PROCEEDINGSOF THE XXIthANNUAL MEETING OF THE SOCIEDADEASTRONÓMICA BRASILEIRA (AUGUST 1995)
EDITEDBY F.JABLONSKI, F.ELIZALDE, L.SODRÉJr. AND V.JATENCO-PEREIRA
.
ISSN 0104-6071
PROCEEDINGS OF THE XXIthMEETING OF THE BRAZILIAN ASTRONOMICAL SOCIETY
July 31 - August 4, 1995 Caxambu, MG, Brazil
SAB** * *
Edited by Francisco Jablonski, LABORATóRIONACIONALDE ASTROFíSICA- LNA/CNPQ C.P. 21, Itajubá, MG, 37500-970,Brazil Flávio Elizalde, INSTITUTONACIONALDE PESQUISASESPACIAIS- INPE/MCT C.P. 515, SãoJosédo Campos,SP, 12201-970,Brazil
Laerte SodréJr. andVera Jatenco-Pereira. INSTITUTOASTRONóMICOE GEOFíSICO- USP C.P. 9638, SãoPaulo, SP,01065-970,Brazil Ficha catalográfica:
SociedadeAstronómicaBrasileira Proceedingsof the XXIlhMeeting of the Brazilian Astronomical Society 205 páginas.
Contribuiçõesde conferencistas.1996. I. Título: Proceedingsof the XXI1'1Meeting of theBrazilian Astronomical Society
ISSN: 0104-6071
Distribuição:
Instituto Astronómico e Geofísico - USP Av. Miguel Stéfano, 4200 - Parquedo Estado 04301-904-São Paulo - SP XXIth MEETING OF THE BRAZILIAN ASTRONOMICAL SOCIETY
i238* c -ff./t Í r - - *-V A«i 'I'? — W.i ».-ÿ
.... • •. • o' • Ç «•»< . f — # r. ' 7*-# vC X
f . - * «** f it fed <>? r k>feO/Lf " f- tx f .7/ r I f «'j < *• Í"ií.*r í;i Í4f«l .ÚJ-; fj ILL1* V *7** -À Jli - 'vÿH
f' i > m .,>m. » w f i * > M * r • 1
July 31 - August 4, 1995 Hotel Glória, Caxambu, MG, Brazil
TABLE OF CONTENTS
Foreword
Section1: Extragalactic Astronomy and Cosmology
What can we learn from near-infraredcolours os spiralgalaxies? R.F.Peletier andM.Balcells 3 UV spectra of early-typegalaxies. C.J.Bonatto, E.Bica, M.G.PastorizaandD.Alloin 15 Redshift surveysof field galaxies. C.N.A.Willmer 27 Magneticreconnectionwithin galaxiesand extragalacticstar forming regions. L.C.Jafelice 37 Gamma-ray burst: A cosmologicalsettling. G.A.M.Tanco 45 Black Holes: From galacticnuclei to elementaryparticles.
J.P.S.Lemos ...... 57 Neutrinos,DM and the RW universe.
J.C.Huang . 71
Section 2: Teaching of Astronomy
Teachingof Astronomyin Brazil. R.H. Trevisan 79 Astronomy educationandtheIAU. M.Gerbaldi 87
Section3: Instrumentation
Improving polarization-sensitiveVLBI images with microwave-linked data transmission system. E.Llidke 97 Astronomyin space.
T. Villela, 105
Section 4: Galactic Astronomy and Interstellar Medium
Chemical composition in pholoionized nebulae: the problem of the temperature determination. R. Gruenwald 117 The interstellarextinction curve andthepolycyclic aromatichydrocarbon.
H.M.Boechat-Roherty,M.B.Fernandes,C.A.Lucas,M.L.RoccoandG.G.B.deSouza 127 Structureof interstellarmedium: a multivariate approach. J.R.Ducati, J.M.G. Fachel 133
Section5: Stellar Astrophysics
The southernhemisphereyoung stars survey . C.A.O. Torres and G.R. Quasi 139 Dynamicsin theequatorialplaneof Be stars. F.X. de Araújo 149 The upper H-R diagram. A. Damineli Neto 159
Section 6: Solar System and Astrometry
Nongravitationalforcesin present andpast solar system.
R.S.Gomes 171 Voyagersimagesrevealingthe structures of theplanetaryrings. S.Giuliatti Winter 181 Is the solardiametervariable?. N.V. Leister, M. Emílio, P.C.R.Poppe 191 FOREWORD
This volume containsthe proceedingsof the XXI1'1Meeting of the Brazilian Astronomical Society ("Sociedade AstronómicaBrasileira" - SAB), held at Hotel Glória, Caxambu, MG, Brazil, on July 31 - August4, 1995.
Only the invited talks given at the conferenceare included, but the meeting also had a large number of poster presentations,whoseabstracts,as usual, have been publishedin a special numberof theBulletin of theBrazilian AstronomicalSociety(vol. 15, no. 1, 1995).
We had all the contributed papersin the form of posters, so the authorshad enoughtime to explain their work, while the talks with duration of 40 minutescovered broad subjectsand reachedmost of the audience.
The meetingwas very fruitful owing to the excellentcontributions of many invited speakers andto the largenumberof poster presentations.
We are very grateful to Fundaçãode Amparo à Pesquisado Estado de São Paulo (FAPESP), ConselhoNacional de DesenvolvimentoCientífico e Tecnológico(CNPq), Coordenadoria e AperfeiçoamentodePessoalde Nível Superior(CAPES), Fundaçãode Amparo à Pesquisado Estado do Rio de Janeiro(FAPERJ) and Fundaçãode Amparoà Pesquisado Estado do Rio Grandedo Sul (FAPERGS) for themeetingfinancial support. 00 fAPEfP We are also grateful to FAPESP for financingthis volumeof the Fundação de Amparo à Pesquisa proceedings. do Estado de São Paulo
FranciscoJablonski FlávioElizalde Laerte Sodré Jr. VeraJatenco-Pereira
Section1 Extragalactic Astronomy and Cosmology
Pi'ocwfcdinÿsof flic XXIth SAB .3 WHAT CAN WE LEARN FROM NEAR-INFRARED COLOURS OF SPIRAL GALAXIES? R. F. Peletier1 and M. Balcells2 (1) - Kapleyn AstronomicalInstitute Poslbus800, 9700 AY Cronin c/en, Netherlandsand Instituto dc Astrofísicade Canarias Via Lactca s/n, 88200La Laguna, Tenerife,Spain. (8) - Kapleyn AstronomicalInstitute Poslbus800, 9700 AV Groningen, Netherlands. Abstract Colours of spiral galaxiesgenerally are affected by the underlying old stellar population, younger stars, and extinction by dust. Old and young populations generallycan be disentangledusing a combination of blue and near-infrared colours. Extinction effects are very hard to take into account, except for galaxieswith special orientations. In this paper we give some results of one of the first modern studies of galactic bulges and disks in various optical and near-infrared bands, and its implications for the stellar populations of spirals. 1 Introduction
Whenone compares spiral galaxieswith ellipticals, one can notice that muchlessis known about the stellar populations of the former than of the latter, When one then looks at a. grand-designspiral, one immediately understandsthe reason for this. Spirals not only seem to have an old, underlying population, but many times they also contain young stars, as well as dust betweenthe stars. For ellipticals one can in generalget a goodidea of the met aliicity, just from an optical colour like B— V7.For spiralswith a combination of young and old stellar populations one colour is not sufficient to be able to separate them. Even a combination of severaloptical colours (e.g. U — V and B —V) is generallynot sufficient, sinceone cannot distinguish betweena lower metallicity of the old population, and a larger fraction of young stars. To overcome this problem one has to go to the fear- in frared. '1'ully cl al. (1982)and especiallyFrogel(1985)hasshown that the U —V vs. V —K diagram is a very useful tool for detectingyoung stellar populations in spirals. SinceV —l\ is a very red colour, it happensthat young stars do not contribute as much in V — l\ as in U V, and that for a given V — l\ the U — V of an old population is very differentfrom the U — V of a combination of old and young stars. On the other handFrogelnoticed that there is a considerablescatter in the U — V vs. V — K diagram lor spirals, much larger than for ellipticals, and that many spirals were redder than the reddestellipticals. Thesetwo effectsvery likely are due to ext inction by dust. Imagesof edge-ongalaxies, like NGC 801,showmany magnitudes of extinction by dust in the central regions of the 4 Proceedingsof the XXIrhSAB disk. Becauseof this, the integratedc:oloui;sof this galaxy are very red (Wirth & Shaw 1983). In Balcells & Peletier (1994,hereinafter called Paper I) we have shown that the colours of the stellar populations themselvesin theseobjects are not red, in generaleven bluer than thoseof ellipticals of the same size. Sincethat paper only discussesdata in the optical, the information presentedabout stellar populations in bulges was limited. For that reason we have obtained images in J and K for the sample, with a new two-dimensionaldetector with a resolution that is sufficientto spatially select regionsthat are not or almost not affected by extinction. The colours (optical and optical-infrared)of theseregions have been analyzed in the way describedby Frogel (1985), and in this paper we discussthe implications for stellar populations in spiral galaxies. In particular, we discussthe differenceÿ between bulges and disks.
2 Sample and observations
We have investigated a sample of early-type spiral galaxies,ranging from SO’sto She. These galaxiesgenerallyhave bulgesthat can be separatedrather easily from their sur¬ roundingdisk. Furthermore,except for the latest type, the amount of extinction by dust in these galaxiesis limited, so that the colours will still be able to contain information about the stellar populations. The galaxieshaveinclinations larger than 50°, and are the brightest galaxiesin a certain part of the sky. These two properties enable us to also on the basis of their morphology,and not just on their surfacebrightnessprofiles,-separate disk and bulge. The samplethat we observed were are galaxiesof Table 1 of Paper I, except for 2, which had declinations larger than 60°, and so could not be observedat UKXR.T.Optical data (presentedin that paper) were obtained for all galaxies. All galaxiesof the dnstfree subsamplewere observedboth in J and K. The other galaxies were all observedin A, but not necessarilyin J. The optical data consist of U,B, R and / surfacephotometry, obtainedin June, 1990, on the 2.5m INT telescopeat La Palma. The data have been describedin Paper 1. The pixelsize of the data was 0.549”, arid the effective seeingon the imageslies between1.2” and 1.6”. The imageswere taken underphotometric,conditions. The near-infrared data were observedin June, 199-1,at the 4m United Kingdom In¬ frared Telescopeat Hawaii, using IRCAM3 (Puxley V" * m ... r- v ...» dÿ'ms#' W- URK Composite U-R R-K Figure 1: Greyscaleimage and two colour maps of IC 1029,showingthe geometry and position of the dustlanesin a typical galaxy (a) andthe lack of colour differencesbetween bulgesand disks (b and c). effectiveseeingon the final frameswas between 0.8” and 1.0”. Here alsothe frameswere takenunder photometric conditions,with maximum zero point errors of 0.1 mag in J and K. In Fig. 1 we show for a typical galaxy a composite U —R —K map, and two colour maps. Theseshow the bulge and disks,and the regions with the largest extinction and star formation. In Fig. 2 we show a cut along the minor axis in the U —R vs. R —K diagram. One side of the galaxy is well behaved,and U —R is increasingfor increasing R —Ar,but on the other sidethe combination of extinction, scattering and star formation makesthis diagram very hard to interpret. 3 Dust in spiral galaxies Colourscontain useful information about stellar populations only when they have been corrected for extinction. Sincethe effectsof extinction can be severe, the errors intro¬ ducedby correcting for it are often so large that it is better to chooseregions of the galaxy that are not, or very little, affected by extinction. In paper I we describethe procedure we employ to find extinction-free regionsof bulges.For disks the situation is more complicated. Since dust and stars here are mixed almosteverywhere,one cannot find dustfreeregions. From a colour map (e.g.Fig. 1) one can see that the extinction in theseearly-type spirals is almost all concentratedon one side of the galaxy. This is the case for all galaxiesthat are not seen close to edge-on. One minimizes the extinction if one measures the disk colourson the other side. Sincewe are interestedin radial and not vertical disk profiles we havemeasuredthe coloursin wedges,at 15° from the major axis, with a width of 10°. For this procedure to work the disk itself should contain relatively little extinction, 6 ProceedingsoITIio XXI11'SAB IC 1029 • • • 3.5 o/ •• n 2.5 s* Dustfree side • Other side 2.5 3 3.5 U - R Figure 2: Color-colordiagramof a minor axis cut of IC 1029. Both sidesare indicated by different symbols. sinceotherwise both sidesare severelyaffected. A wa.y to measure the extinction in the disk is to measure its radial colour gradientbetweena band that is affectedby extinction and one that is not, e.g. B and K (Peletier et al. 1995).If the densityratio of dust and stars is more or lessconstant the centralextinction will bemuchlarger than the extinction in the outer parts, purely becauseof the differencein stellar density. In dusty galaxies this will cause the scalelengths in B to be much larger than those in I\ . In Fig. 3 we plot scalelength ratios betweenB andK as a function of galaxy type. The open squares are points from Peletier et al. (1994). In that paper also details are given about how the radial scalelengths have been determined. Fig. 3 is very instructive. One can see that the scalelength ratios for galaxy types up to 2 (Sab) are almost, never larger than 1.4. Scalelength ratios for galaxiesof type 6 a.ndlarger alsoare small. Only galaxiesof type 3-5 are sometimes very dusty. This is in good agreementwith the surface-brightnessvs. inclination test (Valentijn1990). An independentestimate for the scalelength ratios due to stellar population gradientsis f.l - 1.2 (Peletier el al. 1995). This means that disks of galaxiesof type 2 or smallerare likely not to be very dusty, and so the effectsof dust in this paper will in generalnot be very large. We may be reasonablyconfident to use the colours for stellar population measurements. 4 Stellar populations from dustfree colours In PaperI a discussionis given of optical (U —R, B —R and R —l) coloursand colour gradientsas a function of otherparameters like bulgeluminosity. It was found that about half the bulgeshave the same, colour as elliptical galaxiesof the same luminosity, and that the other half is bluer in U —R.and B — R. This was interpreted as a sign of the Proceedingsof l.lio XXI11'SAB 7 • This sample PVMF 94 2 o 1 5 ! ! ° D s ié -2 2 Galaxy type Figure 3: Scale, length ratios between13and K ot‘ galaxy disks as a function of galaxy type. presence of younger stars, under the assumption that all ellipticals are old (15 17 Gyr). Now, with the combination of optical and near- infrared colours,we don’t have to make that assumption any more and can directly determine the average age of the stars. First we present, in Fig. 4, some colour-colourdiagramsfor the bulges.30 galaxiesare plotted, at i’eir/2,or 5”, if reir < 10”, but only 20 in the diagramin which J is included, since10 galaxies were not observedin J. Also included are lines of singleburst models of constant age, with metallicity varying. Thesemodels are from Vazdekiset al. (1995) describedelsewherein theseproceedings.The modelsare madeby converting theoretical parameters like effective temperature and gravity to colours,choosingan initial mass functionand integrating along a theoreticalHR diagram, similar to Peletier (1989)and Worthey (1994). In this case however,we have taken a lot of care with the temperature- metallicity-colour calibration that the integrated models fit the data for giant elliptical galaxies.For example,for a given B —V the V —K givenby Worthey (.1994)is much too red to fit the giant elliptical galaxies.In Fig. 4 modelsare plotted for agesof 17, 12, 8, 4 and 1 Gyr. Alongthe lines metallicity ranges from Z=0.05 to Z=0.0004. Onecan see that only the diagramsthat contain optical and optical-infrarecl colourscan separate age and metallicity. — and are both sensitive more or lessthe same so R K J —K to kind of stars, in this diagrammetallicity and age cannot be separated.The same holds for the U —R vs. B —R.diagram. The systematic error in the model colours,due to uncertaintiesin the stellarevolution theory, or lack of template stars, seems to be on the order of 0.1 or 0.2mag, although this valuecan be smaller,since the reddest bulgesnot necessarilyhave to be 17 Gyrs old. Despite theseerrors some galaxiesin the U —R vs. R —K diagram cannot be fitted by a model of 17 Gyr, and for some the bestfitting model only hasan age of 1 Gyr. This doesnot mean that thereis no underlyingold stellar population present, 8 Proceedingsof the lXI*h SAB but just that a much younger stellar population is dominant. We see that most of the reddest,and also largest (PaperI) bulges are old, and that some have to be younger. It is alsonoteworthy that all bulgeshereare redder than R —K = 2.4. This means that the metallicily of all thesebulgesis larger than about 0.3 Z0.This is probably due to the way the sample was chosen- bright nearby galaxies,automatically excluding all dwarfs,and galaxieswith low metallicities. We now compare the colours of bulgeswith those of disks. Even though both bulges and disks are generallybluer with increasingradius (cleJong 1995;Paper I), the gradients are smallenoughthat we can assignrepresentativevaluesfor the colouroftaaahcomponent. For bulges,we have taken the colour at 0.5 r ejj or at 5 arcsec, whichever is larger. For disks we use the colours at 2 major axis scalelengths. We find that disk colours,while somewhat bluer, are very similar to bulge colours for all the galaxies(seeFig. 5). Here the diagonalline indicates the locus whereboth colours are equal. The averagedifferences betweendisk and bulge and bulge colours are 0.126 ± 0.165 for U —R, 0.045 ± 0.097 for B — R, 0.078± 0.165 for R — K and 0.016± 0.087for J —K. Unlessthere is a conspiracy betweenmetallicity and age, this diagramimplies that age differencesbetweenbulgesand disks are small. Our result is in disagreementwith Bothun & Gregg (1990) who for a sample of SOgalaxies,using near-infrared aperture photometry, found that for a given J-K disks were on the average 0.4 - 0.5 mag bluer in B — H than bulges. We claim that the errors in their measurements are muchlarger than they claim. In our data, the range in colours betweenthe galaxiesis much larger than the averagecolour differencebetween bulgeand disk of the same galaxy. Accordingto the modelsa bulgeis at most 30%older than its disk, or much less,if metallicity gradients are taken into account. What can we learn from this about the formation of spiral galaxies?The data of this paper put the strong constraint that disks of early-type spirals must have been formed together with, or just after the bulge. It does not endanger the traditional model of Eggen,Lynden-Bell & Sandage(1962) of galaxy formation, but on the other hand implies that it is very unlikely that there were three discrete events of massive star formation: for the bulge, the disk, and the thick disk. The properties of bulge and inner disk seem to merge smoothly in terms of for examplekinema.tical properties and kind of stars. The continuous infall model of Gunn (1982),wherethe age of the stars is determined by the free-fall collapsetime of the infalling gas, would produce such a, system, since these time scaleswill be similar for bulge and inner disk. An alternative way of forming bulgeshas beenproposedby Gombesel al. (1.990)and others. They form a bulge through the formation,and later desctructionof a bar due to dynamicalinstabillities. This methoddoesnot predict any age differences,and is in good agreement with our results here. However, galaxieswith large bulgeslike the Sombrero cannot be formed in this way, and also one would not expect a continuous changein colours,colour gradient and surfacebrightness profile shapefrom large bulgesof SO’sto small Sb bulges(Andredakis et al. 1995). Apart from the colours themselveswe also have measuredsome colour gradients. In Fig. 6 a plot is given of R —K colour gradientsin the bulgesas a function of total A'-band magnitude. SinceR — K is an indicator of old stellar populations measuring the bulge Proceedingsof the XXIth SAB 9 gradientsin this colour is a. good way to measure the metallicity gradients in bulges. Two different symbols are used for our galaxies.The filled circles indicate the galaxiesof the dustfreesample of Paper I. The filed squares are thosethat may contain some dust. There are no significant differencesto be seen betweenthe R—K gradientsof both groups, showingthat dust extinction is not a major factor influencing the gradients.The gradients are somewhatlarger than the visual-infraredgradientsof elliptical galaxies(Peletier et al. 1990,Silva& Elston 1994). Our averageR - K gradientis A(R - K )/A(logr) = -0.232, and for U —R -0.399. For ellipticals average valuesare -0.14 for R — K and -0.20 for U—R, although not much data are availablefor R —K. In Paper I we concludedfrom the similarity between gradientsin the optical betweenbulgesand disks that this is a strong argument to show that bulgesand ellipticals formedin the same way, and that the subsequentdisk-formation did not substantially affect the bulge. We can say the same now usingthe R —K colour. A recent review by Minniti et al. (1995)showsthat possibly also the bulge of our own galaxy has a similarly small colour/metallicitygradient, even though earlier measurementsby Terndrup (1988)were indicating large values.Fisher et al. (1996)find that for a sampleof SOgalaxiesthe vertical gradientsin thebulgein B —R and in absorptionline strength are much larger than the radial gradients. Although we don’t have many galaxiesin our samplethat are really edge-onwe alsosee hints for the same effect, and this might be the reason of Terndrup (1988)’slarge metallicity gradient in our Bulge. Using colour - metallicity conversionsby Vazclekiset al. (1995)we have convertedour average R —K and U —R gradientsto metallicity gradients,assuming constant age. From R — K we find that on the average A(logZ) = -0.41 per dex in radius,and from U —R we find a valueof -0.42. Also plotted in Fig. 6 is the sampleof Terndrup et al. (1994)of similar kinds of galaxies.Even though almost all R —K gradientsof our sampleare negative,thoseof Terndrup et al. can be both positive and negative. We presume that the errors in that sampleare larger than indicated by the errorbars,somethingwhich is possible,sincethey useda small detectorcoveringa small field. As a final point we would like to talk a bit about colours of disks as a function of galaxy type. Here the colours are the ones that have been determined in such a way as to minimize the effectsof extinction (seeabove).4 different colours are plotted in Fig. 7. The J — K and R —K colours here are indicators of the old underlying stellar population that constitutes most of the mass. B — R andU — R alsodependon additional young stars. Noteworthy is the fact that there is not much changein the colour, when varying galaxy type. One can say that ellipticals, SO’sand spiralsup to type 2 (Sab’s)have more or lessthe same colours,indicating few young stars and similar metallicities. Somegalaxies of type 3 obviously suffer from extinction effects,while in all colours,but especiallyin U — R galaxiesof type 4 (Sbc’s)are bluer than the others. This figure showsagain that a considerablefraction of the stars in Sbc’sand probably alsolater type spiralsis young. I 10 Proceedings of Lhe XXI h SAII T1- 2 T 'a? /. A? • op 3 2 .* " - m i m 1-5 w "ÿ 5* K 1.5 oz Sjt 1 3 CQ I I ., I i I .., 1 .... 2 2.5 3 2 2.5 3 R - K R - K T T T T T1" A? A? 1 2 r 2 _Ofl m m 1 1.5 Í2 I / 0.8 “ 3 w .1 1 1 1.1 1 1 . 1 ... i. 0 6 L 1 1.2 1.4 1.6 1.8 2 2.5 3 B - R R - K Figure 4: Colour-colourrelations for the bulges. Also drawn are lines of constant age by Vazdekiset nl. (1995). Plotted are ages17, 12, 8, 4 and 1 Gyr. Bulge 1 1.5 2 2.5 OB 1 1 2 T /ÿ 1 r i 1.2 U - R J - K 2 A 1 1.5 • 0 8 1 % .‘eruni, a + i T +7 1.8 - B - R / - R - K 3 1.6 1.4 - 2.5 1.2 /m m 1 0.B x | x X x 1 1 2 0.0 1 1 2 1.4 1 6 1 8 2 2.5 3 Bulge Figure 5: Disk coloursas a function of bulge,colours H Proceedingsof the XXIthSAB 11 • Dustfree Dusty x Terndrup etal. 0 5 1 S 0 § -0.5 -ií-r-í -1 -22 -23 -24 -25 -26 -27 Figure 6: R —K gradientsfor the bulges. Also plotted are the data of Terndrup et al. (1994) 5 Conclusions The most important results that havebeendiscussedin this paper are: • By investigating colour gradientsin disks betweenB and K for a sample of 70 galaxieswe find that galaxiesof type 3-5 are muchmore affectedby extinction than are others. In the galaxiesof other types, it is possibleto find regionsthat are only slightly affectedby extinction, and for which the colourscan beusedto study stellar populations. • Usingcolour-colourdiagramsof optical andoptical-infraredcolours,and comparing them with single age, singlemetallicity stellar population models,we find that a considerablefraction of the bulgesin the samplehas to be younger than elliptical galaxies,or contain a large fraction of stars that are younger than thosein elliptical galaxies. The differencesin colour from one galaxy to another can be much larger than the colourdifferencebetweenthe bulge and disk of the same galaxy. For that reason the inner disksof thesegalaxiescannot be more than 2-4 Gyrs younger than their bulges. • RadialR —K colour gradientsof bulgesare negative and small. Bulgesgenerally are redderin the center than in their outer parts, and the average colour gradient is about 0.2 mag per dex. These valuesare due to metallicity, or possibly age gradients, but not due to' extinction by dust. Metallicity gradientsinferred from R —K gradients are very similar to those calculatedfrom U —i?,,indicating that the contribution from very young stars in thesegalaxiesis small. 12 Proceedingsof the XXIthSAB 3 3 £ 2.5 p-rrjTTTjT 0) TT W _ "w 1.2 •ã 5 1 ==“ Í2 1-5 - : : . T3ÿ 3 0.8 s". K 1 - a i liu! l 0 6 -2 0 2 4 -2 0 2 4 Type Type S 2 1 1 1 1 l ' I 1.8 " •a* 1.6 14 _! m 2.5 1-2 3 - 0£ 1 2 i 0.B * iii i.t.ii r 0 2 4 0 4 03 -2 OS -2 2 Type Type Figure 7: Disk coloursas a. function of galaxy type. • Dustfree colours in disks are remarkably constant. Only loi galaxiesof type 3 and later disks are significantly bluer tlia.ii those ol SB’s. Acknowledgment s We acknowledgethe help Massimo Stiavelli lor help wiili observationsand Alejandro Vazdekismaking availablehis stellar population models before p 111>Ii References Andredakis,Y.C., Peletier,IFF. G.Baleells.M.. 1995. .\I\IF\S 275. ST1 Balcells,M. G Peleiier, H.F., 191)1, AJ 107, 135 (Paper I) Bothun,G.D. G Gregg,M., 1990, ApJ 3-70.73 Combes,F., Debbaseh,F., Friedli, I). G Pfenniger. I).. 1990. AGA 233. >2 de Jong, R.S., 1995,Ph.l). Thesis, llniv. of Groningen Eggen,()., Lyndeu-Bell, I). G.Sanda.ge,A.IF, 1902.ApJ 130. 7IS Fisher,D., Franx, M. G Illingworth, (}., 1990. ApJ. in press. Frogel,J.A., 1985, ApJ 298, 528 Gunn, J.E., 1982, in Astrophysical ( 'osutoloyi/.ed. II.A. Brink. C.Y. Coyim. and iVI.S. Longair; Vatican City: Pontifica Academia Sricni ianim. p. 233 Proceedingsof the XXIUlSAB 13 Minitti, D., Olszewski,E., Liebert, J., White, S.D.M.,Hill, J.M. & Irwin, M.J., 1995, ESOpreprint 1088 Peletier,R.F., 1989,Ph.D. Thesis,Univ. of Groningen Peletier,R.F., Valentijn, E.A. & Jameson,R.F, 1990,A&A 233,62 Peletier,R.F., Valentijn, E.A., Moorwood,À.F.M. & Freudling,W., 1994, A&A Suppl. 108,621. Peletier,R.F., Valentijn, E.A., Moorwood,A. & Freudling,W., Knapen, J.H. & Beckman, J.E., 1995, A&A 300,Ll Puxley, P. &; Aspin, C., 1994,Spectrum2, 4 Silva,D,R. & Elston,R., 1994,APJ 428,511 Terndrup,D.M., 1988,AJ 96, 886 Terndrup,D.M., Davies,R.L., Frogel,J.A., Depoy, D.L. & Wells,L.A., 1994, ApJ 432, 518 Tully, R.B., Mould, J.R. & Aaronson,M., 1982,ApJ 257,527. Valentijn, E.A., 1990,Nature 346,153 Vazdekis,A., Casuso,E., Peletier, R.F. & Beckman,J.E., 1995,in preparation Wirth, A. & Shaw,R., 1983,AJ 88, 171 Worthey, G., 1994, ApJ Suppl 95, 107 Proceedingsof the XXI1*SAB 15 UV SPECTRA OF EARLY-TYPE GALAXIES1 Charles J. Bonatto*, Eduardo Bica*, Miriani G. Pastoriza* and Danielle Alloin** (*) - Departamento de Astronomia, IF-UFRGS, CP15051, Porto Alegre91501-970, RS, Brazil (**) - Observatoirede Paris/Meudon,URA 173 CNRS, F-92195 MeudonPrincipal Cede.x,France Abstract We analyze the UV spectra of the early-type galaxiesin the IUE library, includ¬ ing both normal and active nuclei. We coaddedthem into groups of similar spectral properties in the UV, also taking into account those in the visible/near-infrared ranges. This procedure provided spectra with a muchhigher signal/noiseratio than in previous studies,which as a rule were basedon individual spectra and therefrom derivedcolour indices. As a consequenceof the coaddingprocedure,information on spectral features can be accessedand analyzed. We obtained several spectral groups of galaxiescharacterizedby red and blue stellarpopulations. The red stellar population groupspresent a far-UV flux with varying intensities. It was possible to infer on the extinction law affecting some internally reddenedgalaxies;there are casesin which the presence of the À2200Â absorption feature suggests reddening similar to the Galactic law, and others without the A2200Âfeature,suggestingthe SMClaw. 1 Introduction The International Ultraviolet Explorer (IUE) Satellite has been collecting fundamental data on a wide variety of astronomicalobjectssince1978. Recently we haveusedthe IUE library to study the spectralproperties of a large sampleof star cluster spectra with an unprecedentedsignal to noise (S/N) ratio by coadclingthem into groups. The grouping procedure allowed us to analyzein detail the far-UV properties of star clusters, with reliableinformation on absorption features. The resultsof this analysisare presentedin Bonatto et ah (1995,hereafterPaperI). In what concerns galaxies,severalstudiesof IUE spectra by object classhave already beencarried out, e.g. the spectralatlasesof Kinney,et al. (1993)for star-forming galaxies, and of Rosaet al. (1984)for ITII complexes. In this work we gatherthe spectra of theearly-typegalaxiesstoredin theIUE database and analyzetogether the ensemblesof normal and active early-type galaxies,in view of betterunderstandingtheir far-UV spectral properties. We also use in the interpretation single-agedUV templatesof stellar populations, i.e. the star clusterspectralgroups from PaperI, which considerablyconstrain the conclusions. 1basedupon datacollected with the International Ultraviolet Explorer (IUE) Satellite, supported by NASA, SER.Cand ESA 16 Proceedingsof the XXIthSAB Someearly-typegalaxies(in this work,ellipticals andS0’s) are known to emit consid¬ erableamounts of energy in the far-UV, producing the so-calledUV turnup (e.g.Bertola et al. 1980;Okeet al. 1981).However,the origin and characteristicsof this emissionare not yet quite well understood.Sofar, studiesof theseobjects haveemployedindividual IUE spectra which usually have a rather low (S/N) ratio, or are basedon colour indices (Burstein et al. 1988; Longo et al. 1991). Bica & Alloin (1988) have used interpolated continuum distribution. The different spectra of early-type galaxies,in addition to the red stellar populations with varying amounts of UV turnup, also present evidenceof bursts of star formation (e.g.Burstein et al. 1988). The analysesof IUE spectra of Active Galactic Nuclei (AGNs) in early-type galaxies havebeenmostly focussedon variability, e.g. NGC3516(Wanderset al. 1993)and Mk 279 (Stirpeet al. 1994). Our goalhere is to study average properties by coaddingspectra of a given AGN, then classifying and coaddingthem into groups of comparable spectral characteristics. So we work with a time averageof the activity states in which the nuclei were observed. 2 Star clusters spectra In order to interpret composite stellar populations, and in particular galaxiesat high redshift, it is necessary to considerboth age and metallicity as essentialparameters. Observationally,one shouldcover a spectralrange as wide as possible:young populations will showup better in the ultraviolet (UV) whereasold components are conspicuousin the visible and infrared (IR) ranges. Galaxies present indeed a wide variety of stellar populations,ranging from those dominated by old components to those with recent star formation events (seefor a review e.g. Bica 1992). The IUE library contains a wealth of information on different classesof objects. In the present study we use all the availableintegrated spectra in the IUE library of star clustersin the Galaxy, LMC, SMC and Fornax, as well as HII regions in M101, M33, LMC and SMC, for a comprehensiveanalysisof their ultraviolet spectral properties as a function of age and metallicity. The approachis similar to previous analysesin the near-UV, visible and near-IR ranges (Bica, Alloin & Schmitt 1.994,and Bica & Alloin 1987),i.e. by defining spectral windowsand continuum tracings for equivalent widths (W) measurements. As the signalto noiseratio (S/N) in individual spectra is in general moderate we decided to group them accordingto similarities in the object class,age and metallicity after information in the literature and, as the ultimate criterion,the UV spectral properties giventhe availableS/Nratio. The following steps were taken to build the final spectralgroups: (i) we averaged individual spectra of a given object working separatelyin the short (SWP) and long (LWPand LWR) wavelengthranges. A few spectra were eliminatedat this stagebecause they presentedproblemssuchas instrumental defects,discordantcontinuum distribution and/orspectralfeatures;(n) using the information on the object class,age andmetallicity, we averagedthe spectra of different objects — the ultimate criterion being the spectral Proceedingsof the XXIthSAB 17 i o I—1 1I Regions Q 6 M101 [ZÿoJ— 0.20 M33 *w~v — (zyznj— o.oo 2 fseÿ-Kol— o-i 200 1 "700 2200 2700 3200 ? A-(A) Figure 1: M 101, M33, SMC and LMC H II region spectral groups; metallicities are indicated;Flux in FAunits normalized at À2646A, while constants have been added to the spectra for clarity purposes I -4 'V' oung and Intermediate Ago O luste ra 1 2 v Ago- 1 2Myr [Z/Ze]— 1 .O *—ÿ***«4 V'-**A — 'w*>V v«.w V V“v N/ 1 o — V’ r KVK 'l\, JV'* V LMCJ A(j»»12Myr [Z/ZJ— 0.20 II Ag« 2BMyr [Z/Z0]— 0.3 8 — .....- Ao— 7BMyr 'Z0] —o.zo e NW\VVA-V -4 2 ° 200 1 "700 2200MA) 2700 3200 Figure 2: LMC and SMC young and intermediateage star cluster spectral groups; ages and metallicities are indicated;units as in Fig. L 18 Proceedingsof Lhe XXIth SAB 1 o Globular Olusters e 6 kTTua [Z/ZQ]— O.TO —— \/A yÿ_ÿr->iÿvvÿdAyVWvJ [z/z0] 1 — 2 “'ÿ'-U -vÿ \T ° 200 2200 2-700---3200 Figure 3: Globular cluster spectral groups; ages are I (I (»'///'and melalliciticsare indi- cated;units as in Fig. 1 similarity; and (m) the short and long wa.velengthdomains were finally connected. The weightsin the averages were given accordingto the square of the S/Nratio. We show in Figure 1 the HII region spectra, of the four groups; the young and inter¬ mediate age cluster groups are in Figure 2; and in Figure 3 we show the globular clusters groups. 3 Early-type galaxies The spectra have been extracted from the 1UE databaseas stored at the Instituto As¬ tronómico e Geofísicoof the Universidadede SãoPaulo (IAG-USP). We have selected the availableSWP,LWP andLWR spectra of early-type galaxiesobtained with the large aperture (10"x20") mode. Galaxiespresent a wide range of spectral properties, but thosepresentingsiinihirities can be grouped, a procedurewhich allowsone to study their average;properties in more detail with better (S/N) ratio data. In order to study the UV spectral li in detail it is necessary first to coaddthe availableIUE spectra of an individual object, andsubsequently coaddinto spectralgroups thoseof different objects whichresult similar. For the.present grouping procedurethe ultimate criterion is Similarity of UV spectra within theavailably (S/N) ratio, but vm also consideredtheir spectral properties in the visible and near- III. ranges found in the-literature, together with their classificationsas AGN, 11II galaxy, Starburst,as well as usualnormal stellar population types. The spectral groups were built in a similar way as the Star cluster groups. We endedup with 8 groups characterizedby a llil Stellarpopulation (Figure 4) and 5 groups with an AGN or a blue stellar population (Figure 5). Froglings of Lli<-XXI"' SAI? 19 12.0 9.0 LI— 8 v/'ÿrfvV’ s 6.0 cc 3.0 G_N1553 G _N221 I FV\G ~ Mgl F«l 0.0 Fqll _ Mgll _ 1200- 1700 2200 2700 3200 X(A) Figure 4: Groups for red stellar populationsin early-type galaxieswith varying contribu¬ tion from the UV turnup. Flux in F,\ units normalized at A2646Â,while constants have been added to the spectra for clarity purposes, except for the lowestone 1 0.0 CIV w Q-O !gf IT Q _Mk279 (Soy1) G_N351G /\ | (Sey1 > S cS? °'°200 -I "700 2200 2-700 3200 A. (A) Figure 5: Groups for active nuclei together with those for blue stellar populations in early-type,galaxies. Notice how the two groups of SeyfertIs have similar emission lines but differin the continuum slope. Flux in F.\units normalizedat A2646Â,while constants l lave, beenaddedto the spectra for clarity purposes, except for the lowestone 20 Proceedingsof the XXIthSAB 30 O SWBV Q.NS102 r° WBIVA• NQC5253 J 10 NGC1705V \A • GN15I0 r RHli 0 47Tuc G2b G_N15|3 O o 30 G_N221• % G3r G_N4849 f G4r„ G.N3115 ° G_N4472 20 G3b G_N4408 % Q4b° 10 G5 ° | • G_N3998 0 0 5 10 15 20 25 30 W(Mgl12g52) Figure 6: Equivalent widths of MgI and MgII for galaxyand/orgroups (filled symbols). Upper panel: blue stellar population groups; lower panel: red stellar population groups. Ws for star cluster templates (opensymbols)are alsoplotted as comparison. Groupsof strong LINERs, G-N3998and GJM4486,have W(MgII) diluted by emission We selectedtwo strong absorption featuresin the LWP/Rrange (MgIIAA2796,2803and MglA2852)and plotted their Ws in Figure 6, in order to investigatethe basicproperties of UV lines for galaxiesand to compare them to the Ws for star clustersfrom PaperI. In the upper panel of Fig. 6 we show W(MgII,U2796,2803)x W(MgL\2852)for the blue stellar population groups and individual galaxies. They occupy similar loci as the blue star clusters SWBI (age« lOMyr) to SWBIVA (age« 200Myr), and the template RHII which is an average of ITII regionsin M101, M33, LMC and SMC. The loci of the HII/ amorphousgalaxiesNGC1705,NGC5253 and the group GJM1510suggest that they are not pure II II regions,but contain a range in age. The Ws for the red stellar population galaxy groups are compared to those of the globular cluster templatesin the lower panel. For the globular clusters,the sequence from G5 to G 2b (and 47Tuc) increasesin metallicity up to [Z/ZQ] PS —0.5(PaperI). As expected,the galaxy groups are closeto the metal-rich globularclusters.In MgI they present larger Ws indicating higher metallicity, but in MgII Ws are comparable,which might be related to a saturation effect in the galaxies,the presence of some emissionand/orsome interstellar absorptionexcess in the line of sight to the globularclusters.Notice that the groups G_N3998 and G-N4486, which are strong LINERS, havesignificantly smaller W(Mg IIAA2796,2803)with respect to the other galaxy groups. This is certainly causedby a contribution of Mg IIAA2796,2803in emission. Proceedingsof the XXIth SAB 21 5.0 4.0 WA G_Mk279 (b) G_N3516 (SMC law. E(B-V),=0. 1 O) LiT*3.0 CO 2.0 (a) G _ N351 6 (Galactic law, E(B-V),=0.15) 1 .o o. 1200 1700 2200 2700 3200 X(A) Figure 7: The redder Seyfert1 group GJM3516lias beenreddeningcorrectedusing (a) a Galactic extinction law with E(B—V)=0.15, and (b) the SMC law with E(B—V)=0.10. As comparedto the bluer Seyfert1 group GJVlk279,the Galactic law underestimatesthe far-UV flux and overestimates the A2200Ã correction, whereasthe SMC law produces a very similar spectrum. Fluxes are in F,\ units normalized at A2646Â,while constants havebeenadded to the spectra for clarity purposes, except for the lowestone. The strong CIV emissionhas been clipped for clarity 4 Seyfert galaxies Weobtainedtwo groups of Seyfert1 galaxies:the spectrum of G_Mk279 is considerably bluer than that of G-N35I6. Otherwise,equivalentwidths and FWHM of the emission lines do not present systematic differences.In principle, three possibilitiesmight explain the continuum differencebetweenboth groups: (i) reddeningdifference;(ii) the galaxiesin the two groups wouldbe in different averageactivity states; and (in) the G_Mk279 group might contain a starburst component. Weappliedseveraltests in order to constrain these possibilities. If differential internal reddeningis responsiblefor the effect,the reddening law must not contain a strong A2200Âbump. This can be seen in Fig. 7 sincethe SMC extinctionlaw (Prévotet al. 1984)with E(B—V)=0.10 applied to the G-N3516spectrum reproducesquite well that of G_Mk279, whereasSeaton’s(1979)law with E(B—V)=0.15 alreadyoverestimates the A2200Ã correction while, at the same time, it underestimates the far-UV flux. Assumingthat emission lines are comparablein both groups of Seyfert1 galaxies, we normalizedthe spectra to equal Mg IIA.\279G,2803and obtained the differencespectrum G_Mk279 — G-N3516.In Fig. 8 we compare the resulting spectrum to that of the typical nuclearstarburst NGC5236, and also to the power-law F,\ oc A-1,8,which provides the 22 I’rocmlings of the XXIth SAIi 3.0 Vi I? . 5 M. f GMK279 G_N35 I6 2.0 - LJ_d CL> A/y 1 .5 I ct A. ' Vn 1 .o / °"l SUM NGC5236 Roll 0.5 civ 0.0 200 1 700 2200 2700 3200 A-(A) Figure.8: The difference sped rum betweenthe bluet (G_Mk279) and redder (G.N3flJo) SeyfertI groups is compared to a power-law continuum and to the nuclear starburst NGC5236. Fluxes are in F.\ units normalized at A26-I6A, while constants have been added to the spectra, for clarity purposes, except Forthe lowest one best fit to the continuum of the difference, spectrum. The power-law fits well the difference continuum, as alsodoesthe starburst continuum. In the latter case, however, the starburst absorption lines are not conclusively present in the difference spectrum. On the other hand, both a power-law and a. starburst continuum with such steep slope would cause a considerable dilution in the equivalent widths of the emission lines of G-MU279with respect, to those of CLN3516,which is not the ca.se. We concludethat internal reddeningappears to be*the cause of the continuum differ ence between the two Seyfert I groups. Within the scenarioof the Unijii d Modd ofAdNs the nuclei of the galaxiesin the group O.N35K) would be seen at. lines of sight which cross larger amounts of the dusty torus ( Antonueei1993). If this is the case, our results from Fig. 7 would suggest that the extinction law in the dusty tori doesnot have a conspicuous A2200Ã bump, resembling more that of the SMC. In Fig. 9 we compare the SWP region of the Seyfert2 group G_Mk573 with that of the Seyfert2 prototype NGC 1068.It is worth noting that individual IFF spectra in the (J_Mk573 group-are very noisy and do not contain much information. We assumedthat they were Seyfert2 galaxiesrelying on visible spectra. The resulting 1.1V spectrum of the group confirms the Seylert2 nature of its membersbecauseol the similarity to the spectrum of NGC 1068. The FWHM for the ensembleof the ('mission linesreinforcesthis Proceedingsof the XXIthSAB 23 12.0 s S' 1 0.0 0.0 5 gj> 0.0 M 5i O _ Mk573 4.0 2.0 0.0 NOC 1 060 200 1 400 1 MA)OOO 1 OOO 2000 Figure 9: The SWP region of the Seyfert2 group G_Mk573 is compared to that of NOSCl 068; emission lines in G_Mk573 have similar width and relative intensities as in NGC 1068. Fluxesare in F,\ units normalized at A2646Ã,while a constant was added to the upper spectrum lor clarity purposes conclusion. In terms of the equivalent width, those of the group GJVIk573are system¬ atically larger than those of NGG1068, which suggests a higher excitation and/orless contam in,ation by a stellar continuum component in the former. 