Hierarchical star formation: Stars and stellar clusters in the Gould Belt

Mon. No . R. As on. Soc. 397, 2–13 (2009) doi:10.1111/j.1365-2966.2009.14465.x Hie a chical s a o ma ion: s a s and s ella clus e s in he Gould Bel F. Elias,1E. J. Al a o2and J. Cab e a-Ca˜ no3 1Ins i u o de As onom´ ıa. Uni e sidad Nacional Au ´ onoma de M´ exico. M´ exico, D.F., C.P. 04510 2Ins i u o de As o ´ ısica de Andaluc´ ıa, CSIC, Apa ado 3004, E-18080 G anada, Spain 3Facul ad de F´ ısica. Depa amen o de F´ ısica A ´ omica, Molecula y Nuclea . Uni e sidad de Se illa, Apa ado 1065, E-41080 Se illa, Spain Accep ed 2009 Janua y 1. Recei ed 2009 Janua y 1; in o iginal o m 2008 Oc obe 14 ABSTRACT We pe o m a s udy o he spa ial and kinema ical dis ibu ion o young open clus e s in he sola neighbou hood, disce ning be ween bound clus e s and ansien s ella condensa ions wi hin ou sample. Then, we disc imina e be ween Gould Bel (GB) and local Galac ic disc (LGD) membe s, using ou p e ious es ima e o he s uc u al pa ame e s o bo h sys ems ob ained om a sample o O-B6 Hippa cos s a s. Single membe ship p obabili ies o he clus e s a e also calcula ed in he sepa a ion p ocess. Using his classi ied sample, we analyse he spa ial s uc u e and he kinema ic beha iou o he clus e sys em in he GB. The wo s a o ma ion egions ha domina e and gi e he GB i s cha ac e is ic-inclined shape show a s iking di e ence in hei con en o s a clus e s: while O i OB1 is ichly popula ed by open clus e s, no a single one can be ound wi hin he bounda ies o Sco OB2. This is mi o ed in he eloci y space, ansla ing again in o an abundance o clus e s in he egion o he kinema ic space popula ed by he membe s o O i OB1, and a ma ginal numbe o hem associa ed wi h Sco OB2. We in e p e all hese di e ences by cha ac e izing he O ion egion as a clus e complex ypically su ounded by a s ella halo, and he Sco-Cen egion as an OB associa ion in he ou ski s o he complex. In he ligh o hese esul s, we s udy he na u e o he GB wi h espec o he op ical segmen o he O ion A m, and we p opose ha he di e en con en o s a clus e s, he di e en heigh s o e he Galac ic plane and he di e en esidual eloci ies o O i OB1 and Sco OB2 can be explained in e ms o hei ela i e posi ion o he densi y maximum o he Local A m in he sola neighbou hood. Al hough mo phologically in iguing, he GB appea s o be he esul o ou local and biased iew o a la ge s a clus e complex in he Local A m, ha could be explained by he in e nal dynamics o he Galac ic disc. Key wo ds: s a s: ea ly- ype – s a s: o ma ion – open clus e s and associa ions: gene al – sola neighbou hood. 1 INTRODUCTION The Gould Bel (GB) was i s disco e ed by John He schel (1847) and Benjamin Gould (1879) as a sys em o b igh s a s inclined wi h espec o he plane o he Milky Way. Fo mo e han a cen u y, many s udies ha e been de o ed o desc ibing i s s uc u e and kinema ical beha iou , as well as o p oposing a eliable global scena io ha would accoun o i s o igin ( o an ex ensi e e iew on he subjec , see P¨ oppel 1997, 2001; G enie 2004). Today, i is conside ed ha , in he scope o he mos ecen heo ies o hie a chical s a o ma ion (E emo 1978, 1995; Elmeg een e al. 2000; Elmeg een 2006), he GB is likely o be ou closes gian s a - o ming complex (Come ´ on 2001). E-mail: [email p o ec ed] (FE); [email p o ec ed] (EJA) The s ella componen o his complex akes he shape o a plana dis ibu ion o b igh and young OB s a s inclined wi h espec o he Galac ic plane (Lesh 1968; S o he s & F ogel 1974; Wes in 1985). Mos o he young OB associa ions in he sola neighbou hood a e known o be pa o he GB (Blaauw 1965; de Zeeuw e al. 1999; Elias, Cab e a-Ca˜ no & Al a o 2006a, he ea e Pape I). Also, a sys em o young, low-mass s a s, de ec ed by c oss-ma ching X- ay and op ical Hippa cos (Pe yman 1997) based ca alogues, appea s o be associa ed wi h he GB (Guillou e al. 1998). As we should expec om a gian ongoing s a - o ming complex, he local in e s ella medium is p ominen ly associa ed wi h he GB. The wo k by an den Be gh (1966) on e lec ion nebulae, by Sandq is (1977) on da k clouds and, mo e ecen ly, by Gaus ad & Van Bu en (1993) on maps o in a ed emission ha e ound a spa ial dis ibu ion o he da k clouds o in e s ella dus compa ible wi h he pa e n shown by he s ella componen o he GB. Tomi a (1986, 1987) and he s udy o he CO molecule by Taylo , Dickman C 2009 The Au ho s. Jou nal compila ion C 2009 RAS Hie a chical s a o ma ion 3 & Sco ille (1987) seem o con i m his inclina ion o he local molecula clouds. Neu al hyd ogen in he sola neighbou hood has also been ela ed o he GB a e he disco e y o Lindblad’s ‘ ea u e A’ (Lindblad 1967; Lindblad e al. 1973), in e p e ed as a ing o gas wi h an expansion mo emen (Olano 1982; Elmeg een 1982). A e Lindblad’s wo k, he gian molecula clouds we e ound o be e- la ed o he mos p ominen OB s a associa ions (Sancisi e al. 1974; Ku ne e al. 1977; de Geus 1992). A ull map o he CO molecule o e he sky la e con i med ha mos o hese clouds wi hin 1 kpc om he Sun ollow he GB pa e n (Dame e al. 1993). Then, i he GB is a s a o ma ion complex composed o bo h young s a s and associa ions and in e s ella ma e ial, we would expec o ind a popula ion o young open clus e s ollowing he GB pa e n. This is ob ious i we conside he concep o young s a clus e in i s wides sense, i.e. ep esen ing he dense inne egions o he hie a chical s uc u e o young s a ields (Elmeg een 2006). None he less, in his s udy we wan o dis inguish be ween a young clus e