatedinsidethehabitablezone.SimulationS1also ThesimulationsofS1-S5exhibitclassicalterrestrialplanetaryac- showssimilarshepherdmechanism,asshowninFigure1.Interest- cretion scenarios in their late stage formation (Chambers 2001; ingly,thetwoinnerterrestrialplanets,aswellasthemigratinggiant Raymondetal.2004,2006;Fogg&Nelson2005,2009). planet,arediscoveredtofinallyforma4:2:1MMRorbitalconfig- uration.However,wedonotobservematerialaroundthehabitable zone. In simulation S1, the massive giant planet sweeps away or 3.2.1 Terrestrialplanetsformationwithonegiantplanet accrete most of the materialsalong its pathway when it migrates Figure 1 shows the orbital evolution of the planetesimals and inward.Furthermore,wepointoutthatsuchmaterialdepletionof the Jupiter-mass planet in the first 5 Myr for simulation S1. At migratinggiantplanetcanbeobservedinoursimulationS5(Fig- an early time, the Jupiter-mass planet migrates inward in the ure7),inwhichthesystemcontainstwomigratinggiants. disk, thereby giving rise to the planetesimals nearby the giant planet to either be scattered outward into high-eccentric orbits (Mandell&Sigurdsson 2003) or shepherded inward by the gi- antplanet’smovingmeanmotionresonances(Tanaka&Ida1999; InitialsinsimulationS2arethesameasthoseinS1,however Fogg&Nelson2005).Thebuildupofinnermaterialinducesrapid inthismodel wesimplydonotletJupiterundergotypeIImigra- growth of two close-in planets within a few Myr. At 5 Myr, five tion.Figure3showstheorbitalevolutionoftheplanetesimalsand terrestrialplanetsforminside0.1AU.Thetotalmassofthesefive theJupiter-massplanetforsimulationS2.Insimilarcase,theplan- planetsis7.95M⊕,correspondingto∼80%oftheinitialbuilding etesimalsarequicklyexcitedduetotheirmutualgravitation,along materialssetinthesimulation.Thereare16planetesimalsthatwere withthatoftheJupiter-massplanetlocatingat5.0AU.Theobjects, scatteredtoorbitsbeyond5AU,thetotalmassofthemis∼0.08 involvedinthe3:1MMRat2.50AU,2:1MMRat3.28AU,andthe M⊕, the final destination of the planetesimals is shown in Table 5:3MMRat3.70AU,arestirredwithin0.1Myr.Thistrendcanbe 2.ThesimulationS1continuesevolvingfor100Myr.Bytheend clearlyseenbytheriseoftheeccentricitiesoftheplanetesimalsat of100Myr,mostoftheplanetesimalintheinitialdiskhavebeen theselocationsasshowninFigure3.Allstuffintheplanetesimal nearly cleared up by ejection or collision scenarios, arising from diskexterior tothe3:2resonance at 3.97AU isquickly removed frequent orbital crossings over the chaotic evolution. Weobserve fromthesystemviacollisionandejectionresultingfromthegiant that therealsoexistacouple of moderate-eccentric planetesimals planet.Inthetimeevolution,theplanetesimals’eccentricitiescan whicharenotinvolvedinaccretionprocessbeyond∼5AUforthe increase and the system becomes chaotic when the eccentricities remainingdisk. arelargerenough. Thebodiesintheinner diskbegintogrowvia Moreover, wefindthattwoterrestrialplanetswithamassof accretionarycollisionswithin1Myr.Thelargerbodiestendtohave 5.48M⊕and1.62M⊕areeventualsurvivorsinthesystem,locat- smallereccentricitiesandinclinations,duetothedissipativeeffects ingat0.069AUand0.104AU,respectively.Thesetwoterrestrial ofdynamicalfriction.At3.9Myr,therearefourlessmassiveter- planetsandtheJupiterforma4:2:1MMRorbitalconfigurationat restrialplanets(ascomparedtoS1)withamassof0.15−0.51M⊕ ∼10Myr,asshowninFigure2.Thediscoveryoftheexoplanets formedintheregion[0.07,0.095]AU.Theentiremassoffourter- showsthattheplanetarysystemsintheuniversearequitediverse. restrialplanetsis1.41M⊕,correspondingto∼14.1%ofthetotal Oursimulationspresentveryinterestingresults,whichmayprovide initialmaterialsasshowninTable2. somecluestofutureobservationsforsuchsystems. OursimulationS1issimilarwiththeworkinRaymondetal. (2006) that investigated habitable terrestrial planet formation un- deramigratinggiantplanet.Intheirmodel,theyadoptedahigher In comparison, we can see that an inward-migrating giant mass of 17 M⊕ and a more