5 Blue groups Wecompare in Fig. 10 the spectrum of the group G-N1510(averageof HII and/oramor¬ phousgalaxies)with that of NGG5253. As expected,the spectra are very similar. Emis¬ sion lines are not characteristic features of the UV spectrum of fill galaxies. We also showin Fig. 10 the HIT region template from PaperI (including SMC,LMC, M 33 and M 101.IIU regions), the star cluster templates SWBI (age ~ J0Myr) and SWBII (age ~25Myr). 'Thus, the presence of strong emission lines in the UV spectra of galaxies, e.g. Mg CI,\.\2796,2808,suchas in those of Seyfert nuclei is not related to conventional stai¬ lorination,indicating the presenceof activity. The different slopebetweenthe H II region template and both ITII galaxy spectra might be explainedby more internal reddening in the latter two. If this is the case, the extinction law would necessarilybe similar to that of the SMC becauseof the absenceof the A2200À absorption in the H II galaxies. Alternatively,the slope can also be changedby a combination of the H II template with other,recent star formationtemplates,e.g. SWBII. In the latter case, suchgalaxieswould have'been forming stars in the last cubOMvr. \Ae compare in Fig. fl the spectrum <§f the group G-N5102 with those of the- star 24 Proceedingsof the XXIth SAB o.o NQC5253 e.o / > 4.0 SWB (1 OMyr) 2.0 SWBII (25Myr) "~~V~ °.o 200 2200 2700 3200 >-CA) Figure 10: The spectrum of the group G-N1510is compared to that of NGC5253 and to those of H II region and star cluster templates of different ages. Fluxes are in F,\ units normalized at A2646Â,while constants havebeen addedto the spectra for clarity purposes, except for the lowestone s.o -<*-o SWB 11I (75Myr) C3-0 2.0 SWBIVA (200Myr) I 1 -O SWBV <1 .S2Qyr) °'°200 1 "700 2200 2700 3200 Figure 11: The spectrum of the blue population group G-N5102is compared to thoseof star cluster templates at different ages. Fluxesare in FAunits, normalisedat A2646Ã. Constantshavebeenaddedto the spectra for clarity purposes, except for the bottom one Proceedingsof the XXIthSAB 25 cluster templates SWBIII (age «75 Myr), SWBIVA (age«200 Myr) and SWBV (age « 1.2 Gyr). Two membersof this galaxy group, NGC5102 and the M31 companion NGC205,were population-synthesizedin the visible and near-IR ranges with a star clus¬ ter basein Bica (1988,NGC5102 therein referred to as the template E8) and Bica et al. (1990),respectively. They concluded that these galaxieshad a strong burst of star formation at ages t « 100 —500Myr. In the UV (Fig. 11)it is confirmed that the cluster template SWBIVA («200 Myr) must be a major contributor to the G-N5102group. 6 Red groups At this moment we are workingon the groups characterizedby the red stellar population. Wehope that the resulting (S/N) ratio is enoughto allow us to find the origin of the UV turnup observedin some spectra of early-type galaxies. Acknowledgments We thank CNPq and FINEP for partially supporting this work;C.B. thanks FAPERGS for coveringtravel expenses to the XXI SAB meeting. References Antonucci,R.R.J., 1993. ARA&A , 31, 473. Bertola,F., Capaccioli,M., Holm, A.V. & Oke,J.B., 1980. ApJL, 237, L65. Bica,E., Alloin, D., 1987. A&A, 186, 49. Bica,E. & Alloin,D., 1988. A&A, 192, 98. Bica,E., Alloin, D. & Schmidt,A.A., 1990. A&A, 228, 23. Bica,E., 1992. in “The Stellar PopulationsofGalaxies”,eds.B. Barbuy , A. Renzini, IAU Symp. 149, p.215. Bica,E., Alloin, D., Schmitt,H.R., 1994. A&A, 283, 805. Bonatto,C., Bica,E. & Alloin, D., 1995. A&ASS,112, 71. Burstein,D., Bertola, F., Buson,L.M., Faber,S.M. & Lauer, T.R., 1988. ApJ, 328, 440. Kinney,A.L., Bohlin, R.C., Calzetti,D., Panagia,N., Wyse,R.F.G., 1993. ApJS, 86, 5. Longo,G., Capaccioli, M. & Ceriello,A., 1991. A&AS,90, 375. Oke,J.B., Bertola,F. h Capaccioli,M., 1981. ApJ, 243, 453. 26 Proceedingsof the XXIth SAB Rosa,M., Joubert,M. & Benvenutti,P., 1984. A&AS, 57, 361. Seaton,M.J., 1979. MNRAS, 187, 73p. Stirpe, G.M. et al. , 1994. A&A, 285, 857. Wanders,I. et al. , 1993. A&A, 269, 39. Proceedingsof I lie XXI lh SAB 27 REDSHIFT SURVEYS OF FIELD GALAXIES Christopher N.A. Willmer ( 'NPq/Obscrvalório Nacional, llua GeneralJose Crislino 77, Rio dt Janeiro, RJ, 20921-030. Brazil ( -mail: enawJvoii.br 1 Introduction hi this contribution I will present some of the main results obtained from redshift surveys of galaxies. Because of space constraints, the coverage will be at best superficial, and omissions,inevitable. For more completereviewsthe readeris veferedto worksby Sandage (1988) for an overview of observationaltests of world models;Hubble (193G),Mill alash Binney (f9Sf) for reviews of the early work; (Jiovanelli Haynes (1991), where many of the main results of redshift surveys are reviewed;Koo N.Kron (1992) who deal with the evolution of faint galaxiesand Peebles(1993), who shows the inter-relation between theory and observations. A few definitions should be explained before proceedingfurther. The first refers to field galaxies. As noted by Koo A:Kron (1992), these are objects which are selected without regard to their environment {t.g., galaxiesin clusters), or to some special char¬ acteristics (e.g., X-ray emission), in general, the samplesof field galaxiesare selected through criteria such as the observedflux in a given wavelengthdomain, for which a selectionfunction can be easily calculated. Redshift surveys of field galaxiesaim at measuringgalaxiesin a volumethat constitutes a fair sample, such that its statistical propertiesmay be consideredas representative of the propertiesof the Universeas a whole (Peebles J980). The surveys can be carried out usingeither full or sparse sampling. Fully sampledsurveys aim at observingall galaxies to a givenlimit, while sparse samplesuse a.random sub-sampleof the catalogueof objects. The latter allowsprobing a larger volume of spacefor the same amount of telescopetime whencomparedto fully sampledsurveys. Another advantageof sparse samplingis that it can reducethe noisedue to the discretenature' of the galaxydistribution whenmeasuring statistics such as the spatial two-point correlation function (Kaiser J9S6). On the other hand, one losesinformation of certain higherorder characteristics of the distribution, even when using an optimal sparse sampling rate (Szapudi N.Szalay 1995). 2 First results Both counts of galaxies and redshift surveys had an irnense boost in the beginningof lis century with the use.' ol photographic plates, and in the 1970’sand 1980’s,with the developmentof solid-statedetectors. Of the eaiier work, we will concentrate on two results. 28 Proceedingsof Lhe XXI11'SAD The first is the demonstration by Hubble (1926) that the number counts of galaxiesin the magnituderange mpa= 8.5 - 16.7,obeyedthe following power-law Log(N(m)) = 0.6m + constant (1) when averagedover many directions. This value is also quite close to what would be expectedin the case of a Euclidean Universe,and suggestedthat averagedover very large scales,the distribution of galaxiesis homogeneous.This also meant that different from stars, there was no “edge” in the distribution of galaxies,so that these objects could be used to trace properties of the Universe. The secondwas the demonstrationby Hubble (1929)of the tight correlationbetweenthe radial velocity and distance,the latter measuredusing distance estimators such as cepheidsand brightest galaxiesin clusters. This relation provided observationalevidencesupportive of Friedmann’s(1922)solution for the field equations of GeneralRelativity Theory, where an expanding Universe was predicted. In the 1930’sgalaxy counts were carriedout coveringlarge extents of the sky in both hemispheres,and in smaller regions,but to much fainter magnitudes using the largest available telescopes(e. 3 The first complete redshift surveys In spiteof the limitations of photographicplates, the catalogueof bright galaxiescovering both celestialhemispherescompiledby Shapleyand Ames(1932)was systematicallyused to measure redshifts and the complete sample was finally publishedby Sandage and Tammann (1981,RSA). Becausethis sampleis dominated by galaxiesbelongingto the Local Supercluster,its use for the.characterization of large-scaleproperties ol the galaxy distribution is somewhatrestricted. For instance,the first determinationsol the redshift- Proceedingsof the XXIth SAB 29 spacetwo-point correlation function by Davis, (Seller & Huchra (1978)using a catalogue with a similar depth as the RSA, produced a correlation length of r0=2.4h_1 Mpc for the galaxiesin the north galactichemisphereand ro=3.6h-1 Mpc in the southern galactic hemisphere(Peebles1980). The first survey to probe beyond the Local Superclusterwas the CfAl by Davis et al. (1982),who useda sampleof 2401galaxiesderived from the CGGG,and showedthat galaxiesform a complexdistribution wherean alternation betweenregionsalmost devoid of galaxiesand clusters and lower density filaments was clearly seen. However,because the Local superclusterand galaxiesbelongingto Coma are contained within this sample, it was possiblethat some physicalparameters that were estimatedusing this sample,such as the mean volume density of galaxies,could be biased.This promptedthe realization of surveys in other directionsof the sky. The southern counterpart of the CfAl, the SSRSl (da Costa et al. 1988), probed a region of the sky were few nearby (andrich) clusters of galaxieswere found, and thus was more representativeof what a typical “field” sample would be. Other important surveys using other wavelengthbands are those using the catalogueof sources detectedby the IRAS satellite to definegalaxy samples(Strauss et al. 1990;Fisher et al. 1995;Saundersel al. 1991)and the surveys which measure the the neutral hydrogen emission at 21 cm (e. t(r) = (ro/fP (2) with i'o ~ 5 h_I Mpc and 7 ~ 1.8, whencalculatedfor all the nearbygalaxy samples.The CÍA2 and SSRS2are deepenoughthat a meaningful estimate of the three-dimensional power-spectrum can be made. If galaxies are good tracers of the matter distribution, the power spectrum should in principle provide information on the fluctuations of the distribution of mass, which can be directly compared with model predictions. For the combinedCTA2-SSRS2it is found that the power-spectrum risesmonotonicallyfrom small scalesup to ~ 200h-1 Mpc, showingonly weaksignsfor a turnover beyond that distance 30 Proceedingsof (.lie XXIthSAP (da Costa1995). Anotherimportant statistic is the luminosity function, which may he parameterized using the Schechter(1976) form: (L)dL = /,///) (3) which,when expressedin magnitudesbecomes: A/*-J\V)(a+I ) The luminosity function hasbeeninvestigatedfor both the CfA2 (Marzke at al. 1994) and SSItS2 (da Costa ei al. 1994). T.he.Schechterparameters differ betweenboth works, while the SSRS2parajneters are consistent with thosemeasuredby deepersurveys (Loveday cat al. 1992;Vettolani el al. 1994). The inconsistenciesfound for the C1A2samplecan be due to a variety of reasons, which could be either systematic errors in the CCCC magnitudes or even incompletenessin the CCCC. The luminosity function when calculated for different morphological types showsthat the.Schechterparameters are generally similar, except in the case of irregular galaxies whosefaint-end slope is much steeper (Marzke cl al. 1994). The SSRS2databasewas combinedwith the ESOSurfacePhotometry Catalogue (Lauberts fc VaJentiju1989) to study the behaviour of the luminosity function with colour. A preliminary analysis carried out by da Costa,et al. (1995) has shown that blue galaxiestend to have a faint-end slope similar to that detected for-Irregular galaxies,suggestingthat a.nearby population orfblue galaxiesis a.possiblecontributor to the faint galaxy counts (da Costa 1995). An issue which is still, not well quantified, refers to tire frequency of Active Calactic Nuclei,and how the presence of this phenomenoncould he related to local environment. The frequency of AGNs lias been estimated to be 2% by Maia el al. (1.995), who — inspected spectra belonging to the combinedSSRSI and SSRS2database. 4 Redshift surveys using sparse sampling While the ESO/SRCsurvey plates were beingtaken,mucheffort was placedon developing fast platemeasuringmachinesthat couldhandle the largeamount of data generatedfrom suchscans. The UKSTplateswere digitized and analysedusingobjective methodsby two independentmachines,the ARM (Kibblewhite d al. 198-1)and COSMOS(MacGillivray & Stobie1984). The APM survey (Maddox el. al. 1990) contains galaxiesclown to b,/=20 for the region 8 <17.5° and b<--40° . A total of about 2 million galaxieswere detected and provided a sample with angular coverage and depth similar to the galaxy counts of Shaneand Wirtanen (1967). A bright (bj <17.15) subsampleof 1769galaxies was selectedto carry out a redshift survey usinga 1 in 20 samplingrate, which constitutesthe Stromlo-APMsurvey (Loveday et al. 1992). The galaxieswere visually inspectedand had their morphologicaltypes classi¬ fied. The characteristic depth of this survey is about 12000kins-1(the CIA2 and SSRS2 Proceedingsof the XXIth SAB 31 have a. characteristicdepth of ~ 7000 kins-1 ). The luminosity function for this sam¬ ple, presents the following Schechterparameters: ci— -0.97±0.15; M* 19.50i0.I3 and 0* = L4O±O.17x}_O-2h-3 Mpc3,and in this calculationthe of Efstathiou,El¬ lis & Peterson ( 1988) were taken into account. Although theseresults were not confirmed by Marzke.at al. (1994) for the CfA‘2,both the shallowerSSRS2(da Costa et al. 1994) and the deeperESO SliceProject (Vettolani el al. 1994) present a similar determination as Loveday el al. (1992). These authors also found that the luminosity parameters for early and late type galaxiesare different. Other results from this survey refer to the clusteringproperties of galaxieswhen selectedby morphologyand luminosity, using the two-point correlation function both in real and reclshift-space(Loveday et al. 1994). The correlation function is quite different whencalculatedin real andreclshift space, the latter usually being flatter. Theseauthors confirm earlier findingsby Davis & Geller (1976)and Giovanelli,Haynes& Chincarini (1986) that early type galaxiespresent a stronger corre¬ lation than later types whencalculatedin real space. On the other hand,in reclshift space, the correlation of early types is much flatter, even when comparedto late-type galaxies, indicating that “early-type galaxiessuffer from enhancedspace distortions”. They also find that lower-luminosity galaxies(L 5 Surveys of faint galaxies Severalsmall areas on the sky were explored using plates taken at the prime-focus of 4m-classtelescopesat the AAT, KPNO and CTIO, and used to make number counts to very faint limiting magnitudes(B~22). At the faint end the number counts are above the expectedvalue for no-evolution (NE) models,the discrepancybeing usually larger in the blue counts. This result implied that significant evolution could have occurred at low reclshift, such that faint galaxieswere more luminous and bluer clueto a higher stellar formation rate. rLoaddressthis question,Broaclhurst,Ellis & Shanks(1988,BES) designeda reclshift survey of galaxiesselectedin the magnitude interval and band where the greatest discrepanciesbetween the observedand predicted NE counts were found. About 230 galaxieswith 20.0 Finally, a survey carried out with the Caua-da-Prance-1la waii Telescope (hilly el al. 1995, and referencestherein) has produced results regarding the properties of the hi- Proceedingsof the XXIthSAD 33 minosity function to z~l for a sampleof galaxiesselectedin the range 17.5 6 Works in progress A follow-up of the CfA2 is beingcarried out by Geller et al. (1995), which will measure redshifts for all galaxiesin a 100° x Io slice centeredat 5=29° . The samplewas defined from photographic photometry of POSSI E plates down to a limiting magnitude of R=16.4. This work showsa distribution wherethe repetition of structures is clearly seen, therebeing no structures larger than the characteristic depth of the survey. Kirshner,Oemler,Schechterand Shectman(1987)had carried out severalpencil-beam surveys.closeto the galacticpolesusingsparse samplingthrough which they discovereda largeunderclenseregion,originally estimated to havea sizeof ~ 80 h-1 Mpc in the Bootes constellation. More recently, theseauthors designeda survey, againusingsparse sampling, in the southernsky, using a sampledefined from CCD photometry, whereslabsof 1.5° by 1.5° are observedthrough drift scans (he., the CCDis read at the same rate as the earth’s rotation) and usedto derivea specroscopicsample,which is observedusing a multi-fibre system installed on the 2.5 m telescopeat the Las CampanasObservatory. This survey aimsmeasuringredshiftsfor about 30000galaxieswithin the magnitude range of 16.0 < r < 17.5. The samplingrate is 6 galaxiesout of 10, and dueto the cut in surfacebrightness, the efficiency (he., gathering a measurablespectrum) is ~ 90%. The distribution of galaxiesshowsmany structures with scales< 50 h-1 Mpc and preliminary results from this survey show that about 20%of the galaxiespresent emissionlines. The two-point correlation function measuredin this survey showsresults similar to those detectedin shallowersurveys. However,whenseparatedbetweenspectralcharacteristics(presenceor not of emissionlines) they find that at small separationsgalaxieswith emissionlines tend to have a lower correlation than galaxieswithout emissionlines (Shectmanet al. 1992), suggestingthat galaxiesin denserenvironmentscould haveundergonegas depletion. A survey attaining a similar depth as KOSSis the ESO Slice Project (Vettolani et al. 1994), probing a 22° x Io slice,limited at bj=19.4, at <5=-40°. This survey uses a sampleof galaxiesoriginatedfrom the Edinburgh-DurhamCOSMOScatalogue(Heydon- Diunbleton et al. 1989). Preliminary resultsfrom this survey indicate that the luminosity function parameters agree with those measuredby Loveday et al. 1992. They alsodetect 34 Proceedingsof t lie XXI1'1SAB some of the southernpeaksof BEKS, though the periodicity is not as strong. A somewhatdeeperworkis the ESO-Sculptorsurvey of deLapparent andcollaborators (Bellanger et al. 1995;Arnouts tt al. 1995)with the aim of characterizingthe large-scale propertiesof the galaxy distribution out to depth of z~0.5. The catalogueof this survey is derivedfrom CCD photometry obtainedin B,V and R in a slice of 1.75° x 0.2° . The number counts in the three coloursshowevidenceof evolution, while the galaxy colours in B-R and B-V showa blueing trend of ~ 0.5magnitudesfor 21.0 7 Concluding remarks Redshift surveys have been the one of the most important means of characterizingthe statistical properties of galaxiesin the nearbyUniverse. Not only havethey haveallowed a quasi-three-dimensionalview of the distribution of galaxies,but more importantly, from the statistical analysesit becomespossibleto placeobservationalconstraints which must be satisfied by successfulmodels for the formation a.ndevolution of galaxiesand large- scalestructures. We could thus summarizesome of the main results from theseworks that we havepresentedabove: 1. Galaxiestend to concentrate on low-projecteddensity structures which surround large (< 50 h-1 Mpc) regionswithin which few bright galaxiesare found. 2. The two-point correlation function presents discrepancieswhenmeasuredin real and redshift space. 3. The scalesizewherethe correlationfunction becomes1 is roughly the same both from optical as well as infrared samples,and correspondsto ~ 5 h-1 Mpc. 4. The two-point correlationfunction presents slight differencesdependingon what kind of sampleis used, fie., different valuesare measuredwhen galaxiesare divided between morphologicaltypes, luminosity classes,colours or presence of emissionlines. 5. The luminosity function parameters measuredfor the nearbysamples(z<0.1) present similar values.When calculatedfor different types and morphologiesit is found that the Schechterparameters are fairly similar for all morphologiesexcept for irregular galaxies and blue galaxieswhich present steeper slopes. 6. The deepersurveys showthat reddergalaxiespresent almost no evidencefor evolution sincez ~ 1, while blue galaxiespresent evidenceof significantevolution sincez> 0.5. 8 References - Abell,G.O. 1958,ApJS,3, 211 - Arnouts, S., de Lapparent, V., Mathez, G., Mazure, A., Mellier. V., Berlin, E., A: Kruszewski,A. 1995, AA, submitted - Bellanger,C., de Lapparent, V., Arnouts,S., Mathez, G., Mazure, A., A:Mellier, Y. 1.995,AASS,110,159 Proceedingsof lhe XXI1'1SAB 35 - Broadhurst,T.J., Ellis, R.S.,k Shanks,T. 1988,MNRAS, 235, 827 (BES) - Broadhurst,T.C., Ellis, R.S., Koo, D.C., k Szalay.A.S.1990,Nature,343,726(BEKS) - Colless,M., Ellis R.S.,Taylor, K., k Hook,R. 1990,MNRAS, 244,408 - Colless,M., Ellis R.S., Taylor,K., k Shaw,G. 1991,253,686 - Colless,M., Ellis R.S., Taylor, K., Broadhurst,T.J., k Peterson,B.A. 1993,MNRAS, 261,19 - da Costa,L.N., Marzke,R.O., Pellegrini, P.S.,k Santarosa,G. 1995,in preparation - da Costa,L.N., Pellegrini, P.S., Sargent,W.L.W, Tonry, J., Davis, M., Meiksin A., Latham,D.W., Menzies,J.W., and Coulson,I. A. 1988,ApJ 327,544 - da Costa, L.N., Geller, M.J., Pellegrini, P.S., Latham, D.W., Fairall, A.P., Marzke, R.O., Willmer, C.N.A., Huchra,J.P., Calderon,J.H., Ramella,M., k Kurtz, M.J. 1994, ApJ, 424,LI - da Costa, L.N., 1995, to appear in “Clustering in the Universe”, Proc, of XXXth Moriond Conference,Les Arcs” - de Lapparent, V., Geller,M.J., and Huchra,J.P. 1986, ApJ 302,LI - de Lapparent, V., Geller,M.J., k Huchra,J.P, 1991, ApJ, 369, 273 - Davis,M., k Geller,M.J. 1976,ApJ, 208,13 - Davis,M., Geller,M.J., k Huchra,J.P.1978,ApJ, 221,1 - Davis,M., Huchra,J.P., Latham, D.W., k Tonry, J. 1982,ApJ, 253,423 - Efstathiou,G., Ellis, R.S.,k Peterson,B.A. 1988,MNRAS, 232,431 - Ellman, N. 1994,Pli.D. Thesis,University of California at SantaCruz - Friedmann,A. 1922,Z. Phys., 10, 337, - Fisher,K.B., Huchra, J.P., Strauss,M.A., Davis, M., Yahil, A. k Schlegel,D. 1995, ApjS, in press - Geller,M.J., k Huchra,J.P. 1989,Science,246, 897 - Geller, M.J., et a.l. 1995,in preparation - Glazebrook, K., Ellis, R.S., Colless,M., Broadhurst,T.C., Allington-Smith, J., k Tanvir, N. 1995a,MNRAS, 273,157 - Glazebrook,K., Peacock,J.A., Miller, L., k Collins,C.A. 1995b,MNRAS,in press - Giovanelli,R., Haynes,M.P. 1991, ARAA, 29, 499 - Giovanelli,R., Haynes,M.P., k Chincarini, G. 1986,ApJ, 300,77 - Groth, E.J., and Peebles,P.J.E. 1977,ApJ 217, 385 - Heydon-Dumbleton,Collins,C.A., k MacGillivray, II. T. 1989,MNRAS, 238,379 - Hubble,E. 1926, ApJ, 64, 321 - Hubble,E. 1929, Proc. Nat. Acacl.15, 168 - Hubble,E. 1936, “The Realmof the Nebulae”,(New Haven: Yale University Press) - Kaiser,N. 1986, MNRAS, 219, 785 - Kibblewhite,E.J., Bridgeland,M.T., Bunclark, P. k Irwin, M.J. 1984,in Astronomical MicrodensitometryConference,NASA CP 2317,277 - Kirshner,R.P., Oemler,Jr., A., Schechter,P.L., and Schectman,S.A. 1987,ApJ 314, - Koo, D.C., Ellman, N., Kron, R.G., Munn, J.A., Szalay, A.S., Broadhurst, T.J., k Ellis, R.S. 1993, ASP Conf. Ser. Vol. 51, ObservationalCosmologySymposium, eel. G. 36 Proceedingsof the XXIthSAB Chincarini,A. Iovino, T. Maccacaro,k D. Maccagni(SanFrancisco: ASP), 112 - Koo, D.C., k Kron, R. 1992, ARAA, 30, 613 - Lauberts, A., and Valentijn, E.A. 1989, “The SurfacePhotometry Catalogueof the ESO-UppsalaGalaxies”,(Garching-bei-Muncheri:EuropeanSouthern Observatory) - Lilly, S.J.,Tresse,L., Hammer,F., Crampton,D., & Le Fèvre,0. 1995,preprint - Loveday,J., Peterson,B.A., Efstathiou, G., k Maddox, S.J. 1992, ApJ, 390,338 - Loveday,J., Maddox,S.J.,Efstathiou, G., & Peterson,B.A. 1994,Preprint - MacGillivray, H.T., k Stobie,1984,Vistas Astr., 27, 433 - Maddox,S.J.,Sutherland,W.J., Efstathiou,G., k Loveday,J. 1990,MNRAS, 243,692 - Maia, M.A.G. 1993,Proc. of the XIXth SAB Meeting, ed. J. Braga, B. Barbuy, N.V. Leister,p. 157 - Maia, M.A.G., Suzuki,J.A., da Costa,L.N., Willmer, C.N.A.,k Rite, C. 1995,AASS, submitted - Marzke,R.O., Geller,M.J., Huchra,J.P.,k Corwin,H.G. 1994,AJ, 108,437 - Metcalfe,N., Shanks,T., Fong,R., k Roche,N. 1995,MNRAS, 273,257 - Mihalas,D. k Binney, J. 1991,“Galactic Astronomy”, (NewYork: W.H. Freeman and Company) - Peebles,P.J.E. 1980,“The Large-ScaleStructure of the Universe”,(Princeton:Prince¬ ton University Press) - Peebles,P.J.E. 1993, “Principles of Physical Cosmology”, (Princeton: Princeton Uni¬ versity Press) - Ramella,M., Geller,M.J., k LIuchra,J.P. 1992,ApJ, 384, 396 - Sandage,A. 1988,ARAA, 26,561 - Sandage,A., k Tammann,G.A. 1981, “A RevisedShapley-AmesCatalog of Bright Galaxies”,(Washington:CarnegieInstitution of Washington,Publ. 635)(RSA) - Saunders,W., et al. 1991,Nature,349,32 - Schechter,P. 1976,ApJ, 203,297 - Seldner,M., Siebers,B., Groth, E.J., k Peebles,P.J.E. 1977,AJ, 82, 249 - Shane,C.D., Wirtanen, C.A., 1967,Publ. Lick Obs.,XXII, Part 1 - Shapley,H., and Ames,A.A. 1932,Harvard Ann., 88, N° 2 - Shectman,S. et al. 1992, Clustersand Superclustersof Galaxies,ed. A. C. Fabian (Dordrecht:Kluwer) - Strauss,M.A., Davis,M., Yahil, A., and Huchra,J.P. 1990, ApJ 361,49 - Szalay,A.S. 1987,in “Large ScaleStructures in the Universe”,ed. L. Martinet, M. Mayor, (Sauverny:GenevaObservatory) - Szapudi,L, k Szalay,A.S. 1995,preprint - Vettolani, G., et al. 1994, to apppear in “Wide Field Spectroscopyand the Distant Universe”,ed. S.J. Maddox,A. Aragon-Salamanca,(Singapore:World Scientific) - Willmer, C.N.A., Koo, D.C., Szalay,A.S.,k Kurtz, M.J. 1994, ApJ, 437, 560 - Zwicky,F., Herzog,E., Karpowicz, M., and Kowal, C. 1961-1968,Catalog of Galaxies and Clustersof Galaxies,(Pasadena:California Institute of Technology),6 vols (CGCG) Proceedingsof the XXIth SAB 37 MAGNETIC RECONNECTION WITHIN GALACTIC AND EXTRAGALACTIC STAR FORMING REGIONS Luiz Carlos Jafelice Departamento deFísica Teórica e Experimental UFRN - CCE - CampusUniversitário C.P.: 1641- CEP 59072-970Natal,RN - BRAZIL PHONE: 55-84-23195 86 branch: 27 - FAX: 55-84-23197 49 email: [email protected] Abstract We discussthe relevanceof the mechanismof magneticreconnection within star forming regions. The role played by sucha mechanismin the process of star forma¬ tion and the observational consequencesare briefly discussedfor both interstellar and intracluster environments. 1 Magnetic reconnection Magnetic reconnection (MR) is a well known process of conversionof magnetic energy into other forms of energy (e.g. kinetic, thermal,of plasma waves,etc) (Biskamp1994). What characterizesthe process of MR is the topologicalchangeit causes in the magnetic field. This aspect is very important in the treatment of transport processes in magnetized plasmas,particularly for collisionlessplasmasas thoseobservedin the interstellar medium and the intracluster medium of galaxies. Althoughsucha process is well knownin laboratoryandastrophysicalplasmas,recent theoretical and experimental results involving tridimensionaltreatment of the problem (Biskamp1994)haverevealedthat the previousknowledgeis still quite incipient. Such results,however,clearly indicate that magnetic energy dissipationrates (or MR rates) are independentof the value of the collisional resistivity and are basically associatedwith the typical Alfvén time scale. Contrary to the first ID and 2D studieson MR, the main different and relevantresults obtained recently come from the use of numericalsimulations (Biskamp1994 and referencestherein) and on the realizationof the importance of taking adequately into account the boundary conditions when dealing with the MR physics (Priest& Forbes1992). The results of the works by Costa & Jafelice1995,Friaça & Jafelice1995,Jafelice & Friaça 1995a and Jafelice L Friaça.1995b are briefly commented on below. 2 Star formation and magnetic fields The process of star formation is intimately dependenton the presence of magnetic fields. That process will proceeddifferently in what concerns, e.g., its rate and the shapeof the initial mass function,whether the magnetic field is present or not. 38 Proceedingsof the XXIth SAB In the scenarioof star formation process usually adopted, one can point out the fol¬ lowingrelevancesof magneticfield features: 1. Magnetic fields,in general,are important in defining (Mestel 1990; Rees.1994): (a) The Jeansmass; and (b) The efficiency of angulai1momentum transfer. 2. Magnetic reconnection,in particular, is important to account for (McKee et al, 1992): (a) Topologicalchangesin the magnetic field; (b) Magnetic field annihilation;and (c) The internal structure of molecularclouds. Topologicalchangeis essentialfor star formation. A molecular cloud forms out of diffuse interstellar medium with a magnetic field which is well connectedto the field of that medium. Nevertheless,when the star forms out of the molecularcloud it has a field essentially detachedfrom the interstellar field (McKeeet al. 1992). Interstellar protostars condensewith too muchinitial magneticflux (liees 1994). Mag¬ netic fields are expectedto increasethe valueof the Jean’s mass and to make it more difficult for the protostar to contract. Furthermore, flux freezingconditions,even if just approximately valid, would imply in an excessof magnetic flux as the protostar con¬ tracts gravitationally. Protostars must get rid of such an excessiveflux, and magnetic reconnectionis a source of magnetic annihilation. Molecularcloudsare clumpy instead of homogeneous.Valuesfor the magnetic field of giant molecularcloudslower than the equipartition valuesadded to not so small clumps will imply in inevitable tangledmagnetic fields within the cloud becauseof the motions of the gas, which favours the reconnection,gradually detachingthe clumps from the interclump medium. Magnetic braking is expected to makean important contribution in transferring angular momentum from the forming star to the surrounding medium. However,such a process may be greatly modified in clumps wherethe field connection with the rest of the cloud is much more weak becauseof the ocurrence of reconnection (McKeeet al. 1.992). 3 MR within tlu' interstellar medium Costa 4 Condensations within cooling flows X-ray observationsof clustersof galaxieshaverevealed that the gas in the core of several clusterslias a cooling time shorter than a Hubble time, thus allowing to inter a cooling flow towardsthe center of cluster, where, interestingly, thereis alwaysa dominant galaxy (for reviews,see Fabian 199-1).Additionally, in somecooling flows,optical emission line filaments are seen around the central galaxy. Theseiilamentsare thought to be a.phase of ('solution of thermal instabilities arisingin the cooling flows. The optical filaments are the most outstanding optical signature of cooling flows. Severalmodels,consideringdifferent mechanismsof ionization and heatingof the gas - shocks,thermalconduction, and photoioilization by soft X-raysproducedin the cooling gas -- have met problems in explaining the observedline ratios and luminosities (see Jafelice& Friaca 1995afor further discussionand references). The possibility of MU as a source of excitation for such filamentshavebeengenerically suggestedby some authors (Johnstone h Fabian 1988; Sokerh Sarazin 1990; Tribble 1991).Jafeliceh Friaça1.995atreat the process of MR in those filamentsin greater detail through a. quantitative approachand the results are compared with the observations.In 1Tiaça h Jafelice1995 we consideranomalouseffectsassociatedwith MU in suchflows, as we summarizebelow. 4.1 MR in the Intracluster Plasma The plasma,of the intraclustermediumol galaxiesis collisionless,magnetizedand has,in general,ft value of flu >> 1 (where the ratio betweenthe plasma kinetic energy density 40 Proceedingsof the XXIthSAB and the magnetic energy density /3pi= STmekBTelB'2, with nebeing the electronnumber density, Tethe electron temperature, B the ambientmagneticfield,and kBthe Boltzmann constant).When we talk below of low- and high-/?regionswe are referring,respectively, to valuesof /?p/< 1 and (3pi> 1 within suchregions. This dimensionlessparameter allows to distinguish between thoseregionswherethe plasma dynamics is determinedmainly by magnetic forces(i.e., where/3pi« 1) and thosewherethe dynamicsis governedby hydrodynamic forces(i.e., where(3pi>> 1). Filamentswith opticalemissionare observedbasicallyin the inner parts of the clusters. In theseinner parts, correspondingto distancesfrom the clustercentre of up to ~ 20kpc, one expects to have f3pi~ 1. Sucha reduction of /?,,/from values>> 1 for the outer regionsto values~ 1 for the inner ones,is mainly becauseof the increasein the valueof the magnetic field which,frozen-in to the ambient plasma,is draggedby the plasmain its process of contraction towardsthe centre of the cluster (Soker& Sarazin1990). In the subsequentprocess of condensationand formation of the filaments (3pigets even smaller,and can reach,in principle, valuesquite smallerthan 1. The filamentsare expectedto be sites of low-/?on the groundsof: 1) the amplification of the magnetic field due to the compressionof the magnetic field frozen-in to the plasma as the gas condensesto form the filament; and 2) the decreaseof the thermal energy density due to the radiative lossestaking placein the densermedium of the forming filament. In the phaseof filament formation, the gas evolution is nearly isochoricand the latter reason is the most relevant for the reduction of /?p/.Anyway,the final result is that both processes contribute to make an ambientICM plasmain which f3pi~ 1 (for distancesto the cluster centre,say, < 20 kpc) easilyachievevaluesof f3pi« 1 within the filaments. Observationalbacking to the abovescenariocomes from the study by Ge (1991)on magneticfieldsin clustersof galaxieswith intensecoolingflows that clearly indicatesthat optical line emitting nebulaeare magneticallydominated,in some cases with (3vias low as ~ 10-2within the filaments(as,for example,for the clusterA1795). The MR process taking place within intracluster emitting nebulaeis a collisionless one. We define, as usual, the Alfvénic Mach number Me= VtjVAt- —1/InRme,which gives the reconnection rate of the MR process (whereRmeis the magnetic Reynolds number). For typical filament conditionsone obtains Rme~ 1,3 x 1019Me, which gives us for the conditions within intracluster filaments the value Me~0.02,whichis in very good agreement with usual valuesfor the reconnection rate expectedto be found in astrophysicalplasmas(Priest 1982). In Jafelice& Friaça 1995athe evolution of the coolingfilamentsis obtainedby solving the hydrodynamical equations of mass, momentum and energy conservation(seeFriaça 1993).The cooling function and the coefficientsof collisionalionization,recombination and chargeexchangeof the ionization equations were all calculated with the atomic databaseof the photoionizationcodeAANGABA (Gruenwald& Viegas1992). It seems that MR cannot be the dominantmechanismpowering line emissionin cool¬ ing flows,as it underproducesHa and [NII]A6583.However,it could be an important ingredient to explain the emissioncoming from low ionization lines,such as [01]A6300, constituting an additional component acting togetherwith other mechanismsproducing Proceedingsof the XXIUlSAB 41 most of the emissionin II rv and higherionization lines (suchas [Nil] À65S3). On the other hand,the study of MR-basedmodelsdemonstratesthat MR heatingconstitutes an alter¬ native to shocksas a source of [01] emissionfor moderatelybright filament systems. MR, however,is not sufficient to power not even the [0/]A6300emissionin the most luminous systems, like Perseus.In this case, the shockmodel representsa best candidateto explain the optical emission. In a consistent picture of the evolution of condensatesin the intracluster medium,MR, allowingthe condensatesto losemagneticfield,is a necessary link in the formation of cold, densecloudsthat are neededto account for the excess X-ray absorption, star formation and presence of dust in cooling flows. As suggestedabove,it is likely that other processes are corroborating with MR in the energizationof intracluster filaments. In this sense we are beginning to study quantitatively the possiblerelevanceof Alfvén waves in such an energization (Friaça et al. 1995). The idea is to compare numerical calculations with observationsconsideringthree generalsituations: cases in which reconnectionacts alone, thosein which Alfvén waves act alone,and thosein which both processes are acting at the sametime. A generalconclusionfrom the study in Jafelice& Friaça 1995a.is that in the formation of condiinsatesout of the cooling flow, large amounts of energy are locked in magnetic fields,and, as the MR is neededto get rid of surplusmagnetic pressure, it has a role to play in the energeticsof cooling flows. Furthermore,estimatesmadein that work strongly suggest that the physical scenario for MR in intracluster nebulae is expected to be very rich in plasma effects,where a wealthof plasma instabilities may be ruling important physical processes. In particular, plasma,effectsmay be governing the magnetic energy consumption, heat conduction and the resulting emission signature emergingfrom these nebulae. The physics and efficiency of the MR process as well as observationalconsequencesin this case can not be properly studied from a magnetohydrodynamicapproach.A specifictheoreticalapproachto treat collisionlessreconnectionin the presence of plasmainstabilities will be done elsewhere. 4.2 Other Consequences of MR in Cooling Flows Friaça f| Jafelice1995 treat, on the one hand, anomalouseffects(followingJafelice 1992) in processes involving the formation andenergizationof filaments within coolingflowsand, on the other hand, the effect of energeticelectronsaccelerateddue to the reconnection process. Tfio.seelectrons, with energiesin the range 1 to 10 keV, may play an important role in the production of forbidden emissionlines for ions like Oil 1 and heavier. In that study the authorscompare line ratios (like, e.g., [0 111]/[OII\ versus [OIII]/ H(3) with the cases they calculatein whichenergeticelectrons generatedby the reconnectionare absentand thosein which they are present. At the moment there is not yet any conclusiveresult; the expectation is to obtain some signature and observationalpredictions from that kind of study. Anomalouseffects are studied aiming lo addressthe fate of the mass which is removed 42 Proceedingsof the XXI*1'SAP (i.e., which condenses)from cooling flows. The flow rale can be very high (up to some hundredsof solar masses per year), neverthelessthe last, destiny of the removedmass is not known (Fabian 1994). On the one hand plasmainstabilities have a great influence in the process of MR which occurs in the collisionl&S plasmaswithin thoseflows(as shown by Jafelice & Friaça 1995a), and such plasmaeffectsmay haveimportant implications also at a macroscopiclevel (as shownin Jafelice1992). On the other hand, the reconuec- tion process is expectedto have a fundamentalimporta.ncein the star formation process (McKee et al. 1992; also the first resultsin Costa fe Jafelice 1995). Therefore, the in¬ terrelationship betweenthoseprocessesis expectedto have important implication in the formation of stars within the intracluster medium. In that work the authors addressthe study of stellar formation in that medium which, in many aspects, have quite different physical conditions from those found in Star forming regions within galactic interstellar media (which increasesthe interest in that kind of study). In Jafelice & Friaça 1995bthe authorsdiscussthe possiblerelationship between the optical signature of the cooling flows and the cluster evolutionary history. They also propose a scenario to take into account the destiny of the mass removedfrom the cooling flow and they make some observationalpredictions. It follows from such a scenario that one must expect that much more filaments are present within the flows than thoseobserved,even for cooling flowswhereoptical filaments have already been detected. In their model the association between the evolutionaries histories ol the cluster as a whole and ol the cluster magnetic field topology is crucial to explain the presence or absenceof optical filaments. Their analysis indicate that the anti-correlation between the rotation measure decre¬ ment and the optical emission observedin clusters with high cooling flow rates suggests that the presenceof laded (i.e.., “invisible”) filaments shondbe searchedthrough rotation measur e observation s. The scenario proposedlead to a consistent picture to the fate of the mass removed from the cooling flow. The final destiny of that mass is naturally either faded filaments (which may aboundin most coolingflows) or a coolingflow population of mainly low-mass dim objects. MR, allowing the condensatesto lose magnetic field, is a. necessary link in the formation of cold, denseclouds that are neededto account for the excess of X-ray absorption, star formation and the presence of dust in coolingflows. 