unde s ood as a me e s ella condensa ion and a g a i a ion- ally bound sys em (i.e. as a condensa ion ha ‘has had su icien ime and g a i a ional sel -a ac ion o ge mixed by s ella o bi al mo ions’, as pu by Elmeg een 2006). Ou wo k ocuses on he analysis o he clus e sys em, and on how hese objec s dis ibu e and beha e in compa ison wi h he GB de ined by he massi e s a s. Since he i s sys ema ic s udies ha led o a disc imina ion be- ween he GB and he LGD (S o he s & F ogel 1974; Wes in 1985), he e has been a g ea imp o emen in he numbe o ca alogued open clus e s, as well as in he a ailabili y o hei as ome ic and physical da a. Su p isingly, i is no possible o ind in he scien- i ic li e a u e any wo k de o ed o he s udy o he open clus e s membe ship o he GB and hei dis ibu ion wi hin his sys em be o e 2006. Thus, o many yea s, i has been accep ed ha he GB did no con ain a signi ican popula ion o bound clus e s. Only e y ecen ly, Piskuno e al. (2006), in hei analysis o he Galac- ic open clus e popula ion, disco e ed an open clus e complex (OCC) ha hey associa e wi h he GB. Al hough hey ind his OCC as a densi y peak wi hin he spa ial dis ibu ion o clus e s in he sola neighbou hood, hey es ima e OCC membe ship p obabil- i ies by kinema ical me hods, h ough he analysis o he angen ial eloci ies. In wo p e ious pape s (Pape I; Elias, Al a o & Cab e a-Ca˜ no 2006b, he ea e Pape II), we ha e s udied he spa ial dis ibu ion and he kinema ic p ope ies o he OB s a s and associa ions in he GB. Ou line o wo k hus leads in a logical way o he s udy o he dis ibu ion o young open clus e s in he GB and hei kinema ic p ope ies. Ou analysis will be cen ed in he compa ison o he GB mo phology as ob ained om he dis ibu ion o massi e s a s and clus e s. This will ep esen ano he s ep in he unde s anding o he na u e o he GB, and will also con ibu e o he knowledge o how s a o ma ion mechanisms p oceed o he o ma ion o s ella clus e s. 2 OPEN CLUSTER SYSTEM 2.1 Associa ions, s ella condensa ions and bound clus e s P io o desc ibing he selec ion c i e ia o ou obse a ional sam- ple, we wan o punc ua e some ideas abou he concep s o hie a - chical s a o ma ion, associa ion, loose g oupings and clus e s as g a i a ionally bound physical sys ems. S a s a e bo n om molecula gas clouds whose in e nal s uc u e can be cha ac e ized by a ac al dimension alue ha appa en ly anges om 2.5 o 2.7 (e.g. S´ anchez, Al a o & P´ e ez 2005, 2007). Simula ions o he collapse o gas clouds hin a a hie a chical s uc u e o he s ella o ma ion, wi h clus e s p esen in se e al dense co es (Walsh, Bou ke & Mye s 2006). Du ing he pas wo decades, a se o obse a ions has been collec ed which indica es ha young s ella g oupings show hie a chical pa e ns ha ange om he la ge scales o locculen spi al a ms and s a complexes o he smalle scales o OB associa ions, OB subg oups, small loose g oups, clus e s and clus e subclumps (E emo 1995). In o he wo ds, he newly bo n s a s seem o ollow he same ac al pa e n as he gas clouds om which hey we e o igina ed. The la ges scale o s ella g ouping, he g ea s a complexes, would be associa ed wi h he gas supe clouds wi h masses o he o de o 107sola masses. Acco ding o some au ho s (e.g. E emo & Elmeg een 1998; Come ´ on 2001), he GB, wi h i s spa ial scale o he o de o 1 kpc, would be he s a complex closes o he Sun, and hus would come om a single gas cloud wi h a mass o a ew million sola masses. Wha happens a smalle scales? We al eady know ha mos o he OB associa ions in he sola neighbou hood, wi h ypical sizes o abou 80 pc, a e mainly dis ibu ed along he plane o he GB. They ep esen he obse able spa ial scale immedia ely below he s a complex. Bu , does i make any sense o alk abou a ypical scale wi hin a ac al s uc u e? The answe should be sough in he obse a ional bias ha is in oduced by he age limi o he sample: OB associa ions a e de ec ed and selec ed as concen a ions o OB s a s wi h a ypical age o abou 10 My . The exis ence o some gene al co ela ion be ween he du a ion o he s a o ma ion and he size o he egion (La son 1981; Elmeg een 2006) implies ha , o a ypical age o 10 My , he ypical size o he egion is 80 pc (E emo & Elmeg een 1998; Elmeg een 2006). S a clus e s a e o med in he co es o gian molecula clouds; hey ep esen he s ella g oupings associa ed wi h he inne and denses egions o he gas, and can be in e p e ed as he una oidable esul o s a o ma ion in hie a chically s uc u ed gas (Elmeg een 2006). Howe e , only a ew o hese condensa ions will s ill be g a i a ionally bound a e 10 My . I has been es ima ed ha 90 pe cen o he clus e s lose a high ac ion o hei s a s in he i s 10 My o hei li es (Fall, Chanda & Whi mo e 2005). Thus, a e 10 My o li e, i is possible o ind s a - o ming e- gions ha main ain a la ge numbe o s a clus e s, while o he s like NGC 604 in M33 do no con ain a single clus e (Ma´ ız-Apell´ aniz 2001). This seems o depend on a ia ions o he mean densi y o he clouds; hose whe e he a e age densi y is low will o m s ella concen a ions in he co es o he clouds, bu hey will no ha e enough binding ene gy o keep a bound clus e when he gas lea es. Thus, he clus e s obse ed wi hin a s a - o ming egion a a ce ain momen could be ep esen a i es o wo dis inc physical sys ems: ei he s a clus e s, g a i a ionally bound and able o su i e galac- ic idal o ces, o a me e s ella condensa ion wi h a mean li e ime o he o de o 10 million yea s o less. Thus, he analysis o he dis ibu ion o s a clus e s (bo h bound sys