extended planetesimal disk that ar- planetcansignificantlyincreasetheaccretionrateintheinnerpart rangesfrom0.25to10AU.Theyfound thathot Earthscanform ofplanetesimaldiskbytheshepherdeffect.Meanwhile,itcanscat- interiortoamigratinggiant duetotheshepherding effect,andin terthebodiesintheinnerdisktotheouterdisk. (cid:13)c 2015RAS,MNRAS000,1–?? TerrestrialPlanetsFormationunderMigration 5 Table3.Physicalparametersofthe3runsthatform4:2:1MMRat100Myrandcomparisonwithexoplanetsystems. Name Mass Radius Period a e Name Massa Radius Period a e M⊕ R⊕ day AU M⊕ R⊕ day AU S1Planet1 5.48 2.47 6.62 0.069 0.001 Kepler-238c 6.0 2.39 6.16 0.069 - S1Planet2 1.62 1.65 12.25 0.104 0.038 Kepler-238d 10 3.07 13.23 0.115 - S1Jupiter 333.55 9.72 24.91 0.167 0.000 Kepler-238e 77 8.26 23.65 0.169 - S4Saturn 95.69 6.41 5.64 0.062 0.001 Kepler-302b 18 4.06 30.18 0.193 - S4Planet1 5.05 2.41 11.72 0.101 0.012 - - - - - - S4Jupiter 333.46 9.72 24.91 0.167 0.001 Kepler-302c 180 12.45 127.28 0.503 - S5Planet1 3.69 2.16 10.02 0.0911 0.142 S5Jupiter 333.02 9.71 20.15 0.1449 0.025 S5Saturn 95.12 6.40 40.79 0.2315 0.014 a KeplerplanetarymassesestimatedwithEq.(1)in(Lissaueretal.2011b). 3.2.2 Terrestrialplanetsformationwithtwogiantplanets theorbitalcrossingbetweentwogiantplanetshappenswithinsev- eral thousand years and brings about fairly chaotic behaviors for In the following, we will present the simulation outcomes of the them. Hence, a sudden jump in the Jupiter’s eccentricity occurs terrestrialplanetaryformationco-existencewithtwogiantplanets upto ∼0.72 at 0.037 Myr, whereas Saturnmay obtain amoder- intheplanetarysystems. ateeccentricity right after theclose encounter but itseccentricity In simulation S3, two giant planets, which bear a Jupiter or isquicklydampedbythegasdisk.Fromthesimulations,wefind Saturnmass,respectively,donotperformtypeIImigrationinthe thatJupiter experiencedcloseencounters withSaturnatthemax- simulation.Therefore,similartosimulationS2,theplanetesimals imum star-centric distance at ∼ 2.41 AU, which leads to strong inthediskcanbeswiftlyexcitedduetogravitationalperturbations interaction between them. From ∼ 0.037 Myr to 0.04 Myr, note fromtwogiantplanets.Inparticular,thebodies,whichareinvolved that the semi-major axis of Jupiter drops down from 3.58 AU to in the MMR with the Jupiter-mass planet, then acquire moderate 1.63AU,whereasthatofSaturndramaticallydecreasesfrom5.75 eccentricitieswithin0.1Myr.AscomparedtoJupiter,theSaturn- AUto0.45AU.Furthermore,suchchaoticmotionsforJupiterand massplanetplaysalessdominantroleinthemassaccretionofplan- Saturntriggercatastrophicfatefortheplanetesimalsinthesystem, etesimalsinformingterrestrialplanets.Inthemeanwhile,owingto where most of their orbits are severely excited and scattered. As theco-existenceoftwogiantplanets,theterrestrialformationpro- Figure6 shown, the eccentricity of the planetesimal (Planet 1) is cesscanbespeededupalthoughJupiterandSaturndonotdeviate firstkickedtobe0.88thenfallsto0.13,whileitssemi-majoraxis much from their initial orbits in this run. Figure 4 shows tempo- cansuddenlyincreaseamountupto5.18AU,thensoondropdown ralorbitalevolutionofalltheplanetesimalsandthegiantplanetsin to0.26AUoverthetimespan. simulationS3.At∼2Myr,threeterrestrialplanetsareyieldedwith As the two giant planets continue migrating inward, the 2:1 amassintherange0.10−0.57M⊕,andtheyorbitinthebroadre- MMRbetweenthemisbrokenup.Subsequently,Saturniskicked gionfrom0.07AUto1AU.Atthetimeof3.7Myr,theentiremass intotheregion∼0.1AUduetodynamicalinstabilities.By10Myr, offourterrestrialplanetsformedinside0.1AUis1.79M⊕,occu- thegiantplanetshavemovedveryclosetotheircentralstar.Inthe pying 17.9% of the initialmass of theplanetesimal disk. Similar late stage, the eccentricity of Jupiter is gradually damped by dy- withthesimulationS2,thereisnoplanetesimalbeyond5AU. namicaldissipation.However,aninnerterrestrialplanet(Planet1) InordertoinvestigatetheroleoftypeIImigration,wehave