5 Final comments on magnetic fields in astrophysics Thereis an important point to be emphasizedwhich the recent researchon astrophysics is makingclearerin the last years: 'Flic magnetic field importance in many astrophysica.1 phenomena,is not necessarilybecause, of its intensity. Originally, mainly becauseof the lack of observationaldata concerningmagnetic fields, the natural difficulty to deal with magnetic fields in analytical studies, and because,in severalsituations the magnetic field was not dynamically relevant, its influenceswqre underestimatedand the magnetic field was simply excluded from the calculations. (Note, Proceedingsof the XXIth SAB 43 however,that those were not the only real motives for that posture; some prejudice or overlookingwas present in some cases.) That procedure is not elsejustifiable in a plenty of astrophysicalcontexts. In many cases the magnetic field is stronger than initially assumed,but that is not the more unexpectednew result. Even dynamically weakmagneticfields revealedto be significant becauseof their influence on microscopicprocesses,like thermal conductivity and other transport processes (Jafelice 1992;Rees1994). Magnetic reconnectionis particularly important in this latter aspect becauseit natu¬ rally implies unexpectedinterrelationships(and many yet unpredictableones!)between: 1. Local and global features of a given system through topologicalchangesassociated with the reconnection;and between 2. Microscopic and macroscopic featuresof the system through the effects of the re¬ connection on transport and energeticprocesses. Transport properties, like thermal and electricalconductions,proceedmuch more ef¬ ficiently along magnetic field lines than across them. This is particularly conspicuousfor collisionlessplasmas as those present in interstellar and intracluster media. On the one hand, reconnection can make regions with quite different physical conditions (like, e.g., temperature, density, chemical composition, etc.), which were initially separated,to be connected;on the other hand it can make a condensationwhich was cooling, for example, through the conduction alongmagneticfield lines to becomesubstanciallythermally insu¬ lated from the surrounding medium. In either case the evolutionary implications for the wholesystem may be quite different than if reconnection was not taking place. (Someim¬ plications of this scenariofor the formation of condensationsand stars in the intracluster medium of galaxiesis discussedin Jafelice& Friaça 1995b.) As one can see, the future work involving magnetized astrophysical media is very stimulating and has many important open questions to be tackled. Magnetic fields may be directly relevant,through their dynamic contribution,or indirectly becausethey affect transport processes, energy input (through MR) and/ortopologicalchanges. In short, magnetic fields can establish unexpected connections between local processes and global phenomena in a given system which are widely unknown and unexplored yet. That constitutes a rich and challengingresearchfield for young (as well as old!) astrophysicists. What are you waiting to set to work,then? Good luck! Acknowledgements The author would like to thank the organizersfor the invitation to present this lecture during the XXI Annual Meeting of the SociedadeAstronómica Brasileira, held in Cax¬ ambu,MG (Brazil; in August 1995)and the Brazilian agency CNPq for partial support. 44 Proceedingsof B* XXIthSAB References BiskampD., 1994,Phys. Rep., 237, 179 Costa,J.R.V., Jafelice,L.C., 1995,work in preparation FabianA.C., 1994, ARAA, 32, 377 Friaça, A.C.S.,1993,A&A, 269,145 Friaça, A.C.S.,Jafelice,L.C., 1995,to be submitted for publication Friaça, A.C.S., Gonçalves,D.R., Jafelice,L.C., Jatenco-Pereira,V., Opher, R., 1995, work in preparation Ge J.P.,1991,PhD. Thesis,New Mexico Institute of Mining and Technology,New Mexico Gruenwald,R.B., Viegas,S.M., 1992,ApJS, 78,153 JafeliceL.C., 1992, AJ, 104,1279 JafeliceL.C., Friaça A. C. S. 1995a,submitted for publication JafeliceL.C., Friaça A. C. S. 1995b,to be submitted for publication JohnstoneR.M., Fabian A.C., 1988,MNRAS, 233,5$1 McKee C. F., 1992,in M. Matthews and E. Levy, eels.,Protostars and Planets III. Uni¬ versity of Arizona,Tucson Mestel L., 1990,in R. Beck,P.P. Kronberg, R. Wielebinski,eds,Proc. IAU Symp. No. 140, Galactic and Intergalactic Magnetic Fields. Kluwer,Dordrecht,p. 259 Priest E.R., 1982,SolarMagnetohydrodynamics.Reidel,Dordrecht Priest E.R., ForbesT.G., 1992,J. Geophys. Res.,97, 16757 ReesM. J., 1994, in D. Lynclen-Bell,eel.,CosmicalMagnetism. Kluwer, Dordrecht,p. 155 SokerN., Sarazin C.L., 1990, ApJ, 348, 73 Tribble P.C., 1991,MNRAS, 248,741 Proceedingsof lhe XXIthSAB 45 GAMMA-RAY BURSTS: A COSMOLOGICAL SETTLING Gustavo A. Medina Tanco instituto Astronómico e Geofísico,USP Abstract A brief introductory review is given on the state of the art of gamma-ray burst (GRB) research,and a model (due to Medina Tanco,Steinerand Terlevich1995) is presentedrelating their sources with a possibleevolving cosmologicalscenario. We assume that the mechanismof GRB generation is related to binary interactions of compact remnants (white dwarfs, neutron stars or black holes) and showthat the densecores of elliptical galaxiesare observationally acceptablesites for GRB production. We use the elliptical galaxy core luminosity function of Terlevichand Boyle (1994)with power law luminosity evolution to calculate theoreticallog JVvs. log S curves which we fit to the observationsof BATSE. Weshowthat for a universe with L = 0, decelerationparameter The existenceof the 7-ra.y bursts (GRB) was suggestedfor the first time by Colgate (1968). He speculatedthe possibility that supernova remnant shock waves in distant galaxiescouldproduce a flashof 7-rays whenhitting stellarsurfaces.Klesbesadel,Strong and Olsen (1973) discoveredthe GRB testing this conjecture. However,they did not discoveredthe GRB in connection with extragalactic supernovas. A new puzzlewas born that, even now, more than 2500 publications (Hurley, 1995)and 140 different theoretical modelslater (Nemiroff, 1994),still remainsmostly unsolved. Basically, the GRB are a burst of 7 photons with approximatelydaily frequencyand energiesvarying betweena few keV and some MeV (but three events are known with energiesin the GeV band). The duration of the burst is very variable,covering5 ordersof magnitudefrom some few msec to 5 minutes. Their temporal structure is fragmentaryand complicated,which hassuggestedto some authors that somethingis being broken,or comingout into pieces, during the process. Two basicdifferent types of GRB can be distinguished,the classicalGRB and the soft gamma,repeaters (SGR). The SGR,as their name proclaims,are softer than the classical ones (with energiesof only a fewkeV) and,principally, they repeat. Three suchbursts are known,two in the generaldirection of the galactic center and one in the directionof the 1,1 46 Proceedingsof the XX 1 SAB 10.00 «to, -1.7Í3SJ 3-U :s;j: >' «N* = 2 1.00 Earw>-2.3*as(M«V) E >~ 0.10 1 J V 5 o.oi CL r S 3 .-9 1-3 1 10 Energy (MeV) Figure 1: typical GRB spectrum LMC. Two of this repeaters are associatedwith young (> 10'1yr) supernova remnants, and the most acceptedmodel for their energy source is magnetic reconnectionnear the surfaceof highly magnetizedneutron stars, with Bsurjase= 101"1Gauss,calledmagnetars (Duncanand Thompson1994). We are interested,however,on the classicalGRB. They are harder than the SGR.and, apparently,they do not repeat on scalesof = 1 yr. They are characterizedby the lack of X-rajr emission. Their spectra are increasingin the hard X-ra.yregion, with a turn-over in the soft 7-rays and a power law extending up to a cut-off in the 7-ray region (seeFigure 1). Thereis no observationallyfavoredmodel for them. The ideas about the nature of the classicalGRB have been strongly influenced by the BATSE (Burst And Transient SourceExperiment) experiment on board the CGRO (ComptonGammaRay Observatory) satellite. During the first few years after the discov¬ ery of the GRB the largest variety of theoretical models was published,including both, galactic and extragalactic astrophysicalobjects (comets, white dwarfs, neutron stars, black holes,AGNs, ultra-relativistic iron dust grains, etc.). Nevertheless,by the end of the 70'sand alongthe 80's,and for reasons that will explainbellow, an implicit consensus was reachedabout galactic neutron stars (thick disk population). This consensus was so strong that, by the end of the decade,the modelswere mostly refinementsof this general idea. I11April 1991BATSE was launchedand,in Septemberof the same year, its first results were releasedto the public. Whatever the consensus at that time, it finishedon the light of the BATSE data. This was due to the fact that, contrary to any expectation,no sign of the galactic disk was found in the observedangular distribution, which was remarkably isotropic (see,Figure 2). Furthermore,the brightnessdistribution (log N vs. logS curve) showeda decreasedof the density of burst at the larger distances(Figure 3). That is, the GRB distribution has an edge. Both factors combinedalmost rule out modelswhich Proceedingsof the XXIthSAB 47 compriseonly a disk population. In the searchfor more isotropic distributions inside the Galaxy, two must be mentioned: (a) a nearby heliocentric distribution, like the Oort’s cometary cloud and, (b) a halo distribution. The last option, a natural extension of the neutron star hypothesis, counts the off- center position of the Sun inside the Galaxy as one of its most mortal enemies. The more isotropic the GRB distribution observed,the larger the halo. Up to the present, compatibility with the observationsrequiresa halo of more than 300 kpc in size,which implies not few problems. For example,the MaguellanicClouds would be completely insidethe halo, whichshouldproduce some observableanisotropy. Moreover,if our Galaxy has a hugeGRB-producinghalo,it is difficult to avoidsuchhalos aroundnearby galaxies. In particular, M31 should have a halo twice as large as ours, and it should be certainly observableby BATSE. It must be noted that this arguments, if not strong enough to completelyrule out the halo models,make them eachtime more difficult to fit into the observationalconstraints. +90 mmm. \ tyw hãá,. .ÿ . •.•so,....'it. mÊÁ, Figure 2: angular distribution of GRB seen by BATSE Oort’s cloud models,on the other hand,haveproblemsof their own. Due to the off- center position of the Earth in the Solar System, the minimum sizeof the source region shouldbe « 5 X 104A.U.. Among the variousproblems,we can mention: (a) the lack of a certain granularity in the angulardistribution that shouldbe observable(Eichler1994) but it is not; (b) nearby stars should produce some observableanisotropy due to their own Oort’s clouds; (c) the logN vs. logS curve deducedfrom known distributions,like the distribution of the comets, doesnot agree with the observedone for GRB. In summary, even if none of the abovedistributions can still be discarded,the cos¬ mologicalhypothesisis gaining strength as the observationsproceed.Note,nevertheless, that a combination of different populations of sources is always possible,as already pro¬ posedby some authors,howeverlogically difficult to justify this strategy may be. In this way not even a disk population can be completelyruled out. 48 Proceedings of the XXIth SAD A key question concerningclassicalGRB is if they repeat on scalesof few months to one year. The answer is, to say the least, tricky. This is due,mainly, to the large error box in angularpositions (whosediametercan be estimated as ft 4°) inherent to BATSE detections. Consequently,repetition is practically translated into clustering on scalesof ft 40.Moreover,different types of repetition producedifferent signaturesin the data (it is not the same, for example,to have 499 sources that not repeat and one that repeats 501 timesin a sampleof 1000 events,than 500 sources that repeat once in the same sample). Theproblemis more severe whenthe low number of availableevents is takeninto account. But, what is the importance of the existenceof repetition ? Basically, the fact that if helps to distinguish betweenlarge groups of models. For example,galactic disk models require,in general,repetition to match the GRB detection rate due to the small number of nearby sources. Cosmologicalmodels,on the other hand, are originated at so large distancesthat the extremely high energiesinvolved needthe completedestruction of the production region during the emission process. Clearly, no repetition is possiblein the latter case. True repetition, at least, since gravitational leasing and missassociation of angularly close events (specially in a small-number- statistics context), could emulate repeaters in various time scales. BATSE data are releasedas discrete packagesor catalogs. The first catalog, BATSE IB, contained 260 events,the second,BATSE 2B, 471 events and, the third catalog (still unreleased,despitealready in use by the BATSE team), BATSE 3B (?), should include ft 1000 triggers. BATSE IB generatednot few discussionregardingrepetition, since it showed(a dubious)clustering on scales< 40. BATSE 2B did not solved the problem, but increasedthe number of adherentsto the no-repetition main stream. At present, the BATSE team, working with data from the still unavailable 3B catalog,claimsthat < 20% of the GRB repeat on scalesof ft 1 yr with 99 % confidence.The most appropriateanswer seems to be that, apparently, classicalGRB do not repeat on time scalesspanning from few months to ft 1 yr. Note, however,that repetition on scalesof secondsto minutes is not ruled out. Another questionof fundamental importance says about the intrinsic luminosity in¬ terval observed. Are GRB approximate standard caudlesor not ? The brightnessdis¬ tribution observedby BATSE extends over nearly 2.5 dex and, schematically,is formed by two brancheswith slopesft 3/2and ft 0.8 respectively.'The transition he) ween both regimesis rather abrupt (< 1 dex), which seems to suggest a limited inliusic luminosity interval. In fact, a. study of the integral moments of the observeddifferential bright ness distribution (Horackand Emslie 1994,Wald 1939) in Euclideanspace showthat 80%of the observedGRB are drawn from a population of sources with luminosities varying by lessthan a factor 6 (Horack,Emsliee Meegan,1991). This would mean, in practice, that, GRB can be regardedas standard candles! Finally, another problem of relevance, is whetherlines actmilh exist in GRB spectra. Two GRB, among 23 GRB observedby the Japanese.satellite ( linga. presentedabsorption lines in the ft 15 — 75 keV range (Murakami ct al.1988:Graziani et al. 1992). In one of the events, the lines constitute a harmonically separatedpair: 19.3 and 3S.6 keV respec¬ tively. In the other event, only one line at 21.1eY is observed. I his sped ml lealures were Proceedingsof'l lio XXI'1'SAB 49 interpreted as cyclolj-on lines produced in magnetic fields of the order of fa 1012Gauss, which was interpreted as a strong indicai ion of a. neutron star involved in the emission process. This lines,however,are not observedby BATSE. This discrepancyis, still now, matter of intense debatedue to the fact that, even if BATSE spectra rarely extend to low enoughenergiesto detect the 20 keV line, the 40 keV line should be clearly visible in principle. However,once instrumentaldifferencesbetweenboth experiments and the pertinent statistical factors are taken into account, it is not easy to concludewhether a discrepancyactually exist. Finally, lines with energiesslightly below 511 keV were re¬ ported in the past, and interpreted as electron-positronannihilation lines,gravitationally reclshiftedwhile escapingfrom the neighborhoodof a neutron star. They are not observed by BATSE either, and a consensus exist that they are actually absentin GRB spectra. The generalpicture is that there is still considerableroom for speculationin the field. This is due,principally, to the fact that the availablestatistics are rather poor at present and the situation will not changein the near future as the gathering information rate is low (one event per day). Even the isotropy, certainly high, is not known yet with precision enoughto completely rule out halo or heliocentricmodelsof the Oort’s cloud type, despitethe increasingpopularity of cosmologicalmodels.It is not even clearwhether mixed models,including disk and/orhalo galactic populations are a plausible way out. Consequently,it must be admited that, at present, the uncertainty in the location of the sources spans 10 orders of magnitude, ranging from some tens of thousands of A.U. to some few Gpc. To this it must be added the uncertainty about the temporal repetition scales(secondsto years - if they repeat at all); the significanceof the taxonomic distinction of two classesof bursts, writh duration above and below « 2 sec; the uncertain existence of cyclotron lines in the spectrum that could point to a particular astrophysical object; the fact that galacticmodelscan reproducethe observedspectra rather easilybut not the spatial distribution of the burst, while exactly the opposite is true for the cosmological models;etc.. 2 GRB in a cosmological environment We can take three facts as an starting point: (a) that the angular distribution of GRB is highly isotropic, (b) that the distribution of sources, as indicated by the brightness distribution function, seems to be inhogeneous,and,(c) that 70 % of the published the¬ oretical models involve neutron stars as the main object associatedwith the generation mechanismof the burst, while 15 % include black holesor white dwarfs,that is, 85 % of the modelscomprise compact objects which we could call generically as ashesof stellar evolution. With this in mind, the most appropriate sites to look for the sources of the GRB seem to be those located at cosmologicaldistanceswhere intense star formation process took placeand a high density of compact objects can be found. A classof astro- Jihysicalenvironmentsthat fulfill theserequirements are the nucleusof elliptical galaxies that underwenta star burst phaseat some moment of their history. Two questions appear immediately: (a) why the preferencefor compact objects and, 50 Proceedingsof the XXIthSAB (b) why a star burst region, and not nuclear region of steadystar formation ? The preferencefor compact objects has basically two contributions, one historic, the other energetic. Historically, the detection of possiblecyclotron lines by Ginga consis¬ tent with TGaussmagnetic fields,characteristicof the surface of neutron stars, and of possibleelectron-positronannihilation lines redshifteclby an amount equivalent to the gravitational redshift resulting from the escapeoff the surfaceof a neutron star, had ini¬ tially a great influence on the theoretical modelingof the physical processes behind the emissionmechanism. On the other hand,in case that the sources were effectively located at cosmologicaldistances,the total energy releasedin a GRB episodemust be of the order of LQRBx ATQRB~ 1051erg/secx30 sec ~ 1%of the gravitational binding energy of a canonicalneutron star, or ~ 0.2%of its rest mass. This, togetherwith the fact that all that energy is packedin a relatively small region of space from whichit can lie promptly freedunder appropriatecircumstances,makesof neutron stars favorite candidatesto GRB progenitors1 . It seems to be an observationalfact the lack of spatial correlation betweenthe galaxies that are undergoinga star burst phase,and the angularpositions of the GRB. Therefore, even if the star bursts are favorable events in the sense that they are able to build up a largedensity of compact remnants in a short time interval (e.g.,~ 10sobjects in a region with a diameter of only & 1 pc in a soundcrossingtime 60 Myr), the GRB production process must be decoupledfrom them. A natural way to accomplishsuch a decoupling is to admit that the GRB are originated by the compact remnants of stellar evolution on a much longer time scale than that of the star burst which originated them. That would be the case, for example, if the GRB were producedby the collision of compact objects, or the merging of binary systems. Then, there would be two relevant time scales: (i) the characteristic time scale of the star burst (i.e., of the formation of the compact objects)tstar burst ~ 60 Myr and, (ii) the characteristic(interaction,merging,etc.) time associatedwith the physical mechanismresponsiblefor the GRB, TQRB ~ few Gyr, i.e., perhapscomparablewith the lifetime of the host galaxies. 3 Analytical modeling and numerical comparison with the observations FollowingMedina Tanco,Steinerand Terlevich (1995),if n'GRBis the number of GRB produced per proper volume per proper unit time, and nsBc is the proper density of galacticcores that experienceda starburstphaseat some time in the past (SBC), we can write: GRB — (aX «C/í)x n-sBC (1) hIt must be noted that, in the case of galactic origin, the temporal variability of the flux is not a strong criteria for restricting the size of the related astronomical object, but only of the emissionregion instead. Hence, for example, the energy could be released in a stellar Hare, in which cage the emission region is compact (the magnetic loop) but not the associated astronomicalobject (the star). Proceedingsof l,he XXI11'SAB 51 where.,nrn is the density of compact remnants (CR) in the nuclear region after the star burst, and a and I) are constants that take into account the physicsof the specific mechanismthrough which the burst energy is released. In this context, the ('volution of the luminosity function of the cores of elliptical galaxies in which the star burst is taking placeshould be related to the evolution of the rate of (!RB per unit proper volume, since the mass processedduring the star burst gives the total mass in compact leftovers alter the A(!N phase. FollowingTerlevichk. Uovle(1993) we assume that the luminosity function of young cores of ellipticals undergoesstrong evolution betweenz —2 and the present epoch,and that its redshifl dependenceis well representedby a power law, k (l>[z)oc (I + z) (2) where/.*lie in the range 3.1 < k < 3.6 and adopt here k = 3.3. At a given redshift z, the density of galactic core that went through a starburst phase, USBC, is built up from the integration from ~ = Z\ to c of the rate of star bursts,nt(z), which we assume evolves in the same way as the luminosity function of SBCs.That is, iit(z) = const for 2 < z < Z\ (3) i„(z) = [l + z)k for 0 < r < 2 The density of galactic nuclei is given by: ( \ i \ diw Bc{z) = Jr nl{z)x—jj—x dz (4) where1(z) is the proper time. In addit ion, in the present approximation the rate of CIRBs per unit proper time, per unit volume is proportional to the proper number density of galactic nuclei that underwent a star burst event at some time in their history, idaiw = £ x 11sue (5) Hence,the number ol GRBs observedper unit time with flux larger than a certain value F is: " UISBC c)r(z) Av(> F) = Í x 47T x r2(z) x x dz (6) Jo (I +.r) dz where r(z) is the proper distance to a redshift z, and ~ is obtained by inverting the equation: LX /’(-) I IS ItH (7) \TTC2Z'*(Z)(1 + ~)2 which gives the llux obs('r\('d from an ('vent of luminosity photons per unit pro[X'r time at. redshift ~,where 52 Proceedingsof the XXIthSAD 1/2 { wherezm{n is the maximum depth at which BATSE can see a GRB. Dependingon the desiredapproximation, either (6) and (7) or (6) and (9) can be used as the parametrical form of the logN vs. logS curve. Figures 4 and 5 showthe fitting of equations(6) and (7) to the logN vs. logS curve observedby BATSE for more than 687GRBs as givenby Horack (1994).Different values of qo and z\ are used,and it can be seen that, despitesome freedomin the adjustment is allowed,a good fitting is obtained for qo = 0.5 and Z[ = 4. If thesevaluesare used for normalization of the curves, then the resultant intrinsic luminosity of the GRBs (flat spectrum)in 50-300keV is LQRB~ 1051erg/secfa 1%of the rest mass of a uoutrou star whenintegrated over a characteristic time scale(7’90)~ 30 sec), while the local rate of GRBs production per unit volumeis n'(niBfa 0.8 j\Ipc_ ! Gyr- 1, or one event each ss 105 yr in the local group. It seems also that BATSE is reaching as deepas z,upih ~ 2.8 —2.9. In Figure 6 we show the logN vs. log S observedcurve superimposedto thecalculated curves for a power law GRBsenergy spectra (eqs. (6) and (9)) of index a = 0.5,1.0,1.5, 2.0 and 2.5, and the same parameters used to fit figures-l.a-b,and = 4. It can be seen that an average spectral index a fa 1.5 can fit the data, despite some dispersion. This result agrees well with fittings to the actually observedGRB’s spectra in the 50-300keV band by Pendelton et al. (1994)(seealso.Fig 9 of Meeganet al. 1994). The luminosity obtained in (he case of a power law spectrum with a = 2 is LGRB~ 1.2 x 1051erg/secin the 50-300keV. In the model presentedhere,galactic cores insidea given solid anglewith the aperture of the error box but at different redshifts could giveplaceto the identificai ion of apparent repeaters when working with small number statistics. The importance of this effectmay be easilyestimated. If GRB can be considereda.sstandard candles,we could expect that different events of the same repeater should be observedwith similar brightness. Hence, the period betweenevents observedinside an error box of aperture AD with flux in the interval (F,F +DF) can be approximatedby: -1 sn dr(z ) F(z)AFrei Trep = — x n'GIW(z)x 4?r x r2(z) x Xw 47r dz dF(z)/dz m whereAFrei = AF/F.It can be seen that, for AFreias high as 100 % for the previous fitting, the period of the apparent repeaters can vary between « 2 and 40 yr for the faintest bursts, dependingon how much restrictive is our definition of repeating events. Proceedingsof the XXIth SAB 53 1000 r£ Z 100 \ O -3/2 5 10 | % g Ci 0/ ••• 1 1 0.1 1.0 10.0 100.0 1000.0 Peakflux (ph cm'2s'1)P Figure 3: GRB’s brightnessseen by BATSE I 1 T 11 T 1 1 II It 1 I I I I 1 i i i I 11 1E+15 z , = = 2 =4 ' 2, =3 2 2.1 1E+14 ,= — -3/2 £ Z * 1E+13 o 00 o o 1E+12 q0=o-5 o mil i 1 111 i i M 1 111 1E-8 1E-7 1E-6 1E-5 Rz,q „,z ,) Figure 4: 54 Proceedingsof the XXI1*1SAB I TTTTTT I TTTTTp i r i 1 1ii | q o=0.01 1E+15 1 q 0=o-5 q.*1-0 1E+14 - -3/2 Q q: & z % 1E+13 %o oo o o 1E+12 o _ XXLX X i I i_ul X 1 I I i I I 1E-8 1E-7 1E-6 1E-5 F(z,q 0.z,) Figure 5: ( I riTTTTV T—r T I l l'l| O (BATSE2B)minusshortbursts - N(>C_/ C„) calculated 100 .o' o Ho q: UL \\ z 10 V\ o cx=2.5 a=0.5 o o _ j J l 1 IX I i x •nil 1 10 100 Figure 6: power law spectrum at the source. Proceedingsof the XXIthSAB 55 Nevertheless,this values are consistent with the present lack of small scaleclusteringor repetition. 4 Conclusions Thereis still considerableroom for speculationin the field of GRB. This is due,principally, to the fact that the availablestatistics are rather poor at present and the situation will not changein the near future as the gatheringinformation rate is low. However,if the GRBs sources are located at cosmologicaldistances,as growing evidencefrom BATSE seems to point, then the total energy channeledinto 7-rays by the mechanismpowering GRBs amounts to a considerablefraction of the rest mass or gravitational energy of a solarmass compact body. This makesappealingcatastrophicprocessesinvolving compact objectslike white dwarfs,neutron stars, or black holes, Therefore,the natural placesto look for the GRB sources are those where it is likely to find a high density of these compact objects. One suchplaceis the core of an elliptical galaxythat, at some time in its early history, underwent a violent star burst phase.The evolution of thesestarbursts can be tracked as a function of redshift by following the evolution of the SBC’sluminosity function. By doing this it is possibleto demonstrate that GRB production in thesecores is compatible with BATSE observationsfor at least q0 = 0.5, and starburst turning on at zi = 4. References [1] CohenE. and Piran T., 1995,A &; A, in press. [2] Fishman G. J., Brainerd J. J. and Hurley K. (eds),1994, Gamma-ray Burst 2nd Workshop, Huntsvile 1993, AIP Conf. Proc. 307, AmericanInstitute of Physics, Ney York. [3] Graziani C. et ah,in Gamma-Ray Bursts, C. Ho, E. Fenimore and R. Epstein (eds.), CambridgeUniv. Press.,p 407,1992. [4] Hartwick F. D. A.and ShadeD., 1990,Ann. Rev. Astron. Astrophys., 28, 437. [5] HorackJ. M., 1994,Science,265,1186. [6] Horack J. M., Emslie A. G. and MeeganC. A., 1994,Ap. J., 426, L5. [7] Hurley K., 1994, GRBs 2nd Workshop,Huntsville, AIP ConferenceProc. 307, G.J. Fishman,J.J. Brainerd and K. Hurley (eds.),pp. 726-729. [8] Kouveliotou et al, 1993, ApJ Lett, 413, 101L. [9] Mao S., Narayan R. and Piran T., 1994, ApJ, 420,171. 56 Proceedingsof the XXIthSAB [10]Medina Tanco G. A., Steiner J. E. and TerlevichJ. R., in preparation. [11]MeeganC. A., Fishman G. J., Wilson R. B., PaciesasW. S.,PendletonG. N.,Horack J. M., Brock M. N., Kouveliotou C., 1992,Nature,355,143. [12]MeeganC. A., Fishman G. J., Wilson R., Brock M., Horack J., PaciesasW., Pendle¬ ton G., Briggs M., Koshut T., Kouveliotou C., TeegardenB., Matteson J. and Hakkila J., 1994,GRBs 2nd Workshop,Huntsville,AIP ConferenceProc. 307,G. J. Fishman [13]Murakami T. et al, Nature 335,234,1988 [14]Nemiroff R. J., 1994,Comments Astrophys.,17,189. [15]Pendelton et ah,1994,in preparation. [16]Terlevich,R., 1992in: Relationship Between Active Galactic Nuclei and Starburst Galaxies,ed. Filippenko, A.V. Pub. Astron. Soc.Pacific,SanFrancisco,p. 253. [17] TerlevichR. and Boyle B. J., 1993,MNRAS, 262,491. Proceedingsof the XXIthSAB 57 BLACK HOLES: FROM GALACTIC NUCLEI TO ELEMENTARY PARTICLES José P. S. Lemos Departamento de Astrofísica,ObservatórioNacional-CNPq, Rua GeneralJoséCristino77, 20921Rio deJaneiro,Brazil, E-mail:[email protected]. Abstract We present a broad review on black holes. We analysesome of the fundamental concepts in black hole theory, the observationaland theoretical status of stellarand galactic black holes,and their appearance as quantum objects. 1 What is a black hole? One of the basic ingredientsof a physical theory is its set of fundamental constants. Thus,for instance,classicalmechanicshas no fundamental constants. The Newtonian theory of gravitation contains one constant alone,the universal constant of gravitation, G. The electromagnetismof Maxwell contains the velocity of light c, whichin vacuum is a fundamental constant. Planck’sconstant h, whichappeareddirectly from an experimental result (the black body spectrum),was immediately taken as the fundamental constant of quantum mechanics,developedamong others by Bohr, Heisenbergand Schrõdinger. Thermodynamics yields the Boltzmann constant which arguably can be considered fundamental. Thesetheories and their constants can then be combined to yield unified theories. If one tries to unite classicalmechanicsand electromagnetismone obtains the theoryof specialrelativity whichhasc as a fundamentalconstant. The electricchargee is a constant of nature. If we join e and c one obtains classicalelectrodynamicsof Thomson and Lorentz, which is of course related to specialrelativity. If one further joins h one obtainsquantum electrodynamicsdue to Dirac and developedby others,afterwards. The unification of the electromagneticand the weakforcesby Weinbergand Salam as well as the inclusion of the strong forcein the grandunified theoriesby Glashowand others alsomix the fundamentalconstants of eachseparate theory. Finally, if one combinesG and c one obtains the theory of generalrelativity of Einstein. On further joining Planck’s constant h one shouldobtain quantum gravity, a theory which is still eluding the realm of physics,although there are some hints as to what it shouldbe. Black holes (BHs) are objects which belong to the theory of generalrelativity, and can be used to explain many powerful phenomenaobservedin the celestialsphere. On the other hand, their (quantum)effectsand the singularitiesthey hide yield an excellent framework to probe into the nature of quantum gravity. A simpleNewtonian argument can lead us to the concept of dark star, the Newtonian closestrelative to the black hole (BH) of generalrelativity. The escape velocity veof an object ejected from the surface of a body, such as a star, of mass M and radius R, is givenby The escape velocity on Earth is 11 km/s, on a white dwarf it is around6000Km/s.A dark star is definedas a star for which the escape velocityis greater 58 Proceedingsof the XXIth SAB or equal to the velocity of light, i.e., for which the following relation holds |e2< However,such a star is not a black hole for two reasons. First, the velocity of light is not a fundamental constant (i.e., it has no fundamental meaning) in Newtonian gravity, and thus other objects thrown with tachyonic velocities can escape and be detected at infinity. Second,the dark star is only dark for distant observers,near its surfacethe star is still bright since it emits light, althoughit cannot escape to infinity. The correct theory to explain the BII phenomenonis generalrelativity. The BH is a “state of the gravitational field”, different from the state of the gravitational field of a star. Wecan understand BHs most easily through the collapseof a star (seee.g. [f] [2][3] [4][5]). As the star collapses,its own radius shrinks. From Newtonian gravity we know that the forcethe star exerts 011 an object goesas r-2. Thus, a contraction by a factor two increasesthe forceby a factor four. In addition, if the star collapsesto a point, the force becomesinfinite at r - 0. GeneralRelativity yields a. different result, the gravitational force increasesmore rapidly than r-2. The force is then infinite when the radius of the star is R = 2 2 Stellar black holes BIIs with stellar masses can form through the collapseof the iron cores of massivestars after they havereachedthe end of their thermonuclearevolution. The outer layersof the star explodein a supernova leavingat its center, dependingon the core’smass,a neutron star or a BH. The maximum mass for a neutron star is still a matter of debate,sinceit Proceedingsof the XXIthSAB 59 dependsstrongly on the equationof state of the constitutive matter. Rhoadesand Ruffini [6] found a maximum mass of 3.2M®,while Hartle [7] can put an upper limit of 5M®, (seealso [8]). However,Bachall et al. [9] arguedthat one can construct a IOOMQstar madeof other types of matter at nuclear densitieswhich they called Q-stars(related in some sense to Witten’s strange stars [10]). Thus the limit of the maximum mass 3.2 or 5M®is still uncertain,although probably correct. The sizesof a BH and a neutron star do not differ much. For a lil/® object, the neutron star has10 Km of radius,whereasthe BH has 3 Km. An isolated stellar BII cannot be seen. A BH can only be observedif it belongsto a binary system [11],and is detectedthrough spectroscopicobservationsof the bright optical companion. The main problem is that the unseen body can also be a. neutron star, and to distinguishbetweenboth possibilitiesone has to follow a complicatedlist of steps. A binary system can evolvein the following way: first, two massive stars, with masses of the order of 20M®form a binary system. Then, in a secondstage, the more massivestar evolvesmore rapidly and soon becomesa compact body with a few M®after havingexplodedin a supernova. Finally, the other star also evolvesto becomea similar compact star. One thus has a binary system of two compact stars, of which the most famousexampleis the binary pulsar of Iiulse and Taylor [12].During the secondstage, whenthe binary is composedof one compact and one giant star there is the production of spectacularphenomenavisible in the X-ray band. The binary systems in this stage are called X-ray binaries. X-ray binaries have a very short orbital period which by Kepler’s third law implies the objects are very close. Sincethe Rochelobe of the binary system (the surface of gravitational neutrality) can be filled in part by the massivecompanion, the outer layers of the massive star are captured by the compact star. The captured gas then forms an accretion disk. The emissionof X-rays can happen through severalprocesses. If the compact star is a neutron star then the gas, througha magneticfield mechanism,hits the crust of the neutron star regularly with the consequentemissionof X-rays. The neutron star is then called an X-ray pulsar, with luminosities of the order 1(Hi®. If, instead of regular, the X-ray emissionis sporadic,then the source is called a burster. Bursters are usually produced through an explosionof the surfaceof the neutron star, but can alsoappear by eruption of a very hot region of the accretion disk [13][14].In this last case the matter from the luminouscompanionspiralstowardsthe unseen compact object and emits X-rays with temperatures of 108—109K.Thesetemperatures are generated throughdissipationby viscousprocessesof the gravitational energy of the infalling matter acceleratedto high velocities. BHs do not have a hard solid surface,the explosionhappensin the disk. Is there any way to distinguishbetweenBH and neutron stars bursts? Onecouldthink that variability would give some clues. If a source (disk, in the case)changesshape,the speedof change cannot exceedthe speedof light. If one detectsvariability in a time At, the sizeof the source is at most / ~ cAt. If the source changesin At ~ 10_3s,then its linear size is / ~ 300 Km. For a stellar BII, its inner edge(defined as the last stableorbit, see e.g. [15]), is of the order of 30 Km in radius say, so its circumferenceis around200 Km. A rotating 60 Proceedingsof the XXIth SAP hot bubble emitting X-rays would have'a varabilily in the miliseeondrange, as observed. This could be a signature for a BH. However.Circinus X-l also showsfluctuation of this order, and it was shown that it also has periodic bursts which characterizesa neutron star [16]. The identification of a black hole through radiation processes is not yet well developed,although it is a field advancingquickly. The best criterion to identify a BH is to find its mass through dynamical studies of the X-ray binaries. The weighingof stars in binaries is a techniquefully understoodnowadays. Knowing the orbital period of the binary and the projected mean speedof the optical star, one can using Kepler’s law to deducethe mass function definedby [17],f{M) = {Mx sint)3 wÿere source and (Mx+Mc)2 ~ P('V-2wGÿ• *Jr and AJCare the masses of the X-ray the companion respectively,i is the orbital inclination angle, P is the orbital period and Vcsini is the projected velocity seiniamplitude of the optical companion. 11one puts Mc —0 and i —90° one gets the mimiimun possiblevalue for Mr which in this case is equal to f(M). What one would really like to obtain is an f(M) greater than 5M@. However,following tin* 1lieory, a valuecloseto 3.2.1/. yields already a goodBH candidate. There are three very strong candidates, two good candidates and a list of possible candidates. We start here giving some properties of the three very strong candidate's. (la) Cygnus X-l - II was discoveredin 1971 [18]and il is a persistent source with Lx ~ 2x1037erg/s. It is a high mass X-ray binary. The orbital period is 5.6 days and Vcsin ? ~ 76Km/swhi<-hyields a low /(.!/) = 0.35.1/,.,. Now. one has to derive a reliable lower limit for the mass, which is a difficult task. 1lie optical companionis a blue giant with mass Mc~ 30.1/,.. The inclination angle is supposedto be i ~ 30° (there are no eclipses). This gives a mass of .!/.,. ~ 1GA7®[19], 'l’he most conservativeassumptions lead to M > 3M®.(lb) LMC X-3 It was discoveredin 1983 [20]. It is a persistent 3xl()i8 source and a high mass X-ray binary with Lx ~ erg/s. P = 1.7 daysand Vcsini ~ 235Km/s.This gives f{M ) — 2.3.1/.. The mass of the companion is estimated to be Mc~ 6MQwhich then yields M.v~ 6/1/®.Since the distance to the Magellaniccloud is known one can use Paczynskimethod [21]to infer Mx~ 4/1/®.(lc) 0620-00 - It was discoveredin 1986 [22].('outl ay to the other two, it is a transient source. It is a low mass X-ray binary with Lx ~ I x 10:,8erg/s.P = 0.32 days and V'.sini ~ 467 Km/syielding f(M ) = 3.18M®.Based solely on the value of the mass function, the minimum mass is alreadyequal to the BI1 threshold mass. It is consideredthe strongest of the very strong candidates.Mc~ 0.7M®which is low, and yields the following lower limit Mx~ 4M®. There are two other candidateswhich havebeen weighed,althoughthe uncertainties are greater than the sources mentionedabove. (2a) CAL87 -- OnehasLx ~ lxl036erg/s. It is an interesting system since it undergoeseclipses.If Mc> 0.4M®if was found that Mx ~ 4M&[23].(2b) LMC X-l - Lx~ 2xl038erg/s.The optical companionis still not identified conclusively. However, there are hints that A/r ~ 3A/®.There are a number of other candidateswhich have been selectedbecausethey show X-ray behavior similar to CygnusX-l, of which the prime exampleis GX239-4,and otherswhich show transient behaviorsimilar to 0620-00,such as GS2000+25,GS2023+33,GSH24-68, 4U1543-47, 4U1630-47,H1705-250[24]. Further dynamical studies are neededto obtain the masses of thesesources . The spectacularsource SS433which showsemissionof jets was recently Proceedingsof the XXIthSAB 61 discardedas a black hole,sinceits mass was shownto be M ~1.4M0[25]. What are then the featuresthat allow us to identify a BH candidate? The classical steps are: 1) The luminosity of the X-ray source has to be high Lx > 1036erg/sand of rapid variability < Is. This implies that the binary system must contain an accreting compact object. 2) The optical companionis identifiedand allowsto measure the orbital period and the projected orbital velocity, to yield f(M). 3) The mass of the optical companionand the inclination of the orbit are inferred or limited, basedon the distance, optical spectrum, Lc, and eclipes. Then using f{M) one deducesMx. 4) If the mass of the object is Mx ~ 3.2MQthen it is considereda BH candidate. There are now three other criteria whichcan helpin identifying a BH candidate:(i) the source has a spectrum with soft X-rays, ~IKev. (ii) Fe emissionlines very near the compact object (will) allow one to measure the velocity of the rotating disk. This then implies a dynamical measure of Mx [26].(iii) Hard X-rays ~ lOOKevare a signatureof BHs, sinceneutron stars have a hard surfacewhoseradiated photonscool the accretiondisk through Compton cooling [27]. These last too criteria are very recent [28],and it is expectedthe situation will improve with work on some other half-dozen sources. Onedrawbackis that all thesecriteria are indirect. Onereally wants to come closeto the collapsedobject. In the long run, one is after clear evidencefor the existenceof an event horizon [29][30].This might be possibleafter the gravitational antennas are fully operating, to detect unambiguouslythe formation of a BH. If the BH is in a binary one expects subsequentX-ray emission.How many BHs there are in the Galaxy? The last estimatesgive 1000-3000BHs, of the same order as the numberof neutron stars [24]. 3 Black holes in galactic nuclei Wehaveseen that in the completegravitationalcollapseof a star a BH can form with mass in the range 3 —20MQ.Yet, the theory of gravitational collapseallows for the formation of BHs with much greater masses,masses that can be in the range 103—1O9M0.These BHs may appear in the core of clustersof stars or in the center of galactic nuclei. If the mass of the original system is very large, there is no uncertainty in the equation of state of the collapsingmatter whenit crosses the event horizon. Indeed,at R = the density of the matter is For a p = 3%fG3 ~1.3xl016(ÿp)2g/cm3.