ems and ansien condensa ions) could gi e us in o ma ion abou he his o y o s a o ma ion in he GB, as well as abou he physical condi ions o he gas om which hey we e bo n. 2.2 Clus e sample We ex ac ou sample om he Ca alogue o Open Clus e Da a (COCD) and i s Ex ension 1, compiled by Kha chenko e al. (2005a,b). This ca alogue has he ad an age o homo- genei y o e o he exis ing compila ions, and since we a e also C 2009 The Au ho s. Jou nal compila ion C 2009 RAS, MNRAS 397, 2–13 4F. Elias, E. J. Al a o and J. Cab e a-Ca˜ no in e es ed in wo king wi h space eloci ies, he ac ha he COCD had ca alogued p ope mo ion alues in he Hippa cos sys em and newly de e mined adial eloci y da a was decisi e in ou choice. As ou aim is o s udy he clus e dis ibu ion in he GB, we es ablish dis ance and age limi s in ou sample. The GB sys em should be well con ained wi hin a heliocen ic adius o 1 kpc (S o he s & F ogel 1974; Wes in 1985; Pape I); and since i s age has been es ima ed be ween 20 and 90 My (To a, Fe n´ andez & Figue as 2000), we only keep clus e s younge han 100 My . Thus, ou i s selec ion ( ha will be educed a e ou lie elim- ina ion, as we explain in he ollowing sec ion) is composed o 93 open clus e s, 83 o which ha e comple e kinema ical in o ma- ion. We calcula e o e e y clus e i s Ca esian Galac ic coo di- na es (X,Y,Z), whe e Xis posi i e in he di ec ion o he Galac ic cen e, Yin he di ec ion o Galac ic o a ion and Zpe pendicula o he Galac ic plane so ha hey o m a igh -handed, o hogonal ame. We also calcula e hei espec i e space eloci ies, (U,V, W), o hose clus e s wi h adial eloci y da a. 2.3 Gould Bel and Local Galac ic Disc clus e s In Pape I, we de eloped a h ee-dimensional classi ica ion me hod ha allowed us o sepa a e he GB s a s om he local Galac ic disc (LGD) s a s by pu ely spa ial c i e ia. This me hod conside ed ha he LGD and he GB could be desc ibed as a dis ibu ion along wo in e sec ing planes; wo king wi h a sample o Hippa cos O-B6 s a s we ob ained an es ima ion o he pa ame e s ha de ined hose planes. In Pape II, we used hose pa ame e s o classi y s ic ly by spa ial c i e ia a new sample o OB s a s wi h kinema ic da a. Simila ly, we now use ou classi ica ion me hod and he pa ame e s ound in Pape I o sepa a e he GB open clus e s om hose o he LGD ield. We wan o s ess ha his sepa a ion is done only by means o hei spa ial posi ion in he h ee-dimensional space. As desc ibed in Pape I, his classi ica ion me hod leads also o he iden i ica ion o ou lie s; i.e. hose objec s ha a e oo a away om he mean planes and hus lie in egions o low densi y o p obabil- i y in he sample space. We ha e ound nine ou lie s in ou clus e sample; a e hei elimina ion he emaining sample consis s o 84 clus e s, 76 o hem wi h ull space eloci y da a (lis ed in Table 1). Fu he in o ma ion on he de ec ion and meaning o ou lie s may be ound in sec ion 2 o Pape I and e e ences he ein. Finally, he indi idual GB membe ship p obabili y (P) o each clus e is lis ed in Table 1. In o al, 40 clus e s ha e a p obabili y P>50 pe cen , and hus, ollowing he Bayes minimum e o a e decision ule, we conside hem o be membe s o he GB. 2.4 Bound and unbound clus e s This sample may con ain bo h bound and unbound clus e s, as we ha e p e iously seen (Sec ion 2.1). A simple c i e ion o selec hose condensa ions wi h a high p obabili y o being bound clus e s is an age cu -o , i we conside ha all he objec s olde han 10 My can be called bound clus e s, in he sense ha hey ha e su i ed o he high a e o in an mo ali y ha happens du ing he i s 10 million yea s in he li e o a clus e . Once his c i ical h esh- old has been su passed, he mean li e o a clus e , al hough e y dependen on he en i onmen al condi ions, is usually la ge han 1Gy . I is e iden , hough, ha such a selec ion is jus an app oxi- ma ion o he p oblem, and ha o de e mine i a s ella sys em ep esen s a bound clus e , we mus compa e i s densi y wi h he idal densi y a i s posi ion in he Galaxy. Howe e , we lack he comple e in o ma ion o pe o m his de ailed analysis, and hus we mus eso o empi ical classi ica ion c i e ia. The COCD ca alogue p o ides h ee a iables o each clus e ha may gi e us some addi- ional in o ma ion abou wha we unde s and as young s ella g oup- ing and clus e ; hese a iables a e clus e adius, co e adius and age. In Fig. 1, we ep esen he clus e sample in he clus e adius– age space. I is in e es ing o no e ha he clus e s olde han 10 My in ou sample seem o g oup a ound an elonga ed s ip wi h a posi i e slope and ha only a ew objec s p esen a clus e adius ha de ia es om his mean dis ibu ion. The mos no o ious case is ha o NGC 2264, which has a adius close o 20 pc. This clus e seems o p esen a high deg ee o subs uc u e, spa ial as well as kinema ical (F˜ u ´ esz e al. 2006), ha i s cen al po en ial has no been able o e ase. Thus, we elimina e his objec om he sample, o i s p ope ies appea o di e in some s uc u al aspec s om ha o ‘classic’ bound clus e s. We ha e also d awn in Fig. 1 an uppe en elope o he main clus e dis ibu ion wi h ages olde han 10 My and younge han 100 My . The e seems o be a na u al sepa a ion be ween he main dis ibu ion o clus e s olde han 10 My and hose ha show a adius la ge han expec ed o hei age. Based on his appa en sepa a ion, we ha e aced his uppe en elope o he main clus e dis ibu ion. The exac loca ion o he line has been d awn by consensus o he au ho s. I simply aims o ep esen a di iding line be ween he la ges concen a ion o ‘p