isexcitedontoahighlyeccentricorbitexteriortotheregimeoftwo performed two additional runs for the systems composed of two giantplanets,andsooncrossesJupiter’sorbit,thenfinallycaptured giantplanets,SimulationS4andS5.Inthistworuns,bothJupiter insidetheorbitsofJupiterandSaturn.ThemassgrowthforPlanet andSaturnundergotypeIImigration.Ourresultsprovideevidence 1startsfromoneplanetesimalat∼0.06MyrtoasuperEarthwith thatterrestrialformationmaytakeplaceintheinnerregionofthe a mass of 5.05 M⊕ at ∼ 0.40 Myr. Finally, Planet 1 remains at planetary systems (Fogg&Nelson (2005); Mandelletal. (2007); 0.101AU,whereasSaturnandJupiterceasemigratingat0.062AU Fogg&Nelson(2009)).However,newrunsfurthershowsomein- and 0.167 AU, respectively. This planetary system shows a very terestingoutcomes-the4:2:1MMRconfigurationsareformedfor interestingconfigurationthatasuperEarthlocatesbetweentwogi- thesesystems. antplanets’orbits,andthreeplanetsaretrappedintoanear 4:2:1 Figure5showstemporalorbitalevolutionofalltheplanetes- MMR (see Figure 8). Accordingly, the residual materials of disk imals and giant planets. According toEquation 6, τmigII1 is pro- areeitherremovedoutofthesystemorscatteredintodistantorbits. portionaltotheplanetarymass,thuswelearnthatSaturnmigrates Theireccentricitiesrisealongwiththeirsemi-majoraxes. fasterthanJupiterdoes,indicatingthattherewouldbeapossibil- ThescenarioofS4showsaresemblancetoNICEModel(e.g., ityfortwogiantplanetstorendezvousinthesystem.At0.02Myr, Tsiganisetal. 2005; Morbidellietal. 2007), which proposes that Jupiter’seccentricityisgraduallypumpedupto0.30duetogravi- theLateHeavyBombardment (LHB)–aspikeintheimpactrate tationalinteractionfromSaturnwhenitmovesinwardclosertothe onmultiplesolarsystembodies that lastedfromroughly 400 un- Jovianplanet. til 700 Myr after the start of planet formation (Teraetal. 1974; At0.036 Myr,Jupiter andSaturnreachtheorbitat 5.75AU Cohenetal.2000;Chapmanetal.2007)–wastriggeredbyanin- and3.58AU,respectively,indicatingthattheypassthrougha1:2 stabilityinthegiantplanets’orbits.IntheNiceModel,theorbits mean motion resonance. This resonance crossings may have ex- ofthegiant planetswouldhavebeeninamorecompact configu- cited the orbital eccentricities of the planets which cross the res- ration,withJupiterandSaturninteriortothe2:1resonance.Inour onance (e.g., Tsiganisetal. 2005; Morbidellietal. 2007). Thus, modelofsimulationS4,theJupiter-massandSaturn-massplanets (cid:13)c 2015RAS,MNRAS000,1–?? 6 Z. Sun et al. canswitcheachplace,similartothecaseofUranusandNeptune giants,andinthiscasethematerialoftheregionbetween0.2and in the NICE Model. Raymondetal. (2008b) showed that planet- 4.5AUofthediskwasfullyclearedbytwogiantplanets. planet scattering could create MMRs, most of these resonances areindefinitelystable.Thepreviousinvestigationsreportedthatin a convergent migration scenario for two giant planets in the so- 4 THEEMERGENCEOF4:2:1MMRSONLYWITH larsystem, the2:1MMR couldbeformedand thendisintegrated TERRESTRIALPLANETS intheevolution(Morbidelli&Crida2007b;Zhang&Zhou2010). Zhang&Zhou(2010b)statedthatwhenJupiterliesoutsideSaturn, Theinvestigationsimplythattheplanetaryconfigurationsinasso- convergentmigrationcouldoccur,Saturnisthenforcedtomigrate ciationwithanear4:2:1MMRswouldbecommonintheuniverse. inward by Jupiter where the two planets are trapped into MMR, Herein,weconsidertwoplanetswithaperiodratiointherangeof andSaturnmaymoveonitstracksofapproachingthecentralstar. [1.83, 2.18] as a pair near 2:1 MMR (Lissaueretal. 2011b). For Furthermore, to apply our model to explain the planetary forma- example, Jupiter’s Galilean moons are known to be locked into tionofthesystems,weextensivelyexaminethepublisheddataby a so-called three-body Laplacian resonance (i.e., 4:2:1 MMR) in Keplermissionandfindarelevant planetarysystem(Kepler-302) our solar system. On the other hand, other evidences are found similartothissimulation.Kepler-302isasystemconsistingoftwo intheexoplanetarysystems,inwhichthreeplanetsoftheGJ876 planets orbiting a star with a mass of 0.97 M⊙, as shown in Ta- system(Marcyetal.2001;Riveraetal.2010;Batyginetal.2015; ble 3. The outer planet of Kepler-302c is more massive than the Nelsonetal.2016)andKepler-79(KOI-152)(Steffenetal.2010; inner companion Kepler-302b, and the two planets have approx- Wangetal. 