~1M0BH the density is very high, abovethe nucleardensity. However,for a ~ 108M0object one has p ~ lgm/cm3. In this case one has ~ 108. This roughly means that for a cluster composedof 108suns, the cluster crosses its own Schwarzschildradius when the suns, uniformly distributed over the volume,are touchingeachother. For a ~ 1O1OM0object one has Po ~ 10~4gm/cm3.In this case, ~ 1010,which roughly means that for a clustermade of 1010suns, it crosses its Schwarzschildradius whenthe suns, distributed uniformly over a sphericallayerof thicknessof one sun diameter,are touchingeachother. In all theselatter cases the physicswhenthe matter crosses the horizon is well known. Theory and numerical simulationsfavor the appearance of a binary system in the 62 Proceedingsof lhe XXIthSAB center of globular clusters. The compact binary scatters any incoming star. There is, in principle no formation of a BH in the core of the cluster. This is supportedby observation. However,central BIIs are not totally excluded[13]. There is controversy about the existenceof a central BII in our Galaxy sinceit was first proposedin 1971 [31]. The Galactic center has the following features: (1) a disk of gas with inner and outer radii given by 5 —30 light years (ly); (2) a cavity interior to the disk with 2xl06 stars; (3) a possibleBH with 2xl0(3Km ~ 2x]0-7ly accreting matter slowly. The evidencefor a central compact source comes from the radio emission of a region as small as the orbit of Saturn around the Sun [32]. This source is called Sgr A* and has L ~ ICP'erg/s~ 107,0(~ 104times the luminosity of a single radio pulsar). From the disk of gas, one can infer (if it is in a Keplerian orbit) a central mass of 5 —8X106M®.Subtracting the mass in the red giants one obtains 3 —6xl06A7®.The evidencefavors the existenceof a central BH, althoughit is not absolutelyconvincing [33] [34].Evidenceagainstthe existenceof a central BH has appearedafter observationsfrom the Sigma/GRANATtelescopeled to the conclusionthat, contrary to expectations, there is no X-ray source coincidentwith Sgr A* [35].However,thereare now modelswhich can explain the phenomenonin a natural way in which the matter is swallowedbeforeit has time to radiate [36]. Thereis alsodynamical evidencethat M31 (Andromeda)has a compact massivesource at its center with M ~ 3x10' M®which favorsthe existenceof a BH [37].M32 alsoharbors a central dark object of M ~ 5xl0(>. There is evidence for very massive nuclei in other nearby galaxies,in NGC 4594 (the Sombrerogalaxy) [38]and in NGC 3115 [39]. The nuclei of most galaxiesare inactive, in the sense that L ~ 10_4Z,gaiaxy.There are active galactic nuclei (AGN) which can shine more than the entire galaxy. Galaxies that have AGN are 1%of all the galaxies.Examplesof AGN are the quasars (of which 3C273has a luminosity equivalent to 103galaxies),blazars,Seyferts,radio galaxiesand other objects. In a spectrocospic.classification these AGN are divided in AGN type 1 which showbroad and narrow emissionlines and AGN type 2 which show only narrow emissionlines. They have some common features: (1) non-thermal radiation, (2) high concentration of mass in a small region, (3) variability in luminosity, (4) ejection of jets at great distancesand(5) similarities with normal galaxies. The idea, is to explain genericallyall different objects and phenomenawith one model. The most favoredmodel invokesaccretion onto central supermassiveBIIs as the ultimate power source for theseluminousobjects whichradiate at the Edclingtomlimit. Even,if one invokesother centralobjects,suchas spinars(yielding their rotational energy)or a cluster of packedstars (supplyingnuclear energy through supernova explosion), the emissionof so much energy from sucha small volume (which is measuredthrough variability) leads inevitably to the collapseto a BII [40]. There is also the possibility that some galactic nuclei may contain two massiveBIIs in orbit around eachother [41].Further out from the central object, thereis a dusty accretion torus whichprovides a mechanismto understand AGN 1 and 2 [42]. The jets, when they exist, point from the central region into two opposite directions aligned with the rotation axis of the torus. The blazars are thought to be quasars with one jet pointing towards us. The spectroscopicdifferencesin AGN are Proceedingsof the XXIuhSAB 63 also due to different orientations of the torus with respect to the Earth. If one can see the inner edgeof the torus one observesboth the broad lines emitted in the inner region by high speedcloudsand the narrow lines emitted in the outer edge.If the torus is seen edge-ononly the narrow lines are observable. The evidenceto detect the central mass, both in AGN and normal galaxies,is basedin most cases on the increaseof mass-to-lightratio in the central region. Only in two cases, M87 and NGC4258, is the value for the central mass based on gas dynamics rotating around the central mass. By using the Hubble SpaceTelescopeit was possibleto measure the Doppler shift of emission lines from doubly ionized oxygen around R ~ 601yfrom the center of MS7 [43]. This implies a rotation velocity of v ~ 550Km/sfor the gas in orbit which then givesM = ~ 2 —3xlO9M0.The mass is so great in sucha small region that is difficult to think of any other explanation than a supermassiveblack hole inhabiting the center of the galaxy.If, for instance,the mass were containedin solartype stars in a densecluster, they wouldbe packed100 thousandmore times closelythan in the solarneighborhood. However,this is discarded,since there is not enoughlight comimg from this region. In the case of NGC4258,recent work [44][45]has alsopointed to the confirmation of two things: 1. Keplerian velocities of ~ 1000Km/sin an inner orbit of very small radius, R ~ 0.1ly, around the central mass have been measuredwhich imply a mass of M ~ 2x107Mm. This work is consideredto provide the strongest case for a supermassiveBII in the center confirming the predictions of Lynden-Bell [46]. 2. The velocities are measured through water masers which are found to come from a torus-like region confirming the unifying model of Antonucciand Miller [42].It has beensuggested [47]that the best one can do for the black hole case is to refute the other models on physical grounds. For NGC4258this has been undertaken [48]. One can thus have a model in which all galactic phenomena are unified, not only within AGN themselves,but alsorelating normal galaxiesand AGN. In AGN,part of the potential energy is releasedwhen matter approachesthe event horizon and the energy escapes as radiation providing the mechanismto power the emission. An accretionrate of a few tens of M0per year, which can be suppliedby surroundinggas and by stars tidally disruptedin the gravitational field, will provide a power greater than 1047erg/s which would explain even the highest quasar luminosities [49]. In normal galaxiesthere is no matter to be accreted.The real differencethen betweenthe nucleiof normal galaxiesand AGN is, in this model, not the mass of the BH, but the phaseof the cycle of accretion. The quiescentnuclei would be BHs starvedof fuel, i.e., dead quasars. 4 Quantum black holes and elementary particles Wehavedescribedin the previous sectionsstellar BHs with masses 3 —2OM0(and 10 — GOIvm),and galactic BHs with 10°— 1O9M0.But therealsoexiststhe possibility of having BHs with much smaller masses. For instance,if the Earth with mass ~ 1027gm~ 1O_6M0 was compressedto a radius of 1cm it would turn into a BH. There is the possibility that primordial BHs with mountain masses,1014gm~ 1O_19M0,and a radius similar to the 64 Proceedingsof the XXIthSAB proton radius 10~13cmcould be formedin the early universe[50]. The smallestpossible BH wouldhavea mass of 10_5gmand a radius equal to 10-33cm,the Planck radius,which is thought to be the minimum possibleradiusthat occurs in nature. Smallermasses would havea Schwarzschildradius smaller than the Planck radius,and thus if compressedinto a BH, thesemasses wouldbe snatchedby the Planckregime,(i.e., by the spactimefoam [51][52]),before they had turned into a BII. Of course, theseBIIs havea totally different interest from the macroscopicand giant BHs. Their physical effectsare of a different kind. Let us take the mountain mass BH, M ~1O-19M0and R ~ 10-13cm.Its gravitationalattraction at a distanceof 10m would be relatively small (~ 0.1m/s2).Its tidal force on a 1cm tight object of lgm wouldbarely be felt at a distance of 10cm. Sucha black hole could cause some damageon nearby objects,but not a lot. For a Planckian10-5gmBH its gravitational atraction would give an accelerationof 10_6cm/s2at a distanceof 10-5m,roughly the sizeof a living cell. On this basis,even if one of theseBHs enters our body we would live without noticing it, the accretion onto it would be vanishinglysmall. However,there are other proceses that would made the BH noticeable,and thesecould cause damagein our body. There was a great turn in BII theory after Hawking found in 1974 that BHs can radiate through quantum effects[53]. There were already hints that BHs have a ther¬ modynamic behavior. If one throws entropy S into the inside of the BH, this entropy disappearsfrom our universein direct violation of the secondlaw of thermodynamics. Sincethere is a theorem [54],within classicalgeneralrelativity, that states that in any process the area A of a BH never decreases,Bekensteinproposedthat SBHOC A, such that the secondlaw is not violated,S +SBH> 0 always [55]. Since,to an entropy one can associatea temperature through the thermodynamic relation S = j,, the BH must havea temperature. Indeed,by complicatedcalculationsof quantum field theory in a BH background,Hawking was able to find that the BH emits blackbody radiation at a tem¬ Sinceso fundamental constants (section perature T = If —6x10-8(ÿp)/v. many 1) enter this formula one can say that quantum mechanics,generalrelativity and thermo¬ dynamicsmust merge together in a unified theory. For M ~ 1M0one has T ~10_7K. When M ~ 1O_19M0implies T ~ 1012K.For a PlanckianBH, M ~ 10_5gm,therefore, T ~ 1032K.Making the translationE = ksT one has that the energiesinvolvedin the evaporation of a Planckian sizeBH are ~ 1019Gev~ lOOwatt-hour. This could do some damagein our body. The mechanismfor evaporationcan be explainedin severalways, the most popular uses the idea that the vacuum is full of virtual particles which are created and annihilated without violation of the uncertainty relation AEAt ~ Ti. Howevernear the event horizon it can happen that one particle enters the BH while the other escapes out to infinity. The net result is blackbodyradiation at the Hawkingtemperature T. Now, the power emitted by a radiating BH is inR2aT4 = , wherea is the Stefan-Boltzmann A-- A above. constant,and = 153607rG2, a valuefound through the equationsgiven Then,one finds that the BH loosesenergy at a rate which can be integrated to give M = (M03- . For an initial mass of M0~1M0one gets that the BH evaporates in 1067years.If one puts M0~ 1O_19M0one finds t ~ lO10years, which means if created in the primeval Universe theseBHs should be evaporatingby now. This couldhappenin Proceedingsof the XXIthSAB 65 a hurst of final radiation after parsing through the Planck scale. Somehave speculated that the observed7-ray bullfil could come from thesemini-BHs,but there are tight limits on their existencefrom gravitationallensing[56]. Thus, classically,Jills are stable, but quantum mechanicallythey are unstable,they slowly evaporate and shrink. One striking effect that arisesimmediately is the violation of baryou number. Baryou numberconservationis a law in elementaryparticle physics. However, if, say, a totally isolatedneutron star of JO5'neutrons (baryons) collapsesonto a 1111,it will evaporate' in a baryon-anlibaxionmanner, actually most of the radiation will be-in photons which carry zéro baryou number anyway. Thus, gravity and quantum field theory produce'violation of baryou number. One pro feten that Hawking radiation givesrise, to is called the information paradox. To describea star completely, one must specifya large ammount of information,suchas, total mass M . total chargeQ, total angulai-momentum J, temperature, pressure, gravitational multipole moments, other chemicalpotentials, and so on, including the quantum states of the 1057protons and neutrons that constitute the star. When the star collapsesto form a BII, the 110-hairtheoremssay that the BH is describedby only threeparameters, .1/, Q and J. All the other information that was necessary to describethe original stal¬ ls now hidden inside the event horizon. Hawking [57]found within his calculation,that the blackbody thermal spectrum of 1,1®emitted flux of particles would not carry the original information out to the exterior region. After the BH completely evaporates, the information that was trapped inside alsovanisheswith the BH. The information paradox for BHs is the problem of explaining what happens to the missing information. It is of great importance because,in usual quantum mechanics,the wavefunction-0evolvesin sucha way that information contained in it is never lost. However,if the picture described hereis correct, then gravitational collapseviolates a fundamentalprinciple of quantum mechanics. Another important reason to study BH evaporation is that the final stagesof the evap¬ oration process involve physicsnear the Planck scale,where quantum gravity is expected to becomeimportant. Thus, BHs provide a theoretical laboratory where one can gain insight into the physics at this minimum scale. All these issuesare highly complicatedin four spacetimedimensions.To understand better theseproblemsone must resort to lower dimensiontheories.In two dimensions(one time plus one space dimension) generalrelativity is trivial, it hasno dynamics. However, if one adds a dilatou scalarfield the theory has many featuressimilar to four dimensional generalrelativity (see, e.g. [58]). There are many different theoriesin two dimensions with interesting dynamics [59][GO][61]. One that has beenextensively studied [62][63] [64]is related 1o string theory (a consistenttheory of quantum gravity, although it has problemsin delivering the other three fundamentalinteractions). In three dimensions general relativity has dynamics, although not much (the theory has no local degrees of freedom). Surprisingly, it has been found that a three dimensionalblack hole in a space with constant curvature exists [65][66]. One can connect these three dimensional theorieswith four dimensionalgeneralrelativity [67][68][69][70]. The results obtained using two, three and four dimensional theoriesto solvethe information paradox are still 66 Proceedingsof Uie XXIth SAB controversial[71]. However,theoretical experiments, involving annihilation of a pair of BH-antiBH, have shown that information can indeeddisappearaltogether,inside an event horizon [72]. Extreme BHs alsoprovide interesting results. A chargedBH is called extreme when Q = M (in geometricalunits whereG = c = 1, otherwise we can write Q = \/GM). (If Q > M then there is no horizon, instead one has a naked singularity, which if it exists complicatesthe thermodynamicpicture. That is one reason why cosmiccensorship[73], whichforbidds the existenceof nakedsingularities,is widely accepted).For extreme BHs the Hawking temperature is zero, they do not radiate. Thus they can be consideredstable particles,that do not decay.If one of theseBHs absorbsan infalling neutral particle, the BH’s mass will be increased,the charge-to-massratio is then loweredraising the Hawking temperature above zero. The BH then emits particles by the evaporation process, and returns to its ground state. This appears as a scatteringprocess, an incoming initial state of one particle is scatteredinto other particlesas a final state. There are other processes that resembleparticle physics or are connectedto other physical branches,e.g., BH-BH scatering, and the statistics a BH gas shouldobey (areBHs fermions,bosonsor neither? [M])- Physicists believe that gravity becomesthe dominant field at the quantum Planck scale 10-33cm.It represents the minimum scaleat which spacetime can be considered smooth. BHs are the objects to test this scale,through Hawking radiation, and related phenomena.Imagine the following futuristic experiment: two incoming particles in a huge acceleratorare set to collide face-on,such that, a center of mass energy of ~ 1019Gevis produced. Then, one might form a PlanckianBH which will evaporate quickly in a burst, allowingus to study the physicsat the Planck scale. Onemight think that by increasing the energy the study of sub-Planckianscaleswould follow. However,this is not the case. By increasingthe energy one would produce a BH with larger mass, which would decay slowly,not allowingany test of Planckianphysics. 5 Conclusions BHs are usedin many different phenomena,from high energy astrophysicsto high energy elementaryparticle physics. The results brought from eacharea of study, either observa¬ tional,theoreticalor experimental, will serve to gain a better understandingof thephysics of thesebeautiful objects. This review is a summary of some aspectsof the nature of BHs. Acknowledgements- I thank Horácio Dottori, Dalton Lopes and Vera Jatenco for inviting me to deliver the seminar to which this article corresponds,in the XXIth SAB meeting, 1995.I thank Verne Smith for his careful reading of the manuscript. Proceedingsof the XXllh SAI3 67 References [1] C. Misner, Thorne, Wheeler,1973, Gravitation, Freeman,S. Francisco. [2] J. P. S. Lemos,1994, CiênciaHoje, 97, 42. [3] D. Lyn den-Bell e J. P. S. Lemos,1988,Mon. Not. R. astr. Soc.,233, 197. [4] JLP. S. Lemos e D. Lynden-Bell, 1989,Mon. Not. R.. astr. Soc.,240, 317. [5] J. P. S. Lemos,1992,Phys. Rev. Lett., 68, 1447. [6] C. E. Rhoades,R. Ruffini, 1978,Phys. Rev. Lett. 32, 324. [7] J. B. Hartle, 1978,Phys. Rep., 46, 201. [8] G. Baym, C. Pethick,1979, Anna. Rev. Astron. Astroph.,17, 415. [9] S. Bachall, B. W. Lynn, S. B. Selipsky,1990,Ap. J., 362, 251. [10]E. Witten, 1984, Phys. Rev. D, 30, 272. [11] J. Michell, 1784, Phil. Trans. R. Soc. (London),74, 35. [12] J. II. Taylor, 1993, in GeneralRelativity and Gravitation 1992,Proceedingsof the Thirteenth International Conferenceon GeneralRelativity and Gravitation held at Cordoba,Argentina,eels.R. J. Gleiser,C. N. Kozameh,O. M. Moreschi,IOP, Bristol, p. 287. [13]S.L. Shapiro,S. A. Teukolski,1983,Black Holes, WhiteDwarfsand Neutron Stars, the Physics ofCompactObjects,Wiley and Sons,New York. [14] J. P. Luminet, 1992,Black Holes,CambridgeUniversity Press,Cambridge. [15]M. Anderson,J. P. S. Lemos,1988,Mon. Not. R. astr. Soc.,233, 489. [16] A. F. Tennant, A. C. Fabian, R..A. Shafer,1985, Mon. Not. R. astr. Soc.,221, 27P. [17] E. V. P. Smith, K. C. Jacobs,1973,Introductory Astronomyand Astrophysics,W. B. Sa.nn.dersand Co., Philadelphia. [18]II. M. Hjellming, A. M. Wade,1971, Ap. J. Lett., 168, L21. [19] E. P. Liang, P. L. Nolan, 1984,SpaceSci. Rev.,46, 273. [20] A. P. Cowley, et ah, 1983, Ap. J., 272, 118. [21] B. Paczyiiski, 1983, Ap. J. Lett., 273, LSI. [22] J. E. McClintock, R. A. Remillard,1986,Ap. J., 308, 110. 68 Proceedingsof the XXIthSAB [23] A. P. Cowley,P. C. Schmidtke,D. Crampton,J. B. Hutchings, 1990, Ap. J. 350, 288. [24] A. P. Cowley,1992, Annu. Rev. Astron. Aslroph.,30, 287. [25]S.D’Odorico, T. Osterloo,T. Zwitter, M. Calvani,1991,Nature,353, 329. [26] A. C. Fabian,M. J. Rees,L. Stella,N. E. White, 1989,Mon. Not. R. astr. Soc.,238, 729. [27]R. Sunyaev,et al., 1991,Astr. Astroph.,247, L29. [28]J. Van Paradijs, J. E. McClintock,1995,in X-ray Binaries, eds: W. H. G. Lewin, J. VanParadijs, E. P. Van denHeuvel,(CambridgeUniversity Press),in press. [29] J. E. McClintock, 1987, in SupermassiveBlack Holes,Proceedingsof the Third GeorgeMason AstrophysicsWorkshop-1986,ed. M. Kafatos,CambridgeUniversity Press,Cambridge. [30]L. Stella,1993,in GeneralRelativity and Gravitation 1992,Proceedingsof the Thir¬ teenth International Conferenceon GeneralRelativity and Gravitation held at Cor¬ doba,Argentina, eds.R. J. Gleiser,C. N. Kozameh,O. M. Moreschi,IOP, Bristol, p. 263. [31]D. Lynden-Bell, M. J. Rees,1971,Mon. Not. R. astr. Soc,152, 461. [32]K. Y. Lo et al, 1985,Nature 315, 124. [33]R. Genzel,C. H. Townes,1987, Annu. Rev. Astron. Astropliys., 25, 377. [34]R. Genzel,D. Hollenbach,C. H. Townes,1994,Rep.Prog. Phys., 57, 417. [35]A. Goldwurm,et al, 1994,Nature 371, 589. [36]R. Narayan,I. Yi, R. Mahadevan,1995,Nature,374, 623. [37]J. Kormendy,1988,Ap. J., 325, 128. [38] J. Kormendy, 1988, Ap. J., 335, 40. [39] J. Kormendy, D. Richstone,1992,Ap. J., 393, 559. [40]M. J. Rees,1984,Annu. Rev. Astron. Astropliys., 22, 471. [41]M. C. Begelman,R. D. Blandford, M. J. Rees,1980,Nature, 287, 307. [42]R. R. J. Antonucci,J. S. Miller, 1985,Ap. J., 297, 621. [43]H. C. Ford, et al., 1994, Ap. J. Lett., 435, L27. Proceedingsof the XXIthSAB 69 [44]M. Miyoshi, et al, 1995,Nature 373, 127. [45]L. J. Greenhill et al, 1995,Ap. J., 440, 619. [46]D. Lynden-bell, 1969,Nature, 223, 690. [47] Kormendy, Richstone,1995, Annu. Rev. Astron. Astroph.,36, in press. [48]E. Maoz,1995, Ap. J. 447, L91. [49]M. J. Rees,1988,Nature, 333, 523. [50] B. J. Carr, S. W. Hawking, 1974,Mon. Not. R. astr. Soc.,168, 399. [51] S. W. Hawking, 1979,in GeneralRelativity, an Einstein Centenary Survey, eds. S. W. Hawking, W. Israel,CambridgeUniversity Press,Cambridge,p. 746. [52] J. A. Wheeler,1963,in Relativity, Groupsand Topology,eds.B. S. and C. M. De Witt, Gordon and Breach,New York,p. 317. [53] S. W. Hawking, 1974,Nature, 248, 30. [54]S. W. Hawking, Phys. Rev. Lett., 26, 1344. [55jJ. D. Bekenstein,1973,Phys. Rev. D, 7, 2333. [56]B. Carr, 1994,Annu. Rev. Astron. Astroph.,35, 531. [57] S. W. Hawking, 1976,Phys. Rev. D, 13, 191. [58] J. P. S. Lemos,Paulo Sá,1994, Class. QuantumGravity, 11, Lll. [59] J. P. S. Lemos,Paulo Sá,1994,Phys. Rev. D., 49, 2897. [60] J. P. S. Lemos,Paulo Sá,1994,Mod. Phys. Lett. A, 9, 771. [61] J. P. S. Lemos,1995, Class. QuantumGravity, 12, 1081. [62]E. Witten, 1991,Phys. Rev. D, 44, 314. [63]C. G. Callan, S. B. Giddings, J. A. Harvey, A. Strominger,1992,Phys. Rev. D, 45, R1005. [64] G. T. Horowitz,1992, “The Dark Sideof String Theory”, hep-th/9210119. [65]M. Banados,C. Teitelboim, J. Zanelli,1993,Phys.Rev.Lett, 69, 1849. [66]M. Banados,M. Henneanx,C. Teitelboim,J. Zanelli,1993,Phys. Rev. D, 48, 1506. [67] J. P. S. Lejnos, 1995,Phys. Lett. B, 352, 46. 70 Proceedingsof the XXIthSAB [68] J. P. S. Lemos, 1994, “Cylindrical Black Hole in GeneralRelativity”, gr-qc/9404041. [69] P. M. Sá,A. Kleber,J. P. S.Lemos,1995,“Black Holesin Three-DimensionalDilaton Gravity Theories”,hep-th/9503089,to appear in Class.QuantumGravity. [70] J. P. S. Lemos, V. T. Zanchin,1995, “Rotating Charged Black String and Three Dimensional Black Holes”,submitted. [71] G. ’t Hooft, 1991,in Proceedingsof the 5th Seminar on QuantumGravity, eds. M. A. Markov, V. A. Berezin, V. P. Frolov, World Scientific,Singapore,p. 251. [72] S. W. Hawking, G. T. Horowitz, S. F. Ross,1994, “Entropy, Area, and Black Hole Pairs”, gr-qc/9409013. [73]R. Penrose,1969,Riv. Nuovo Cimento, 1, 252. [74] A. Strominger, 1993,Phys. Rev. Lett.,71, 3397. Proceedingsof the XXIthSAB 71 NEUTRINOS, DM AND THE RW UNIVERSE Justin C. Huang1 DepartmentofPhysicsand Astronomy, University ofMissouri, Columbia,MO 65211, USA My topic in this talk will focus on the formal problem of neutrinos and the RW universe. The results of this discussionare basedon work that I havedone with N.O. Santosand A. Kleber (1995). Sinceneutrinos obey Dirac’s equationand the RW universe is governedby Einstein’s equation, the system of equationsthat I will be dealing with is the Einstein-Dirac system of equations. The problemis not just of academicinterest,but can haveimportant physicalconsequences with respect to our present day understanding of dark matter in the universe.There will alsobe some ambiguitieswhich originate from assumptionsmadein the analysis. Let me begin with the RW spacetimewhichdescribesa spatially homogeneous,isotopic universe.There are strong observationalevidencesfor the spacetimedescribedby this metric. The three major ones are: 1. Hubble expansion(1929) 2. Cosmicmicrowavebackgroundradiation (1965) 3. Nucleosynthesisinvolving the relative abundanceof the light elements (1965- present ) The fundamentalequation of gravity is the Einstein equation which has the form lfn,-(\/2)gin;R = S7rC'Tin, (1) whereRIWis the Ricci tensor and lx is the scalar curvature. A cosmologicalterm which is proportional to can be added to the left, hand side of the aboveequation, but we will ignore this possibility in this talk. The RW metric has the form for the invariant distanceels (ds)2 = glivd:v,'d.r1' (2) = [tit)'2- a2[(dr)7( 1 - hr2)+r2(d,0)2+r2sin20{d <ÿ[2/-y+ 3H2+k/a2]= SirGTg (4) and RlA= SWGTA= 0 (5) 1Internet: physjch'fsmizzoul.inissouii.edu 72 Proceedingsof the XXIthSAB where the subscripts A and B can be r, 0, or ,H is the Hubble variable which is representedin terms of a by H = à/a (6) and wherethe dot stands for differentiation with respect to t. Using this sphericalbasis,we see that all the off diagonalelementsof T[?are zero: Tl= Tg=T* = TTg= T* = Tf= 0. (7) The spatialcomponents are all equal, Trr= Tjj= Tf (8) and the time component Tfand the diagonalspatial components must only be functions of t since the cosmic scalefactor a is only a function of the time t. The constraints on all the elementsof are rather severe and it is not at all obviousthat all fundamental sources of nature havethe necessary properties which can satisfy theseconstraints. Let us next turn to the Dirac world. In flat spacetime,let us write down the equation for a field with a conservedcurrent associatedwith chargeconservation, Í7°(da - (9) where Aarepresents the electromagnetic4-potential and for simplicity, a unit coupling has beenchosen.In group theory language,this is the field equation for a system with U(l) gauge symmetry. The conserved4-current associatedwith this equation has the form Ja = (10) where 'L is the adjoint field. There are various possibleinterpretation of this equation in the literature. We can interpret the aboveequation as a single particle quantum mechanicalequation. We can also normalize the field to represent N fermions and considerthe equation as alV particle classicalequation.It is of course not classicalin the same sense as Maxwell’s equationas we cannot form a continuumof states with fermions. We couldalsotreat the aboveequation as an operator equation within the frameworkof a quantum field theory. What happenswhenwe transfer to curvedspacetime?There are some straight forward modifications which must be performed. An easy way to understandthe adjustments is to replacethe local chargesymmetriesby the localLorentz symmetries. The electromagnetic 4-potential Aais then replacedby a gravity related quantity which we shall write as T/t. The flat spacetimeDirac matrices 7“ are replacedby curved spacetimeDirac matrices eU Weuse the Greekindicesa and (i to indicate flat spacetimequantitiesandthe Greek indices fi and v to indicate curved spacetimequantities. The curved spacetime Dirac equation can be written as i(A(dÿ—rft)'Ir = m’P. (11) Proceedingsof the XXIthSAB 73 In the group theory language,the U(l) gauge symmetry has now been replaced by the SL(2,C)gauge symmetry. The curved spacetimee1'is definedby the equation = V". (12) The curved and flat spacetimeDirac matrices are related by the tetrads e" = <7“- (13) In order to maintain the gauge symmetry or covarianceof the equation, there must be a relationship betweenthe tetrads and TM.It has the form r„ = (14) where (16) and TIU,= (i/ 4)[(®eÿ® + - (dÿeÿ +dv9eÿ) (17) — + T„eit)\I/— +rMe„)'ir]. In the RW spacetime,we can write the tetrads as eto = (1,0,0,0) (18) 6ra = (0,—a/v/l —kr2,0,0) (19) eoa = (0,0,-ar,0) (20) erj>a = (0,0,0, —arsing) (21) and the gauge quantity TMas rt = 0 (22) Tr = -(l/2)[á/Vl - fcrV (23) Tg = —(l/2)[àra2+Vl —kr2ctlct2] (24) I* - -(1/2) [hr sin0a3- JT- kr* sin 6a1a3— COS 0a2a3]. (25) Let me return to Eq. (1) and the Einstein-Dirac system of equations. The left- hand sideis a completely classicalquantity. Therefore,the right-hand sidemust also be completely classical which it is not in an unambiguousway. In the literature, there are 2 ways to handle this delicatematter. If we are working with quantum fields with being a field operator, we must take some expectationvalueof T sincein some sense,average valuesare classicalquantities. A secondmethod is to consider $ as a classicalfield with 74 Proceedingsof the XXIth SAB again the understandingthat it is not classicalin the same sense as the Maxwell field. This seems to us to be the lessambiguousapproach. If we use Eqs (18) to (25), the energy-momentum tensor for a Dirac source can be constructed in a completely consistent manner with the RW spacetime. Using the separationof variable method, we found a generalsolution for the mass m not zero which has the form +CC —imt = CSC £ Gnein<1'[(tan 0/2)in( 4 mjfi U-2 {v, -v) (28) U3= (to,to) V.l= (to,—w) with v — (T 1) to = (1,-1). (29) We next tried both axially symmetric and non-axially symmetric boundary conditions for the solution to T and searchedfor agreement wil.h the conditions imposed by the RW spacetime on Tp as specified by Eqs (8), (9) and the time dependenceof tlie diagonal elements. We found that the only class of solutions possibleare solutions with zero energy-momentum tensor 'i\anbut nonzero valuesfor the current as determined from Eq. (16). Theseare known as ghostsolutions and were alsofoundby Davis and Ray (1974)for neutrinosin a plane symmetric universe.The ghostsolutions represent a peculiarsituation where a nonzero field cannot be distinguishedfrom a vacuum situation gravitationally. However,in the RW case, they also correspondto the situation when the RW spacetime reducesto a flat spacetime. For completeness,we studiedalsothe masslesscase. Asis wellknown,Dirac’s equation decouplesinto two component equations with solutions which were straight forward to obtain. The results were more drastic for this case. Becauseof 7sterms appearingin there were now no solutions which could satisfy the constraints imposedon T# by the RW spacetimefor both axially symmetric and non-axially symmetric solutions. Finally, let me turn to a possiblephysicalapplicationof theseresults. We have seen that in a world of just neutrinos andgravity, the standardmodelof cosmologyas embodied in the RW spacetimecannot be madeconsistent with eachother. Can we have a world of predominantly neutrinos and gravity? bet me turn to the questionof dark matter in the universe. From the searchof mass using dynamical means, we find that there may be much missingmass in the universe. This is most obviously seen in spiral galaxieswhere instead of the damping curvesof rotational velocity as a function of distance predicted from Kepler’s third law applied to luminous matter, we have flat curves. The dynamical effectshavebeen observedfrom rotational velocity'curves at nearly all distancescalesand Proceedingsof the XXIth SAB 75 havealsobeenreported in gravitational leasing observations. The ratio from theseeffects of luminous to missingmass is around one to ten. There have been a number of possible explanation for the dark matter question. Some have argued that the force laws are incorrect for large distancesand that there is no missing mass. Most explanationinvolves missingmass of various forms. Let me mention four that are frequently mentioned. They are:(i) empty space contains mass in the form of a nonzero cosmologicalconstant, (ii) massive neutrinos (hot dark matter), (iii) axions (cold dark matter), and (iv) machos (baryonicdark matter). It is convenient to introduce an important parameter known as Í) which is a ratio of the mass density to the critical mass density (for a flat universe).D is greater than one for a closeduniverse,equalto one for a flat universe and lessthan one for an open universe. Let me quote some approximate values for Í1. For the luminous part of the universe, flium< .01. From nucleosynthesis,the baryon contribution is givenby .01 < flbaryon < .1 with the uncertainty driven by the uncertainty in the valueof the Hubble constant. The favoredchoicetoday for the total valueof 0 is one sinceit is consistent with the inflation¬ ary modelsof the early universe.If we accept the í) = 1 value,then most of the universe must be composedof nonbaryonic,nonluminousmatter. The approximateamount would be around 90 to 99%.If neutrinos are the answer to the dark matter dilemma,then the dominant equations governingthe universe will be the Einstein-Dirac system of consis¬ tent equations. Using the classicalfield interpretation of the Dirac equation, our results indicate that theseequations are inconsistent with the standard model of cosmologyas representedby the RW spacetime.This suggest that a more likely answer to the missing mass problem will residein other explanations. Recent detailed structural simulations of galactic formations seem to favor models of mixed cold and hot dark matter. These calculationsare consistent with our results. References [1] J.C. Huang, N.O. Santos,and A. Kleber, Class.QuantumGrav. 12 (1995)1245. [2] T.M. Davis and J.R. Ray, Phys. Rev. D9 (1974)331. Section 2 Teachingof Astronomy Proceedingsof the XXIthSAB 79 TEACHING OF ASTRONOMY IN BRAZIL Rute Helena Trevisan Departamento deFisica, UniversidadeEstadualdeLondrina CampusUniversitário, CEP 86.051-970,Londrina,Pr 1 Introduction Astronomy,one of the most old science,hasleadedthe basesof all human knowledge.As the time pass, the man starts a blindly searchfor the generalobjective laws,renouncingto the philosophyand to the human being. Looking for a return to a.scienceteachingwhere the historical necessitiesof looking for comprehensionof the laws that move, make and govern the natural phenomenaof the universe,the most of the curricula of the primary schoolof Brazil following the international tendencyof the most schoolsof the world, are introducing Astronomy in the Scienceclasses. 2 The professional astronomer in the basic teaching of astronomy Feeling the necessity of the improvement of the primary schoolsteachers( Bisch, 1993; Lattari et al. ,1993;Lattari and Trevisan,1993 a,b;Pereira,1993;Trevisan et al. 1993a,b) and some professionalastronomers who teach in universities,started a work in national level, having as principal objective, the teachingof astronomy in the primary and high schools,beginningwith the teachingof astronomy to the teachers.Initially, it was created the Astronomy TeachingGroupformedby some professionalastronomers, during a double¬ annual Symposium of PhysicsTeaching( X National Symposiumof Physics Teaching- SNEF)occurredin Londrina- Parana, South of Brazil (1993). It was the beginningof a large movement in the sense of teach astronomy to the children and to teachin a new way ( Buso and Canalle,1993;Crispin and Canalle,1993; Daminelli,1993;; Espositoand Canalle,1993;Nozawa,1993;Oliveira,1993;Souza,1993). The initial objectives of the Astronomy TeachingGroup were : 01) developmentsof didactical equipment and proceduresfor astronomy teaching; 02) incentive to the editions of astronomy books; 03) to offer improvement courses for teachers; 04) to orientate students of graduation in Astronomy,at ScientificInitiation level; 05) to developscientific divulgation activities; 06) to produceeducative videos 07) to interact with other groups in Brazil and in the world. Historically, we can say that the principal events and activities in teachingof astron¬ omy, after the creation of this Group, occurred during the following events: a) XIX ANNUAL MEETING OF BRAZILIAN ASTRONOMICAL SOCIETY - SAB 80 Proceedingsof the XXIthSAB Caxambu,August, 1993 * Teachingof Astronomy Round Table;* Teachingof Astronomy Poster Section;* Assemblysuggestionof Creation of the Commissionof Astronomy Teaching b) XX ANNUAL MEETING OF BRAZILIAN ASTRONOMICAL SOCIETY - SAB Camposdo Jordão,August, 1994 * Teachingof Astronomy Round Table * Teachingof Astronomy Poster Section( Afonso,1994; Araujo, 1994; Bisch, 1994; Canalle,1994; Livi, 1994 ; Matsuura, 1994; Rocha, 1994; Trevisan, 1994a,b; Varella, 1994a,b,c.) * Creation of the : Commission of Astronomy Teaching Brazilian Astronomical Societyduring the GeneralAssemblyof SAB. The initial objectives of this commission were: 1) To make contacts with the Physics TeachingSecretary of the Brazilian Society of Physics, to changeideasabout teaching; 2) To increment the activities of Astronomy Tea,ching(Courses,Hands on, Talks,etc.) in Brazilian Symposiums,in the next years, 3) to follow all the objectivesinitially proposed( listed above). c) XI SNEF ( National Symposiumof Physics Teaching) Niterói, RJ, January, 1995 (seeAtas do XI SimpósioNacional de Ensino de Física 1995) * Poster Section (Jafelice,1995a;Lattari and Trevisan,1995a;Makler,1995;Nunes, 1995; Pereira,1995; Pereira and Barros, 1995, Trevisan et ah, 1995a,b; Trevisan and Lattari, 1995) * Coursesand Hands-onfor Teachers(Canalle,1995a;Martins, 1995;Trevisan,1995a) * VideosSections;* Showof Astronomy; * Theater; * Oral Communication (Araújo, 1995; Makler and Martins, 1995; Nascimento and Bittencourt, 1995;Nascimento,1995; Nevesand Pereira,1995;Neves d ah. 1995;Nunes and Zanon,1995b;Tignanelli, 1995;); * Meeting on Teachingof Astronomy(Livi and Bisch, 1995) d) ANNUAL MEETING OP BRAZILIAN SOCIETY TO THE PROGRESSOF SCI¬ ENCE (SBPC) : * Hands-On ( Canalle, 1995b;Jafelice, 1995b) e) INTERNATIONAL SCHOOL FOR YOUNGERAS TRONOMERS- ISYA - IAU Belo Horizonte, MG. .July. 1995 * Course of Teachingof Aslronomy( Percy. 1995); * Talks of Teachingof Astronomy (Guerbaldi, 1995a;Livi. 1995: Trevisan.1995b); ' Teachingof Astronomy Round Table f) XXI ANNUAL MEETING OF ASTRONOMICAL BRAZILIAN SOCIETY Caxambu, MG, August, 1995 * Talks of Teaching of Aslronomv (Guerbaldi, 1995; Trevisan,1995c); * PosterSectionabout Teachingof Aslronomy ( AlonsoandNadal, 1995; Canalle,!995b; Campos,1995; Faraco et ah. 1995: Sou/.ad ah. 1995and Trevisan, 1995 Talks,Hands On and Showsin a MovablePlanetarium.(Boczko,1995; Canalle,1995c; Jafelice,1995b;Magalhães,1995). Besidesthese,some important projects of teachingof Astronomy occurredin Brazil. They are describedbelow. 3 Some important projects of teaching of astron¬ omy in Brazil 3.1 The Reformulation of the Curriculum in Elementary School The most important project of Teaching of Astronomy in Brazil in the last years has been the reformulation of the curriculum of the Primary Schools,including big contents of Astronomy, generally,in Science,in the most states of the country. The TeachingAstronomy Commission workedin this project together with the Gov¬ ernment Education Technician, whenit was possible. The TeachingAstronomy Commissionanalyzedthe curriculum of 15 Brazilian States (Trevisanet al. 1995c),including the FederalDistrict (Brasilia). It was made a survey of the extension whereAstronomy appearsin the curricular structure of the primary and high school. This study tried to identify the problemsand the deficiency,to suggestthe way to use astronomy as a manner to stimulate and to improve the Teachingof Astronomy. Thiswork detectedthat Astronomy is present in ElementarySchools( Children Ages: 6 to 13 ) in some of the eight years of school. In High School,Astronomy is absent of the most curricula, with only some topics suggested( Gravitation and Cosmological Concepts).We found the followingproblems:curriculadeficient approaches;badprepared teachers;serious mistakes in the books and no pedagogicalmaterial. There are only one exception where we have Astronomy in all the series of the Elementary Schools. This exception is the curricular reformulation occurredin Parana State. They made a reformulation in the primary school,in the nine first years, with basein the contents of pedagogicalideas. It was a work that began in 1987 and evolvededucatorsof all TeachingGovernments Groups. This project has as a start point the implantation of a BASIC ALPHABETIZATION CYCLE, that permits the systematic progress of the child in the knowledgedomain. This curriculum proposed three axes of knowledge: Health, Matter/Energy and Astronomy. A resume of this curriculum is below. The Paraná Basic School Curriculum The three axes proposed,must give an opportunity to appropriate of contents in a per¬ spectiveof totality; they must developthe work with the fundamentalconcepts and their inter-connections. The contents must enable the discovery of relations between an axis with the other axes, permitting to forme one link of contents, giving a total perspective of the reality. The content of the nature sciencemust have the fundamentalin the multiple relation of 82 Proceedingsof the XXIthSAB interdependenceof the elementsthat constitute the ecosystem and betweenthe ecosys¬ tems. The objective is to give the opportunity of a clear idea, of the dynamism of the various elementsof the systems: physical, chemical and biological,having as a guide the transformer action of the man, interfering in nature. The contents of Astronomy are includedin sciencediscipline, alwaysin relation with the other axes, never isolated. To havean idea of the contendsof astronomy, it is listed below. Paraná Curriculum - Astronomy Contents GRADE 7 - Sun: primary source of energy. Day and Night. GRADE 8 - Sun: primary source of energy. Earth Movement: a) Referential;b) Rotation: day and night Orientation: a) CardinalPoints GRADE 9 - Sun: primary source of energy: a) Heat source; b) Light - Solar Spectrum. Earth Movement: a) Referential;b) Translation;b) Rotation: Gravity. Another Celestial Bodies: a) Illuminated: moon/planets/asteroids;b) Illuminated: Stars. GRADE 10 - Sun: primary source of energy: a) energytypes and transformations;ID) IR/UV/Itsinfluenceon biosphere. GRADE 11 - Astronomy: historical aspects. Matter and Energy: basic elements of Universe- General considerations. Solar System: Sun; Planets and satellites. Sun and Moon influence on living things. Translations Movements: a) differencesof Sun/Moon trajectories ( apparent movement); b) night and day: differencesof duration. GRADE 12 - Solar System: basic and generalphysicaland chemicalconditions for the analysis of the possibletransformations of matter and energy. Physical and Chemical conditionsof planetspermitting or not the life. GRADE 13 - Earth Planet. Earth position in Solar System. Inorganic Sphere:a) Hydro¬ sphere;b) Lithosphere;c) Atmosphere.Tides. GRADE 14 - Sun: source and heat and energy. Astronautic: developmentand applica- tions. 