obable’ bound clus e s ( ep esen ed as illed ci cles in he igu e) and hose ha in spi e o hei age show a clea ly dis inc beha iou . The objec s loca ed unde his en elope line can be conside ed, om a conse a i e poin o iew, as highly p obable bound clus e s. Using he clus e adius and co e adius da a, we ha e es ima ed a pseudo-concen a ion pa ame e in he o m log(Rcl/Rco), ha we ep esen in Fig. 2 e sus he clus e adius. Ci cled squa es indica e hose clus e s ha , acco ding o he p e iously exposed c i e ia, can be conside ed as p obable bound clus e s. We no e ha mos o hese objec s a e dis ibu ed, as in Fig. 1, along a s aigh line. Howe e , wo o hese objec s (Pla ais 6 and NGS 2546) show a clea sepa a ion om he gene al endency. Using he online ools a WEBDA (WEB Base Donn´ ees Amas) da a base (Me milliod 1995), we ha e ecalcula ed he pa ame- e s o hese wo clus e s om hei pho ome ic da a. This anal- ysis indica es ha he pho ome ic diag ams o Pla ais 6 show a good i ing o a dis ance modulus o 7 and log( )≈6.5. Simila ly, a isual inspec ion o he colou –magni ude diag am o NGC 2546, as ca alogued by WEBDA, seems o indica e ha i is a e y young clus e [log( )≤7] wi h a ich popula- ion o p e-main-sequence s a s. This diag am is simila o ha shown by NGC 2362 and o he young clus e s loca ed in he hi d Galac ic quad an (Delgado e al. 2006; Delgado, Al a o & Yun 2007), meaning ha we a e no acing an objec almos 100 My old. Thus, we conside hese wo objec s as clus e s wi h ages in e io o 10 My , and consequen ly hey do no belong – acco ding o ou c i e ia – o he g oup o clus e s wi h a high p obabili y o being g a i a ionally bound. In Table 1, we p esen he classi ica ion o he sample ha ol- lows om his easoning. The las column shows an indica o o he p obabili y o being a bound clus e , acco ding o he c i e ia p e iously discussed (1 s ands o hose objec s ha a e ‘p obable’ bound clus e s, and 2 s ands o hose we conside as ansien s ella condensa ions). C 2009 The Au ho s. Jou nal compila ion C 2009 RAS, MNRAS 397, 2–13 Hie a chical s a o ma ion 5 Table 1. Ca alogue o s ella clus e s used in his s udy. (X,Y,Z) a e he spa ial coo dina es o he clus e s in he di ec ions o Galac ic cen e, Galac ic o a ion and no h Galac ic pole. (U,V,W) a e hei espec i e heliocen ic space eloci ies, and (U,V,W) hei esidual eloci ies co ec ed om sola mo ion and Galac ic di e en ial o a ion. Rcl and Rco s and o clus e adius and co e adius, espec i ely. Tis he clus e age. Pis he GB membe ship p obabili y calcula ed wi h ou me hod exclusi ely by spa ial c i e ia. Class desc ibes whe he a clus e is p obably bound (1) o unbound (2). COCD Name XYZ U V W U VWRcl Rco log(T)PClass (pc) (pc) (pc) (km s−1)(kms −1)(kms−1)(kms −1)(kms −1)(kms −1) (pc) (pc) (y ) (pe cen ) 2 Be keley 59 −471 878 87 20.8 3.1 6.8 2.0 15.7 13.1 3.0 1.0 6.80 10 2 32 S ock 7 −491 496 1 10.5 9.7 −7.9 3.9 22.3 −1.6 2.2 0.9 7.13 7 1 35 T umple 2 −476 438 −45 22.1 −34.9 −11.3 17.3 −22.3 −5.0 5.7 1.7 7.93 54 1 41 S ock 23 −291 244 14 13.4 −9.9 −3.6 14.9 2.7 2.7 1.2 0.5 7.51 17 1 42 Melo e 20 −159 102 −21 −12.9 −26.4 −7.5 −6.8 −13.8 −1.2 20.8 4.3 7.55 50 2 46 IC 348 −354 125 −121 −16.9 −11.6 −8.3 −11.6 1.0 −2.0 0.9 0.4 7.79 94 1 68 Collinde 65 −301 −47 −58 −14.8 −11.3 −6.9 −4.0 1.3 −0.6 13.0 5.7 7.41 67 2 72 Collinde 69 −413 −113 −92 −28.3 −11.7 −7.9 −15.4 0.9 −1.6 4.2 1.4 6.76 84 2 73 NGC 1981 −334 −178 −130 −24.6 −11.9 −6.5 −9.6 0.7 −0.2 1.7 0.9 7.50 95 1 74 NGC 1976 −329 −183 −132 −23.2 −16.3 −7.1 −8.1 −3.7 −0.8 2.8 1.4 7.71 95 1 75 NGC 1977 −415 −225 −164 −18.7 −17.7 −5.4 −2.2 −5.1 0.9 1.8 0.8 7.08 97 1 76 NGC 1980 −451 −255 −184 −20.3 −13.6 −7.1 −2.9 −1.0 −0.8 2.0 1.2 6.67 98 2 77 Collinde 70 −338 −158 −117 −16.3 −9.2 −5.8 −1.9 3.4 0.5 10.2 2.7 6.71 95 2 80 Sigma O i −340 −172 −119 −25.1 −16.4 −3.5 −10.3 −3.8 2.8 2.8 0.7 6.82 95 2 91 Pla ais 6 −313 −148 −38 −21.2 −13.1 −12.5 −7.2 −0.5 −6.2 9.5 3.3 7.79 27 2 93 Collinde 89 −789 −120 53 −24.8 −11.8 −0.7 −11.7 0.8 5.6 7.0 2.1 7.50 0 1 95 NGC 2232 −265 −183 −42 −12.6 −9.0 −9.7 2.5 3.6 −3.4 3.4 0.7 7.49 38 1 107 NGC 2264 −607 −257 25 −14.4 −12.5 −11.8 3.1 0.1 −5.5 8.4 1.8 6.81 1 2 125 Alessi 21 −363 −344 0 −35.3 −19.1 −4.3 −15.0 −6.5 2.0 4.4 0.9 7.47 6 1 126 Collinde 132 −183 −362 −66 −24.5 −17.5 −10.4 −3.6 −4.9 −4.1 2.2 0.7 7.51 74 1 133 Collinde 135 −112 −292 −62 −17.9 −7.0 −13.9 0.7 5.6 −7.6 6.1 1.1 7.54 82 1 136 Collinde 140 −168 −361 −55 −21.3 −13.4 −14.3 −0.4 −0.8 −8.0 3.5 1.4 7.57 62 1 143 Bochum 4 −528 −694 12 −13.8 −9.7 −4.9 17.7 2.9 1.4 2.0 0.9 7.25 1 1 155 Ha ne 13 −301 −646 −46 −52.5 −49.7 −14.4 −22.6 −37.1 −8.1 6.0 1.2 7.51 12 1 159 NGC 2451A −56 −178 −25 −26.2 −12.8 −13.4 −11.2 −0.2 −7.1 5.4 2.5 7.76 53 1 162 NGC 2451B −132 −406 −50 −20.4 −6.0 −15.3 1.9 6.6 −9.0 3.0 1.5 7.88 58 1 182 Vel OB2 −48 −404 −57 −22.6 −12.6 −3.0 −0.3 0.0 3.3 8.6 3.6 7.26 78 2 183 NGC 2547 −44 −450 −68 −18.9 −11.1 −13.8 4.8 1.5 −7.5 2.4 1.0 7.70 78 1 186 NGC 2546 −240 −886 −33 −37.4 −26.8 −9.3 0.2 −14.2 −3.0 9.8 4.0 7.92 7 2 190 dBe gh-Hagen 23 −120 −420 −8−24.8 −10.9 −5.0 −2.1 1.7 1.3 2.4 0.8 7.14 25 1 202 IC 2391 1 −175 −21 −27.7 −15.6 −6.1 −12.8 −3.0 0.2 3.8 1.1 7.88 57 1 204 Mamajek 1 37 −90 −39 −13.5 −18.6 −10.6 −1.3 −6.0 −4.3 0.7 0.2 6.90 57 2 205 IC 2395 −42 −706 −44 −16.8 −23.6 −8.1 15.1 −11.0 −1.8 4.3 1.2 7.08 41 1 210 T umple 10 −53 −414 5 −25.8 −12.0 −10.6 −3.3 0.6 −4.3 5.5 1.5 7.38 35 1 213 dBe gh-Hagen 56 −65 −677 17 −23.3 20.3 0.3 7.7 32.9 6.6 3.6 1.1 7.24 26 1 216 Pla ais 8 20 −147 −20 −13.0 −18.8 −3.7 