2012) arereported tobe close to4:2:1MMR. Inthis imate sizes as compared with those of the giant planets in simu- work,wehaveprovidedsubstantialevidencesonthisfromthesim- lation S4, whereas their orbital distances fromthe host star area ulations. For example, in simulation S1, S4 and S5 of Section 3, bit farther than those of the planets in the present run. Interest- weshow that the systems initiallycomposed of one or two giant ingly, the two giant planets of Kepler-302 are nearly close to a planetscanfinallyproduce4:2:1MMRs,inwhichtheformedter- 4:1MMR(Roweetal.2014),suggestingthatthesystemmayhave restrialplanetsareinvolvedin.Naturally,aquestionarises-what gonethroughsimilarmigrationprocessasinoursimulation.There- ifthesystemsonlycontainterrestrialplanets? fore,ourmodelhereinpresentsalikelyformationscenariofortwo As mentioned previously, the released Kepler outcomes re- planetsthatareclosetoa4:1commensurability,andalsoprovides portthatagreatnumberofplanetarysystemsharboringterrestrial evidencethatterrestrialplanetformationundertheinfluenceoftwo planetsinvolvedinanear2:1resonance.Furthermore,aportionof giantplanetsinthecompactsystem. the terrestrial planets discovered by Kepler are in near 4:2:1 res- InsimulationS5,althoughSaturnmigratesfasterthanJupiter, onance.Herein,wefurtherinvestigatetheconfigurationformation thereisnorendezvous betweenSaturnandJupiter.Instead, when ofthosesystemsonlywithterrestrialplanetstounderstandwhether Saturnentersthe2:1MMRorbitofJupiterat0.02Myr,thesetwo the4:2:1MMRscanbeproducedinsuchsystems.Withthesepur- giantplanetswerelockedinthe2:1MMRconfigurationandthey pose,weperformadditionalrunstoexplorethesystemsthatsimply migrateinwardtogether.Figure7showstemporalorbitalevolution consistofterrestrialplanetsthroughsimulations. of the simulation S5 for the first 5 Myr. Jupiter’s eccentricity is Inouradditional simulations,therearethreeterrestrialplan- graduallypumpedupto0.30duetogravitationalinteractionfrom etsinitiallywithmasseslessthan30M⊕ aroundcentralstar,and Finally,aterrestrialplanet of 3.69 M⊕ forminside0.1AU. This they are assumed to migrate from the outer region of the system newterrestrialplanetandthetwogiantplanetsforma4:2:1MMR undertheeffectofthegasdisk.Thegasdensityprofileispropor- orbitalconfigurationat∼5Myr.AlthoughJupiterkeepsmigrating tionaltor−1,andanemptyholelocatedintheinnerregionofthe until 10MyrinS1,Jupiterceasesitsmigrationat1.5Myrand3.5 system.Duringtheformationprocess,theterrestrialplanetssuffer MyrinS4andS5,respectively.Thisisprobablyduetothetorque fromtypeI migrationand gasdamping. Thetimescalesof typeI fromPlanet1,andthetorquedependsontheedge. migrationandgasdampingareshowninEquation(4)and(5),re- UnlikeinsimulationS3,insimulationS4andS5wecansee spectively.Intotal,wecarryoutsevengroupsofsimulations,and planetesimalsbeyond5AUduetoscatteringofinward-migrating eachgroupcontainsfiverunsbyconsideringdifferentspeedoftype giantplanets.At5Myr,thereareabout1.48M⊕materialsbeyond Imigration.WeaddanenhancefactorηtothetimescaleoftypeI 5AUinsimulationS4,correspondingto14.8%oftheinitialtotal migration.Thefiverunscorrespondtoηtimesoftypicaltimescales mass.WhileinsimulationS5,thereare∼0.59M⊕beyond5AU. of type I migration as shown in Equation 5. The values of η are Figure8showsthefinalconfigurationintheinnerdiskforthe 100, 33.3, 10,3.3, and1,respectively. Figure9showsoneof the threerunsthathaveturned ontypeIImigration. Interestingly, all runswithtypicaloutcomes.ConsideringvariousspeedoftypeImi- theserunsyield4:2:1MMR.InsimulationS1,twoterrestrialplan- gration,threeplanetsarelocatedat140,500,and1450dwhenη