3.2 The Divulgation Project of Total Solar Eclipse of Novem¬ ber, 3, 1994 Oneof the most important projects of Teachingof Astronomy,in the last times, was the Project of Divulgation of the Total SolarEclipseof November,3, 1994,that occurredin the Southof Brazil. It was createdby AstronomicalBrazilian Society,under the recommendationof Inter¬ national AstronomicalUnion, a commission called ”Eclipse94Executive Commission ”, (Matsuura et al. 1994)tocoordinate the assistancein arrangements for the observation of the total solareclipseof 3rd of November,1994,that in Brazil, was total in the South of Brazil.. Professionalastronomers from Brazil and from severalparts of the world were Proceedingsof the XXIthSAD 83 mobilized in observingthis eclipse. The generalcoordination was done by. Dr. Oscar Matsuura from the Astronomicaland GeophysicalInstitute of University of SãoPaulo ( IAG/USP), which workedwith State Coordinators, eachone of one state of the South and South-WestStatesfrom Brazil. Following the suggestionof the Working Group on Eclipsesof the International As¬ tronomical Union, this commissiondecidedamplify their action, assumingthe coordina¬ tion of a large elucidation campaign about eclipses,closeto the common people. Such campaign was aimed at giving technical and astronomicalinformation and to prevent ophthalmologicalaccidentsin the peopleduring the eclipse.Utilizing this fact, the Exec¬ utive Commissiondecidedto use this campaignto collaboratewith the teachers,mainly in Elementary Schools,in science classes,motivating the students to observethe sky phenomena. With the help of Regional Supervisorsand SpecialMonitors , chosenbetweenteachers from Universities, from High Schoolsand from ElementarySchools,educationprofession¬ als from State GovernmentEducation Department, City Hall, Cultural Center, Second Grade Schoolsand Physics Students,Planetarium Technitian,and others,the Executive Commission,promoted: courses, talks, fairs, interviews in newspaper, radio, television, divulgationspapers, posters, folders, elaboratedand distributed didactic experiments and teachinginstruments for children. It’s impossible to cpiantify the total population attended in the area of this cam¬ paign, but as indicated the cpiantificationmadein Parana State,where all studentsof the Elementary and High Public Schools (1.207.832students); all teacherstoo (2.040);the studentsof Primary Municipal Schoolsin all Cities of the State; journalists(60),Educa¬ tion Technician from Official Organisms(412)where attended, we suppose that 80 % of population from South of Brazil was reachedby the National.Project Eclipse 9j. The Commission of Astronomy Teachingworked directly in this project, making the coordi¬ nation in some states of Brazil. 3.3 Courses for Elementary Schools Teachers. SeveralUniversities,together with Education Technicianof the City Hall, including some Planetariums, are working with teachersof Elementary and High Schools,teachingsome basic concepts of Astronomy. In general, the courses focalize the content, considering the most practical wa.y to teach it to the children,what is make with the help of simple experiments that show to the student the fundamental relations between the men and the space/time/energyof the ambiencethat involves him. The Universities that offers regular courses of Astronomy for teachersare: University of Londrina-Pr; Federal Uni¬ versity of Espírito Santo- ES; FederalUniversity of Rio Grande do Sul - RGS;Federal University of Viçosa- MG; FederalUniversity of Minas Gerais- MG; FederalUniversity of Rio Grande do Norte- RGN; State University of SãoPaulo- SP;University of Campinas- SP; FluminenseFederalUniversity - R.J;State University of Rio de Janeiro - RJ, among others. 84 Proceedingsof the XXIthSAB 4 Conclusion The Commissionof Astronomy Teaching,workedhard in the three last years, presenting some goodresults,mainly of interest of the teachersof the Elementary and high school. As a matter of fact this work was an answer to the longing of the Elementary and High Schoolteachersthat haveto teachbeforelearningastronomy. This work is not only a responsibility of a few astronomers. There is a lot of work to be done and it must be done by the community of astronomers, that can collaboratein many ways. Our proposal is to chancethe actual situation where we found students of physics graduationthinking that Venusis a star, the Earth orbits is a very eccentricellipse and that the moon is illuminated in different ways in eachphase,for example. We are convincedthat the level of popular ignorance about astronomy nowadays is, at least in part, responsibility of all that work with astronomy. The community of professionalastronomers can changethis situation. We noted also that, the initiative of Brazilian Astronomical Society to help the Elementary schoolsteacher is very well acceptedby them,and they really hope that we carry on this work,becausethere are, as we said,a lot of work to be done. The author would like to thank to the Commission of Teachingof Astronomy of the Brazilian Astronomical Society that makethis goodwork. We also thank to JoãoBatista Canalle(\]ERJ) for his comments on the manuscript and to Cleiton Joni BenettiLat- tari(DEL), for useful discussions. 5 References - Afonso,G.B. 1994,Boletim da SociedadeAstronómica Brasileira Vol. 14, 1,128 - Afonso,G.B.; Nadai, T.M. 1995,Boletim da SociedadeAstronómica Brasileira Vol. 15,1, 42 - Araujo, N.M.1994,Boletim da SociedadeAstronómica Brasileira Vol. 14, 1, 129 - Araújo, N.M. 1995, Atas do XI SimpósioNacionalde Ensino de Física SBF, 397 - Bisch,S. 1993,Atas do X SimpósioNacionalde Ensino de Física SBF, 575 - Bisch,S. 1994,Boletim da SociedadeAstronómicaBrasileira Vol. 14, 1,131 - Boczko,R. 1995,XXI ReuniãoAnual da SociedadeAstronómicaBrasileira - Buso,S.J.;Canalle,J.B. 1993,Atas do X SimpósioNacionalde Ensino de Física SBF, - Campos,J.A.S. 1995,Boletim da SociedadeAstronómicaBrasileira Vol. 15, 1,46 - Canalle,J.B.1994,Boletim da SociedadeAstronómica Brasileira Vol. 14, 1, 131 - Canalle,J.B. and Jafelice,L.C 1995. 27“ Reunião Anual da SBPC, - Canalle,J.B. 1995a,Atas do XI SimpósioNacional de Ensino de Física SBF,95 - Canalle,J.B. 1995b,Boletim da SociedadeAstronómica Brasileira Vol.15,1,42 - Canalle,J.B. 1995c,XXI Reunião Anual da SociedadeAstronómicaBrasileira Proceedingsof the XXIUlSAB 85 - Crispin, S.C.; Canalle, J.B. 1993, Alas do X SimpósioNacionalde Ensino de Física SBF, - Daminelli Neto, A. 1993,Atas do X SimpósioNacionalde Ensino deFísica SBF, - Espósito,I.M.; Canalle,J.B. 1993, Atas do X SimpósioNacional de Ensino de Física SBF, - Faraco,S.A.;Trevisan,R.H.; Sanzovo,G. Boletim da SociedadeAstronómicaBrasileira Vol.15,1,45 - Guerbaldi,M. 1995a,IAU International Schoolfor YoungAstronomersin Brazil - Guerbaldi,M. 1995b,XXI ReuniãoAnual da SociedadeAstronómicaBrasileira - Jafelice,L.C.1995aAtas do XI SimpósioNacionalde Ensino de Física SBF,179 - Jafelice,L.C.1995b, XXI Reunião Anual da SociedadeAstronómicaBrasileira - Lattari, C.J.B; Trevisan,R.H.; Romano,E.B. 1993, Atas do X SimpósioNacionalde Ensino deFísica SBF 584 - Lattari, C.J.B.; Trevisan,R.H. 1993a,Atas do X SimpósioNacionaldeEnsino de Física SBF,487 - Lattari, C.J.B.; Trevisan,R.H. 1993b,Atas do X SimpósioNacional deEnsino de Física SBF,501 - Lattari, C.J.B. and Trevisan,R.H. 1995a Atas do XI SimpósioNacional de Ensino de Física SBF,164 - Lattari, C.J.B. and Trevisan,R.H. 1995bAtas do XI SimpósioNacional de Ensino de Física,SBF,166 - Livi, S.H.B. 1994,Boletim da SociedadeAstronómica Brasileira Vol. 14,1, 132 - Livi, S.H.B. 1995,IAU International Schoolfor Young Astronomersin Brazil - Livi, S.H.B. and Bisch, S. 1995, Atas do XI Simpósio Nacional de Ensino de Física SBF,106 - Magalhães,A.M. 1995, XXI ReuniãoAnual da SociedadeAstronómica Brasileira - Makler,M. e equipe,1995 Atas do XI SimpósioNacional de Ensino de Física SBF 177 - Makler,M. and Martins, R.A. 1995,Atas do XI SimpósioNacionaldeEnsino de Física SBF,395 - Martins, R.A. 1995,Atas do XI SimpósioNacional deEnsino de Física SBF,94 - Matsuura,0.1994,Boletim da SociedadeAstronómicaBrasileiraVol. 14,1, 130 - Matsuura, O. et al. , 1994,Apostila do Eclipse,Editora IAG/USP - Nascimento,S.S.1995,Atas do XI SimpósioNacionaldeEnsinode FísicaSBF, 397 - Nascimento,S.S. and Bittencourt, M.K.C. 1995, Atas do XI SimpósioNacional de Ensino de Física SBF, 396 - Neves,M.C.D. and Pereira,J.R.D. 1995,Atas do XI SimpósioNacional de Ensino de FísicaSBF,403 - Neves,M.C.D.; Costa,L.G.; Ichiba, R.; Lima, S. 1995, Atas do XI SimpósioNacional de Ensino de Física SBF, 403 - Nozawa,A. 1993, Atas do X SimpósioNacional de Ensino de Física SBF, - Nunes,M.R. 1995 Atas do XI SimpósioNacional de Ensino de Física SBF, 158 - Nunes,M.R. and Zanon,R. 1995, Atas do XI SimpósioNacional de Ensino de Física SBF,403 86 Proceedingsof the XXIthSAB Oliveira, I.A.G.; Canalle,J.B. 1993,Atas cioX SimpósioNacional deEnsino de Física SBF, - Percy, J. 1995,IAU - International Schoolfor Young Astronomersin Brazil - Pereira,O.S. 1993, Atas do X SimpósioNacionalde Ensino de Física SBF, - Pereira,O.S. 1995,Atas do XI SimpósioNacional de Ensino de Física SBF, 181 - Pereira,O.S.and Barros,M.P. 1995 Atas do XI SimpósioNacional de Ensino deFísica SBF,182 - Rocha,J.F.V.1994,Boletim da SociedadeAstronómicaBrasileira Vol. 14, 1,133 - Souza,M.O.;. 1993,Atas do X SimpósioNacionalde Ensino de Física SBF, Souza,E. ;Trevisan,R.H.; Nabarro, R. 1995, Boletim da SociedadeAstronómica BrasileiraVol. 1,15, 45. - Tignanelli, H. Atas do XI SimpósioNacionalde Ensino de Física SBF, 394 - Trevisan,R.H; Romano,E.B.; Lattari,C.J.B.,. 1993a.,Atas do X SimpósioNacionalde Ensino de Física SBF, 579 - Trevisan,R.H.; Dotta, E.; Pedron,I.; Machado;Richter, M.; Lattari, C.J.B. 1993b, Atas do X SimpósioNacionalde Ensinode Física SBF, 397 - Trevisan,R.H.1994a,Boletim da SociedadeAstronómicaBrasileiraVol. 14, 1,128 - Trevisan,R.H.1994b,Boletim da SociedadeAstronómicaBrasileira Vol. 14, 1,130 - Trevisan,R.H. 1995a,Atas do XI SimpósioNacional de Ensino de Física SBF,93 - Trevisan,R.H. 1995b,IAU InternationalSchoolfor Young Astronomersin Brazil - Trevisan,R.H. 1995c,XXI ReuniãoAnual da SociedadeAstronómicaBrasileira - Trevisan,R.H. 1995c!,Boletim da SociedadeAstronómica Brasileira Vol.15, 1, 43 - Trevisan,R.H.1995e,VIII Latin- AmericanRegionalMeeting of Astronomy - Trevisan,R.H.; Lattari, C.J.B. 1995 Atas do XI Simpósio Nacional de Ensino de Física,SBF,170 - Trevisan,R. H.; Bruno, A.T.; Faraco,S.A.1995a Atas do XI SimpósioNacional de Ensino de Física SBF,155 - Trevisan,R.H.; Souza,E. de;Nabarro,R.A. 1995bAtas do XI SimpósioNacional de Ensino de Física SBF 174 - Trevisan,R.H.; Biscli, S.; De Campos,J.A.S.;FonsecaRodrigues,O.P.; Jafelice, L.C. 1995c,VIII Latin- AmericanRegionalMeetingof Astronomy - Varella,I.G.V.1994a,Boletim da SociedadeAstronómica BrasileiraVol. 1-1,1. 134 - Varella,I.G.V. 1994b,Boletim da SociedadeAstronómica Brasileira Vol. 11,1, 131 - Varella,I.G.V. 1994c,Boletim da SociedadeAstronómicaBrasileiraVol. 14, 1, 135 Proceedingsof the XXI1'1SAB 87 ASTRONOMY EDUCATION AND THE IAU Michele Gerbaldi Institut d'Aslrophysique, 98 bis Bd Arago, 75014Paris France and Universilê de Paris Sud XI Introduction The importance of Astronomy Education has been stressedmany times (seefor exam¬ ple the first Colloquium on TeachingAstronomy (IAU Colloquium 105 edited by Jay M. Pasachoffand John Percy, Cambridge University Press,1988) or the Newsletters published by the Commission Teaching of Astronomy. We shall point out here only some of the arguments developed. Astronomy Education is important in two different domains : - the quality of astronomy in the next century rely upon the quality of education given to-day to the undergraduateand graduatedstudents. - the degreeof support of the public for astronomy rely upon the public’s under¬ standing and appreciation of our science. This one is comingpartly from the teaching of astronomy in the schools,the collegesand the university specially to the non-scientific students and also through to the many channelsto the public. Why the non-scientific student ? very often the leadersof our countries are graduated in non-scientific domains. There is also another reason to developastronomy education: ’’math phobia” and fears of scienceare endemic everywhereand astronomy taught at basic level can be of great help because this science is very attractive to the youth. So, astronomy can inspire school children to study math and sciences,but it works only if the teacher is inspired first. To fulfill this role for astronomy, very often, one first think of the activities developed by the amateur astronomers. In some domains they can play an important role, but as soon as we are dealing with astronomy education of the school teachers,professional astronomers must be present. This is not a problem for the countries where astronomy is developedon a large scale through the universities, observatories or dedicated centers for research. So, we shall focus our attention on the countries with a ’’lone astronomer” that is the astronomer who facilitates astronomy at any level,being the only one in the country. For example the achievementsof Mazlan Othinan in Malaysia, Maria Luisa Aguilar in Peru or Alexis Torche Borgnigno in Paraguay are remarkable, just to take a few examples. 88 Proceedingsof the XXIth SAB It exists another situation which is an extended country with few well developedastro¬ nomical centers as well as severalsmall astronomical groups spreadall over the country, often far from eachother : India, Mexico, Brazil are such examplesamong others. In thesetwo situation : lone astronomer or small group of astronomers what are their needs? - to meet with other astronomers - needs for books and journals : a small group must face to a competition with other larger departmentsin physicsof the same university in order to buy the needed publications in astronomy. - a goodINTERNET link : to-day it is as important as a library. - the training as .astronomers, of their undergraduatedstudents that is MSc and then PhD : collaborations with other institutions are neededbecausevery often their university doesnot offer a degreein astronomy. The main programmes developedby the IAU in thesedomains are presentedbelow,but the efforts in those directions are not exclusively done by the IAU. Countries play also an important role in these domains,Brazil is one of them : at the universities of Sao Paulo or Rio de Janeiro, students coming from Uruguay, Argentina, Peru for example, are preparing their MSc or PhD. The International Astronomical Union The IAU is a non-governmentalscientific union founded in 1919 to ” provide a forum where astronomers from all over the world can developastronomy in all aspects through international cooperation”. There are currently 60 countries adhering to the Union and 7839 individual members. Individual membership is by nomination,and no feesare to be paid to becomea mem¬ ber. A global contribution is paid by eachadhering country. The nomination is based on qualification - usually a Phcl plus some years of experience. The nominations are proposedon a national basis,by eachNational Astronomical Committee which are ruled by the SciencesAcademy. Individual membershipcan be applied directly to the Union. The scientific activities of the Union are organizedthrough the 40 Commissionswhich refer to all astronomicaldomains,one commissionis concernedwith education. The headquarter is based at Paris. Most of the IAU funds are coming from adhering countries as well as some International Society such as UN. GeneralAssembliesare held every 3 years and the IAU support symposia, colloquia. Many other meeting are co¬ sponsoredwith other international scientific Unions and Committee within the structure of the International Council of Scientific Unions (ICSU). The IAU is governedby an Executive Committee , composedof a President, six vice- Presiclents,a General Secretary and an AssistantGeneralSecretary. The IAU publishes a semi-annualInformation Bulletin which is send to eachmember as well as the libraries. Proceedingsof the XXIth SAB 89 The IAU is the sole authority for assigningdesignationsand names to celestial bodies and the surface features there on. The IAU is primary an associationof researchastronomers and programmes are devel¬ oped by the Union in order to promote international cooperation in various domains. In the framework of the international cooperation the isolated astronomers must not be forgotten : this is the role of the Commission38 ’’Exchangeof Astronomers”;there is also a Working Group for the World Wide developmentof astronomy. The Commission 46 ’’Teaching of Astronomy” act’salsoin this line. Commission 38 : Exchange of Astronomers This Commission was createdin order to help isolated astronomers to collaborate in various domains. For example,in the period 1988-90, 35 travel grants were given for that purpose. For the period 1991-93, 38 were given which represent 10 percent of the budget allocated during the same period for the symposia and colloquia. Commission 46 : The Teaching of Astronomy This Commission was establishedrecently in 1964, as a special Committee of the Exec¬ utive Committee of the Union. The main aim of this Commissionis to act for the developmentof astronomy education at all levels. This Commissionhas about 100 members including representative from each country adhering to the IAU, consultants from countries not yet part of the IAU. The most important membership of the Commission consists of national representa¬ tives approved by adhering organizationsof respective countries. Their duties are to disseminateall the informations on the activities of the Commissionin their countries, as well as to give to the other membersof the Commission informations concerning the developmentof astronomy teachingin their countries. This Commission as well as the others is governedby a President (currently J. Percy, Canada)a vice-President (currently Julieta Fierro, Mexico) and an Organizing Com¬ mittee of individuals responsiblefor scientific projects and programmes. A Newsletter is published. It is a link between all those interested by education in Astronomy. This Newsletter has been gradually transformed into an electronic one except for thosefor which the e-mail doesnot exist or is prohibitively expensive. To receive the electronic version of the Newsletter contact Armando Arellano Ferro : armandoli'astroscu.unam.nix In 1988. for the' first time an international colloquium on Astronomy Education was organized. A second OIK- will be lielded at London in July 1996. During the IAU 90 Proceedingsof the XXIth SAB GeneralAssembly and the RegionalMeetings, sessionsdealingwith different aspects of astronomy education are helded,as well as one-clayworkshopfor local schoolteachers. Thesesessionsprovide the opportunities to establish closercontacts between teachers and astronomers. The last suchmeeting helded‘in Utretch, Netherland,in August 1994, attracted more than 180 teachers. Concernsis felt that many scienceteachersare not trained to deal with the introduction of astronomy at School. Obviously this is also the case with primary school teachers having a curricula in social sciencesdomain,more than in physics. The organization of suchmeeting is not enough; teaching material and ideas debated at a meeting will only be usedif there is a follow-up. Such follow-up can be organizedif a suitable connection is established between as¬ tronomers and schoolteachers. This can be done through the creation of a non-profit organization in order to establish and maintain suchlinks. The objectives and methods of suchorganization can be summarizedas the following : - to give access to theoretical knowledgethrough practical activities - to increasethe mood of observingand experimenting - to push the teachersworking at various level or on different subjects to exchange their experimentation and to dialoguein order to overcome the barriers between disci¬ plines and teachingorder - to produce and distribute : low cost and good quality educational material easy to use not extremely spendersof time well tested from an educationalpoint of view. The driving idea in this domain is the following : to create a network of ’’resource persons” among the teachers,the starting point being the ’’training of the trainers” by astronomers. The feedbackcomes from the network of the Association’smembers. In France,we have developedAstronomy among the schoolteacherswith suchmethods, launchingin 1977,the CLEA : Comité de Liaison Enseignants Astronomes; after several years it proved to be an efficient way. Visiting Lecturer Programme (VLP) The Visiting Lecturers Programme (VLP) was developedin order to provide a series of visiting lecturers to astronomically developing countries to establish on a longer term basis astronomy. The IAU provided the lecturer’s travel and the host institution supported the lecturer while giving the courses. The students attending thesecourses were receivingan academiccredit. The coordina¬ tor of the VLP was ProfessorDonat G. Wentzel. Proceedingsof the XXIth SAB 91 Why such a programme : simply becausewhen the ”lonely astronomer” disappearsfor any reason, astronomy disappearstoo. Astronomy can exist only if there is at least a small but strong group of astronomers in a country. The Visiting Lecturer Projects started in two countries where the universities agreed to a noticeable growth of astronomy. The countries were : Peru and Paraguay. In each case, there was already an astronomer teachingat the university. The aim was to provide an astronomy course annually in each country during several years. Then in a secondphase,students shouldbe sent abroadfor advancedstudies and they would return. In such a way an astronomical community should be created. In Peru,the programme started in 1984,and in Paraguayin 1988. The lecturers had to speakSpanishand it has been extremely difficult to find Spanish speakingastronomers who can afford to spendabout three months in Peru or Paraguay. So,the courses havebeen delayedin time. Nevertheless9 lecturers went to Lima , and among them Prof. Horacio Dot tori. Now, astronomy is very active at the Universidad Nacional Mayor cle San Marcos : there is a Seminário de Astronomia y Astrofísica, and young astronomers are studying in Bresil either for a MSc or a. Phd. The VLP in Paraguay at the University Nacional de Asuncion started in 1988 and ended in 1994. 6 visiting lecturers participated to it. Severalformer students now teach astronomy in Paraguay schools. Two students are continuing their training abroad in Mexico and in Argentina. The VLP has been cancelledand another project created : TAD ” Teachingfor As¬ tronomy Development” more flexible. It should start at the end of 1995 and will be widely announced. In common with the VLP , the goal is to enhance astronomy at selectedhost institutions on a long-term basis. Requestsfor more informations on this new programme should be sent to : Dr. Donat G. Wentzel,Dpt of Astronomy, Univer¬ sity of Maryland, CollegePark MD 20742,USA, e-mail : [email protected],fax : 1 301 314 9067. The International School for Young Astronomers (ISYA) The ISYA is one of the most important project of the Commission46. It started in 1967 with Josip Kleczek(Czechoslovakia)as the first secretary now retired and followedby Donat G. Wentzel(USA) and Michele Gerbalcli(France)as Assistant Secretarysince 1991. The dominant goalof ISYA is to advanceastronomy in astronomically developing countries. This is accomplishedby increasing astronomy in the host institution. 92 Proceedingsof the XXIth SAB An ISYA lasts 3 weeksand eachschoolmust have at least a third of its students from outside the host country, but very often it is half and half. Participants’ background ranged from just-finished bachelor’sdegreeto a PhD already started . They come usually from small universities with essentially no astronomy resources to establish researchcenters. The Schoolinclude lectures,starting for some of them at a level of undergraduate up to a graduatelevel including survey providing ’’researchflavor”. The ISYA includes also as far as possible,observationalwork and data analysis. The scope of the ISYA are : - to broaden the academicexperienceof the participants - to show the students the broaden aspect of astronomy - to enlargestudent’s scientific perspectives It is important to have faculty membersable to stay if possible,for the whole duration of the ISYA in order to give to the participants the opportunity of many informal scientific individual discussions;anotherimportant factor is to alwayschoosethe school’s participants from different countries,of different background,to get the discussionmore lively. Most studentsgive a brief seminar about their own university or observatory and they present their own research. It is necessary to organizethe ISYA in developedinstitutions with access to local li¬ braries, computers. Observingfacilities are useful even if it is a small telescope.For such purpose, when necessary the Traveling Telescopecan be used. This Celestron-8 telescopehasbeenfundedby the CanadianCommittee for UNESCO with some other in¬ stitutions and is usedin countries where astronomical researchis developing.Currently the photometer of this telescopeis usedin Paraguay on a telescopelocated there. An ISYA cannot take place in an university which are far away from astronomical centers. Travel cost of all faculty and participants are coveredby the IAU, while the local costs, must be coveredby the host institution. That is the reason that not all institut can apply for an ISYA, since they must have some local financial support. The budget of the IAU for the ISYA is about 16 percent of the budget for Colloquia and Symposia. There are about 2 ISYA every 3 years. The faculty members are chosenby the host institution according to their dominant research line. Very often they are foreign astronomers having already well established collaboration or with the aim to developit. This year the ISYA was held at Belo Horizonte (Minas Gerai) July 9-29. The Brazilian sponsors were the UFMG, CNPq (CentreNacionalda Pesquisa) and the Fapemig. Proceedingsof the XXIth SAB 93 The local organizing Committee was chaired by Prof. Renato Las Casas.The members of the LOG were : Tulio Jorge dos Santos,Rodrigo Dias Tarsia, Domingo S. de Lima Soares,Luis Themystokliz SanctosMendes,Bernardo Reidel. For the first time lectures on teachingastronomy has been organized as well as a panel discussionon the astronomy at university level. The first two weeks,we met at the University Federalof Minas Gerais(UFMG), Belo Horizonte,the third weekwas at the Observatory SerraPiedade,runs by the Astronom¬ ical Group of the ICEx (Departmentof Phisics). The Observatory is located at some 50 km from Belo Horizonte,at an altitude of about 1800m. The foreign faculty invited were : Dr. Craig Gullixson (National Solar Observatory - USA) CCD Detectors Dr. Jens Knude (CopenhagenUniversity Observatory - Denmark) Interstellar Medium Dr. Reynier Peletier (Kaptein Astronomical Institut - Netherlands)Galaxies Dr. John Percy (Toronto University - Canada)Teachingof Astronomy - Variables stars. Dr. J. Percy is the President of Commission46 ’’Teachingof Astronomy” Dr. Bo Reipurt (ESO - Chile) Star Formation Dr. Silvia Torres Peimbert (ObservatórioAstronomico Nacional - Mexico)Practical Astronomy, Observations Dr. Michèle Gerbalcli (Institut d’Astrophysique de Paris, Université de Paris Sud XI France) Teachingof Astronomy, Practical Observations. The Brazilian faculty were : Dr. Silvia Livi (Porto AlegreUniversity) Teachingof Astronomy Dr. Rute Trevisan (Londrina University)Teachingof Astronomy Dr. Luis Paulo Vaz (UFMG Belo Horizonte)Stellar Evolution Two seminars were given by the Brazilian faculty of Belo Horizonte University : Dr. Quirogaand Dr. Soares. 38 students participated to this ISYA, among them 19 came from Argentina,Bolivia, Colombia,Paraguay,Peru, Spain and Uruguay. The feed back of these Schoolsare measured 2 years letters through a questionnaire sent to all the participants. The result for the previous ISYA were : - new scientific collaboration betweenthe participants - strong influence on the choiceof the subject for a MSc or PhD - broader view of sciencein general and how sciencecan serve educational purpose - spreadingthe friendship among young scientists 94 Proceedingsof the XXIth SAB Conclusion Many initiatives have been developedby the IAU but much more are neededand they could be achievedonly if more professionalastronomers are involved in it. Acknowledgement. I greatly appreciated the invitation of the SAB to participate to this Annual Meeting Section3 Instrumentation Proceedingsof the XXIthSAB 97 IMPROVING POLARIZATION-SENSITIVE VLBI IMAGES WITH MICROWAVE-LINKED DATA TRANSMISSION SYSTEMS Everton Liidke [email protected] UFSM-LACESM,Centrode Tecnologia 97119-900,SantaMaria/RS,Brazil Abstract The MARK Ilia data recording system has become standard in VLBI acqui¬ sition systems for radioastronomy and geodeticmeasurements. Only recently the system has been usedfor interferometric circular polarization measurements which is a powerful tool to study magnetic fieldsin radio sources. In this talk,I review the recent advancesin the field and and present recent results of an extensive investiga¬ tion on a new techniquewhich uses microwaveradio links to improve the quality of radio maps by using joint observationswith the availablehardware of Multi-Element Radio-Linked TelescopeNetwork (MERLIN) and the European VLBI Network. 1 Introduction Circular polarization measurements of microwaves emitted by cosmicradio sources are a very powerful tool to study the dynamics of luminous radio sources in active galactic nuclei and the nature of magnetic fields in space. The advent of long-baselineinterferometry technique in which antenna separations can reach up to an Earth radius has made high sensitivity and resolution polarimetric observationspossible. Technical limitations of previous digital data recording systems such as the now-extinct NRAO MARKII system (Clark, 1973)did not allow to record both handsof circular polarization (LCP,RCP)simultaneously.The MIT/NASA/NRAO MARKIIIa data acquisitionsystem (Clark,1985)hasbeenintroducedin 1980with a faster data rate of 112 Mbits/sallowing a maximum 56 MHz bandwidth with severalstandard setups for data recording which are used in radioastronomicalobservationsto obtain continuum and spectralline information. ThemodeC is normallyusedin polarizationbut with reducedbandwidth and tape usage and therefore,reducingthe imaging sensitivity by ~ 30%from the full capability. A descriptionof signalflow and frequency conversion in suchsystem can be found elsewhere. In this review paper, I present the results of an extensiveinvestigation on a new ob¬ servational techniquewhich uses the longestand intermediate spacingsof the MERLIN array (Thomasson,1986)as an additional source of information to improve the polariza¬ tion measurements accuracy and imaging fidelity. The method discussedhere has been applied to data obtained with MARKIIIa recordersfrom the European VLBI Network since 1993 and the MERLIN array and is currently offered as standard joint MERLIN and EVN observationsin polarization mode. 98 Proceedingsof the XXIthSAB c a 1. JodrellMark 2 (25m) U 2. Cambridge(32m) o 3. Defford(25m) 4 4. Westerbork(25m) 5SK 5. Onsala(20m) 6. Effelsberg(100m) & 7. Medicina(32m) 6 rÿO 0 - 8. Noto (32m) S : \ -~7 Figure 1: Telescopelocations of the European VL131 Network (EVN) which are rec¬ ommendedfor VLBI polarization measurements. The Cambridge telescopeis addedby microwave link networksto improve the polarization imaging quality. The correlationsand cross-correlationsto obtain interferometric visibilities havebeen obtained after the observations with a MARK Ilia complex correlator such as the Max- Planck Institute fur Radioastronomie,Germany (Alef, 1989) and in real-time with the MERLIN correlator, UK, allows to produce polarization maps which are good enough for most astrophysicalpurposes. The main results from this paper are that microwavelinks from the MERLIN network are ideal to improve short-spacingcoverage in the UV plane allowing polarized emission to be well detected at larger distancesfrom the radio map phase centre. 2 The imaging problem Radio interferometers employingVLI3I techniquesconsiston an ensembleof antenna pairs with separationslarger than ~ 10 Km which are used to make high-resolutionmeasure¬ ments of brightnessdistribution and position of cosmicradio sources of strong astrophys¬ ical interest, radar imaging of planets and asteroidsand tracking deep-spaceprobes. In particular, the European VLBI Network (Figure 1) is a consortium of European radio observatoriesbut with the eventualinclusion of Chinese(Shanghai),.Japanese(Kashima.) and South African (Hartbeestoek)antennas to increasethe resolution and improve the synthesizedbeam shapefor targets at low declinations. The current EVN data record¬ ing system are MARK Ilia and VLBA data recorders with the Bonn Correlator used in part-time to process the data to obtain the complex visibilities, hi the correlator each Proceedingsof the XXIth SAB 99 antenna output voltages (Eij(u,v )) are amplified, multiplied and integrated to obtain complexcorrelation coefficients, after digitally compensatingthe wave paths due to the earth surface geometry and electronic phasedelays: Eij{urv) = (l, m)e-27ri(ul+vmhlldm (1) J oo —ooI r Jr — The radio image are the source brightnessdistribution 1(1,m) which is obtained by inverting the above Fourier transform from gridding and interpolating the data in both visibility plane (/,???.)and the sky baselineprojection plane (u,v). The Fourier trans¬ form of Eij(u,v ) is a ‘dirty’ map which is contaminated by phase errors and spurious sideloberesponse since it is a convolution of the true source brightnessdistribution and the interferometer difraction pattern. The true source image can then be recoveredby using the non-linear complexCLEAN algorithm(Cornwellh Wilkinson 1991)which sub¬ tracts iter actively a beam pattern from the dirty map, obtaining the intensity at each pixel of the (l,m) plane. The final map is obtainedby convolvingthe pixel valueswith a Gaussianfunction with same profile as the synthesizedmain beam. CLEAN conver¬ gence is guaranteedby the fact that radio sources are normally simple and that both beam and dirty map are obtained with the same locus in the (u;v) plane. The algo¬ rithm is not efficient in recovering extended,diffuse structure in the (l,m) plane but it works fine for compact and bright features in the dirty map, which are of astrophysical interest. CLEAN also converges with negative patterns so that the StokesParameters Q(l,m) and U(l,m) can be mapped with the same algorithm since they are related to the across correlations of LCP and RCP signals(Conway& Kronberg, 1969). The po¬ larization position anglesx(/,??r)= 1/2atan(U(l,m)/Q(l,m)) and fractional polarization m(l,m ) = lOOy +U(l,m)2/I(l,m ) are obtained from simple map arithmetics and are quantities which can be obtained by numericalsimulations. It is out of the scope of this paper to make a detailed analysis of polarization imaging algorithms for VLBI (Conway& Kronberg, 1969;Cotton, 1993)but an important consequence is that the ad¬ dition of long and intermediate spacingsfrom MERLIN telescopes(10-212Km) provides more constraints which helps in improving the definition of imageswith normal EVN capabilities. The high-frequencypart of VLBI data acquisition systems consist on Cassegrainian radiotelescopesof at least 20 m diameter,a corrugate horn antenna mountedin the sec¬ ondary focus,a polarizer quadripole(septum,1/4-wave plate or orthomodetransducer) and a low-noise cryogenicreceiver (Tsys~ 35 K) (Rohlfs,1990).Instrumental polariza¬ tion occurs mainly due to wave reflectionsbetweenthe antenna surfaceand subreflector, leakagedue to the cooling which can damagethe electric shieldingbetweenpolarization channelsas well as imperfections in the polarizer and horns. Further electronic stagessel¬ dom contribute to crosstalk betweenLCP and RCP signals(< 0.1%,Ludke,1994). The Linear approximationwhich considersa simplecontamination of LCP and RCP channels by an arbitrary quadripoleis employedto extract the receiver instrumental polarization and it works well for polarimetric.observationswith the Very Large Array and MERLIN. In both methods,the apparent source rotation due to alt-azimuthal telescopemountings 100 Proceedingsof the XXIth SAB is usedto calibrate the amplitude of instrumental polarization factors(D-terms, Conway & Kronberg, 1969)but it is unlikely to work when instrumental polarizationis very strong since only the lowest-ordercontamination terms are used. Another way of solving for instrumental polarization apart of this linear model is a more generalschemeproposedby FomalontandWright (1974).The Ellipticity-Orientation Model (EOM) assumes an elliptical feedresponse which can be known a priori from feed measurements in a antenna testing range (Cotton, 1993). The EOM model has been built on more realistic and physicalassumptionsabout the feedpolarization response and takeinto account the electric characteristicsof suchsystems which is the factorization of spatial responses basedon parallactic anglevariations. As suggestedby Fomalont and Wright, the feedproperties are representedby elliptical response functions with feed el- lipticity and orientation as main variables.Liidke (L994) has shown that both methods can be related by a simple transformation,if the instrumental contamination is lessthan about a dozenpercent. This procedurehas beenrecently implemented in the AIPS task PCAL by W.D. Cotton (1993),who was able to obtain a sensiblepolarization map of the jet in 3C138with global VLBI (Cotton, private comm.) at (i cm. Further investigations in this field by myself and collaboratorsis that it is better to calibrate MERLIN and EVN data sets independentlywith the linear and EOM model and then to combine the data Sets tliail to use a single phasecalibration routine to be applied in the whole visibilities (Liidke, 1994). 'I'll® main restriction here is that there is a polarization position angle calibration for each telescope,which is called AC phase offset,which is not stable for some EVN antennas while the phasestability is better for MERLIN telescopes. 3 The MERLIN hardware The transport of information from a remote telescopeby microwave links is a difficult task. At high frequencies,transmitter and receiver towers can vibrate due to the wind and variations in the troposphererefraction index can inducefast phasevariations which can decorrelatethe signalproducingmap noiselevelslarger than the theoreticalNyquist noiselevel which is determined by antenna system temperatures, number of telescopes, bandwidth and data averagingtime. At 5 GHz the MERLIN telescopesare remarkably goodbut the polarization sensitivity of the Defford telescopeis not so good as it is at L band due to surfaceerrors. The basic MERLIN signal flow architecture which is used in joint observationsis summarizedin figure 2. The 2-bit digital correlator givesthe visibilities A,-,(v,,u ) for all n MERLIN telescopeswhile the polarization data from Cambridgeis also recordedin a VLBA-type tape, before the stages. Phasevariations due to the K-band microwave link paths are monitored by a pulse modulation schemefrom which transmission and arrival times are measuredby a logic unit and phase delays can be obtained and correctedon-line by phase rotator circuit boards before the data, enters (he correlator. The receivers are coherentsince the phase Proceedingsof the XXIthSAB 101 CAMBRIDGE32m FRONT-END i— — ráj SYNTHESIZER F K-BAND LINK 5Jtar MASER LO. 1 , . , e-p0-£ÿ]---- — I 1 -joENEÿonr I •sn 10RRf 2-BIT CORRELATOR RCP.LCPANT»» - BJW -VISIBILITIES VISIBILITIES Figure 2: Summary of the MERLIN polarizationarchitectureusedwhenthe Cambridge Telescopeis included in the EVN network. The system is designedand maintained by the Nuffield Radio Astronomy Laboratories/UK. of local oscillatorsin the front-end are phase-lockedto the carrier of a secondaryL-band link which transports the phaseof a MASER clock to all telescopes.Similarly, all local oscillatorsfrom themicrowavelink network whichare important to analogAM modulation and demodulation,are alsophased-locked.Double-sidebandAM modulation schemesend LCP and RCP over a ~ 30 MHz bandwidth in a ~ 7GHz carrier wave with no channel crosstalk,allowing a wider bandwidth (15 MHz) than the original architecture usedin the late eighties. This automatically allowson-line polarization data recordingwhichis a requisite of MARKIIIa and VLBA systems. Fortunately, theseimprovementswere such that both hands of polarization from the Cambridgetelescopecan be recordedon-line by a VLBA data acquisition rack and makingsuitable to add the Cambridge32-m telescope from the MERLIN array to the EVN network. In figure 1, this correspondsto (1-2) baselinewhich fills in the shortest EVN spacingsadding more polarization sensitivity. 4 Defining the observing mode The observationswere planned in 1992but startedin 1993due to delaysin the hardware commissioning.They employedMark III Mode A 28D,in whichLCP is connectedto the data acquisitionrack IF1 input and RCP is connectedto the IF2 input of the MARKIIIa system. In this mode,a largerbandwidthis availablefor polarization (28MHz), whichis almost twice the band allowedby the standardMode C recordingemployedin Cotton’s 3C138observations. Within the data acquisitionrack, the bandpassis split into several 102 Proceedingsof the XXIthSAB / / rf- +4ÿ/ & m \l a ij a >o 2 ° *9 2 jo Figure 3: Polarization maps of the quasars 3C286. (TOP) is the MERLIN map at 5 GHz and 55 mas resolution;(MIDDLE) is the EVN+MERLINradio map at 5 GHz showing flow details at 12 mas resolution;(BOTTOM) is the EVN polarization map at 5 GHz showingthe polarization position angleswhich are related to the magneticfield direction. sidebandsof 2 MHz each. The signalsare recordedin sucha manner that LCP lies along even numbered videoconverters and the RCP is processedby odd videoconverters. No polarization crosstalk are expected from the tape recorder due to random longitudinal tape motions, since the tracks are more than 600 fim apart and the tape can run with enoughmechanicalaccuracy. The instrumental polarization can be obtained for the calibrators 0552+398and OQ208and the weighted average valuesof the baseline-basedpolarization contamina¬ tion values are about 3% or lessfor MERLIN and 10 % or more for the EVN network showinghigh feedellipticity for a large numberof antennas. The parallel (LL,RR), cross (LR,RL) and autocorrelations(LR per telescope)were obtained with the Bonn correlator. The autocorrelationshave shown to be useful to measure directly the Figure 3 shows the results of the first polarization experiment with the EVN and MERLIN held in 1993. The MERLIN polarization image of 3C286 at 5 GHz and 50 mas resolution showsthe central component with the bulk of polarized emissionand an extension to the southwestwhere the jet is seen. The combinedMERLIN and EVN total intensity map (1(1,m)) shows fine details of the jet with 10 mas resolution. The central region is expandedshowingthe EVN-only polarization map with electric vectors which are perpendicular to the magnetic field vector direction. From this exampleone can see clearly the improvements in the image quality sincethe magnetic field lines can be tracers of fluid motion within the Hist few kiloparsecsfrom the central engine. The astrophysicalimplications of the data which suggests a connection between3C286and extendedFanaroff-Riley objects will be the subject of a forthcoming paper (Jiang et al, 1996). 