1.0 −6.2 2.6 7.9 1.6 7.75 62 1 255 dBe gh-Hagen 99 151 −509 −5−28.7 −14.6 −16.2 −3.1 −2.0 −9.9 3.0 1.2 7.86 68 1 259 IC 2602 53 −150 −14 −8.2 −23.1 −0.4 5.9 −10.5 5.9 7.0 1.4 7.83 62 1 261 Alessi 5 123 −378 −14 −22.3 −20.5 −7.2 −0.9 −7.9 −0.9 2.1 0.7 7.71 66 1 357 dBe gh-Hagen 164 382 −392 −58 −13.4 −29.2 −19.1 8.5 −16.6 −12.8 3.8 1.3 7.14 13 1 366 NGC 6025 623 −444 −79 −12.5 −12.2 3.4 11.0 0.4 9.7 5.1 1.3 7.96 3 1 371 NGC 6087 758 −479 −85 −13.9 −4.7 −1.9 10.7 7.9 4.4 7.5 2.2 7.93 1 1 395 NGC 6322 961 −252 −53 −56.4 11.3 −6.8 −39.1 23.9 −0.5 2.1 1.0 7.16 0 1 402 NGC 6383 982 −74 1 3.6 −2.2 −10.5 15.3 10.4 −4.2 4.3 1.4 6.71 0 2 408 NGC 6405 486 −29 −6−12.3 −12.1 −4.4 −2.1 0.5 1.9 3.4 0.9 7.91 6 1 412 IC 4665 290 171 103 −1.4 −14.2 −7.5 2.4 −1.6 −1.2 6.1 1.8 7.63 92 1 423 Collinde 359 542 310 140 10.4 −21.4 −13.6 9.8 −8.8 −7.3 12.3 4.5 7.45 99 2 425 NGC 6514 810 101 −4−2.4 1.3 −10.8 3.7 13.9 −4.5 3.3 1.7 7.28 0 1 449 IC 4725 601 146 −48 2.2 −16.6 0.6 6.8 −4.0 6.9 6.5 2.2 7.83 2 1 456 S ephenson 1 142 331 99 −4.5 −19.4 −10.2 −5.8 −6.8 −3.9 5.7 1.3 7.69 83 1 479 Roslund 5 133 396 2 −5.8 −18.9 −6.9 −9.2 −6.3 −0.6 3.9 1.1 7.77 28 1 484 Collinde 419 153 723 36 23.2 −12.9 −6.7 9.3 −0.3 −0.4 3.2 0.9 6.85 47 2 500 IC 1396 −134 822 54 19.4 −11.9 −7.2 2.4 0.7 −0.9 9.5 3.6 6.69 56 2 501 NGC 7160 −191 766 89 19.0 −22.4 −1.0 3.8 −9.8 5.3 3.2 0.7 7.66 43 1 506 Pismis-Mo eno 1 −258 858 83 6.4 −22.0 −4.3 −11.7 −9.4 2.0 2.5 1.1 7.55 34 1 510 Cep OB3 −253 651 36 12.0 −11.3 −4.8 0.4 1.3 1.5 12.2 3.7 7.44 31 2 1013 ASCC 13 −765 228 45 5.7 −5.3 −6.4 7.7 7.3 −0.1 9.8 3.5 7.71 1 2 1016 ASCC 16 −408 −156 −145 −0.6 −1.5 0.9 13.7 11.1 7.2 5.0 2.0 6.93 97 2 1018 ASCC 18 −439 −178 −159 −11.9 −6.0 −2.5 3.1 6.6 3.8 5.4 1.7 7.12 98 1 C 2009 The Au ho s. Jou nal compila ion C 2009 RAS, MNRAS 397, 2–13 6F. Elias, E. J. Al a o and J. Cab e a-Ca˜ no Table 1 –con inued COCD Name XYZ U V W U VWRcl Rco log(T)PClass (pc) (pc) (pc) (km s−1)(kms −1)(kms −1)(kms −1)(kms −1)(kms −1) (pc) (pc) (y ) (pe cen ) 1019 ASCC 19 −299 −139 −117 −16.0 −9.0 −6.1 −2.3 3.6 0.2 4.9 1.2 7.64 94 1 1020 ASCC 20 −399 −158 −136 −16.1 −5.3 −5.0 −1.7 7.3 1.3 5.9 2.0 7.35 97 1 1021 ASCC 21 −451 −163 −142 −17.4 −8.3 −5.3 −2.9 4.3 1.0 7.0 2.6 7.11 98 2 1024 ASCC 24 −318 −236 −57 −10.2 −10.3 −15.1 6.6 2.3 −8.8 2.4 0.8 6.96 47 2 1033 ASCC 33 −434 −661 −121 −23.6 −1.0 −8.8 6.8 11.6 −2.5 12.6 6.3 7.26 81 2 1047 ASCC 47 −183 −881 5 −41.9 −2.6 −17.6 −4.4 10.0 −11.3 7.9 2.4 7.88 10 1 1050 ASCC 50 −111 −843 22 −30.4 −13.5 −9.6 5.9 −0.9 −3.3 5.9 1.8 7.48 17 1 1058 ASCC 58 122 −587 14 −33.0 −15.3 −13.0 −4.9 −2.7 −6.7 4.2 1.1 7.04 60 1 1069 ASCC 69 482 −867 −126 −29.8 −17.5 −9.4 7.2 −4.9 −3.1 7.0 2.1 7.91 11 1 1076 ASCC 76 376 −465 −44 −19.7 2.7 −5.4 4.5 15.3 0.9 3.7 1.1 7.45 15 1 1079 ASCC 79 612 −513 −40 −8.1 −15.3 −7.9 17.7 −2.7 −1.6 7.3 3.5 6.86 3 2 1084 ASCC 84 721 −532 −85 −16.2 −11.7 −10.7 10.2 0.9 −4.4 3.9 1.3 7.68 1 1 1104 ASCC 104 460 654 −22 9.3 −13.7 −8.2 −2.3 −1.1 −1.9 6.7 2.1 7.71 2 1 1105 ASCC 105 229 444 19 −2.5 −19.5 −7.0 −7.4 −6.9 −0.7 5.2 1.8 8.00 31 1 1114 ASCC 114 −68 546 10 15.5 −20.2 −0.8 7.4 −7.6 5.5 1.5 0.8 7.75 62 1 1118 ASCC 118 −224 869 66 27.4 −35.6 −11.6 8.9 −23.0 −5.3 3.3 1.6 7.02 41 1 1127 ASCC 127 −132 323 25 −5.5 −10.5 −8.4 −6.6 2.1 −2.1 4.4 1.1 7.82 42 1 Figu e 1. Clus e adius e sus age. The illed diamonds ep esen ou p obable bound clus e s in he age ange om 10 o 100 My . 3 ANALYSIS AND DISCUSSION 3.1 Spa ial dis ibu ion A wo-dimensional p ojec ion o he spa ial dis ibu ion o open clus e s in he sky is shown in Fig. 3 ( op panel). The e we see, as we commen ed in Sec ion 1, ha GB clus e s ( illed ci cles) a e mos ly concen a ed owa ds sou he n Galac ic la i udes, and ha only h ee o hem (395 =IC 4665, 423 =Collinde 359 and 456 =S ephenson 1) clea ly ise abo e he Galac ic plane (b>10◦). Acco ding o ou c i e ia, 395 and 456 a e p obable bound clus- e s, while 423 could a he be a ansi o y s ella condensa ion. We mus no e ha hese h ee objec s a e loca ed in he i s Galac- ic quad an and ha , appa en ly, hey a e no ela ed o any OB associa ion. We also ep esen in his igu e he OB s a s used in ou analysis o he spa ial s uc u e o he GB (Pape I). The associa ions Sco OB2, O i OB1, Pe OB2 and Lac OB1 a e classically hough o be Figu e 2. Pseudo-concen a ion pa ame e , log(Rcl/Rco) e sus clus e a- dius. Filled diamonds ep esen ou p obable bound clus e s. componen s o he GB (e.g. Olano 1982), and s udies ha ollowed he Hippa cos mission sugges ha Vel OB2, T 10 and Collinde 121 also belong o he GB (de Zeeuw e al. 1999), al hough hei po- si ion close o he line o nodes whe e he GB in e sec s he Galac ic disc adds qui e some unce ain y o his membe ship assigna ion. In any case, i is e iden ha he Sco OB2 and O i OB1 associa- ions, espec i ely, ma k he sou h and no h Galac ic ex emes o he inclined plane which bes desc ibes he s ella sys em known as he GB, and ha in a ce ain sense he geome ical cha ac e i- za ion o he GB is de ined by he ela i e posi ion o hese wo associa ions. Thus, we ha e ma ked wi h a ed c oss he s a s belonging o he Sco OB2 associa ion, ollowing he coo dina es as gi en by de Zeeuw e al. (1999), ha si ua e he complex in he ange l= 290◦–360◦,b=−10◦–30◦and D=100–220 pc, and ha we ha e agged as Sco pius-Cen au us in he igu e. In he same manne , we ha e selec ed he s a s belonging o O i OB1 as hose in he ange l=197◦–215◦,b=−12◦ o −26◦and D=300–550 pc (de Zeeuw C 2009 The Au ho s. Jou nal compila ion C 2009 RAS, MNRAS 397, 