etsformedinteriortotheJupitermassplanet.Themigratinggiant islargerthan33.3,otherwisetheyarelocatedat100,250,and600 increasedtheaccretionrateintheinnerdiskandclearedallmate- d,respectively.Themassesoftheplanetsinfivegroupsare5,10, rialinitspathway.Asaresult,thereisnothingleftbetween0.2to and15M⊕,respectively,whileinothertwogroupstheyaresetto 4.5AU.InsimulationS4,therewasanorbitalcrossingbetweenthe be5,5,and5M⊕,respectively.InFigure9,theformationofthe migratingJupitermassandSaturnmassplanets,andsuchachaotic innermost,themiddleandtheoutermostplanetsaremarkedbythe event changed the orbital configuration of the two giant planets. black,redandbluelines,respectively.Panel(a),(b)and(c)show Hence, we note that a close-in terrestrial planet formed between the evolution of the orbital periods, eccentricities, and the period twogiant planets. Thereare some residual materialsbetween 0.2 ratios,respectively.Panel(d)and(e)displaythe2:1meanmotion to5AUinthisrun,andthiscouldbeaconsequenceofthequickly resonanceangles.Inthecasewithterrestrialplanetsinthesystem, orbitalchangeofthetwogiantsat0.1Myr,asshowninFigure6. theyallundergotypeImigration.Duetotheshorttimescalesforthe While in simulation S5, there is no chaotic orbital change of the firsttwoplanets,theyaretrappedinto2:1MMRfirst.FromPanel JupitermassandSaturnmassplanets.SimilartosimulationS1,we (c)inFigure9,wefindthatthefirstandsecondplanetsareinvolved seethatonlyoneterrestrialplanetformedinteriortotwomigrating into2:1MMRat∼0.5Myr,andthenthreeterrestrialplanetsare (cid:13)c 2015RAS,MNRAS000,1–?? TerrestrialPlanetsFormationunderMigration 7 trapped into 4:2:1 MMRs at ∼ 1.3 Myr. From Panel (b), we can 5 CONCLUSIONSANDDISCUSSIONS learnthatwhentheplanetpairsenterintoMMRs,theireccentrici- In this work, we have investigated terrestrial planetary formation tiesareexcited,butduetothedampingofgasdisk,theycannotbe ,especiallytheconfigurationformationofanear4:2:1MMRsby stirreduptohighvalues.The4:2:1MMRisusuallyproducedwhen carryingouttwosetsofsimulationswithandwithoutgiantplanets. thefirstplanetistrappedintheedgeoftheholesinthegasdisk.The From the results we find that 4:2:1 MMRs configuration can be formationprocessissimilartothatmentionedinPierens&Nelson formedinthesystemsbothwithandwithoutgiantplanets.Herein (2008),theMMRformedwhentheplanetiscapturedinthegapof wesummarizethemajorresultsasfollows. theotherplanet.Inaddition,weinvestigatetheresonanceanglesof eachpairwhichareshowninPanel(d)and(e)ofFigure9andfind (i) In the first sets, we place one or two giant planets into the thatθ21 =2λ1−λ2−̟1libratesat∼0◦,θ32 =2λ2−λ3−̟2 initialsystemsaswellastheplanetesimals,tounderstandterrestrial libratesat∼ 0◦,andθ33 = 2λ2 −λ3−̟3 libratesat∼ 180◦, formationinsuchcasesinS1-S5. wherethesubscript1and2representtheorbitalelementsofPlanet InsimulationsS2andS3,thegiantplanetsdonotundergomi- 1 and 2, respectively. At the end of the simulation, three planets gratinginthedisk,andtheterrestrialformationprocedurecanbe locate at about 11.52, 23.26, 47.17 days, respectively. This con- expeditedbecauseoftheexistenceofJupiter-massandSaturn-mass figurationissimilartothesystemKepler-92whichcontainsthree planets.Infinal,severallow-massclose-interrestrialplanetscanbe planetsintheorbitalperiodsof13.75, 26.72, and49.36 days, re- formedasaresultofradialmixingandtypeImigration.Inthetwo spectively. Infinal, by examining all runs fromseven groups, we runs,thereisno4:2:1MMRsintheevolution. find that 6 out of 35 runs (17.1%) show three planets are finally InsimulationS1,S4,andS5,weplaceoneortwogiantplanets involvedin4:2:1MMRs,whereas10runsreportthatonlythetwo in the system , which undergo type II migration during the evo- innerplanetsenterinto2:1MMR,andtworunsfortwoinnerplan- lution. Thus the planetesimals are shepherded inward, cleared or etsin3:2MMR.Abouthalfofthesimulations(48.6%)arenotin