5 Summary In this talk, I havereviewedthe techniquesfor polarization-sensitiveVLBI and reported unpublished results from recent combinedMERLIN and EVN observationsof compact relativistic jets in compact quasars which are part of a major project to study the magnetic fields and Faraday rotation in relativistic jets within the narrow-line region. The basic engineering strategy describedhere is the best to the moment to produce radio polarization maps at about 10 milliarcsecondsresolution with the European VLBI Network at 5 GHz, which matchesthe VLBA at S-bandand MERLIN at 22 GHz. I am grateful to the EVN board of directors for making available their hardware to tests of polarization, the European VLBI friends for their support and the Max-Planck Institut fur Radioastronomiefor hospitality during my stay at the Bonn Correlator. It has been a pleasure to work with Drs. H. Sanghera,VV.D.Cotton, M. Rioja and D. Dallacasawhich are my collaborators in this project. In particular I wish to thanks the the director of MERLIN National Facility for support and encouragement. CAPESand CNPq partially funded this project and without their help, this paper could never be written. References Alef W., 1989,in VLBI Techniquesand applications,eds. Felli M., SpencerR.E., NATO ASI Series,Kluwer, p.97 ConwayR.G., KronbergP.P., 1969,MNRAS, 142,11 CornwellT.J., Wilkinson P.N., 1981,MNRAS, 196, 1067 Cotton W.D., 1993, AJ, 106, 1241 Clark B.G., 1973,Proc. IEEE, 61, 1242 Clark T.A., 1985, IEEE Trans. GeodesyRem. Sens.,GE-23,438 Fomalont E.B., Wright M.C.IL, 1974, in Galactic and Extragalactic Radio Astronomy, Springer(Berlin) 104 Proceedingsof the XXIthSAB JiangD.R., DallacasaD., SchilizziR.T., Liidke E., SangheraH.S.,Cotton W.D., 1996, A&A, in press Liidke E., 1994,PhD thesis,Victoria University of Manchester,UK RohlfsK., 1990,Toolsof Radio Astronomy,P45,Springer-Verlag Proceedingsof the XXIthSAB 105 ASTRONOMY IN SPACE: SATELLITES AT INPE Thyrso Villela Instituto Nacional dePesquisasEspaciais - INPE Divisão de Astrofísica Caixa Postal 515 12210-970- SãoJosédosCampos,SP e-mail: [email protected] Abstract The ongoingscientificsatelliteprogrammes at INPE andthe existing Brazilian satel¬ lite testing facility are described. The need for building space-basedastronomical experimentsis discussedas well as the chancefor the Brazilian astronomicalcom¬ munity to use the existing INPE’s facility and acquiredexperiencewith the design, building and testing of scientific satellitesto carry out astronomical experiments in space. 1 Introduction Astronomy and Astrophysicsin Brazil have reachedsucha level that are demandingnew instruments, including space-basedexperiments. Many astrophysical queries can only be answeredwith data obtained by space-basedinstruments on board balloons,rockets and satellites. As balloons and rockets are, in general,dedicated instruments with not much observationaltime, satellite-basedinstruments seem to be a natural way to obtain these data. Although the majority of Brazilian scientists in the field is composedby individuals with little experimental backgroundor expertise, the identification of the necesssary experimentsto answer thesequestionsdoesnot dependon this as it is dictated only by the scientific driven goals. The Brazilian astronomicalcommunity has already proven itself capableof suchidentification. It is imperative for the Brazilian astronomical community to collect its own original data from space platforms and that thesedata are competitive and of good quality. But, in order to do that, it is necessary to haveboth human and technical capabilities. The technical capability is somewhat consolidated as far as spacecraft design,building and testing are concerned. The payload building capability is the one that needsto be strengthenedas it involvesboth humanandtechnical developments. Someactions have to be taken in order to achieve this goal, such as personneltraining and laboratory upgrades. Overthe years, severalcountrieshaveput in orbit experimentsthat drasticaly changed the way we view the cosmos. The examplesare multiple and range from the small dedi¬ catedexperimentsto the so-calledgreat observatories.The scientific knowledgeobtained with suchinstruments was fabulous and the advancementin severalastronomical fields was enormous. As the economical situation worldwide worsenedlately, more and more scientific collaborationsare expectedin order to carry out experiments capableto fulfil the scientific needs.Any international cooperation must, despitethe fact of inherent mutual 106 Proceedingsof the XXIth SAB scientific interest, rely on the technical capabilitiesof the partners, and Brazil, nowadays, can join these collaborations as an equalin severalaspects, namely the scientific insight and the spacecraftand payload design,building and testing know-how. 2 The present It is possibleat the moment, in Brazil, to design,test and operate a.satellite,as the first Brazilian satellite SCD-J (Satélite de Coleta cleDados - 1) has shown. The SCD-1 was launchedin February 1993 and is still operating in orbit with no failures. The estimated lifetime was one year. This satellite was totally designed,built and tested at INPE. In order to inform the Brazilian AstronomicalSociety of the technical capabilitiesof INPE as far as satellite technologyis concerned,it is describedthe availablefacilities at INPE in SãoJosédos Campos,SP,and the ongoingsatelliteprogrammes. The descriptionsbelow are reproductionsof the availabletechnicalinformation about the existing testing facility at INPE and the SAC-B and SACI-1 projects, availablefrom INPE and CONAE (see references),and can be obtainedupon request from INPE and CONAE headquarters. 2.1 The Existing Facility in Brazil: LIT/INPE INPE’s Integration and Testing Laboratory (LIT) was specially designedand built to attend the necessitiesof the Brazilian SpaceProgram. It represents one of the most so¬ phisticated and powerful facilities in the qualification of industrial products which demand a high degreeof reliability. Environmental testing is conductedin a 100000classcleanarea of 1600 square meters, with a useful overhead crane hook height of 10 meters. The types of tests performed include thermal vaccum, climatic, vibration and shock,eletromagneticinterference and compatibility. The 450 square meters integration area has a useful crane hook height of 6 meters and operates under a class10000 cleanlinessstandard. Found in this part of the laboratory is a specially designedequipment for alignmentmeasurement and three axis dynamic simulation. Theseactivities, together with the measurement of mass, center of gravity, moment of inertia and product of inertia, form the basic nucleusof the functional testing and integration activities of space systems. The satellite check-outstation, designedto govern the functional tests of space hard¬ ware, as well as the other test control rooms, are adjacent to the clean areas with a full unobstructedview over all activities. With the intention of assuringthe precision of all measurements,a sensor calibration laboratory for accelerometers,thermocouples,and pressure transducerswas installed within LIT along with a global electrical calibration laboratory under rigid environmental conditions. To allow the wideningof the scope of activities of LIT, a reliability laboratory is availablefor the reception, inspection, and qualification testing of electronic devicesin accordancewith specifiednorms. Comple¬ menting the environmentaltest facilities, there are specialequipmentinstallations which are designedto support the developmentof other space related activities. This is the ProroBdingsof the XXI1'1SAB 107 case, Forexample,of the explosionproof hydrazinecompat ible micropropulsion chamber and the 80 meter range antenna testing laboratory. With a total constructed area of 10000square meters, LIT containsa completeutility infrastructure which includes a filtered compressedair system, a closedcircuit industrial water system, an electricalgroundingsystem with lessthan 1 Ohm of resistance,a labora¬ tory access control system, a fire detection and extinction system and a filtered stabilized electrical system complete with an uninterruptible power supply and emergency genera¬ tors which help guarantee the quality, continuity, and security of all tests performed in the laboratory. 2.2 The satellites 2.2.1 SAC-B The SAC-B (Satélite de AplicacioncsCientificas) satellite is an exampleof what can be donein terms of international cooperation. It is carried out as an international coopera¬ tive project between the Argentina’sNational Commissionof SpaceActivities (CONAE) and the U.S. National Aeronautics and SpaceAdministration (NASA). The Secretariatof Scienceand Technologyof Argentina participates in the financingof the project through the National ResearchCouncil (CONICET). CONAE is responsiblefor the design and construction of the SAC-B spacecraft,the HXRS instrument, the ground station opera¬ tion, and the scientific data distribution; NASA will provide two scientific instruments, launch serviceson a Pegasusvehicle,and support for initial orbit monitoring and emer¬ gency backupthroughout the missionlife. By a separate agreementestablishedArgentina, the Italian SpaceAgency(ASI) will provide the solar arrays plus a scientificinstrument. The Brazilian National Institute for SpaceResearch(INPE) is providing the facilities for the system qualification tests of SAC-B at LIT and a Brazilian co-investigatorwith the Argentineinstrument (IIXRS). The IIXRS science team is composedby II. Ghielmetti (Argentina), A. Hernandez (Argentina), M. Machado (Argentina), C. Mandrini (Argentina),P. Marias (Argentina), M. Rovira (Argentina), B. Dennis (USA), U. Desai(USA) and T. Villela (Brazil). SAC-B is designedto advancethe study of solarphysicsand astrophysicsthrough the examination of solar flares,gamma ray bursts,diffuse X-ray cosmic background,and energeticneutral atoms. The scientific payload of SAC-B comprises an Argentineinstrument (IIXRS), developedby the Institute of Astronomy and SpacePhysics (IAFE), to measure the temporal evolution of X-ray emissionsfrom solar flaresand non-solar gamma ray bursts; a combination of detectors to measure soft X-ray emitted by solar flares and gamma ray bursts provided by NASA (Goddard X-Ray Experiment - GXRE); a diffuse X-ray background detector using CCD technology(CUBIC) from Penn State University; and an Italian instrument to measure energeticneutral atoms (ISENA). 108 Proceedingsof the XXIthSAB Scientific objectives: Solar Flares Solar flares are the most dynamic and energeticphenomenain our Solar System. On time scalesof a few minutes or even seconds,stored magnetic energy is suddenly releasedin thesetransient events, in the form of acceleratedparticles, radiation in the whole range of the electromagneticspectrum, bulk heating and hydrodynamic shocks. Combinedobservationsof the hard and soft X-ray emissionsfrom solar flares, as those to be provided by the IiXRS and the solar detector of GXR.E,will help us to further our understandingof the interactions betweenacceleratedparticlesand the ambientsolar atmosphere. Gammaray Bursts Intense transient emission of gamma radiation - the gamma ray bursts - doesappear in some completely unpredictableregionsof the sky. In spite of the enormous amount of energy involved in eachof theseevents, the extensiveinvestigationshavenot beenable to explain them. To discriminate among different suggestedmodels,new observations,such as those to be performed by the HXRS and GXRE, are needed. DiffuseX-Ray Background Below 1 keV the soft X-ray diffuse backgroundis believed to be largely Galactic in origin. Thermal emission from hot gas is the likely source of this flux, and for this reason its spectrum is dominated by emission lines from highly ionized species.CUBIC should be able to distinguish between the strong line groups of oxygen, neon, iron, carbon, etc. In the X-ray band aboveI keV, the diffuse emissioncomes predominantly from sources of extragalactic origin. CUBIC studies will help to elucidate the origin of the diffuse X-ray background. Energetic Neutral Atoms Energeticneutral atoms are generatedthroughchargeexchangein the radiation belts and in the ring current region, wherethe trapped ion populationsinteract with the gas of the geocorona, and then travel down to the satellite’s altitude. By the measurements of precipitating and escapingneutral atoms, ISENA will imagethose areas of the Earth’s environment wherethe neutral atoms are coming from. Mission: SAC-B, a 181 kilograms classspacecraft,will be injected into orbit by a NASA pro¬ vided PegasusXL rocket. The Pegasuswill carry SAC-B to an altitude of 550 km and a circular orbit inclined 38 degreesto the Earth’s equator, which ensures an orbital lifetime of at least three years. A high-tech Argentine enterprise, INVAP S.E., is building the spacecraftand the Hard X-Ray Spectrometer(IIXRS). It is expected to fly in 1996-97 dependingupon Pegasusavailability. The SAC-Bmissionwill be controlled from a groundstation locatedin Argentina. This facility comprisesthe Telemetry, Tracking and Control (TT&C) Station and the Mission Proceedingsof the XXIthSAB 109 Operations Control Center. The TT&C Station has a 3.6 m autotracking antenna, and it enablesrange and Doppler measurements and Master Clock Reception (GPS). The Control Center consistsof two fully redundant Sun workstations linked to the receiving and transmitting equipment as well as PC terminals,which will control the satellite. Spacecraft configuration: The spacecraftbody, built in Alluminum alloy, is a rectangularparallelepiped62 cm by 62 cm wide by 80 cm high. Four deployablesolar panels,coveredby GaAs cells, generate 210 W to supply the electrical power to the different subsystems: Attitude Control, Thermal Control, Commandand Data Handling, and RF Communications,as well as the instruments. The solar instruments (HXRS and GXRE solar detector) are mounted on the Sun-facingend of the spacecraftalong with the Fine Sun Sensor,with fieldsof view looking toward the Sun. The GXRE gamma ray detectors are mounted on the solarplatform, but they have fields of view on opposite sidesof the spacecraft,set on normal to the Sun line. CUBIC and ISENA are mounted on the top, or the anti-Sun facingsideof the spacecraft.CUBIC’s field of view is normal to the Sunline. Three-axisstabilization of the SAC-B will be accomplishedby two momentum wheels in a “V” configuration for roll (Sun direction) and yaw axis (CUBIC boresight)control. Pitch axis control is effectedwith air core magnetic torquer coils. The two-axis Fine Sun Sensorand a three-axis magnetometer will allow attitude determination to an accuracy of 3 degreesin real time and 2 degreesby post-facto analysis. The temperature will be kept within the -10 degressCelsius to +40 degreesCelsiusrange by means of a semi¬ passivethermal control, using radiators and electrical heaters. Up to 100 Mbits per day of scientific data will be dumped to the ground station through the RF downlink on S- band (2255.5MHz). Two quadrifilar helix antennae are mounted one at either end of the spacecraft,one right hand circular polarized, the other left hand circular polarized, yeldingan omnidirectional pattern. 2.2.2 SACI- 1 Five experimentshavebeenselectedto be performedon board the first Brazilian scientific micro-satellite SACI-1 (Satélitede AplicaçõesCientíficas): a) PlasmaBubbles: to study plasmabubblesin the ionosphere; b) Airglow Photometer: to measure the intensity of the terrestrial airglow emissions; c) GeomagneticExperiments: to study the Earth’s magneticfield; d) GPS:determinationof the orbit and attitude of the SACI-1using a GPS system; e) ORCAS:to measure anomalouscosmic rays. This latter experiment (ORCAS, which stands for Observaçõesde Raios Cósmicos Anómalos e Solares)is the only one related to astrophysicsand it is the one that should be viewed here. The scientific motivation for this experiment is outlined below. The proposers of this experiment are Dr. A. Turtelli (UNICAMP) and Dr. U. B. Jayanthi (INPE) with collaboratorsfrom INPE (T. Villela), UNICAMP, NRL (USA) and RIKEN (Japan). 110 Proceedingsof Hie XXIthSAB Scientific objectives: In the interplanetary space, heavy nuclei of lie ami even Ne with energy below 50 MeV/amu,called AnomalousCosmicRadiation (ACK). were discoveredin 1970-73by the space probesIMP and Pioneer 10 and 11. While the relation between the Carbon and Oxygenquantities (C/O) in the cosmic and solar radiation (COR, SCR) is lessIlian 0.5, in the ACR the C/Oquantity is lessthan 0.1. The theoretical explanation for this is that the interstellar neutral atoms are capturedby the Solar Systemin its movement . then are ionized and accelerated,and are transformedinto ACR. Starting with the discoveryby theSkylabIII in 1973, a seriesof experimentson board Spacelabsand Sovietsatellites(in orbits of about 500km), with recoverable,passiveplastic detectors,has detectedparticles with energy and composition similar to the ACR near the Earth. The presence of theseparticles, at thesealtitudes, despite the existenceof the Earth magnetic field, is only possibleif they are partially ionized and “trapped” by the magnetosphere. For the first time, in 1993, a very sophisticatedexperience(MAST, in the satellite SAMPEX), using active solid state detectors, resulted in the measurement of theseparti¬ clesin the magnetosphere,as well as registering the time, position and directionof arrival of the particles. A new ion radiation belt was then found aroundEarth (L=2), composed of ACR,similar to thosewell-knownVan Allen belts. This hasa fundamentalimportance, as it allowsthe study of interstellar atom samplescloseto the Earth. The telescopeORCAS,on board the Brazilian ScientificSatellite to belaunchedduring the period of minimum salar activity, will have the capacity to measure fluxes,spectra and composition, temporal and spatial variations of ions from He to Ne, and protons and electrons with energy less than 50 MeV/amu.The main objective of the ORCAS telescopeis to measure ACR fluxes, from C to Ne, “trapped” in the belt, using solid state detectors,identifying the time and direction of arrival of the particles. Although the ORCAS telescopeis 3-4 times smallerin geometric factor, the flux to be measured will be better by a factor of 2 to 10 in comparisonwith MAST, as the intensity of the ACR trapped, expectedfor the period of 1996-98,will be at least 20 to 30 times greater than the measuredby the SAMPEX in 1992-94. Aclcltionally,the MAST experiment determinedthe ion belt of L=2.0 ± 0.2. Is this due to the fact that the SAMPEX-MAST have, measuredonly 5 to 10%of the radiation population? In conditions of minimum solar activity with intense fluxes, the ORCAS telescopewill allow studies of the space distribution and the dynamicsassociatedof the population of the trapped particles. The secondimportant aspect is the presence of lie ions, as if is not expected to find He in the population of ACR in this altitude. Only during solar eruptions the He was observedby the satellitesINJUN and OHZORA. SAMPEX observedagain a thin belt of He at L=1.8. The source of the He and the mechanismsof its entrance and trapping in the magnetosphereis an interestingproblem. The ORCAStelescopeas well as measuring the He flux will determine its source, as during the ORCASexperiment few solar eruptions are expected. Proceedingsof lhe XXI'1'SAB 111 Mission: The SAC'I spacecraftis a low orbit satellite for scientificpurposes. The satellite will be put in a circular orbit inclinedabout 98degreesto theEarth equatorialplaneat an altitude near 750km. It is scheduledto be launched, togetherwith CBERS(China-BrazilEarth ResourceSatellite), by a ChineseLong Marchrocket. It is expectedto fly in 1997. During thepassesvisibleeitherfrom the trackingstationsor user’sgrounddata collectingstations, scientific data will be transmitted to ground. The data can be producedin a longertime1 andstoredin the on board computer. A communicationssystem is responsiblefor sending to the ground the housekeepingtelemetry, receivingthe telecommand,and will also send scientificdata to the ground. Dual coldredundanttransmitters providesstandardS-bancl (2255.5Mllz) telecommunicationsatellite to groundwith BPSKmodulation and 128,256 or 512 kbits/sselecteddata, transmissionrate. The developmentof the microsatelliteshall be done in less than 18 months. The total cost, including bus and scientific payloads, shall be lessthan II, S$ 4.6 million. Spacecraft configuration: The total satellite mass is less than 60 kg and the payload mass doesnot exceed20 kg. Thedimensionsare 600 x 400 x 400mm (parallelepiped).The conceptionis modular,with simple technical solutions. The thermal control shall be passive. The satellite expected lifetime is 18 months. The payloadpower requirementis 30 W and a.solararray generator, whichis composedof four deployablepanelsassembledon the solarsideof the spacecraft structure, provides the necessary electrical power. The solar cells are made of GaAs/Ge with cover glassmicrosheetand silver plated invar interconnectionsand are expected to generate about 125 W. The SAC1-1satellite attitude control system combinespassivespin stabilization pro¬ cedurewith active spin rate and precessioncontrol. The actuators are three magnetic torque coils which interact with the Earth magneticfield to generate the necesary control torques. The pointing accuracy is 1 degree. 3 The future 3.1 The coming opportunities It is expectedin the forthcoming years that a largenumberof opportunities for launching a satellite becomeavailable. The newborn Brazilian SpaceAgency(AEB) is trying to stimulateinternational collaborationsin order to broaden the space activities in Brazil, so the potential partners are Argentina, China,France,Russia,United States,and other countries as well, specially latiu american countries. It is important to stress that in¬ ternational collaboration is the best way to conducta satellite project, since it lowers the costs and adds technologyto the country in a direct way, as simple technological transfer from developedcountries, or indirectly, stimulating local industriesto produce space qualified systems and components. The latter situation is alsovalid for a genuine Brazilian experiment. 112 Proceedingsof the XXIthSAB 3.2 Strategy for getting in space In order to carry out experimentsin space using the already existing Brazilian facilities and expertisein this area, it is imperative to coordinateactions that will lead to the design,building, testing and operation of the experiment. First of all, the scienceis, of course, the main factor to be considered.The mission technicalconstraints as mass limit and pointing accuracy are critical and very limiting points to be considered,since a dedicatedlaunchingwill be very difficult for the first missions. After that, there is a needto searchfor funding. This is a crucial problem,sinceit doesnot makeany sense to drain all the availableresources from the financing agenciesand stop other astronomical projects, namely the groundbasedexperimentsor other space-basedexperiments.So,it is important to obtain independentsupport from the fundingagencies(e.g. CNPq, FINEP, FAPESP)and guarantee the continuity of the ongoingnon-satellitebasedprojects. It is alsoimportant to bear in mind that one needsto use the availableknow-how to build both the spacecraftand payloadin order to estimulatethe training of individuals. This action will be extremelyhelpful for other missionsor even non-space projects like new optical or radio telescopes. On the other hand, using the available human and technicalresources helpsto lowercosts and developmenttime. It becamea tradition in the astronomicalcommunity to propose experimentsin which the scientific instruments are built abroadwith no (or almostnone)interaction of Brazilian scientistsand technicians. The reasons are obviousand residemainly on the lack of human resources, but this needs to be changed.Thereis no tradition in the developmentof astronomicalinstrumentation in Brazil and most of the scientific work produced here is basedon data analysis or theoretical approaches.A space-basedscientificexperiment has the merit to promote the advancementof the experimentalastronomicalactivities in Brazil. It will be very difficult in the first missionsto achievethis goal,but it is extremelyimportant to stimulate such a changein our community. 4 Conclusion It is necessary to start a long term program to capacitateBrazilian scientists,engineers and techniciansto acquirespacetechnology.Thegraduateprograms shouldtry as much as possibleto involve the new studentsin experimentalprojects aiming satelliteinstruments. It seems that this is the only way to get a deepinvolvementwith this new opportunity opennedup to Astronomyand Astrophysicsin the country. This program can no longer be delayedif the Brazilian scientists want to take a step forward to advanceAstronomy and Astrophysicsin Brazil. The response for the Announcementof Opportunity for the first Brazilian scientific micro-satellite showedthat the astronomicalcommunity in Brazil is not yet prepared to answer these calls. Apart from the ORCAS experiment and an X-ray polarimeter (whichdid not meet the mass limit requirement),no one elserelated to the astronomical community submitted a proposal. Severalreasons lead to this situation, such as the very low mass allocated to the payload and the very short time for the preparation of Proceedingsof the XXIthSAB 113 the proposals.A new Announcementof Opportunity, probably for a joint Brazil-France experiment, will be deliveredlate this year or early next year and we hopethat several Brazilian proposalswill be submittedin order to start the long term project proposed above. Acknowledgement: I thank Alex Wuenscheand João Braga for the reading of the manuscript and suggestions. References Laboratóriode Integraçãoe Testes- LIT, folder, INPE. SAC-B - Satélite de AplicacionesCientíficas,folder,CONAE,1993. Satélitede AplicaçõesCientíficas- SACI-1,TechnicalReport1, INPE and AEB, 1995. Section 4 Galactic Astronomy and Interstellar Medium . Proceedingsof the XXIthSAB 117 CHEMICAL COMPOSITION IN PHOTOIONIZED NEBULAE: THE PROBLEM OF THE TEMPERATURE DETERMINATION Ruth Gruenwald Instituto Astronómicoe Geofísico,USP Abstract A good determination of the chemicalcomposition in nebulaeis important for an evolutionary study of a given class of objects, and for the knowledgeof the chemicalevolution of our, as well as of other galaxies.Wedescribehere the methods generallyusedfor the chemicalabundancedeterminationin photoionized nebulae. In particular, the methods for the calculation of the gas temperature, which is an important parameter for the determination of the abundances,are discussed. 1 Introduction Basically, there are two methods for the abundancedeterminationin gaseous nebulae. Onemethod is the use of a detailed photoionization model,applied to a specificobject, which fits a great number of observational data. This method requires a good set of observeddata, from a great number of lines,for eachparticular object. When a great number of objects is treated,an empirical methodis generallyused.This method follows the schemaproposedby Peimbert (1967)and Peimbert & Costero(1969).In this case, the abundanceof an ion which producesan observedline A can be calculated,relative to H+, from the measuredintensities of the line A and of H/?,knowingthe gas density and temperature. Applying correction ionization factors in order to take into account ions which do not have observedlines, the correspondingelementalabundancecan then be calculated. The abundanceof an ion X', relative to the ion H+ can be calculatedfrom the equation n{Xl) _ £{TH/3,ne) ~ I\ (1) n(H+) IH/3e(T\,ne) where n(X‘) is the abundanceof the elementX in the ith ionization state; n(H+) is the abundanceof ionizedhydrogen; is the I A measuredintensity of the line with wavelengthA,producedby the ion X'; e(TA,ne)is the emissivity of the line Ain a giventemperature TAand a given electron density ne. The temperature TAis the gas temperature in the region where the line is produced. The electrondensity in a object with constant densityis practically constant inside the ionizedregion, and generally do not present any problem. So,the empirical method is basedin a good temperature determination. 118 Proceedingsof Lhe XXIth ,SAB There are severalmethods for the determination of the temperature in a gaseousregion. For example,the temperature can be determinedfrom the intensity ratio of auroral to nebular forbidden lines, from the ratio of the intensity of free-boundcontinuum to the intensity of a hydrogen line, or from the ratio of the free-freecontinuum intensity to the intensity of a hydrogen line. The problem is that, in general,different methods provide different valuesfor the temperature. For example,in many cases the temperature given by the intensity of 0++ forbidden lines (T[0///]) are higher than temperatures given by the Balmer discontinuity (TBa;). This fact is generally explained by the presence of temperature fluctuations in the gas. Thus, if there are temperature fluctuations inside the gasthe obtained valuefor the gas temperature from the observationaldata is strongly dependentof the method usedand consequentlythe abundancewill also dependon the method usedfor the determination of the temperature. 2 Temperature determination Peimbert and Costero (1969)suggesteda quantitative definition in order to caracterize the temperature fluctuation (t2) and obtained also relationships between the different definitions for the temperature and t2. If an average temperature is definedfor the region where the ion X* is present Te(r)n(Xt\r)ne(i')dV T0(X{) = f (2) / n{X\r)ne{r)dV the temperature fluctuation for this ion is defined by rJ?{v) 7'2(.V1)| u(X\r)u, (r)dV t2(X‘) ./'I - = T*(X<)fn(X\r)nc(r)dV (3) The integrals are calculatedover the total volume of the ionized region, r corresponding to a givenpoint inside the region. A temperature obtained from the intensity ratio of two lines produced by a given ion can be written as a function of the average temperature Ta and of t2 for the ion. For example,the following relationship is valid for the temperature obtained from the intensity ratio of forbidden lines produced by the ion 0++: Tlo„n= r„[i +(?ÿ2-3)ÿ] U) valid for low valuesof t2. In this equation, T0and t2 correspondto the ion 0++. Peimbert and Costero(1969)also defined a characteristic value for the temperature (TA) of the region where a line with wavelengthA is produced by an ion X*. This tem¬ perature can be written as a function of the average temperature T0 and of t2 for the ion. The valueof TA, called the “line temperature”, would then be the most appropriated value to be usedin the equation that determine the ionic abundancefrom the measured intensity of a line producedby this ion (Eq. 1). Briefly, for the determination of the Proceedingsof the XXIthSAB 119 abundanceof a givenion, the line temperature must be calculated.This temperature,in turn, can be obtainedfrom the averagetemperature of the region wherethe ion exists, and from the temperature fluctuation of the ion. Sinceit is not possibleto determinet2and T&for eachion from the observations,some approximationsare usually donein the literature. In general,no temperature fluctuation is considered(t2=0), e.g., the nebulais isothermal.In this case, the averagetemperature of the ions and, consequentlythe line temperatures, are taken equalto the temperature obtained from an observedforbiddenline intensity ratio, generallyT[0///j. When both T[/v//]and T[o///] are available, Tpv//]is taken as the average temperature of the low ionization ions, while T[OIII] is consideredto represent the average temperature of high ionization ions. These temperatures are then consideredto represent the line tempera¬ tures. Using some appoximations,a value for t2 can be obtained from the observations. In this case,it is assumedthat T0and t2 are the same for all ions. The valuesfor T0and t2 can then be obtained from two equationsrelating theseparameters and observedtem¬ peratures. When more lines are observed,theseparameteres can be obtained separately for the low and high ionization zones. Thesemethodsprovidehigh valuesfor t2(> 0.01) whenapplyeclfor some HII regions(Torres-Peimbert& Peimbert1977),or whenaverages are obtainedfrom HII regionsandplanetarynebulae(Peimbert& Torres-Peimbert1971). In the literature,thesevaluesobtainedfor t2are appliedin the determination of chemical abundancesin HII regionsas wellin planetary nebulae. In a detailed study for the planetary nebulaNGC 7662,Harrington et al. (1982) obtained theoretical valuesfor t2 for this nebula;they argue that the valuesobtained for t2 by Peimbert and collaboratorsare too large. However,in an observationalstudy of severalPNe, Liu & Danziger(1993)also find high valuesfor t2. 2.1 Temperature fluctuations in HII regions Temperaturefluctuations have been discussedin detail for HII regions (Gruenwalcl& Viegas1992). Using a self-consistentphotoionization code, they obtained theoretical valuesfor the temperature fluctuation in a gas with typical physical conditions of HII regions. They also tested different hypothesesassumedin the empirical methods. The ionizingcentral source usedin the calculationsrepresent typical main sequencestars, with temperatures between30900K (B stars) and 50000K (04 stars). Modelswith black-body spectra as well as with differentmodel atmosphereswere alsoanalysed. The gas density, also typical of HII regions,ranges between1 and 103cm-3,while the abundancesvaries betweenthree times the solar to Z®/100.The resultsshowthat the valueof t2 obtained from theoretical modelsranges from 0 to 0.21 for a central source emitting as a black- bocly,andits exact value dependson the consideredion, gas abundance,gas density,and stellar ionizing flux. In particular, t2 decreasesrapidly with increasinggas density. Ions of the same ionization degreedo not havenecessarilythe same valueof t2. The stellar temperature is also an important parameter, but the t2 dependenceis not monotonic. Regionswith higher abundanceshavehigher temperature fluctuations. Different model atmosphereslead to different t2 for a given ion. To illustrate the results of Gruenwalcl 120 Proceedingsof the XXIth SAB 0.035 2A MODELS 0.030 Ne++ 0.025 0“ 0.020 s+ 0.018 0.010 0* 0.008 N* 38 40 48 80 T. (10*K) Figure 1: Temperaturefluctuation versus black-boclystellar temperature. Labelsrefer to the name of the correspondingion. Models use typical parameters for HII regionsand its ionizingstar h Viegas(1992),Fig. 1 showsthe behaviourof t2 with the stellar temperature for ions whichproduceobservedlinesin the spectrum of HII regionsand which are commonlyused for abundancedetermination. In the figure, the gas hasa solarabundanceand a density equal to 100 cm-3. A comparisonof the theoreticalresults with thoseobtained from the empirical relationshipsof Peimbert and collaboratorsshowedthat theserelationships are better than 10%for t2 less than 0.02; for greater valuesof t2, the error increases steeply. Empirical ionizing correction factors were alsoanalysed;a factor up to 3 is found betweentheoretical and empirical results for the elementalabundances(up to 10 for S). An estimate of the effect of t2in the calculation of the abundancesis not straightfoward, since there is the effect of the temperature fluctuation of the ions (including H+) on the ionic abundancesand consequentlyon the ionization correction factors (evenif their empirical formula are correct). Regardingthe effect on the ionic abundancesrelative to H+, the use of t2 = 0 can lead to an error up to 3 (greaterfor oversolar abundances). 2.2 Temperature fluctuations in planetary nebulae As said above,from a modelobtainedfor the planetarynebulaNGC 7662,Harrigton et al (1982)questionedthe valuesobtainedfor t2from the observations.In their modelfor NGC 7662they obtained valuesfor t2 much smaller then thoseprovided by the observations. As also discussedabove,theseobservedvalues are in generalobtainedfor HII regions,or from an averagefor planetary nebulae(PNe) and HII regions.PNe havein generalhigher density,lower metalicity, and higher stellar temperature, when comparedto HII regions. The results of Gruenwald& Viegas’(1992)analysisshowthat, for HII regions,t2 is lower for high density objects. Although PNe are denserthan HII regions,their central stars are usually hotter and t2 may be significant. Proceedingsof the XXIthSAB 121 Dinerstoin at al. (1985)from T[0///] obtainedfrom two line intensity ratios produced by the ion 0++ in the optical and in the infrared, obtained for 4 PNe high valuesfor t2(0++), greater than 0.03. Liu and Danziger (1993),in an observationalstudy for severalplanetary nebulae, assumedthat t2 and T0are the same for 0++ and H+. High valueswere obtained for t2, of the order of 0.03;for some objects t2 are greater than 0.10. In order to analysevaluesfor temperature fluctuations specificfor planetary nebulae, Gruenwald&; Viegas(1995)obtainedtheoreticalvaluesfor t2in physical conditionstyp¬ ical of this kind of objects. The centralionizing source was representedby stars with temperature in the range 30000K to 300000K and luminosities variyng from 10 to 20000 L0. The gas density ranges between102to 106cm-3. Thesevalues,following Pottasch (1984)are typical values for planetary nebulae and their central stars. The assumed abundancesare thosefrom Stasinska& Tylenda (1986). The obtained valuesfor t2 range from ~0 to 0.50, and depend,as for HII regions,on the specificion, and on the nebular as well as on the centralstar parameters. For the ions generallyusedfor abundancedeterminations,t2ranges from ~ 0 to 0.15. Figs. 2 (a), (b), and (c) show t2 versus the effective temperature of the central star with a luminosity of 3000 L0. In thesefigures, the gas density is, respectively,102,104,and 106cm-3. The behaviourof t2 for stellar temperatures lessthan 50000K, for a givengasdensity,obtained for HII regions (Gruenwald& Viegas1992)differs from those obtained for PNe due to differencesin the central stellar luminosity betweenthesetwo kind of objects. Part of the t2 differencemay be due to the adopted chemical abundancefor PNe which are sligthly different from those typical of HII regions. As shown in the figures, ions of different elementsin the same ionization stage can have different values of t2. It can be also be notice that t2(S+)is very sensitive to the stellar temperature. While t2(H+) is strongly dependenton the stellar temperature for high densities,t2(0++) is not important. In the literature the temperature fluctuation for these two ions are generally assumedto be the same. From the figuresit can be alsonoticed that t2 generallyincreaseswith the stellartemperature, andfor high stellar temperatures the temperature fluctuations can be important even for high densities.Low ionizations ions,like N°, havehigh t2(up to 0.20). If theselow ionization lines are used,misleadingvaluesof abundancescan be obtained. Figs. 3 (a) and (b), showfor densitiesof 102and 106cm-3,respectively,the behaviour of T.Hail T[o///] ) and the line temperature for the [OIII]A5007,T50o7,with the stellar temperature. Thesesfiguresshow the results for a stellar luminosity of 3000L0. It can be noticed from the figuresthat for high stellar temperatures and high gas densitiesthe line temperature can not, for example,be taken to be equal to the temperature obtained from the forbidden intensity lines as it is usually done. Comparingthe results for t2of Gruenwald& Viegas(1995)with thosefrom the liter¬ ature, for NGC 7662they agree with thoseobtained theoretically by Harrington al (1982) for the same nebula. Regardingthe valuesobtained observationallyby Liu & Danziger (1993),valuesfor t2 for 7 out of 11 objects, are in agreement. Notice,however,that these authors consideredthat the average temperature and the temperature fluctuation are the same for 0++ and H+ ions. A comparisonwith the valuesobtained from the observations 122 Proceedingsof the XXIth SAB HY /S« 0.03 s*y 0.03 s*y NeV 0.02 HJ. 0.02 Jr" | | 0* 0.01 7 0.01 "o*1 0.00 0.00 50 100 150 200 250 300 50 100 150 200 250 300 T. (10a K) T. (10» K) JT rs* 0.03 s; I 0.02 0.01 0i Figure 2: Temperature fluctuation versus stellar temperature for stellar luminosity L/L0 = 3000. Figures (a), (b), and (c), correspond to densitues of 102, 104, and 106cm-3, respectively. The parameters are typical of PNe and their central ionizing star 20.0 30.0 1 0.0 1 s.o > - 4 £ 10.0 I <ÿ*> o.o 60 1 OO lOO 300 360 300 60 1 OO 160 300 360 300 T. (10° K) T. ( 1 Oa K> Figure 3: Valuesfor the gas temperatures versus stellar temperature. Solid lines refer to T 5007 1 dotted lines to TBah and dashedlines to T[0ur\- densitiesfor Fig. 3 (a) and (b) are, respectively, 102and 106cm-3. The luminosity is 3000 L0 Proceedingsof the XXIthSAB 123 Table 1, Parameters of the studied Planetary Nebulae Object T. U/L0 n NGC 650 150000 300 4780 NGC 2438 145000 120 500 NGC 2440 280000 a 5400 NGC 2452 121000 1300 2130 NGC 2818 210000 220 800 NGC 2867 130000 2500 4510 NGC 3211 150000 1900 900 NGC 3587 108000 44 4570 NGC 3918 141000 4200 6200 NGC 6302 224000 1780 25520 NGC 6309 100000 4300 2710 NGC 6439 147000 3600 4000 NGC 6445 191000 255 2470 NGC 6537 200000 745 15490 NGC 6565 130000 930 300 NGC 6720 134000 530 600 NGC 6741 175000 3180 11600 NGC 6818 191000 1820 2050 NGC 6853 147000 110 4540 NGC 6886 143000 6300 7180 NGC 6905 103500 1350 870 NGC 7027 174000 12700 82800 NGC 7293 110000 36 225 NGC 7662 112000 10700 3250 IC 2165 158000 5200 3930 Hu 1-2 130000 b 4100 J 900 129000 4700 3200 124 Proceedingsof the XXIthSAB by Dinerstein et al (19S5)show an agreement for 1 out of 4 objects. However,there are great incertantities in theseinfrared observations,ancl at least 2 of these 4 objects are clearly asymmetrical, while the theoretical models assume a sphericalsymmetrical geometry. As discussedabove,temperature fluctuations can be important for planetary nebulae whosecentral star has a high temperature (> 100000K). An applicationfor planetary nebulaewith a high temperature star was also done in Gruenwald& Viegas (1995). Planetarynebulaewith high temperature were selectedin the literature, and modelswere performed for those with known stellar luminosity and density. The selectedobjects are given in Table 1. In the table, the columnscorrespondto the name of the object, its central stellar temperature (in K degrees)and luminosity (in units of L0), and gas density (in cm-3). Using the model with the most similar input parameters, temperature fluctuations for eachion were then obtained. With thesevaluesfor t2(X8)and observed valuesfor the line intensities obtained from the literature, ionic abundanceswere then calculated. When no temperature fluctuation is considered,lower valuesfor the ionic abundancesare obtained. However,for the analysedobjects, the differencesfor the ionic abundancescalculated with and without considering temperature fluctuations, are in generallessthan 10%.For nineteen out of 27 nebulae,one or more ions have a difference of up to 20%.The exceptionis S+,with a differenceof up to 40%. 