2–13 Hie a chical s a o ma ion 7 Figu e 3. Dis ibu ion o open clus e s on he sky (Galac ic coo dina es). Open and illed diamonds ep esen clus e s classi ied as LGD and GB mem- be s, espec i ely. Small ci cles s and o he O-B6 GB s a s om Pape I. Red c osses and g een iangles ma k, espec i ely, he s a s belonging o Sco OB2 and O i OB1, acco ding o de Zeeuw e al. (1999). Top panel shows all he clus e s in ou sample, whe eas bo om panel displays only he p obable bound clus e s. e al. 1999). These s a s appea in Fig. 3 as g een iangles, and he egion is agged as O ion. Es ima ions o he age ange o hese associa ions indica e ha Sco OB2 membe s a e be ween 5 and 20 My old (Sa o i, L´ epine & Dias 2003), while he ypical age o O i OB1 membe s is be ween some 105y and11My (B ice ˜ no e al. 2005; He n´ andez e al. 2006). Hence, we a e acing wo associa ions wi h sizes and ages el- a i ely simila , and acco ding o de Zeeuw e al. (1999), wi h a numbe o p obable Sco OB2 membe s la ge han ha o O i OB1. Wi h hese da a, i sounds easonable enough o hink ha he s a clus e popula ion ela ed o Sco OB2 should be la ge han ha ela ed o O i OB1. Bu wha we obse e in Fig. 3 ( op panel) is exac ly he opposi e. The e is no a single clus e wi hin he on ie s o Sco OB2, while we ha e de ec ed 11 objec s associa ed wi h O i OB1 (73, 74, 75, 76, 77, 80, 1016, 1018, 1019, 1020, 1021), six o hem being ‘p obable’ bound clus e s (73=NGC 1981, 74=NGC 1976, 75 =NGC 1977, 1018, 1019, 1020). This is be e obse ed in Fig. 3 (bo om panel), whe e we ha e ep- esen ed only hose clus e s ca alogued as ‘p obable’ bound ones, agged wi h numbe 1 in he las column o Table 1. This igu e shows how almos he o ali y o he ‘p obable’ bound clus e s as- socia ed wi h he GB a e loca ed in he O ion and he Puppis-Vela egions. Mo eo e , almos all o he o he objec s show no ela ion- ship wi h o he OB associa ions in he GB. Figu e 4. Spa ial dis ibu ion o he sample clus e s in he XY plane. Symbols a e as in Fig. 3. Figu e 5. Spa ial dis ibu ion o GB clus e s ( illed ci cles) in he XZ plane. O he symbols a e as in Fig. 3. The same phenomenology can be obse ed in Figs 4 and 5, whe e we ha e ep esen ed he GB clus e dis ibu ion in he spa ial p o- jec ions XY and XZ. The LGD clus e s end o dis ibu e uni o mly ac oss he Galac ic plane in Fig. 4, bu ha is no he case o GB open clus e s. We obse e how he GB’s dis ibu ion is qui e clumpy; many o he clus e s a e loca ed in he egion o Vela, and we specially no e ha he egion o O ion (in he hi d quad- an o he XY plane, and a ound he mos nega i e alues o Zin Fig. 5) p esen s an impo an concen a ion o membe s. Mo eo e , i we conside only clus e s wi h a highe membe ship p obabili y (75 pe cen , as no ed in Fig. 6), and hus elimina ing mos o he clus e s a ound he line o nodes in which he GB plane in e sec s he LGD plane, he GB is p ac ically educed o he O ion egion, as i ha we e solely he clus e popula ion o he GB. C 2009 The Au ho s. Jou nal compila ion C 2009 RAS, MNRAS 397, 2–13 8F. Elias, E. J. Al a o and J. Cab e a-Ca˜ no Figu e 6. Same as Fig. 4, bu including only GB clus e s ( illed ci cles) wi h a membe ship p obabili y highe han 75 pe cen . O he symbols a e as in Fig. 3. This analysis shows as an e iden ac ha while he O i OB1 associa ion is ela ed o an impo an popula ion o s a clus e s (be hey ansi o y s ella condensa ions o g a i a ionally bound sys- ems), he Sco OB2 complex does no include a single s a clus e wi hin i s on ie s, ye i is ex emely ich in massi e OB s a s and p e-main-sequence s a s (Sa o i e al. 2003). In o he wo ds, inside an appa en ly single s a o ma ion complex as he GB, and sepa- a ed only by abou 500 pc, we ind wo s a - o ming egions, well de ined bo h spa ially and kinema ically, ha p esen wo clea ly dis inc modes o s a o ma ion. O ion is an OB associa ion ha p esen s a la ge numbe o s ella condensa ions seen as clus e s, some o which appea o be g a i a ionally bound. On he o he hand, Sco-Cen seems o be o ming only isola ed s a s o loose g oups ha do no p esen he shape o a s a clus e , and hus ha e no been de ec ed as such. Bu , how does his ac ansla e o he eloci y space? 3.2 Kinema ic beha iou I seems ob ious ha he di e en spa ial posi ion o he cen oids o he wo s ella g oups analysed in Sec ion 3.1 implies also a di - e en loca ion in he eloci y space, specially in he Vcomponen due o hei sepa a ion o almos 500 pc in he X-axis. The eloc- i y cen oids o hese wo associa ions, O i OB1 and Sco OB2, as de ined by he GB s a s, a e loca ed a (−16.4, −9.5, −5.0) and (−8.0, −19.4, −6.0) km s−1, espec i ely. This can be clea ly ob- se ed in Fig. 7 ( op panel), whe e we ha e ep esen ed in he UV plane he isodensi y lines o he GB as de ined by massi e s a s, as well as he membe s a s o O i OB1 and Sco OB2, acco ding o Figu e 7. Veloci y dis ibu ion o GB clus e s ( illed squa es) agains ha o he GB s a s (isodensi y lines) om Pape I. C osses and open iangles ma k, espec i ely, he s a s belonging o Sco OB2 and O i OB1, acco ding o de Zeeuw e al. (1999). Top panel ep esen s he o ali y o GB clus e s, whe eas bo om panel displays only hose classi ied as p obable ansien s ella condensa ions. he c i e ia by de Zeeuw e al. (1999) explained in Sec ion 3.1, and he GB s a clus e s. Once mo e, as i was expec ed, we obse e a clea co ela ion be ween he clus e s’ dis ibu ion and he O ion eloci y cen oid, while he numbe o GB clus e s associa ed wi h he Sco-Cen complex is me ely ma ginal. The di e ence