ejectedalongwithmigrationofgiantplanets.Theclose-interres- MMR.Basedonoursimulations,theproportionthatplanetspairs trial planets are yielded with a mass up to several Earth-masses. trappednear4:2:1MMRsappearstobehigh.Thereareseveralrea- ForsimulationS1,wefindthattwoshort-periodterrestrialplanets sonsthatcouldleadtosuchresults.Firstly,wesettheinitialorbital andonegiantplanetarefinallyinvolvedin4:2:1MMR,whereasin separation of planets at ∼ 20 Hill radius, which is a little larger therunsofS4,andS5,weshowthatoneterrestrialplanetandtwo thantheempiricalvalueof12Hillradius(Ida&Lin2004).Insuch giantplanetsenterintothe4:2:1MMR.TheresultantresultsofS1 cases,mostoftheplanetsarefirsttrappedbythe2:1MMRsdur- andS4aresimilartoKepler-238and302systemsinorbitalspac- ingtheirevolution.However,asacomparison,iftheyareinitially ing.WeexaminethemassconcentrationstatisticS andtheorbital c inacompactconfiguration,theprobabilityoftheformingplanets spacingstatisticS (Chambers2001;Hansen&Murray2012)for s in2:1MMRswillsignificantlydecrease.Asshownintheworkof thesesystems.ComparingthesystemS1andKepler238, wecan Wang&Ji(2014),theyfoundthatiftheplanetspairswereinitially notethattheorbitalspacingstatisticsareclosetoeachother.Their incompactconfigurationattheseparationsof∼15Hillradius,the valuesfortwosystemsare1.8and2.76,respectively,indicatingthat probabilityof2:1MMRsdecreasedto10%in50runsandtherewas S1andKepler238aresimilarintheorbitaldistribution.However, no simulation that three planets were evolved into 4:2:1 MMRs. the mass concentration statistics of them differ from each other, Secondly,inoursimulationsweslowdownthespeedoftypeImi- and their values are 125.7 and 30.3, respectively. One reason for gration for low-mass planets, and this could further explain why thisdifferenceinmassconcentrationstatisticisthatthemassesof twoplanetsarefairlyeasiertobetrappedinto2:1MMRs.Finally, planetsinour simulationsarequiteuncertain. Hereinweadopted thefinalorbitalperiodsoftheplanetsinallsimulationsarelessthan anestimatedmassesfromEq.(1)inLissaueretal.(2011b).Ifwe 50dayswhichmaybeaffectedbythetidalforcethatarisesfrom can obtain the true masses of planets in S1, the difference of S c thecentralstar.Thetidaleffectmayleadtothedeviationfrom2:1 couldbenarroweddown.Moreover,wefurthercompareS andS s c MMRs(Leeetal.2013).Insummary,theabove-mentionedfactors ofthesystemS4withthoseofKepler302,whereS foreachsys- s mayplayamajorroleincausingtheplanetspairsevolvedinnear temis5.43and4.0,respectively,andS fortwosystemsis10.4and c 4:2:1MMRsatarelativelymoderatepossibility.Thesimulationsin 23.3,respectively.ThismayimplythatthesystemS4andKepler Raymondetal.(2006)utilizedthetypicalspeedoftypeImigration 302 have similar orbital distribution. Therefore, wecan conclude whichwasveryfast,andtheembryoswereinitiallylocatedclose that S1 and Kepler 238, S4 and Kepler 302 have similar orbital toeachotherincompactconfiguration.Thesemaybetheleading configurations. differencesproducingdiverseresultsbetweenoursimulationsand Thepresentobservationsshowthatthereisalackofcompanion thoseofRaymondetal.(2006). planetsintheobservedhotJupitersystems.However,fromoursim- BycomparingS1,S4andS5simulationcontaininggiantplan- ulationresults,wecanlearnthatterrestrialplanetscansurvivein- etsandtherunsthatsimplyconsistofterrestrialplanetsinthissec- sidetheinnerregionofhotJupitersystemsbasedonourformation tion, we find that the timescales of the planet pairs trapped into scenario.AsshowninthesystemS1,thefinalformedconfiguration MMRs are various for different cases. For the systems that only issimilartothatinRaymondetal.