3 Conclusions and perspectives The resultsof the precedentsectionsshowthat the temperature fluctuation for a givenion can beimportant dependingon the physical conditionsof the nebula. A high temperature fluctuation can affect the determination of the abundancesin a nebula. An analysis for some planetary nebulae with high stellar tempaerature, showed,however,that the combiningeffect of temperature fluctuations in the variousparameters that enter in the calculationof ionic abundancesaffect the resultsin lessthan 40%. The theoretical calculationsup to now assume a sphericalsymmetricalgeometry for the ionized gas and a homogeneousdistribution for the density. However,in addition to the ionization distribution, some other factors can produce spatial fluctuation of the temperature. The effect of density variations on T[o ///] and TBai were studiedby Viegas & Clegg(1994). Regionswhich present chemicalabundanceinhomogeneitiesas well as multiple interconnectedHII regionscan alsopresent temperature fluctuations. For a more detailedstudy of asymmetrical and/orinhomogeneousregions,a 3-dimensionalcode was developed,and an aplication to bipolar planetary nebulaeis under progress (Gruenwald & Viegas,in preparation). References [1] Dinerstein,H.L., Lester,D.F., Werner,M.W. 1985,ApJ 314,551 Proceedingsof the XXIthSAB 125 [2] Gruenwald,R., Viegas, S.M. 1992,ApJS 78,153 [3] Gruenwald,R., Viegas,S.M. 1995,A h A 303,535 [4] Harrington, J.P., Seaton,M.J., Adams,S, and Lutz, J.H. 1982,MNRAS 199,517 [5] Liu, X., Danziger, J. 1993,MNRAS 263, 256 [6] Peimbert,M. 1967,ApJ 150,825 [7] Peimbert,M., Costero,R., 1969,Bol. Obs. Tonz. y Tac. 5,3 [8] Peimbert,M., Torres-Peimbert,S. 1971,ApJ 168, 413 [9] Pottasch,S. 1984,”Planetary nebulae”,D.Reidel [10]Torres-Peimbert,S.,Peimbert,M. 1977,Rev. Mex. Astron. Astrofis. 2, 181 [11]Stasinska,G., Tylenda. R. 1986,A & A 155,137 [12] Viegas,S.M.,Clegg,R.E.S.1994,MNRAS 272,993 Proceedingsof the XXIth SAB 127 THE INTERESTELLAR EXTINCTION CURVE AND THE POLYCYCLIC AROMATIC HYDROCARBON Heloisa M. Boechat-Roberty (Observatório do Valongo - UFRJ) Marcelo B. Fernandes (Observatório do Valongo - UFRJ) Carlos A. Lucas (Instituto de Química- UFF) M. Luiza Rocco (Instituto de Química- UFRJ) G. Gerson B. de Souza (Instituto de Química- UFRJ) Abstract The interestellax extinction curve at ultraviolet region has a broaclbump centered at 2175 Â(5.7 eV) with a width at about 480 Â. It hasbeen sugestedthat the absorption peakis causedby smallgrains andby polycyclic aromatic hydocarbonmolecules(PAHs). Electron energy-lossspetra were obtainedin laboratory at vacuum ultraviolet region in severalscattering anglesfor naphthalene and benzenemoleculesin order to determine cross section and oscillator strength of transitions in this region. Near zero scattering angle the incident electronsbehavior like pseudophotons and the electron energy-loss spectra are consideredas photoabsorption spectra. By comparision of the energy-loss spetrum with the interestellar extiction curve, a strong absorption at 5.9 eV (2138 Â) was observedthat can be confirm the contribution of this moleculesfor the interestellar extinction. 1. INTRODUCTION The interestellar extinction of starlight is due to the absorption and scattering processes by the interestellar matter. The extinction curve showsthe behavior of the 1 normalized color excess in magnitudesE{\ —V)/E(B —V) as a function of 1/Ain (j,m~ . Averagenormalized curve obtained from a number of determinationsis illustrated in Figure 1 (Savageand Mathis 1979). At ultraviolet region a broad bump centered at 2175Â (5.7 eV) with a width at about 480 Â lias beensugestedthat is causedby small particles (Boulanger et al. 1994) and by polycyclic aromatic hydorcarbon molecules PAHs (Joblin et al 1992). The PH As moleculeshave been found in severalobjects such as planetary nebulae,reflection nebulae. The infrared emissionbands at 3.3, 6.2, 7.7, 8.6 and 11.3 /«n known as the unidentified infrared band were interpreted as beeingdue to PHAs. The aromatic structure of this moleculesprovidesthe stability necessary for a long life-times in the interestellar medium, and allows for the conversion ultraviolet energy into infrared emissonbands by the fluorescencemechanism (Leger and Puget 1984). Electron energy-lossspectroscopy has beenused for determination of excitation energies,cross section and oscillator strengthsof valence(Boechat-Robertyand de Souza 1992) and core-levelexcitations (Boechat-Robertyet al 1991). As part of a systematic study of the angular behavior of electron impact excitation, we present in this paper the generalizedoscillator strength (GOS) for the 1 Ag transition (5.9 eV) in the naphthalene moleculeand the electronenergy-lossspectra for the benzenemolecule. 128 Proceedingsof the XXIth SAB 12 Savage & Mathis (1979) lO 2 1 75 A § . (naplithalcne absorption) £ =ÿ 2 138 A O t TJV V '» - IR o 2 6 lO l/X (urn)-1 Figure 1: Interestellar extinction curve (Savageand Mathis 1979). 2. THEORY The differential cross section, (da/dQ,),for the excitation of a molecule with N electronsand M nuclei by an electron with kinetic energy E is given by the following expression: (da\ JL Jkfl..(JOi>)] (1) ) 1 \dtt n |K|4 |k,| Heren representsthe final excited state, K is the momentum transfer,K = ki —kf, and ki and kf are, respectively,the initial and the final momentum vectors for the scattered electron and e(K) the scattering amplitude. A very useful quantity is the Generalized Oscillator Strength(GOS),fn(K), definedas 2E /4*0 = iiq7 MW (2) whereE is the excitation energy and |K|2 is given by |K|2 = |ki|2+|kf|2 —2kikfcos0 (3) where 6 is the scattering angle. Substituing |e„(A’)|2 from equation (1) into equation (2) gives (4) The GOS,useful in the electron-scatteringtheory, is independentof the incident energy at energiesfor which the First Born Approximation is valid and approaches Proceedingsof the XXIth SAB 129 N aphthalenc 3.50° m 1.75° 1.00° o 10 20 30 40 Energy Loss (eV) Figura 2: Electron energy-lossspectra for naphtalenemolecule measuredat scattering angles1.00°,1.75°,2.75°,3.50°. the familiar Optical Oscillator Strength (OOS) as K —» 0. It means that, at small scattering angleor at higth impact energiesthe electron energy-lossspectra are similar to photoabsorption spectra. A very interesting review of the GOS properties can also be found in Inokuti (1971). Ben2sene 4.0° -E3 A 3.0° .s- 1J3 2.S° 2.0° O 5 10 -to 20 25 30 30 40 46 Energy Loss (eV) Figure 3: Electron energy-lossspectra for benzene molecule measuredat different scattering angles. 130 Proceedingsof the XXIth SAB 3. EXPERIMENT The same basic experimentalprocedure previously adoptedin the determination of the oscillator strength for other excitation processesin different targets was usedin the present work. Briefly, electron energy-lossspectra using a crossed-beamspectrometer were measuredfor a givenset of scatteringanglesat a fixed impact energy. The crossed- beamspectrometer usedin the present workhasbeendescribedbefore(BoechatRoberty et al 1991).In order to obtain a higherluminosity and to work in a constant transmission mode, a newly developedWien Filter type velocity analyzer was incorporated into the energy-lossspectrometer. The Wien Filter analyzer features two three-elementzoom lenses. The first one brings the scatteredelectronsto a fixed pass-energy, for example 30 eV, resulting in an energy resolution of 0.8 eV, as determined by the full width at half maximum for the elastic peak, measuredin the same conditions. The second zoom lens re-acceleratesthe electronsin order to increase the detection efficiency. The energy resolution of the apparatus is limited to 0.5 eV by the thermionically produced incident electrons. In the present case, though, a compromise was attained between resolution and count rate. The velocity-analyzedscatteredelectronsare detectedby an electron multiplier (Spiraltron, Galileo Electro Optics). The pulses coming out of the detector are fed to a standard electronics (preamplifier, amplifier, discriminator) and are then then transferred to the memory of a 16-bits personal microcomputer which is also responsiblefor the scanning of the voltagesin the zoom lenses. In the present measurements the incident electron energy was 1000 eV and the energy-lossspectra were obtained in the angularrange of 1.0° - 4.0°, with energy resolution of 0.8 eV. The scattering zero-degreeanglewas checkedby measuring the elastic peak area over a 15° range on both the right hand and the left hand side with respect to the electron analyzer main geometricalaxis. Although a precisionmechanismallowsthe determination of the scattered angle with an accuracy of 0.02°,our angleresolution, defined by the set of apertures in front of the electron analyzer,is limited to 0.23°in the present experiments. The area for the peak correspondingto transition was determined using a Gaussian fitting procedure previously discussed.The background contribution was determined by measuringthe spectra at the same scattering angles,but allowing the gas to enter the scattering chambertrough a side flange,located far from the collision center. This contribution was then subtracted from the raw data. As the main interest consistedin the determination of the behavior of the area of the peak as a function of the scattering angle, great care was taken in order to make sure that all the spectra were obtained in the same experimental conditions. The experimentaluncertainties are estimated in the following manner. The maximum statistical uncertainty was 3.5%as at least 2.300 counts were accumulatedat the peak of interest for eachscattering angle.Fluctuations in the primary beam current and on the samplepressure were of the order of 1% and 0.5%, respectively. The main source of error is the limited angular resolution (0.23°) which dependson the scattering angle. The determination of the area of the peak is 1/2 , also subject to an error of 2%. The overall uncertainty, 6, defined as 6 = (£ 6;) , IS equal to approximately 6%for small angles,and 4%for large angles. Proceedingsof the XXIth SAB 131 OOS 1 1 ---- IVaphthalene T imsi-tiun c_r> 0.00 0.05 o. io O. 1 5> 0.2.0 K2(au) Figure 4: : Generalizedoscillator strength ( GOS) as a function of square 1 momentum tranfer, A”2, for the 1Blu<—Ag transition in the naphthalene molecule. 4. RESULTS AND DISCUSSION The electron energy-lossspectra for naphthalene and benzenemoleculesmeasured at different scattering angleare shownin Figure 2 and Figure 3, respectively. The spec¬ tra, where the energy-losscorrespondsto excitation energy, exhibit strong absorption in 5.9 eV (2138Â) for the naphthaleneand 6.8 eV (1823Â) for the benzenemolecule,both due to dipole-allowedtransitions 7r and n* . As a result the spetra obtained at small scattering anglesresemble that measuredby other absorption techniques. At highter electron scattering anglesdipole-forbidden processes can also be observed,a major ad¬ vantage of this technique over other conventional absorption techniques. The strong naphthalene absorption is near of the A = 2175Â feature seen in the UV extinction 1 1 curve. We determined the GOS as a function of K2 for the BUi<— Ag transition in the naphthalene molecule shown in Figure 4. The value of the optical oscillator strength (1.3) was obtained by the GOS curve extrapolation to zero scattering angle or IO — 0. 5. CONCLUSION Wehaveobtainedelectron energy-lossspetra in laboratory at vacuum ultraviolet re¬ gionin severalscattering anglesfor naphthaleneand benzenemoleculesand determined 1 1 cross section and generalizedoscillator strength (GOS) for the Biu<— Aatransition in the naphthalenemolecule. Near zero scattering angle the electron energy-lossspectra are consideredas photoabsorption spectra. By comparision of the energy-lossspetrum with the interestellarextiction curve, we suggest that the strong naphthalene absorption at 5.9 eV (2138 Ã) contributes to the UV bump and benzene molecule can contribute for the interestellarextinction. 132 Proceedingsof the XXIth SAB 6. REFERENCES Boechat-Roberty H. M., BielschowskyC.E., de SouzaG.G.B.,1991,Phys. Rev. A 44, 1694 Boechat-Roberty H. M., de SouzaG.G.B 1992,J. Phys. B:At. Mol.Opt.Phys. 25, 4641 Boulanger F., Prévot M.L., Gry C. 1994,Astron. Astrophys. 284, 956 Inokuti M., 1971,Rev. Mod. Phys. 43, 297 Léger A, Puget J.L., 1984,Astron. Astrophys. 137, L5-L8 SavageB. D., Mathis J. 1979,Ann. Rev. Astron. Astrophys. 17, 73 Proceedingsof the XXIthSAB 133 STRUCTURE OF INTERSTELLAR MEDIUM: A MULTIVARIATE APPROACH Jorge R. Ducati Jandyra M.G. Fachel Institutos deFísica e Matemática, UniversidadeFederaldo Rio Grandedo Sul, Porto Alegre,Brasil Abstract Sometechniquesfrom multivariate data analysisare presented.Applications to infrared data of stars show how the method helps to identify classesof stars and anomalousefFectslike circumstellar emission,and to detect observationalerrors. Key words: statistical methods - interstellar medium - infrared 1 Introduction New instruments and techniquesto collect astronomicaldatahavegeneratedan increasing inflow of information availableto investigators,to the point that presently only a small fraction of the stored information in all regionsof the electromagneticspectrum, or from cosmicrays, is effectively analyzed. Effects of observationalbias are present not only from the persistent tendency of astronomers to observeobjects which,in some way, are more conspicuous,but alsofrom the analysisof vast amounts of digitized data. We can wonder about discoveriesuntil now deeply buried (perhapsforever!) in shelvesof stored data, after analysiswhich looked only for a precisephenomenon;but we have to recognizethat it is impossible,from an human point of view (and even for computers ?) to face the asymptotic growth of available data. Projections were made on the geometricalincrease of the number of volumes and pages of The AstrophysicalJournal, from year to year, and the resulting rate of linear occupation of shelvesin libraries. It turned out that this speed,in a foreseenfuture, will exceedthe speedof the light ! To this predictions some critics retort, saying that there is no physical deadlock,since the Ap. J. carries no information... Nevertheless,the present trend to publish resultson electronic support is already an option which will replace, to an unknownpoint, the ancient use of solidly printed information. The problemsof big, multiple data, andof the difficulty to handle and correlatemulti¬ classinformation, are already set. An analysisin many wavelengthsis cumbersome,even if only one object is studied;the problem is amplified when we are confrontedwith large lists, database-size.Facing thesechallenges,it is useful to turn to more powerful tools, brought from other sciences.We want in this article to provide an exampleof the use of statistical techniquescolletively calledmultivariate data analysis,appliedto an extensive set of stellar measurements,mainly in the infrared. For thosemore interested,an approach on applicationsof this method in astronomy may be found in Murtagh and Heck (1987). More comprehensivetexts are Hair et al. (1987)or Chatfield and Collins (1980) 134 Proceedingsof the XXIthSAB 2 Applications The catalogueof stars observationsin theUBVRIJHKLMNcolors(Ducati 1993)contains data for about 4000 stars. Besidesthe colors(U—V) to (N—V), other informations are provided, including coordinates and spectral types. Associated, derived lists, provide distancesand color excesses. Thosedigitized files fill approximately1.5 MB each;they allow either searchesfor particular objects, or a vast range of studiesof many possible correlations. Sucha largelist contains many classesof objects; furthermore, the volume of space sampledby theseobservationscovers different regionsin the Galaxy, regarding stellar density, metalicity, or properties of the interstellar medium like reddeninglaws. The span of possiblecorrelations,and the sizeof the data file, makean ordinary analysis by standard methodsa tremendoustask. Besides,one should allow for the probability of surprising, unexpectedcorrelations (or non-correlations), to be uncoveredby special methods. The multivariate data analysisis a tool to deal with theseproblems,using techniques like : - principal components analysis : where a set of variables is transformed in a much smallerset of factors; eachfactor is a linear combination of variables,weightedby coeffi¬ cients,and contains the maximum of original information. Graphical representationsof the transformed data are possible. - cluster analysis : subdividesthe data in groups or clusters,according to similarity criteria, which are specifiedby the user, who also sets the number of clusters. - discriminantanalysis: discriminates individual objects into predefinedgroups, ac¬ cordingto a function which is a linear combination of severalvariables. - multiple regressionanalysis: finds an equationor model that predictsthe dependent variable,as a function of one or more independentvariables. We applied the cluster analysis to our file of infrared data, taking as variablesthe color excesses E(U—V), E(B—V), E(J—V) and E(K—V), for which data for more than 1300 stars exists. To cluster analysisit is crucial the choiceof the number of clusters, so that the maximum information is preserved. A number too small will associatestars with only coarse similarities,while too many groups will put in different groups stars with subtledifferences,possiblyof the orderof observationalerrors. For a sampleof 1613 stars of all spectral types, an analysisgeneratingten clusters was a fair compromise. A goodexampleof how stars are partitioned in groups is as follows: by ordinary sorting we selectedthe eight most distant stars in the catalogue,all at distancesgreater than 20 000 parsecs; this group was comparedwith one of the ten generatedclusters,containing 18 stars. All those eight distant stars belong to this group createdby cluster analysis, togetherwith other four stars with distancesbigger than 17 200 pc; the remaining five objects are main sequence stars, which are certainly in the solar neighborhood,even if the actual distance is a missing data in the catalogue. But from their spectral types we detect binarity, variability, emission or peculiar features. This illustrates how the method can either group similar stars, or reveal cases of anomalousobservationaldata, from circumstellar emissionor other physicalreasons. Proceedingsof the XXIthSAB 135 Regardingnon- hierarchicalcluster analysis,two points shouldbe stressed.One,the concepts of centroidand distanceto the centroid. In the methodeachcluster is computed, after the pivotal variables(for example,distance,E(B—V), E(K—V), E(L—V)). For each cluster definedunder theseconstraints, there is a center of coordinatesin the multi¬ dimensionalspace of the variables,calledthe centroid. Eachstar belongingto the group has a distancefrom the centroid. Examination of thosestars looselyassociated(with the larger distances)gives clues to either observationalerrors, peculiar stars or anomalous extinctions. Socluster analysisis a tool to searchfor outliers, when working with large lists. This leadsto the secondpoint. The partition of a cataloguein a numberof clusters (ten, in our case) may generate one or two groups with cpiite a few stars. Thesein generalare either objects with gross observationalerrors (like misidentifications),bias (like circumstellarmatter), or are specialcases of great interest. Wehavealsoto refer to themultiple regressionanalysiscapability to providefunctional relationshipsbetweencolor excesses,either bi-dimensional(likeE(B—V) vs. E(K—V)), or multi-dimensional(like E(B—V) as a combinationof severalinfrared excesses).The im¬ portant point is that not only this is possiblefor a wholelist, like our completecatalogue, but this alsocan be doneinsideeachcluster,whichis a much smallerset. Clusterstend to associatesimilar stars, or stars under the same environmentalconditions. Transformation equationsfor eachcluster providemeans to evaluateparameters for stars with incomplete (missing)data, with better accuracy than if relations derivedfrom entire samplesare used,sincein this last case all kinds of stars have their data considered. Clearly, this is valid only if the number of stars inside a cluster is big enoughto provide a good statistics. TABLE i. E(U-V) E(B-V) E(J-V) E(K-V) E(L-V) E(M-V) E(N-V) E(U-V) 1 E(B-V) 0.9588 1 (1370) E(J-V) -0.8029 -0.8726 1 (1250) (1299) E(K-V) -0.8186 -0.8883 0.975*1 1 (1366) (1*121) (1458) E(L-V) -0.8001 -0.8795 0.9378 0.9818 1 ( 883) ( 906) ( 9*12) (1042) E(M-V) -0.5695 -0.6928 0.8349 0.9467 0.9763 1 ( 198) ( 213) ( 208) ( 226) ( 195) E(N-V) -0.4402 -0.5536 0.5117 0.8703 0.9304 0.9708 1 ( *1*1) ( 45) ( 45) ( 47) ( 44) ( 34) Other interesting conclusionsmay be drawn from regressionanalysis. The table I showsthe correlationcoefficientsbetweensome selectedcolor excesses (regressionsper¬ formed for completelist of 1613 stars; numberof common stars below coefficient): This givesthe level of reliability wheninfrared excesses are used to evaluateexcesses in the blue region, for very red or very absorbedstars. We can see that observationsof the N band at 10/iare prone to contain a largerscatter, frequently due to the difficulty of observingat thosewavelengths,or to differencesin filters from variousobservers.We see 136 Proceedingsof the XXIthSAB also that correlations for E(L—V) are of the same order than for J or K, in spite of the circumstellaremissionoften associatedwith observationsin L. This is due to the large amount of data involved;in fact, closeexamination showsthat approximately1%of the L data containsnon-stellaremission. 3 Conclusions Our concludingremarkspoint out to the needof usinglargedatabases;this allowssubdivi¬ sionsof the samplein spatialzones andstar populationswhichare statistically meaningful, whichin turn givessolid groundto apply multivariate analysis.It is clearthat the space is not homogeneouswith respect to the distribution of interstellar matter or its properties. Certainclassesof stars havetheir own circumstellarcontribution to observedphotometric indices,which are usually supposedto provideinformation on interstellar processesonly. To distinguishtheseeffectsin largeamounts of data requirespowerfultechniques.Facing databasesof astronomicalsize,multivariate analysisand other statistical methodsare fundamental tools from now on. 4 References Chatfield,C.,Collins,A.J. 1980, Introduction to Multivariate Analysis, Chapman and Hall, London Ducati,J.R. 1993, A Catalogueof 11-Color Observationsof Stars, Wisconsin Astrophysics 504 Hair, J.F.,Jr., Anderson,R.E., Tatham,R.L.1987,Multivariate Data Analysis, Macmillan,Lon¬ don Murtagh,F., Heck,A. 1987,Multivariate Data Analysis, Reidel,Dordrecht Section5 Stellar Astrophysics Proceedingsof the XXIth SAII 139 THE SOUTHERN HEMISPHERE YOUNG STARS SURVEY C.A.O. Torres and G.R. Quast LaboratórioNacional, de Astrofísica/CNPq/MCT Abstract The discoveryof isolated young stars lias shown the need of reconsideringthe spatial distribution of theseobjects traditionally characterizedby their proximity to denseinterstellar clouds, where they are frequently found in associations of similar objects. In order to investigate the subject we have carried out a survey selecting 4041 sources from the IRAS Point Source Catalog by their spectral indexes. The final sampleof 502 optical objects south of 30°N to be observedwas definedbased on imagesfrom the Digitized Sky Survey. The nature of the possibleoptical associ¬ ations with the IRAS sourceshasbeentestedusing spectra at the II cvregion. A star has been classified as T Tauri (TTS) whenit has presentedboth Ila emissionand the Lil A6707absorptionline and as Iierbig Ae/Bestar (HAEBE) if hotter than F5 with Ila emission. The reduction of data is now ready for 95%of the survey. Until now there were discovery of 50 new TTS (plus 11 to be confirmed), 112 possible Iierbig Ae/Bestars, 3 probablepos-fuoristars and,surprisingly,14lithium rich cool giants. During the preliminary stage of the survey 12 other TTS havebeen found (12 others awaiting confirmation), 3 possibleIIAEBE and 1 cool lithium rich giant star. Some young objects found in the survey are located far from star forming regions. 1 Introduction During the pre-main sequencephasestars are usually placedin associationsrich in gas and dust and this environment is usedfor the traditional definition of young stars. This is the case not only for TTS - young stars havingmasses around1 A/®and spectral types G, K, M - but alsofor HAEBE, stars whosemasses are greater than 2 Af®and earlierspectral types. The discoveryof a young double-linespectroscopicbinary star, V4046Sgr (de la Rezaet al., 1987),challengesthis definition. The presenceof Lil and strong Ha emission indicatesthis star to be in the pre-main sequenceevolutionary stage. But in its spatial neighborhoodthereseem to be neither clouds nor stars of the same age, compellingit to be an “isolated TTS” (Torres et al. 1987a). Other isolatedTTS were found (Torreset al. 1987b),but in a reducednumber. It is important to find other examplesin a systematic way in order to analyzethe frequencyof this occurrence. TTS surveys havebeenlimited to denseinterstellar cloudsand generallyusedobjective-prism plates in the Ha region. Even though the Ha emission is generally strong in TTS, an all-sky survey may not be cost effective. An independentandpotentially efficientselectorhasbeenpresentedby Gregorio-Hetem et al.(1988).They noted that TTS in ChamaeleonCloudshave all similar infrared colors as measuredby the InfraredAstronomicalSatellite (IRAS). Analogousresult was presented 140 Proceedingsof the XXIthSAB by Emerson(1988)for TTS in general.From theseresultsemergedthe idea of making a young star survey without positional bias - the Pico dosDias Survey (PDS) - using the selectorbased on IRAS colors. The infrared emission associatedto TTS is supposedto originate from an accretingdust disk surroundingthe central stellar object (e.g.Hartmann & Kenyon, 1988). The spectral distribution of this process will dependmainly on the temperature distribution in the disk, defining a box in the IRAS color-colordiagrams. The use of IRAS emissionto searchfor young stellar objects restricts the PDS to TTS havingsuchdisks. The TTS may be dividedin classicalT Tauri stars (CTTS)or weakT Tauri stars (WTTS)if the equivalentwidth of the Ha emissionline is greater or smaller than 10Â, respectively. It has been proposedthat only CTTS would have such disks. In this case the formation of intenseHa emissionline would dependon thesedisksin some way and the geometry of the disk wouldbe an important factor in the emissionand detection of Ha. Dependingon the line-of-sight,the diskitself may block the view of the Ha emitting region. Thus, genuineCTTS could be detectedeven though their Ha is in absorption or faintly in emission. A searchbasedspecificallyon IRAS emissioncan also shedlight on this question. 2 Observations As the survey extendedover severalyears, some changesin methodologyhaveoccurred. Even new results causedsome feedback.Thus,the PDS may be dividedin 4 stageswhich will be called “Eras” (besidesanother one called “pre-history”).In all of them the objects were selectedfrom the IRAS Point SourceCatalog (IPSC). The observations were made with the coudéspectrographand 1.6m telescopeof Observatóriodo Pico dosDias (OPD) operatedby Laboratório Nacional de Astrofísica (LNA). The spectra haveresolution of 0.7Â and include the region of Ha and Lil A6707lines. Two different GEC CCDs were used,with 578 or 1152 pixels in the dispertion direction. This spectralregion, near the maximum of the detector sensibility, and the coudé focus of the OPD, equipped with specialhigh reflectivity mirrors, givesan efficient combination.The resolutionallowsthe detectionof spectroscopicbinaries as well as discerningbetweendwarf and giant stars. A star was considereda TTS if it presentsboth Ha emissionand the Lil A6707absorp¬ tion. This definition is practical, objective and can be usedin any region of the sky, making the classificationmore reliable. Its main drawbackis the limiting magnitude, not allowing the inclusion of extremely embeddedobjects presentingstrong extinction and highermagnitudes. On F stars, whereconvectionis not important, there is no efficient mixing betweeninternal and external layersand, as a consequence,the photosphericLi abundanceis not a good indicator for stellar ages. Although the proposeddefinition for TTS is usedin the PDS for F stars, their classification must be acceptedwith caution and needsother evidencesfor confirmation. However,there are only a few cases in this situation. The question of the HAEBE definition is more complex. There is no such a similar diagnosis.The time of evolutionto the main sequenceis very short,not onlyin comparison Proceedingsof the XXIthSAB 141 to the dispersionof ages in the forming regions, but also with respect to the probable lifetime of the circumstellardisk. The star rapidly passesthrough distinct evolutionary stages,makingit difficult to characterizethe group accordingto spectralproperties. Only a set of severalcriteria will produce a reliableclassification.Oneof the strongest criteria proposedby Herbig(1960),their locationin an obscuredregion,is uselessfor the purposes of the PDS. A discussionabout difficulty can be found in The et al. (1994;hereafter TWP). Emission line stars earlier than about F5 with no lithium line were tentatively classifiedas possibleHAEBE if their spectra are not very peculiar. So,at this stage of PDS, as a first approach,the possiblyyoung stars were classifiedas TTS if they are later than F5, and as HAEBE if they are earlier. For massiveobjects, the disk can survive the HAEBE phase. Among the selectedstars one can find ft Pic which doesnot have Ha emission and must be closeto the main sequence. If in the PDS a (3Pic-like object was observed,it wouldbe taken for an optical- IRAS coincidence.Nevertheless,TWP have consideredthem possiblyrelated to HAEBE. Anomalieson the Ha profile may indicate the presenceof shells.There are other possible complications. Most of the objects thought to be in a pos-main sequencephase,such as classicalBe, B[e],LBV, etc, havesome HAEBE-like characteristics. TWP prefer to considerthem as possibleHAEBE. It is not only a confusionof spectralfeatures,but there are alsodoubts about their evolutivephase.There are alsoambiguitiesin distinguishing protoplanetary nebulaefrom HAEBE. Their spectralfeatures are similar and PDS and surveys on protoplanetary objects havefound the same stars due to the superposition of the IRAS color boxes. Other parameters are neededfor the classificationof early type emissionline objects. Another group of blue stars showedup in the survey: very bright stars, belongingto the Bright Star Catalog(BSC).Almost all of the 31 BSC stars observedin the PDS are hotter than AOwith 4 exceptions:one A3, one A5 (f3Pic), one F0 and one G3. The histogram of the V-[12] color index showsa bimodal distribution and all “blue” V-[12]stars belong to to form an easily distinguishableclassof objects. 3 The Eras During the executionof the PDS the selectioncriteria and the observingprocedureshave changeddependingon the availablefacilities. For instance,in order to observea target object it is not enough to point the telescopeat the IRAS coordinates due to the low positionaluncertainty of the IRAS and to the telescopepointing errors. The procedures to overcome theselimitations changedfor eachEra. 3.1 The Early Era The selectionof the target has followedthe criteria below: i - IRAS positionssouth of declination+30° ii - fluxes in 12, 25 and 60/imabovethe detection limits in the IPSC 142 Proceedingsof the XXIrhSAB iii - absenceof associations with galaxiesor planetary nebulaein the IPSC iv - infrared color box definedaccordingto: -0.02 < [12]-[25]< 0.53 -0.30 < [25]-[60]< 0.52 A list with 888 target sources was built using the IPSC version 1.0. The telescopewas pointed to the coordinatesof the targets and the field checkedfor an observableoptical object nearby. 'Thepointing of the 1.6m telescopewas not very accurate and the available field at the coudé focus being very small forced a wanderingaround in the vicinity of the aimed position. Besideswasting time with sources without observableoptical objects. one could not be sure if the objects found were inside the IltAS error ellipse,and stars far from the IPSC position with very low associationprobabilities could be observed. Anyway, there were interesting discoverieswhich can be classified as sereudipities. At this Era the GEC P8603385x578pixels CCD was used(Gregorio-Metem et al. 1992). In the PDS pre-history, targets were selectedwith no restriction in the IRAS fluxes around TW Hya, a high galactic latitude TTS, to test its “isolated” status. Two new TTS were discovered,one of them being a visual binary. Hen 600 (de la Reza et al. 1989). Since the distancebetweenoptical and IRAS positions was not checkedduring observations, only Hen 600 was actually associatedwith an IRAS source and belongsto the list of 888 sources. 3.2 GSC Era The use of the Guide Star Catalog (GSC) (through an optical disk kindly provided by NASA) allowed the correlation betweentarget sources and optical objects, avoiding the blind searchat the telescope.It was possibleto know the relative positions of the optical objects and which ones were included in a. chosenerror ellipse. With the help of G.L. Vieira the 888 positions were checkedin the True Visual Magnitude PhotographicStar Atlas (Papadopoulos& Scovil.1980), that hasa similar limiting magnitude,beingpossible to showthat the incompletenessof the GSC for this sampleis less than 13%(Gregorio- Hetem et al. 1992), which was consideredacceptableby the gain in efficiency. The limiting magnitudeis almost identical to the instrumentalone used in the PDS. The missingbright stars can be obtained from catalogsas the SAO. Resultsof the Early and GSC Eras appear in Gregorio-Hetemet al. (1992)and Torres et al. (1995). 3.3 Weintraub’s Era Weintraub(1990),making use of a larger number of young stars, mainly belongingto the Third CatalogofEmission-Line Stars ofthe Orion Population(Herbig & Bell 1988;IIBC), has obtained more reliable limits for the color box. Although it represents an increase in the number of sources, and possibly also in the number of spurious associations,the use of GSChelped in keepingthe work feasible. The IPSC version 2.0 was usedleading to the exclusion of a few sources of the previousEras. Conditions (ii) and (iii) from the Early Era were kept and the color box definedby: Proceedingsof the XXIthSAB 143 -0.95 < [12]-[25]< 0.11 -0.95 < [25]-[60]< 0.32 The new list had 3540 sources all over the sky. An IRAS error ellipse of 3.3ct was used to include 99,9%of the associatedoptical objects. Through the use of GSC and SAO, 702 sources were selected,540 of them south of +30°. Asmuch as possiblethe GECP882001152x770pixelsCCD was used,resultingin doubling the previous spectral coverage. The reduction of thesespectra was donewith the IRAF softwareinstalled in the LNA workstations. 3.4 DSS Era In 1995 it was possible to make use of the optical disks from the Digitized Sky Survey (DSS). Even though the quality and the type of plates used to produce the DSS vary through the sky, its use allowedbetter homogeneity,completenessand efficiency.A more consistentsamplecould be formed,making possiblethe verification of proposedassocia¬ tions to galaxiesand planetary nebulae.The 4041 Weintraub all-sky IRAS sources prior to condition (iii) of the Early Era were examined on DSS images. The eventual optical objects inside the error ellipseswere classifiedaccordingto their magnitude, appearance and environment. For a more complete exclusion of quasars, active galactic nuclei and planetary nebulae,as they can have stellar-like appearance, the A CatalogueofQuasars & Active Nuclei (Véron-Cetty & Veron, 1993) and The Strasbourg-ESOCatalogue of Galactic Planetary Nebulae(Acker et al. 1992) were used. In this way, 551 sources were eliminated, 388 being galaxiesand 163, planetary nebulae. The number of remaining positions, 3490, was somewhat smaller than that in Weintraub’s Era. Eliminating the sources in the direction of the MagellanicClouds (57) and those associatedto HBC ob¬ jects (147), the other 3286optical imageswere examinedfor stellar objects inside the 3.3<7 error ellipse. In this way, 689 sources were selectedamong objects of stellar appearance with magnitudesup to 14, stellar objects with magnitudesbellow12 “near” to the ellipse and bright diffuse objects. The definition of “near” was somewhatpersonal and took into account the brightnessof the star and the field density. Bright diffuse objects with saturatedimageswere included for the possibility of a hiding star. The environment was alsotakeninto account considering the presence of generalnebulositydiffuse object. The sources south of +30° are 502 (40 of them having bright stars “nearby”) and define the final PDS sample. There are 248positions which were observed(or at leastpointed at) in the previous Eras, although not belongingto the final sample. The DSSuses both visual and blue plates, not allowing consistent magnitude estimates. Worse,the observationsof the PDS being in the red and since differencesin color index may be important, the quality of sample completenessis poor. During the examination of the plates, star magnitudeswere eye- calibrated by comparison with GSC fields. A more refined calibration would be time consumingwithout making the the sampleany better. The selectedDSSplate was the one in which the source was nearest to the center. The higher quality of DSS maps compared to GSCresulted in a better reliability in the 144 Proceedingsof the XXIth SAB Table 1: Results for sources not included in the final sample Type of stars Quantity Visual binaries “Normal” stars 18 TTS 10 2 ProbableTTS 11 PossibleHAEBE 03 Cool Li-rich giant stars 01 Other Horemissionstars 05 identification of the associatedobjects. It was necessary to reobservesome positions in confusedfields and in other ones havingcandidatesfainter than thoseformerly measured. 4 Results a) In all Eras, after each observation,the spectrum was displayed with the acquisition program, and as much as possible,classifiedaccordingto its main characteristics(emis¬ sions,presence of Li line or TiO bands,etc) and a preliminary type was proposed for the object. When an object was suspectedto be a TTS or other “interesting” star, more spectra were obtained. For TTS at least 3 spectra were taken in order to detect possible radial velocity variations. As the project was goingon, new and unexpectedtypes of stars showedup, like giant stars with strong lithium line, despitetheir rarity so far. Others are similar in the observedspectralrange to old fuori stars. During cloudy nights standard stars were observedfor spectralclassification. The chosen spectralregion is not normally used for that purpose and the narrow range limits the numberof usefullines. But anyway it seems possibleto make a spectralclassificationin this region with a reasonableprecision,mainly for advancedspectraltypes. That means that the TTS may be fairly well classified,unlessstrong veilingis present. Frequently emissionlines appear on the sky, mainly Ha, but also some forbidden ones, indicating the presence of an HH-region. The observedoptical object may not haveany relation to the HH-region.Probably in thesecases the infrared emissionis related to the HH-region,whoseexciting source is not visible sinceit is very embedded.Theseregions are potentially very interesting. For this reason during reduction the sky spectrum was preservedand, wheneverpossible, the non-stellaremissionwas measured. Theseresults are not presentedhere. b) A more detailed analysisof the spatial distribution will be done later. A preliminary analysisshowsthat PDS young objects havea lessconcentrateddistribution of distances to the nearest neighborthan those belonging to the HBC (Torres et ah, 1992). Small groups of young stars have been discoveredhaving no apparent connection with known Proceedingsof the XXIthSAB 145 Table 2: Resultsof PDS Type of stars Quantity Visual binaries Faint objects/underexposed 018 TTS 044 6 ProbableTTS 009 2 PossibleHAEBE 110 2 Possibleold fuori 003 1 Shell stars 008 CoolLi-rich giant stars 014 Probablecool Li-rich stars 002 Hot bright stars 026 2 Other Ha emissionstars 028 2 “Normal” stars 240 Samplesize 502 star forming regions,some of them at high galacticlatitude,as the TW Hya group or the triple star HD141569,formed by one HAEBE and two TTS. However,for the goalswhich motivated the PDS, the results are rather frustrating, since few stars were found that can be consideredisolated. If confirmed,isolated young stars would be statistically of little importance. c) UBV(RI)c photometry being made for the PDS stars, using the FOTRAP (Jablonski et ah, 1994)installed at the 60cm Zeisstelescopeat OPD, is revealing some trends for HAEBE stars. With 60%of the possibleHAEBE measured,broadly two conditionscould be discussed:first, stars with low reddeningwhich,whencorrected,showtypical colorsof dwarf stars hotter than F5, consistentwith their spectra; second,stronglyreddenedstars wherecorrection is difficult becausethe standardreddeninglaw cannot be applied and other alternatives,like an UV-excessin a cold photosphere,can not easilybe discarded. Almost alwaysthe spectra present only the Ha line. If a standardreddeninglaw is used, extinctions greater than 10 magnitudesare obtained. They may be strongly embedded stars with high intrinsic luminosity. Infrared observationscouldtest this hypothesis. d)After establishingthe final samplein the DSSEra,it was found that 248sources pointed at in previous Eras were no longer included,but for 111 of them no observableobject was found. Other 30 sources have counterparts in HBC, one belongsto the objects were outsidethe error ellipseand showedno peculiar characteristic. Thus,for only 49 cases we haveoptical objects that actually could be examined.Table 1 summarizesthe results for thesesources. In generalthesestars havemagnitudesabove14 on the DSS,explainingthe largenumber 146 Proceedingsof the XXIthSAB of probable TTS. Actually this happens for all TTS and HAEBE (2 of them appear in TWP) in Table 1. The above-mentionedcolor differencesbetween DSS plates and PDS have their nature defined. Further, young stars can show strong photometric variability and there may be magnitudedifferencesbetweenDSS and the PDS observations. Although some young stars were found outsidethe error ellipselooking like casualdiscov¬ eries,this is not as unpredictableas it seems. Severalsources without apparent optical association may be very faint or embeddedobjects. As stars are generally formed in associationsit is not surprising to find other young objects nearby. e) Table 2 summarizesthe results of the PDS final sample, with 95%of the spectra reduced. One of the difficulties in summarizingthe results is the presence of more than one optical candidate inside the error ellipse. Severaltimes just one of the objects is a presumedassociation. Wheneverpossible,even if all objects were observed,only the “best” candidate is displayedin Many possibleHAEBE are associatedwith cloudsor star forming regions,strengthening their classification. Among the 113 possibleIIAEBE presentedin Tables1 and 2, 18 are classified by TWP as being “HAEBE members and candidatemembers”and 15 more ar 5 Conclusions Although more definite conclusionsmust await the completion of the PDS, some of them may be advanced : i - Youngstars may be divided into two groups: a) distributed along the galacticplane without clustering b) clusteredin star formingregionsaboveor below the galactic plane. ii - Truly isolated young stars are rare. iii - There are many young stars in the periphery of associations,not detectedyet. A goodalternative to explain thesecharacteristicsseems to be Lépine & Du vert’s model (1994)relating stellar formation in the regionsout of the galactic plane to the infall of high velocity clouds. The stars on the galacticplane could be formed accordingto more traditional ideas,as for instance,spiral waves. 