is e en mo e e iden i we limi ou sel es o he younges clus e s, which could be ep esen a i es o ansien s ella condensa ions (Fig. 7, bo om panel). As we ha e commen ed abo e, he di e ence obse ed be ween he eloci y cen oids was p edic able due o he di e en ial Galac ic o a ion and o he ac ha bo h g oups we e qui e sepa a ed in space. Howe e , we wonde i he o a ion ield in his egion o he Galaxy can comple ely explain he kinema ic beha iou o bo h he s a s and he clus e s in he GB. Mo eno, Al a o & F anco (1999) analysed he eloci y space o a sample o OB s a s belonging o he GB, and ound ha he obse ed eloci y ield was no compa ible wi h ha ob ained om s a o ma ion models, as well as wi h i s dynamical e olu ion a e he injec ion o momen um and ene gy in he p ime al cloud. In o he wo ds, he esidual eloci ies o he s a s showed a highly nega i e e ex de ia ion (Filin 1957; Mihalas & Binney 1981) ha could no be explained by he dynamical model, because a e a ime in e al qui e in e io han he age o he GB, he di e en ial o a ion p e ailed o e he mo emen s o igina ed by he ‘ad hoc’ ene gy and momen um injec ion, and ga e place o a sligh ly posi i e e ex C 2009 The Au ho s. Jou nal compila ion C 2009 RAS, MNRAS 397, 2–13 Hie a chical s a o ma ion 9 Figu e 8. Dis ibu ion o he esidual eloci ies, co ec ed om sola mo ion and di e en ial o a ion, o GB clus e s ( illed squa es) agains ha o he GB s a s (isodensi y lines) om Pape I. C osses and open iangles ma k, espec i ely, he s a s belonging o Sco OB2 and O i OB1, acco ding o de Zeeuw e al. (1999). de ia ion. Mo eno e al. (1999) also ound ha i he s a s belonging o he Sco-Cen associa ion we e elimina ed, he e ex de ia ion became posi i e. Tha is, hose models designed o explain he o igin o he GB om supe no ae explosions a e a p e ious p ocess o s a o ma ion, o om he impac o a high- eloci y cloud on he Galac ic disc (see P¨ oppel 1997, 2001; Sa o i e al. 2003 o e iews abou possibly o ma ion mechanisms o he GB), a e no able o explain he obse ed esidual eloci y ield, unless he Sco-Cen s a s a e neglec ed. Now we wan o e alua e he esidual eloci y ellipsoid o he GB clus e s; in o de o do so, we ha e co ec ed he eloci ies om sola mo ion and di e en ial o a ion using he Oo cons an s: A= 16 km s−1kpc−1and B=−16 km s−1kpc−1(Pape II; M´ endez e al. 2000). These clus e esidual eloci ies a e ep esen ed in Fig. 8, along wi h he esidual eloci ies o OB s a s, whe e di e en ma ks indica e hose belonging o O i OB1 and Sco OB2. Then, we calcula e he longi ude o he e ex, l , o bo h sys ems, he LGD and he GB clus e samples. The esul , i we cu a GB membe ship p obabili y o 50 pe cen , is ha l =9. ◦3±8. ◦3 o he LGD, and ha l =−1. ◦3±15. ◦5 o he GB. This is a alue e y a om he GB e ex de ia ion o l =−47◦ ound in Pape II o he OB s a s belonging o he GB. This is undoub edly caused by he absence o clus e s in he Sco-Cen associa ion, ha was esponsible o he la ge e ex de ia ion o he GB (Mo eno e al. 1999; Pape II). Mo eo e , i we keep only GB clus e s wi h a membe ship p obabili y highe han 75 pe cen , he e ex de ia ion o he sys em is l =9. ◦7±16. ◦3, which is p ac ically he same as ha o he LGD. The e ex es ima ion o he s a clus e sys em in he GB hus gi es us double in o ma ion. Fi s , om a kinema ic poin o iew, he lack o s a clus e associa ed wi h he Sco OB2 complex is demons a ed. Secondly, he di e ence in he eloci y space be- ween he O i OB1 and Sco OB2 associa ions canno be comple ely explained by he Galac ic di e en ial o a ion. As we may see in Fig. 8, he esidual eloci y dis ibu ions o hese wo associa ions show a di e en beha iou . While he s a s belonging o O i OB1 p esen a main eloci y dispe sion axis wi h a posi i e e ex, he Sco OB2 s a s p esen a main axis ha is almos pe pendicula o he o me , wi h a clea ly nega i e e ex de ia ion. The e o e, as i had al eady been no ed (Mo eno e al. 1999; Pape II), he e ex de ia ion in he sola neighbou hood will depend on he selec ion o he sample. I mos o he sample s a s belong o Sco OB2, he e ex de ia ion will undoub edly be nega i e. I we ex end ou sample a he away om he Sun in o de o include O ion s a s, we will each some balance, and hus he e ex de ia ion will u n owa ds alues close o ze o. The la e is p ecisely wha we obse e in he s a clus e popula ion associa ed wi h he GB: he lack o clus e s wi hin he Sco-Cen complex makes he e ex de ia ion close o ze o. 3.3 Clus e complexes and scaled OB associa ions Wha we ha e obse ed when compa ing he dis ibu ion o s a clus e s and OB s a s in he GB is ha he wo g ea complexes ha seem o de ine he no h and sou h Galac ic ex emes o his la ge s ella s uc u e show a di e en beha iou acco ding o he scene y o hie a chical s a o ma ion. O i OB1 shows a conside able po - ion o i s s ella popula ion as g ouped and o ming s a clus e s, hal o hem being p obable bound clus e s wi h ages la ge han 10 My . This kind o s ella sys em has been de ec ed, obse ed and analysed in bo h he Milky Way (E emo & Si nik 1988; Al a o, Cab e a-Cano & Delgado 1991) and o he galaxies (e.g. 30 Do adus in he LMC; Hun e e al. 1995; Walbo n, Ma´ ız-Apell´ aniz & Ba b´ a 2002); i no only con ains a ich s a clus e popula ion, bu also is no mally imme sed inside a s ella halo. On he opposi e side o he s a - o ming egions’ concen a ion ange, he e can be ound he scaled OB associa ions (SOBAs; e.g. Ma´ ız-Apell´ aniz 2001), o which NGC 604 in M33 is a good example. Al hough he s a o ma ion a e in NGC 604 is much highe han