(2006).Moreover,ifthegasdisk contain terrestrial planets, the two inner planets are found to be issmallenough,thehotJupitercouldhaveterrestrialcompanions trapped into 2:1 MMRs within 1 Myrs, and the two outer pairs formedinthesystem(Ogiharaetal.2013).Nevertheless,currently are in resonance at approximately 2-3 Myrs. Therefore, we may suchterrestrialcompanionsinthehotJupitersystemshavenotdis- concludethatthreeterrestrialplanetsarein4:2:1MMRswithin3 coveredyet,thisisbecausetheymaybeunderthedetectionlimit. Myrsonaverage.Incomparison,forthesystemsconsistingofone Intheforthcoming,wehopethatthehotJupitersystemswithlow- ortwogiantplanets,the4:2:1MMRsemergesatover2Myrswhen mass companions could be discovered with the assistance of the theterrestrialplanetscompleteitsmassaccretion.Additionally,the improvementofobservationalprecisionandtechniques. eccentricities of planets which is in MMR with giant planets are (ii) Furthermore, in the second sets, we also explore the 4:2:1 alwaysexcitedtobehighvalues,whiletheeccentricitiesofplanets MMRsformationinthecaseofthreeterrestrialplanets.Undertype insystemswithoutgiantplanetswillbesuppressedtolowvalues. I migration, ∼ 17.1% of all runs, indicate that terrestrial planets (cid:13)c 2015RAS,MNRAS000,1–?? 8 Z. Sun et al. are evolved into 4:2:1 MMRs. However, as aforementioned, the Fabrycky,D.&Tremaine,S.,2007,ApJ,669,1298 probability that planets are trapped into MMRs depends on their Fogg,M.J.,&Nelson,R.P.2005,A&A,441,791 initial separation and the speed of type I migration. Comparing Fogg,M.J.,&Nelson,R.P.2009,A&A,498,575 the processes with and without giant planets, the timescales that Ford,E.B.,&Rasio,F.A.,2006,ApJL,638,L45 three planet are in 4:2:1 MMRs in systems without giant planets Fressin,F.,etal.,2013,ApJ,766,81 are shorter than systems with giant planets. Additionally, eccen- Genda,H.,Kokubo,E.,&Ida,S.,2012,ApJ,744,137 tricities are exited to higher values in systems with giant planets Goldreich,P.,&Tremaine,S.,1979,ApJ,233,857 thanwithoutgiantplanets. Goldreich,P.,&Tremaine,S.,1980,ApJ,241,425 Haisch, K.E.,Jr.,Lada,E.A.,&Lada,C.J.,2001, ApJL,553, However, in the treatment of accretion scenario in MER- L153 CURY package (Chambers 1999), the collisions between two Hansen,B.M.S.,&Murray,N.,2012,ApJ,751,158 bodies are assumed to be perfect gravitational aggregations. Hansen,B.M.S.,&Murray,N.,2013,ApJ,775,53 Leinhardt&Stewart (2012) pointed out that real collisions could Hayashi,C.,1981,Prog.THeor.Phys.Suppl.,70,35 suffer from partially accreting collision, hit-and-run impacts, or Howard,A.W.,etal.,2010,Science,330,653 graze-and-merge events,ratherthanacompletemerger.Recently, Howard,A.W.,etal.,2012,ApJS,201,15 Chambers (2013) showed that fragmentation does have anotable Ida,S.,&Lin,D.N.C.,2004,ApJ,604,388 effect on accretion under consideration of fragmentation and hit- Ida,S.,&Lin,D.N.C.,2008,ApJ,673,487 and-runcollisions(Leinhardt&Stewart2012;Gendaetal.2012). Ida,S.,&Makino,J.,1992,Icarus,96,107 The final stages of accretion are lengthened by the sweep up of Izidoro,A.,Morbidelli,A.,&Ramond,S.N.,2014,ApJ,794,11 collisionfragments.Theplanetsthatformedwithfragmentationin Ji,J.H.,Jin,S.,&Tinney,C.G.,2011,ApJL,727,L5 thesimulationshavesmallermassesandlowereccentricitieswhen Jin,S.,&Ji,J.H.,2011,MNRAS,418,1335 comparedtothosesimulationswithoutfragmentation.Inthissense, Kokubo,E.,&Ida,S.,1998,Icarus,131,171 thescenariosthatincludeimperfectmergercollisionwillnotalter Kokubo,E.,&Ida,S.,2002,ApJ,581,666 the conclusions of this study. In forthcoming work, we will take LeeM.H.,PealeS.J.,2002,ApJ,567,596 into account hit-and-run and fragmentation process in our model Lee,M.H.,Fabrycky,D.,&Lin,D.N.C.,2013,ApJL,774,L52 forfurtherinvestigation. Lega,E.,et.al.,2014,MNRAS,440,683 Leinhardt,Z.M.,&Stewart,S.T.,2012,ApJ,745,79 Lin,D.N.C.,&Papaloizou,J.C.B.,1986,ApJ,309,846 ACKNOWLEDGMENTS Lin,D.N.C.,&Papaloizou,J.C.B.,1993,in:E.H.Levy&J.I. Lunine(eds.),ProtostarsandPlanetsIII,(Tucson:Unv.Arizona) Wethanktheanonymousrefereeforinsightfulcommentsandgood Lissauer,J.J.,etal.,2011,Nature,470,53 suggestions that helped to improve the contents. This work is fi- Lissauer,J.J.,etal.,2011,ApJS,197,8 nancially supported by National Natural Science Foundation of China(GrantsNo.11273068,11473073,11503092,11573073), the Lovis,C.,etal.,2011,A&A,528,A112 StrategicPriorityResearchProgram-TheEmergenceofCosmolog- Mandell,A.M.,&Sigurdsson,S.,2003,ApJ,591,L111 ical Structures of the Chinese Academy of Sciences (Grant No. Mandell,A.M.,Ramond,S.N.,&Sigurdsson,S.,2007,ApJ,660, XDB09000000), the innovative and interdisciplinary program by 823 CAS(GrantNo.KJZD-EW-Z001),theStrategicPriorityResearch Marcy,G.W.,etal.,2001,ApJ,556,296 Program on Space Science, CAS (Grant No. XDA04060901), Marcy,G.W.,etal.,2014,ApJS,210,20 the Natural Science Foundation of Jiangsu Province (Grant No. Mart´ı,J.G.,Giuppone,C.A.,&Beauge´,C.,2013,MNRAS,433, BK20141509), and the Foundation of Minor Planets of Purple 928 MountainObservatory. Mayor,M.,etal.,2011,arXiv:1109.2497 Morbidelli,A.,etal.,2007,AJ,134,1790 Morbidelli,A.,&Crida,A.,2007,Icarus,191,158 Nelson,B.E.,etal.,2016,MNRAS,455,2484 REFERENCES Ogihara,M.,&Ida,S.2009,ApJ,699,824 Adachi,I.,Hayashi,C.,&Nakazawa,K.,1976,Prog.Theo.Phys. Ogihara,M.,Inutsuka,S.&Kobayashi,H.2013,ApJL,778,9 56,1756 Ogihara,M.,Morbidelli,A.,&Guillot,T.2015,A&A,578,A36 Batalha,N.M.,etal.,2013,ApJS,204,24 Pierens,A.,&Nelson,R.P.,2008,A&A,482,333 Batygin,K.,Deck,K.,&Holman,M.J.,2015,AJ,149,167 RaymondS.N.,QuinnT.,LunineJ.I.,2004,Icarus,168,1 Beauge´,C.,&Nesvorny´,D.,2012,ApJ,751,119 Raymond,S.N.,etal.,2006,Science,313,1413 Bonfils,X.,etal.,2013,A&A,549,A109 Raymond,S.N.,etal.,2006,Icarus,183,265 Brunini,A.,&Cionco,R.G.,2005,Icarus,177,264 Raymond,S.N.,etal.,2008,MNRAS,384,663 Cresswell,P.,&Nelson,R.P.,2006,A&A,450,833 Raymond,S.N.,etal.,2008,ApJ,687,L107 Chambers,J.E.,1999,MNRAS,304,793 Raymond,S.N.,&Cossou,C.,2014,MNRAS,440,L11 Chambers,J.E.,2001,Icarus152,205 Rivera,E.J.,etal.,2010,ApJ,719,890 Chambers,J.E.,2013,Icarus,224,43 Rowe,J.F.,etal.,2014,ApJ,784,45 Chapman,C.R.,etal.,2007,Icarus,189,233 Safronov,V.S., 1969, Evolution of the Protoplanetary Cloud and Chiang,E.,&Laughlin,G.,2013,MNRAS,431,3444 FormationoftheEarthandthePlanets(Moscow:Nauka) Cohen,B.A.,etal.,2000,Science,290,1754 SchneiderJ.etal.,2011,A&A,532,A79 Cossou,C.,etal.,2013,A&A,553,L2 Steffen, J.H.,Batalha, N. M.,Borucki, W.J.,et al.,2010, ApJ, Cossou,C.,etal.,2014,A&A,569,A56 725,1226 Dong,Y.,&Ji,J.H.,2013,MNRAS,430,951 Tanaka,H.,&Ida,S.,1999,Icarus,139,350 (cid:13)c 2015RAS,MNRAS000,1–?? TerrestrialPlanetsFormationunderMigration 9 Tanaka,H.,Takeuchi,T.,&Ward,W.R.,2002,ApJ,565,1257 Tera,F.,etal.,1974,EarthandPlanetaryScienceLetters,22,1 Terquem,C.,&Papaloizou,J.C.B.,2007,ApJ,654,1110 Tsiganis,K.,etal.,2005,Nature,435,459 Udry,S.,etal.,2007,A&A,469,L43 Wang,J.,Xie,J.W.,Barclay,T.,&Fischer,D.A.,2014,ApJ,783, 4 Wang,S.,Ji,J.H.,&Zhou,J.L.,2012,ApJ,753,170 Wang,S.,&Ji,J.H.,2014,ApJ,795,85 Ward,W.R.,1997,Icarus,126,261 Wetherill,G.W.,1980,ARAA,18,77 Wetherill,G.W.,&Stewart,G.R.,1989,Icarus,77,330 Zhang,H.,&Zhou,J.L.,2010,ApJ,714,532 Zhang,H.,&Zhou,J.L.,2010,ApJ,719,671 Zhang, N., & Ji, J. H., 2009, Science in China Series G, 52(5), 794 Zhang,N.,Ji,J.H.,&Sun,Z.,2010,MNRAS,405,2016 Zhang,X.-J.,Li,H.,Li,S.-T.,&Lin,D.N.C.,2014,ApJL,789, L23 Zhou,J.L.,Aarseth,S.J.,etal.,2005,ApJ,631,L85 (cid:13)c 2015RAS,MNRAS000,1–?? 10 Z. Sun etal. 10.0 1.00 5.0 ] 2.0 U A [ s 1.0 y Axi cit r 0.10ntri o e aj c m 0.5 c E - i m e S 0.2 0.1 0.01 0.05 6 6 6 6 6 0 1x10 2x10 3x10 4x10 5x10 Time [ Year ] Figure1.OrbitalevolutionoftheplanetesimalsandJupitermassplanetinsimulationS1.Thesolidpointsshowtheplanetesimals.Theradiioftheplanetesimals areproportionaltotheirmass.Thecolorofeachpointshowstheorbitaleccentricity.Theopencircleshowsthegiantplanet.AstheJupitermassplanetmigrating inward,theplanetesimalsattheMMRorbitswereexcitedtohigheccentricorbits.Duetotheinward-migratinggiantplanet,planetesimalswereshepherded inwardandhencetheaccretionrateintheinnerpartoftheplanetesimaldiskincreases.Therearealsosomeplanetesimalswerescatteredtotheouterpartof thedisk.Severalterrestrialplanetsformat5Myr. (cid:13)c 2015RAS,MNRAS000,1–??

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.