ha obse ed in he Sco OB2 associa ion, hey bo h ha e in common hei lack o s a clus e s. Thus, sepa a ed only by 500 pc, he e a e wo OB associa ions ha , appa en ly, ha e been bo n om molecula clouds unde di e en ambien condi ions, bu ha always ha e been conside ed as he undamen al pa s o a single s ella sys em known as he GB. Howe e , O i OB1 seems o ep esen he s ella halo associa ed wi h a clus e complex, while Sco OB2 appea s o be a clea example o an OB associa ion, no ela ed o clus e o ma ion. I we in e p e his esul in e ms o a hie a chical s a o ma ion p ocess (Elmeg een 2006, 2008), and conside ing ha he age o he s a s in bo h associa ions p esen simila anges, we should conside ei he ha he densi y maximum in he pa en al gas dis ibu ion o Sco OB2 was in e io han he densi y peaks in he clouds ha o med O i OB1, o ha he idal o ces in he Sco-Cen egion we e in ense enough o des oy any subs uc u e obse able as a s a clus e , in an in e al o ime lesse han 10 My . Any o hese wo possibili ies equi es some a ia ions o he ambien physical condi ions in spa ial scales smalle han 500 pc, be hey due o an ex e nal di e ence o p essu e ha caused highe densi y peaks in he O ion egion, o /and due o local g a i a ional po en ials o shea o ces ha sho en he li e imes o he ansien s ella condensa ions in he Sco-Cen egion. C 2009 The Au ho s. Jou nal compila ion C 2009 RAS, MNRAS 397, 2–13 10 F. Elias, E. J. Al a o and J. Cab e a-Ca˜ no Figu e 9. Densi y map o he s a clus e s younge han 10 My , wi hin a box 4 kpc o side cen ed in he Sun ( ed do ed ci cle). The O i OB1 and Sco OB2 associa ions ha e been supe imposed on he map, along wi h hei espec i e esidual eloci y ec o s (black lines). 3.4 The O ion A m Since he pionee ing wo k o Becke (1956), who aced he local spi al s uc u e om he young s a clus e s dis ibu ion wi hin a adius o 2 kpc, each ime ha a new ca alogue o s a clus e s has been ailo ed, he co esponding map o his local spi al s uc u e has been d awn again (e.g. Janes & Adle 1982). The esul s o doing his show a se ies o clus e g oupings wi h ypical sizes o he o de o 1 kpc (E emo & Si nik 1988; Al a o e al. 1991; Al a o, Cab e a-Cano & Delgado 1992) ha seem o ollow h ee segmen s o spi al a ms which ha e ecei ed he names o Ca ina–Sagi a ius A m, Pe seus A m and O ion A m (also known as Local A m). The inclusion o hese op ical segmen s wi hin he gene al scheme o he spi al s uc u e o he Galaxy is con o e sial, and depending on he spi al ace s and he analysis echniques employed, di e en solu ions o he numbe o a ms, hei pi ch angle o he eloci y o he densi y pa e n ha e been ound (see Naoz & Sha i 2007 and e e ences he ein). In Fig. 9, we ep esen he densi y map o he s a clus e s younge han 10 My , inside a squa e o side leng h 4 kpc, cen ed in he Sun. The da a ha e been ex ac ed om he COCD ca alogue. In he igu e, we obse e i e main concen a ions o young clus e s ha had p e iously been de ec ed by o he au ho s (e.g. E emo & Si nik 1988). In pa icula , he O ion and Cygnus complexes seem o align, delimi ing he local op ical segmen o he O ion A m. Supe imposed o e he young clus e s densi y map, we ha e d awn he O i OB1 and Sco OB2 associa ions. The lines ha depa om he cen oids o bo h associa ions ep esen hei esidual eloci ies, co ec ed om sola mo ion and di e en ial o a ion. O i OB1 appea s associa ed wi h he densi y maximum o he O ion A m de ined by he young clus e s; on he con a y, Sco OB2 is loca ed in he inne im o he A m, close o he Sun whe e he clus e densi y is lowe . I we conside ha a highe densi y o clus e s is ep esen a i e o a highe ambien p essu e in he o iginal Figu e 10. Densi y dis ibu ion o he CFI in he GB plane. The open ci cles ep esen he cen oids o he O i OB1 and Sco OB2 associa ions. The solid con ou line ep esen s he CFI alue 0.5, and he dashed con ou lines a e sepa a ed by a alue o he CFI o 0.05. gas, he ela i e posi ion o hese associa ions wi h espec o he main locus o he O ion spi al a m could explain hei di e en con en o s a clus e s. To be e illus a e his, we de ine a Clus e Fo ma ion Index (CFI) ha desc ibes he ela i e con en o clus e s wi h espec o he OB s a s ha shape he GB: CFI =Clus e densi y Clus e densi y +OB s a densi y .(1) The spa ial densi ies o clus e s and OB s a s ha e been ob ained by using Gaussian ke nels in he GB plane (XY), wi h a σ= 200 pc, and no malized so ha he o al sum o he densi y be equal o 1. Then, Fig. 10 shows he dis ibu ion in he GB plane o he CFI pa ame e in he egion ha con ains bo h OB associa ions. A line ha joins he cen oids o bo h associa ions has been ep esen ed o e he densi y map, and a cu along his line (Fig. 11) clea ly shows a CFI g adien ha anges om 0.62 a he maximum nea O ion o 0.39 in he icini y o Sco pius Cen au us. Since he dis ance o he O i OB1 cen oid is a good es ima o o he dis ance o he O ion spi al a m, his indica es ha as we mo e away om he a m, he clus e ed s a o ma ion is less e icien . 3.5 The na u e o he Gould Bel The i s de ec ion o he GB was based on he ac ha he b igh es s a s in he sky, especially in he Sou he n Hemisphe e, p esen an eccen ic posi ion wi h espec o he g ea ci cle o he Milky Way (He schel 1847; Gould 1879). Tha is, i was a me e mo pholog- ical ma e . La e s udies de e mined he main s ella and gaseous componen s o he GB, and om he analysis o hei spa ial and kinema ical p ope ies, he shape, size and kinema ics o he GB we e es ima ed. C 2009 The Au ho s. Jou nal compila ion C 2009 RAS, MNRAS 397, 2–13