6 Acknowledgements The authors acknowledgethe collaboration of the other members of the PDS group: Ramiro de la Reza,JacquesLépineand Jane Gregório-Hetemand the helpful discussions with many other Brazilian astronomers who cannot be mentioned individually. The Proceedingsof the XXI-1'SAB 147 References [l] Acker, A., Marcout, J., Ochsenbein,F., Stenholm,B., Tylenda, R., Schon,C. 1992, Strasbourg-ESO Catalogueof GalacticPlanetary Nebulae, European SouthernObser¬ vatory [2] de la Feza, R., Quast, G.R., 'Porres,C.A.O., Mayor, M., Meylan, G., Llorente de Andrés,F. 1987in New Insights in Astrophysics, Proc. Joint NASA/ESA/SERCCon¬ ference,University College,London, 106 [3] dela Reza,IL, 'Porres,G.A.O., Quast,G.R., Castilho,B.V., Vieira, G.V. 1989,ApJL, 341,L61 [4] Emerson,J.P. 1988, in Formation and Evolution ofLow Mass Stars, Proceedingsof the NATO-ASI, ed. A.K.Dupree k M.T.V.T. Lago, 193 [5] Gregorio-Hefeern,J., Lépine, J.R.D., Quast,G.R., Torres,C.A.O., dela Reza,R. 1992, AJ, 103,549 [6] Gregório-IIetem,J., Sanzovo,G.G., Lépine, J.R.D. 1988, A&ASS, 76, 347 [7] Hartman, L., Kenyon, S. 1988 in Formation and Evolution ofLow Mass Stars, Pro¬ ceedingsof the NATO-ASI, ed. A.K.Dupree k M.T.V.T. Lago, 163 [8] Ilerbig, G.H. 1960, ApJS, 4, 337 [9] Ilerbig, G.IL, Bell, K.R. 1988,Lick Obs. Bull., 1111 [10] Jablonski,F., Baptista, R., Barroso Jr., J., Gneiding, C.D., Rodrigues,F., Campos, R.P. .1984,PASP,106, 1172 [11]Lépine, J.R.D., Duvert, G. 1994, A&A, 286,60 [12]Papadopoulos,C., Scovil, C. 19S0, True. VisualMagnitude PhotographicStar Atlas, PergamonPress [13]Reipurth, B. 1990,I AU SymposiumNo.137 on Flare Stars in Star Cluster, Associ¬ ations and the Solar Vicinity, edsL.V. Mirzoyan, B.R. Peterssenand M.K. Tsvetkov, 229 [14] Thé, P.S., de Winter, D., Pérez,M.R., 1994, A&ASS,104, 315 [15] Torres, C.A.O., Quast, G.R., de la Reza, R., Lépine, J.R.D., Gregório-Hetem,J. 1995. AJ, 109, 2146 [16]Torres, O.A.O., Quast, G.R., de la Reza,R.., Mello, G.F.P. 1987a,Rev. Mex. de Astron. y Astro!., 14, 360 148 Proceedingsof the XXIthSAB [17]Torres,C.A.O., Quast,G.R., de la Reza,R., Mello, G.F.P. 1987b,Rev. Mex. de Astron. y Astrof.,14,361 [18]Veron-Cetty,M.P., Veron,P. 1993,ESOScientificReport No. 13 [19] Weintraub,D.A., 1990,ApJSS,74,575 Proceedingsof the XXIthSAB 149 DYNAMICS IN THE EQUATORIAL PLANE OF Be STARS F.X. de Araújo ObservatórioNacional Abstract Accurate numerical solutionsfor the radial motion equation in the equatorial plane of rotating B-type stars are presented.Wehaveconsideredan outflow driven mainly (or uniquely!) by optically thin lines. This is doneby decreasingthe value of the radiative parameter a in the domain (0.5 > a > 0.0). It was obtained a net shallowingof expansionwith lower valuesof a. On the other hand reasonable mass fluxes require large valuesfor the radiative parameter k. Our results give support to the idea that the dynamicsin the equatorial region of Be star envelopes is dominated by a radiative force due to a great number of thin lines. 1 Introduction Be stars are non supergiantB type stars which present (or havepresented)Balmer lines in emission. Their other major observationalcharacteristics are: a largebroadeningof the photosphericlines; an important infrared excess;the presence of asymmetric absorption lines from ionized speciesin the UV; a degreeof linear polarization which may reach,in some cases,1.0 - 1.5 % in the visible range. The ad-hoc scenariomost acceptednowadaysfor these objects is that of a fast rota¬ tor surrounded by a non-spherical circumstellar envelope. In a first approximation the envelopewould be symmetric with respect to the rotation axis. Accordingto this picture a weak (not dense)and fast wind is present in polar regions - where would be mainly formed the UV asymmetric lines. On the other hand the equatorial plane seems to be much more denseand slowly expanding(but with an important rotational velocity).The emissionlines would originate in this last region. Sucha scenariois generallysummarized by assuminga two-component envelope:a “polar wind” and an “equatorial disc”. In the last decadesome attempts havebeenmadeto understandBe envelopesin the context of the radiation-drivenwind theory. Amongother works the reader is referred to Poe k Friend (1986),Araujo k Freitas Pacheco(1989)and Koninx k Hearn (1992). More recently Bjorkmann k Cassinelli(1993)and Owockiet al. (1994)haveconcluded that a meridionalcurrent may be responsablefor the concentrationof matter towardsthe equator. However,the most important problemwith all thesemodelsis that they produce too strong an equatorial expansion.Not only is the terminal velocity very high (about a thousandkms-1)but also there is too rapid an increaseto thesevelocities.Contrary to it the fitting of H„ line profile requiresterminal velocitiesof about 200kms-1. (e.g.Poeckert k Marlbourough 1978).A detailed comparisonbetweeninterferometric measurements of the star 7 Cas and theoreticalvisibility curves performed by Steeet al (in preparation ) reinforcesthis conclusion. Thoseattempts were basedon Castor,Abbott k Klein (1975,hereafter CAK) theory. The radiative force in “CAK-type” modelsis essentiallydue to strong (opticallythick) 150 Proceedingsof the XXIthSAB lines. It describesan expansionwhich is able to reproducemany properties of hot star winds,andprobably the “polar wind” of Be stars. But, as alreadysaid,it can not explain the properties of the equatorial planeof Be stars. Thepossibility of an outflow driven mainly by thin lines was raisedseveralyears agoby Lamers(1986)in his study of P Cygni’s wind. SubsequentlyPauldrach& Puls(1990)and Lamers & Pauldrach (1991)concludedthat the dominant driving lines may be optically thin linesin winds with largeenoughoptical thickness.Boyd & Marlborough (1991)also suggestedweak lines as a mechanismwhich could possibly act in the equatorial disc of B[e]supergiants. Very recently Araújo et al. (1994; see also Stee& Araújo 1994) have presentedsome results obtained with lower valuesof the radiative parameter a. (This parameter givesthe importance of thick lines in driving the wind: a = 0 correspondsto no thick lines at all whereascv = 1 represents no thin lines). Although they obtained some encouragingresults,technicaldifficulties preventedthem from obtaining solutions for a < 0.4. Chen&; Marlborough (1994)haveencounteredanalogousproblems. The aim of this paper is to present accurate numerical solutions for the expansion equation in the equatorialplane of rotating B-type stars, consideringvaluesof a in the range 0.5 to 0.0. This correspondsto an outflow driven mostly (or uniquely!) by optically thin lines. Whether or not this is the major mechanismacting in the equatorial plane of Be stars remains to be seen. Neverthelessthe solutionsclearly allow a wind with the desiredproperties: low terminal velocitiesand large mass fluxes. In Sect. 2 the dynamical equations are presentedand we discussour method of solution. The specialcase a —0 is treated separatelyin Sect. 3. The method is applied in Sect. 4. We have considered two ensemblesof parameters: one to represent a Be early-type (B0-B1 V-IV) and the other one a more late Be (B4-B5 IV—III ). The results for a standard model are presentedin sub-sections4.1 and 4.2 and some conclusionsare outlined in last section. 2 The expansion equation and it’s solution 2.1 Assumptions and equations Let us consider a sphericalcoordinatesystem in which the centre of the star is at 7- = 0 and the rotational axis coincideswith the 2 axis. The equatorialplaneis at 0 = w/2.By hypothesisour envelopeis in a steadystate situation. With theseassumptionsthe radial motion equation in the equatorialplaneis ()vr 1 dP GM(l-T) _1 Or-r- + F<= 0 (1) or p Or i - P HereP is the gas pressure, G is the gravitational constant, M is the stellar mass and F is the ratio betweenthe radiative accelerationdue to electron scatteringand gravity. Fl is the radiative force due to line opacity. FollowingFriend& Abbott (1986) we may write Proceedingsof the XXIth SAB 151 /•' a, L 0 dvr kl ry(r, ) f2) f> 17rev2 dr The lad, or g(i\ o,., ) Lakesinto account the sizeof the star (g = 1 for a dimensionless point source). I. is an optical depth variable and k and a are the radiative parameters. k is related to the total numberof contributing lines and cv, as said previously, measures the relative importance of thin and thick lines. Other symbolshave their usual meaning: L is star’s luminosity, a, is the electron scattering coefficientand c is the velocity of light. The mass conservation equation can be immediatelyintegrated,leading to (f> = r2pvr (3) where $ is the mass flux per unit of solid angle. Regardingthe rotational velocity we have adopted the following expression: CM ( - = x( i r)y/2ÿy' (4) R Here is the stellar rotation rate (given in terms of ”break-up” speed) and R is the star radius. 6 is an empirical viscosity parameter which may vary from —1 (solid body rotation) to +1 (angular momentum conservation). In addition we have restricted our study to the case of an isothermal envelopeand we have consideredan ideal gas law P = a2p (5) where a is the (isothermal) sound speed. The equation which results for expansion (combining Eqs. (1) to (5)) is ;M b-£|f + • 2n2 C / o dvr\ dvr. + -i(rV-p) *3(r,v, — ) (6) - r rz K clr ' dr C is an eigenvalueof the solution. It is related to the mass flux by YClMk C — (7) 2.2 The method of solution The first step to obtain the numericalsolutionof this type of equation consistsin comput¬ ing the velocity and its derivative at a specialpoint named critical radius. This point is definedby regularity and singularityconditions(seee.g. CAK, Araujo & Freitas Pacheco, 1989). From them we can perform the numerical integration in both directions (inward and outward) obtaining a whole solution. However,it is not straightforward to obtain accurate (and quick) estimates of at eachstep of integration. We havere-written (6) in the form 152 Proceedingsof the XXIthSAB H(x) = /iijlfps + h.2{v,r):va+ h3(r) = 0 (8) with x = As usual,only the r-dependenceof g{r,vr,ÿ) is retainedin a first iteration. In the subsequentones a polynomial fitting of it is employed. The problem consists now in finding the roots of H(x). A simple analysis of this equation showsthat if h\(v) .h2(v,r) > 0 then H(x) is monotonically increasing (or decreasing).On the other handif 1H(V).h2(v,r) < 0 the function has one (andonly one!) point of maximum/minimum.In addition it must be kept in mind that H( 0) = /ÿ(r) and asymptotically H(x ) oc hi(v).x. It is easy to concludethat when hi(v).h3(r) < 0 the function has one (and only one)root. Let us analysenow the case h\(v) .h3(r) > 0. In this case the function H(x) may haveroots if and only if it has a maximum or minimum. When this inflection point is of same sign of hi(v) or h3(r) there are no roots. Whenit is of the opposite sign it has two roots (in this last case one must be careful of trapping the good one: the greater for the outward integration and the smaller for the inward direction). On the basisof the aboveanalysiswe havedevelopeda codewhich determinesnumer¬ ically the derivative of the velocity in eachstep of the integration. Firstly is determined a range in which the function H(x ) changessign. Secondly standard procedures(Press et al, 1992)are employedin order to obtain the root with the desiredprecision. 3 The case a = 0. In this case the line force is uniquely driven by optically thin lines. It has a simple r 2 dependence Fl kaeL (9) P 4ircr2 The expansionequationin the equatorialplaneis reducedto GM(i-r) K-SPr + ,.2 2«2 C (10) + r2 Now,C is given by C = YGMk. So,Eq. (10) may be re-written as 1 dvr h-“2] dr 2a2 GM (ii) r2 [(i- r)[i -x ’(vf1] - *rl This equation strongly resemblesthe solar wind equation (Parker,1958;1960).In this paper we will restrict our solution for a = 0 to the case of Keplerian rotation i.e. 8 = 0.5 (seeEq. 4). Hence one obtains simply Proceedingsof the XXIthSAB 153 1 clvr a2 GM [iv- a2] [(i-r)(i-x2) - *r] (12) v dr ?.2 This is a solar-type wind modified by rotation (y), a continuum radiative force (T), and a thin line force (kT). Similar to solar expansiona transsonic wind must obey v = a at the critical (sonic)point definedby GM rs = - (13) 2«2 [(i-n(i-x') tr] 4 Application to Be stars In order to obtain the solutionsof the equation we must chooseseveralparameters. The stellar parameters are the mass,radius,luminosity (or effectivetemperature)and rotation rate. Concerningthe envelopeitself the eletron temperature must be assumed(sincewe do not considerthe energy equation),also the rotational law and a boundary condition. The following strategy was adopted. Our “standard model” is definedby X = 0.66 Tenv= 0.8 Teff vÿ(r) = vTOt(R/r)'/2(keplerianrotation) and a boundary condition givenby the density at the basisof the wind (r = R) p0 = 1.0 x 10-11gem-3 This last value is suggestedby the work of Waters et al. (1991). The results obtained with thesestandardparameters, whenthe model is applied to stellar parameters typical of Be stars, are shown in next sub-sections4.1 and 4.2. 4.1 A B0-B1 V-IV star In this case it was assumedthat M = 15 MQ, R = 10 R©, L = 3.5 x 104L©(and Teff= 25000 K). These values were adopted following an analysis of several works in the literature: Poe & Friend (1986),Waterset al. (1991),Lamers & Pauldrach(1991), Waters& Marlbourough (1992),Koninx &; Hearn (1992),Chenet al. (1992),Bjorkmann & Cassinelli(1993).They are not addressedto any specificobject but it is supposedthey are representativeof a typical hot Be star. Table 1 summarizesthe main results. Columns1 and 2 give the adopted radiative parameters a and k. Column 3 is the mass flux per unit of solid angle (givenin solar masses per year-steradian)and column 4 is the velocity at 100 stellar radii (givenin kms-1). Vioois not the terminal velocity but it is closeto Vÿ,. In order to justify it we showin column 5 the V100/V300ratio for some models. In almost all cases Vioois greater than 95%of V30o.The only exception is a = 0. In this case we expect that an isothermal wind will expandwithout limit. So, the more important increaseof v(r) beyond Vioois not surprising. When a = 0.5 (i.e. the driving contribution of thin and thick lines is of same importance)Voois about 1200 kms-1. As a is decreasedthere is a continous 154 Proceedingsof lhe XXIth SAP FableI: Results: B0-B1 star, standardmodel a k 4)(4/,.)i/ hsr ') View(kins ') luoo/V300 0.5 0.3 3.G10"9 1184 yy.5% 0.4 0.3 4.710“ 10 853 0.5 1.710 858 0.7 3.S10-9 801 99.4% 0.3 0.3 1.810-“ 618 _ 0.7 2.910 10 624 1.5 3.710-° 630 99.3% 14 0.2 0.3 3.210_ 427 1.5 y.5io-“ 438 3.2 -1.110-9 445 99.1% 0.1 0.3 3.3ir21 217 3.2 5.810-12 265 6.0 3.010-° 272 98.2% 0.05 1.5 6.010-24 159 6.0 6.110-'2 172 8.3 3.810-° 178 97.0% 0.025 8.3 1.710-“ 119 0.0 10.51 3.010-° 68 , decreaseof Vi00,reaching Fio0~ 150 - 250 kins-1 lor cv % 0.05 -0.1. However, when o is small (< 0.2), reasonablemass fluxes require rather large valuesof k. This point will be discussedin the last section. Furthermore, for very small tv there is no solution at all for k smallerthan a. certain value and in the case tv = 0 there is only a.restricted range of k which leads to physical solutions (seeEq. (12)). Figure 1 showsthe expansion velocity profiles we have obtained for several valuesof cv. In the models plotted here the mass flux is in the range (3.0 ± 1.0) 10-9 sr~l . 4.2 A B4-B5 IV-III star In this secondcase we have assumedM = 7.547,.,,R = 6.0/7,.-,,L = 1.65103/,,.,(and Teff= 15000/v). 'File remarks concerningstellar parameters made in the previous sub¬ section are valid as well now. Table 2 gives the results. The notation employedis the same as in 'Fable1. Again, the expansion velocity strongly decreaseswith cv. Fiouof about 200 kins-1 was ob¬ tained for a in the domain 0.05 — 0.1. The model a = 0.5 leads to a mass flux $ = -1 .-1 2.010-11M&y S'l . In order to have solutionswith similar mass fluxes for decreasingcv models we must admit high valueslor k. This requirement is even more strong than in the previous case (Sect. 4.1). For instancethe model cv = 0.1 requiresk = 40.0 to achieve $ = 1.8 10-11M&y~lsr~l. The radial velocity profiles correspondingto suchmodelsare shownin Fig. 2. Proceedingsof the XXIthSAB 155 a = 0.5 0.4 I5 0 5 0.2 0.1 0 05 0 025 0,5 1.5 2 log r/R Figure 1: Expansion velocity (B0-B1 star) , for severalvaluesof a Table 2: Results: B4-B5 star, standardmodel 1 1 a k $ (Mey lsr ) Vioo(kms ) 0.5 0.3 2.010-11 1121 0.4 0.3 7.910-13 80S —12 0.5 2.S10 811 1.0 1.610—11 815 0.3 0.3 4.210-15 585 13 1.0 2.310— 595 3.5 1.510“11 597 0.2 0.3 1.510-19 404 17 1.0 6.010— 410 3.5 3.no-14 416 13.0 2.HO”11 422 0.1 3.5 5.310-22 245 13.0 2.510-16 252 40.0 1.810“11 259 0.05 40.0 3.910-16 162 0.0 110.2 2.no-11 46 156 Proceedingsof the XXIthSAB a = 0.5 04 1 0.3 *ÿ § 0.2 0-1 0-05 0.5 1.5 Figure 2: Expansion velocity (B4-B5 star), for severalvaluesof a 5 The rotational profile vÿr) A consistent treatment of the conservationequationsinvolvesobtaining also a solution for vÿr). However,this demandssome hypothesison viscosity. Theoreticalmodelsoften invokeconservationof specificangularmomentum per unit of mass (Boyd& Marlborough, 1991;see also Chen & Marlborough, 1994). On the other hand Keplerian rotation has been widely employedin ad-hoc scenarios. Our treatment consists (seealso Araújo & Freitas Pacheco,1989; Araújo et al, 1994;Stee& Araújo, 1994)in adopting a viscosity parameter which may simulate different conditions,from solid body rotation to angular momentum conservation. In Sect. 2 it was noted that this parameter (Ú) should vary from -1 to +1. However, we managedto get solutionsfrom our codeonly in the range 8 = 0.0 —1.0. The results we couldobtain with 8 = 0.25;0.5; 0.75and 1.0 are expressedin Table3. Onceagain we haveexpansiondecreasingwith decreasingas. But what must be emphasizednow is that only for 8 = 1/2(Keplerianrotation) are there solutionsfor very small valuesof a (0.05 - 0.025)and we are able to obtain Vjoo~ 100 —200 kms-1! When 8 = 0.25 the model a = 0.1givesVjoo= 542 kms-1while for greater ús(i.e. approachingangularmomentum conservation)it becamesmore and more difficulty to obtain numericalsolutions. This result may possibly indicate that in Be envelopesthe rotational profile must be closeto a Keplerian law. 6 Conclusions Our computationshave shown that decreasingexpansionvelocities are obtained as de¬ creasingvaluesof the radiative parameter a are used. However,it seems that the rota¬ tional field vÿ(r) shouldnot be very far from Keplerianrotation. Proceedingsof the XXIthSAB 157 Table3: Theinfluenceof vÿr) Mr ) a; k V',oo(kms~') QiMoy-'ar-1) or,k VWQ(krns-1) QiMey-'sr-1) (R/r)0'25 0.5;0.3 1327 3.610-9 0.5;0.3 1252 2.010"11 0.4;0.7 1026 3.910-9 0.4;1.0 967 1.610-11 0.3;1.5 820 3.710“9 0.3;3.5 773 1.510“11 0.2;3.2 667 4.410-9 0.2;13.0 628 2.310-11 0.1;6.0 542 3.910-9 0.1;40.0 508 2.410“11 (R/r)05 0.5;0.3 1184 3.610“ 9 0.5;0.3 1121 2.010-11 0.4;0.7 861 3.810-9 0.4;1.0 815 1.610“11 0.3;1.5 630 3.710“9 0.3;3.5 597 1.510"11 0.2;3.2 445 4.110-9 0.2;13.0 422 2.110-11 0.1;6.0 272 3.010-9 0.1;40.0 259 1.810“11 0.05;8.3 178 3.810“9 0.05;40.0 162 3.910“16 0.0;10.54 65 3.510-9 0.0;110.2 46 2.110-11 (R/r)075 0.5;0.3 1165 3.510-9 0.5;0.3 1057 2.010-11 0.4;0.7 765 3.710“9 0.4;1.0 726 1.610"11 (R/r)10 0.5;0.3 1080 3.510-9 0.5;0.3 1024 2.010-11 On the other hand large valuesfor the radiative parameter k are requiredif we want to obtain mass loss rates comparableto observationalones. Let us comment briefly on this point. When a = 0.0 (no optically thick lines at all) k gives the ratio between the radiative accelerationdue to the thin lines and the electronscatteringforce (seeEq. (9)). So,we shouldhave a thin line forceof about ten times stronger than the electron scatteringone for the B0-B1 star. In the secondcase (B4-B5 star) this ratio shouldbe even greater, of about two ordersof magnitude. The abovediscussionallows us to concludethat a radiative force due basically to optically thin linesmay leadto an outflow with the desiredproperties- a shallowexpansion and a large mass flux -, but indeedlargenumber of theselines is required. Is this the main mechanismpresent in the equatorialplane of Be star envelopes?We cannot give a definitive answer to this question now. More sophisticatedmodels shouldbe developed: the radiative force must be calculatedself-consistentlyrather than consideringa and k as ’’free” parameters. Despite the above remark a first sucessfullapplication of such a model is currently being done by Steeet al (in preparation). They haveperformed a detailed reproduction of the H„ line profile and the correspondingvisibility curves of the star 7 Cas. From this work they could infer the acceptablevaluesof a: a —0.05 for the equatorialplane while a = 0.5 for the polar axis of that star. References [1] Araújo, F.X., Freitas Pacheco,J.A. 1989,MNRAS,241,543 [2] Araujo, F.X., Freitas Pacheco,J.A., Petrini, D. 1994,MNRAS,267,501 158 Proceedingsof the XXIthSAB [3] Bjorkman, J.E, Cassinelli,J.P. 1993,Ap.J,409, 429 [4] Boyd, C.J., Marlborough, J.M. 1991, Ap.J,369, J91 [5] Castor,J.I., Abbott, D.C., Klein, R.I. 1975, Ap.J195, 175 (CAK) [6] Chen,II., Marlborough, J. M., 1994,Ap.J,427, 1005 [7] Chen,H., Marlborough, J. M., Waters, L.B.F.M. 1992, Ap.J, 384, 605 [8] Drew, J.E., 1989,ApJS, 71, 267 [9] Friend,D.B., Abbott, D.C. 1986, ApJ, 311,701 [10]Kogure, T., Ilirata, IE, 1984,Bull. Astr. Soc.India, 10, 281 [11]Koninx, J.P.M., Hearn, A.Cl. 1992, ANA, 263, 208 [12]Lamers,H.J.G.L.M. 19S6,ANA, 159, 90 [13] Lamers, H.J.G.L.M., Pauldrach, A.W.A. 1991, ANA, 244,L5 [14]Marlborough, J. M., Peters,G.J. 1986, ApJS, 62, S75 [15] Owocki,S.P.,Cranmer, S.R.,Blondin, J.M. 1994, Ap.J,424, 887 [16]Pauldrach,A.W.A., Puls, J. 1990,ANA, 237, 424 [17] Parker,E.N., 1958, Ap.J,128, 664 [IS] Parker,E.N., 1960, ApJ, 132, 175 [19]Poe,C.H., Friend, D. 1986, Ap.J,311,317 [20] Poeckert,R., Marlborough,J. M. 1978, ApJ 220, 940 [21] Press, W.H., Teukolsky, S.A., Vetterlin, W.T., Flannery, 13.P., 1992, Numerical Recipes,CambridgeUniv. Press,Cambridge [22] Stee,Ph., Araujo, F.X. 1994, ANA, in press [23] Waters,L.B.F.M., Taylor, A.R., van den Ilanvel, F.P.J., Taylor, A.R., Marlborough, J.M., Dougherty, S.M., 1991, ANA, 214, 120 [24] Waters,L.B.F.M., Marlborough, J. M. 1992, ANA, 253, L25 Proceedingsof the XXIUlSAB 159 THE UPPER H-R DIAGRAM Augusto Damineli Neto - IAG/USP Abstract We revise the status of massive stars knowledge. We describethe main families of stars in the upper II-R diagram and their interconnections. The LBV stage is focusedin some detail, as it plays a crucialrole in the mass lossand as a consequence, in the evolutionary tracks. 1 Massive stars in context Many astronomers are not familiar with the upper part of the H-R diagram (UHRD). The blue part of the UHRD is yet imagined as populated by peculiar, extremely rare stars, not atractive to understand the the stellar phenomenonas a whole. This point of view has been drastically changedin recent times. For one hand, the observations revealleda large number of high mass stars in nearby galaxies,specially in starbursts. For another,theoretical simulationsshowedthat they play an important role in the first stages of galaxy’s life, by injecting energy, momentum and processedchemicalelements (He and N) in the interestellarmedium (ISM). Nowdays,modelsandevolutionarytracksof massivestars are able to explain the main featuresof the UHRD. Evolutionary synthesis calculationsfits CIV, SilV and Hell lines of the Wolf-Rayet (WR) galaxies. But the agreement is only qualitative. Somegalaxies,like He‘2-10,are dominatedby massivestars. A population of thousandsof WR stars can be derived from the Hell 4686 emission line strenght in that and in other similar objects. Conti (1994)derived the masses of 9 knots in the nucleus of He‘2-10,containing the WR feature (Hell emission)and showedthat they are compatible with young (10 Myear) globular clusters. This is a very interesting opportunity to study the evolution of theseobjects, that in our neighborhoodare tousancl times older. The look back time for the galaxieswith z=3-4 is suchthat we are seen them in a early stage of evolution, when they were plaintifull of massivestars. Massivestars are of great cosmologicalinterest, in spite of being not completely understood. We refer as massive to a star of M > 8M0, early than B5 on the main sequence (MS) and as supermassive to a star of M > 15M0, O-type on the MS. The data of distant objects axe interpreted by calibrations basedin our Galaxy and in the LargeMagellanicCloud (LMC). At these distances,it is possible to study both the stellar and gaseouscontents of the giant HII regions(GIIIIR), to derive the Initial MassFunction (IMF), Star Fromation Rate (SFR), agesand to model the nebular spectrum. Someof them, as 30 Doradusin the LMC, have a high concentrationof stars, 1.8xlO5df0pc“3.GIIIIRs in starburst galaxiesseems to be at least 10 times larger than 30 Dor. NGC3603,the largest young stellar cluster in our Galaxy, is muchsmallerthan 30 Dor but has an even densercore (Hoffmann et al. 1995). It appears that larger clusters are located near the galactic center (GC), but we need a 160 Proceedingsof the XXIthSAB Log LBV'* L/L0 •+ LBV Eruptions UCar X CoolHypargiont* - 6.5 -II 1 izo'r ____ ». 60 «r RM3 +VarB 0-1° Iget10420 IRASÿ 2 A eo\- VarAX— X—It* -Xy \? pOosX x \ HRCoi fJCepXÿVtCMo--- -9 \ R7I m, 40r HDI60529*- Rsd \ Sun \ _\eSNI987A NM -8 I 5.0 4.8 4.7 46 45 4.4 4 3 4 2 4.1 4.0 39 3.8 37 3.6 3.5 Log T Figure 1: The UHRD (Humphreys,1993) new generationof large telescopes,optimizedfor the near infrared, to get observationsin full details. 2 Main features of the UHRD The main featuresof the UHRD, displayedin fig. 1 are: a) Luminoushot stars are variablein severaltime-scales. b) Masslossrate grows towardshigherluminosity, although the relation is not linear. c) Luminousstars are evolved,the majority of them in the post Helium burning stage. d) There is a well defined upper luminosity limit, abovewhich no star is found. In the blue sideit is known as Humphreys-Davidson’slimit and in the red, as de Jager’slimit. Irregular and low amplitude variability of the a Cygni type is ubiquitousamong hot luminousstars. The amplitudeis AV=0. 1-0.3mag andthe timescale10-50days.In longer timescales:years to tens of years, the LBVs (LuminousBlue Variables)showvariations of the S Doradustype, whoseamplitudeis AV=0.5-2.0mag, but at constant bolometric luminosity. In timescalesof centuriesor millenia the LBVs undergoesgiant bursts, with AV > 2 mag, releasingenergy of the order of that of a supernova in a few decades. Wolf-Rayet and B[e]stars are not intrinsecally variables. There are some other types of stars that seems to berelated to the evolvedsupermassivesones,as some highly luminous OH/IRand A hypergiantsin nearbygalaxies. Stars with L > 10sL0 do showline profile variabilty whenits temperature is high enough:40 000K > T > 20 000K. The amplitude of the line profile variations decreases towardsthe MS (Fullerton1990,Fullerton et al. in preparation).Absorptioncomponents in the blue side of the line profile is an evidencethat a star is undergoingmass loss, specially when the velocity is greatherthan the scape velocity. In many cases,this is Proceedingsof the XXIthSAB 161 7 INSTABILITYREGION 9EFF< 0 / STABLEREGION LEODFORSTARS 9EFF>° DIRECTLYFROMMS / BUTCAN'T ~o / BEREACHED / N / / o s 60Mo /# HD-LIMIT 40M© LEDDFORSTARS STABLEREGION WITHREDUCEDM á 5 1 «EFF>° KFMAXIMUM 4.6 44 42 4.0 38 3.6 log Teff Figure 2: The upper luminosity limit decreasesa) towardslower surfacetemperatures and b) whenthe star evolvesat constant luminosity and loosemass (Lamers1995). corroboratedby the presence of expandingnebulaesurroundingthe stars. In the UHRD, it is a common place to find overabundanceof He and N and underabundanceof 0, as expectedfrom CNO cycled material. The mass loss rate is a crucial parameter to calculateevolutionary tracks. But, in the UHRD, this parameter is up to now, introduced by hand in the models,on the basis of empirical determinations. Typical figures are: 10~7to lO-9M0?/ear-:1for classicalBe stars, 10-5to 10-6MQyear~lfor 0 type, 10-4to 10~5M@year~lfor WRs and 10-3 to lO-5M0?/ear_1for LBVs (Damineliand de Freitas Pacheco1982, Crowter and Willis 1994, Leitherer and Lamers 1994 ). The proximity of the LBVs to the Eddington limit indicates that the low gravitacional force plays an important role in the mass loss. A generalizedformula of the Eddington limit is givenby: kL Y = (1) 4TTGM where Y is the ratio betweenthe radiative and the gravitational forcesand k is the opacity coefficientofthe gas. A star is unstable for Y > 1. Around 40 000K the gas is fully ionized in order that eletronic scattering is the main source of opacity. For lower temperatures, the degree of ionization decreasesand the opacity increases,as it comes mainly from the metallic lines. The maximum opacity occurs around 10 000K. For cooler stars, the atmospheric turbulencealsocontributes to the pressure, reducingthe Eddington limit for a givenmass and luminosity (fig.2 ). In fact, there is an upper luminosity limit also for cool stars, known as de Jager’slimit (Lamers1995). 162 Proceedingsof the XXIthSAB 3 HD 5980 - 12 Z-0.002 0.01 h Car 0.02 \ m - M S \ l R127 10 - S Doi 1 ** \ -9 i 4.8 '4.6 4.4 4.2 4.0 3.8 L°9 Teff Figure 3: Effectofmetalicity on the Eddington limit (Barba et al 1995) . Contrarely to the case of low mass stars, the massiveones evolveat constant luminosity. As they loosemass, the effective gravity becomeslower and lower,in order that a. M < 50MQstar, after becoming a red supergiant, reachesthe Eddington limit and becomes unstable. This seems to be the case for the lower luminosity LBVs. For M > 50M0, however,the star reachesthe Eddington limit before the RSG branch. The evolutionary track bends towards the left, mantaining the star allways in the blue side of the HRD. As the opacity is a function of the chemicalabundance,the Eddington limit is dependent also on the metal enrichment of the star, as can be seen in fig. 3 (Barba et al. 1995). It is important to note that the majoriy of stars in the UHRD are near but not on the Eddington limit. A mass ejection mechanismis neededto put the gas outside the stellar gravity well. The radiative forceis not powerfull enoughto account for the mass lossrates in the LBV and some other phasesin the UHRD. 3 LBV stars LBV is a group of luminous, variable,evolved, generally blue stars, since beforeknown as Huble-Sandagevariablesin M31 and M33 galaxies,S Doraclusvariablesin LMC and P Cygni, rj Car, AG Car, HR Car in our Galaxy,listed in TableI. Spectroscopically,they do not form a homogeneousclass.The degreeof excitation is generallylow, with strong lines of H, Fell, Hel and seldomHell in emission.Their P Cygni profiles showsmoderate to low velocities (100-1000Km/s).In the eruptive phasesthe spectrum can show purely absorptionlines. The most characteristicphotometricvariability of LBVs are the S Doradusincoherent Proceedingsof lhe XXI' h SAB 163 cycles. During the low brightnessphases,called "quiescence”,the star is hotter, bluer and the higher ex ciliatedemission lines are stronger. During the photometric maxima, called"bursts”, the star is redder and the line excitation degreeis lower. The bolometric magnituderemains nearly constant, implying in a reprocessment of the energy from the UV towards the longer wavelengths(Caputo and Viol, ti 1970). The name "burst” is not appropriate to the S Dor cycle, as there is no real variation of the energy release.In the quiescentphase,the surface temperature of the most luminous S Dor reachs30 000 K and that of lessluminous 10 000 K. In the eruptive phase,however,all S Dor sharethe same limit around 8 000K. This is due to the fact that there is no way to the wind to cool below 7 000 K (Davidson 1978). The instability that drives the wind cannot be of photosphericorigin, as the mass involvedin the SDor oscillationis high: 10-2to 10-3M* (Darners1995). LB Vs can show giant eruptions in timescale of centuries to millenia, in which there is a real variation of bolometric luminosity. Examplesof this behavior are the bursts of SN19G1V(that reaching Mbol=-17) that of P Cygni in the XVII century and that of ij Carinae in the past century. These events are acompaniedby hugemass ejection, contrarely to the S Dor cycles,wherethereseems to be no variation of the mass loss rate (Leitherer et al. 1994). The necessityof a phaseof very high mass lossrate is deducedfrom the simple fact that an O type star loosemass at a rate of IQ~6Aíÿyear ~ 1, and at suchrate it is able to eject only around lA/0during their lifetime (1 Myear). This is far below what is neededto expose the CNO cycled material and to transform an O type into a WU type star. There are indications that such phasecorrespondsto that of LBVs, that would be very short, due to the scarcity of this kind of stars. The time spent in this phasemay be evaluated by comparing the number of LBVs with that of their progenitors, the WRs. In the LMC. there are 115 YVRand 6 LBVs, giving a evolutionary time of : N(LBV) tiBV — t\vn = 25 DOthvi/AVi (2) N(\VR) 4 Wolf-Rayet stars WR are stars with strong emission lines of He, N and/orC. The equivalent width of Hell lines may reach 100 A and that of (‘ill up to 1000 A. The line profiles are top Hated, with 1000 Km/s in the base. 'TheIlel lines frequently showP Cygni profiles,indicative of densewinds. The N and C lines are indicative of (!NO processedmaterial. There are two sequences of WR stars, the WNs, dominatedby Nitrogen lines and WCs, dominated by Carbonlines. Theluminosity span the range 1015to 10 for WNs and 10‘17to 105,5 lor WCs (Ilamman et al. 1983, Koesterke and Ilaminan 1995). The WR phaseduration is estimated to be around 180 000 years, including both WN and WC phases.WRs are believedto be the endingproducts of the massive stars evolution, before to explode as SN.In this scenario, the WRs are very usefull to cense the massivestar formation,as they 164 Proceedingsof the XXIthSAB Table 1: Characteristics of known LBVs Name Mbol Tmax(K) Tmin(K) M (M&year Galaxy Eta Car -11.6 30000 10-3 Milk Way PCyg -9.9 19000 2xl0"5 AG Car -11.0 30000 9000 3xl0-5 HD160529 -8.9 11000 8000 HR Car -8.9 14000 S Dor -9.8 25000 8000 5xl0"5 LMC R127 -10.5 30000 8500 5xl0"5 R71 -8.8 13500 9000 6xl0-5 R143 -10.0 20000 8500 R110 -8.9 10000 7500 3xl0~6 R40 -9.0 < 10000 8500 8xl0-6 SMC HD5980 -12.8 < 35000 W AE And -10.0 30000 2xl0-5 M31 AF And -11.4 < 30000 M31 Var B -10.2 9000 lxlO-5 M33 Var C -9.8 25000 8000 5xl0-5 M33 Other LBVs- Milk Way: He3-519, Wra 751, G79.29+0.46, g25.5+02, LMC:HDE269582,R84, R99, S61, S119,- M31: Var A-l, Var 15- M33: Var 2, Var 83, Romano’sstar - NGC2403: V12, V22, V35, V37, V38- M81: II, 12,13- M101: VI, V2, V10-NGC1058:SN 1961V. Proceedingsof the XXIthSAB 165 are much easierto detect than their progenitors,the 0 type stars and their descendants, the SNs. They enablethe determinationof the slopeof the IMF and of the Mu The number ratio WC/WNis indicative of the chemicalenrichmentof the ISM (Conti 1994). Highermetalicity produceshigher opacity,resultingin stronger winds and increasingthe range os masses to whicha star endsit evolution as a WC. This is in accordto the fact that WC/WNis much lower in SMC and LMC than in the Milk Way. In our Galaxy, there is alsoa radial gradient fo this ratio. WC/WN=0.74in external zones to the solarorbit and WC/WN=0.95towardsthe galacticinner regions(Conti and Vacca1990).Thesefigures, howeverare not definitive,as the incompletenessof the sampleis of the order of 50%in the solarneighborhood,due to the fact that Av=7 (van denBergh 1991)for WR stars for the D < 3Mpc. Moreover,the 0 type stars ard born in the inner core of the GMCs, and take 1 million years to escape from it, what is of the same order of its evolutionary time. The majority of them explode as SNbeforeleavingthe most obscuredzones of the MC. It is possiblethat the sampleof WR stars we know,are not representativeof the massivestars, but only of the high speedmembers. 5 Evolutionary tracks in the UHRD The most recent modelsof supermassivestars agree qualitatively with the observations (Maederand Meynet 1987,Maeder1991).Oneof the most striking featureis the predic¬ tion that a star with M > 50M0never reachesthe RSG branch. However,for 0 stars, the predicted luminosity and He contents are higherthan observed,for Of stars, the pre¬ dicted mass loss rates is lower than observedand the LBV phaseis not accountedfor. The theory of radiation driven windsis able to explain the accelerationof the wind,but not to start it. The inclusionof Iglesiaset al. (1992)opacitiesandof the violent coupling modesof oscillation of Kiriakidis et al. (1993),led to surprisinglybetter evolutionary models (Langer et al. 1994). This kind of instability is basedin the annalisis of the stellar structure stability. The instability is originated near the Fe recombinationzone and affectsall massivestars. The prediction of the strange modepulsationsmatchesvery well the observations(fig. 4). The strange, or violent mode pulsational theory predicts alsoinstability of He cores,in accordwith the observationsthat showa lowerlimit to the luminosity of WC stars. This mechanismis also attractive due to the fact that it occurs in deeplayers, as requiredby the S Dor oscillations. Fig. 5 showsthe mass loss rates of a 60M©star caculateclby Langeret al. (1994),appropriate for P Cygni. Fig. 6 showsthe evolutionary track calculatedby Langer et al. (1994),for P Cygni. The main conclusionsof the Langer et al. model are: a) H-rich WNL are in the phaseof central core H-burning and their mass lossrates are explainedby the violent couplingmode instability. b) The high mass loss rate on the MS result in a lower ending mass in the WR phase, explainingthe observedlimit of luminosity of the H-free WN stars. 166 Proceedingsof the XXIthSAI3 80 Mo 6.0 o • • O o s 60M£ -J 5.5 o « O) 40MX o o o 30 M0 5.0 °B ZAMS 20 Mo 4.5 I 50 40 30 20 Teff/kK Figure 4: shadedzones: phasesin which instabilities are predictedby Kiriakidis et al. (1993)model. Circlesandsquares: measuredamplitudeofline profile variability byFuller¬ ton (1990,1991)and Fullerton et al. (in preparation). I _ 10 \>» ss i 10 ~ 5 \ 0 Normal superface composition WNL-troll WC , Figure 5: Mass loss rate for a star of6OM0(Langeret al. 1994)- The peakin the LBV phasereachesbxlO~3Moyear-1. The total mass loosedin the LBV phaseis 6M,.,. Proceedingsof the XXIthSAB 167 T.ff/kK 200 100 50 30 10 3 6 PCygni N i _i / QQa O 5 f jti °Vo° ; Pre-WR - "WNL" - LBV "WNE" ZAMS - WC 4 4.4 - 3.6 5.2 4.8 4.0 log( Teff/K) Figure 6: Evolutionary I rack for a (i()A/. star by I anyrr ft oi (I !)!)f j. Trianglesreferto mesuredWNLs, circles to faintlined I VNE,squares to strong lined WNE, diamondsto WC.Filled dotsrepresentpositive deleccionofH and unfilled, non detection ofIT. c) LBVs are lie-rich, in accordwith the observations,and are near their Eddington limit, due to the violent pulsations. d) Jnthe phaseof core He burning, after the LBV phase,the star pass through a second WNL phase,but now, H-poor. The detailed sequenceof evolution of a 60M0stars is, by this modelis: 0> II- richWN > LBV > II - poorWN > II - freeWN> WC > SN The generalschemepresentedpreviouslyby Maeder (1991)is: 0 > RSG> SN 35A/0> M > 15M0 O > BSG > YSG > RSG > WR > SN 5OA/0> M > 35A/0 O > Of > BSG > LBV >WR> SN M > 5OA/0 The recent discoverof Barbaet al. (1995), showingthat the spectrum of HD 5980 changedfrom WN3.5+WN4.5to that of an LBV, increasingby A V=2.3 mag, indicates that the evolution is not allwaysin the same sense. The star seems to perform incursions back and fort betweenthe LBV and WR stages. Only long term observationalprograms of LBVs and WRs can providethe lackingof informations about the mechanisminvolved in this kind of behavior. We do not know if the giant eruptions are connectedto the S Dor cycles,that have much smaller timescales:5.5 years for ?; Car (Damineli 1996)and 168 Proceedingsof the XXIth SAB up to 40 years for other stars. We do not expect to pick up a giant eruption by surveying a single star. A large sample,say 50 S Dor stars, including the Galactic Center and other galaxiesup to the Virgo cluster,has a great probabilty of recordingone great burst in a 10 year survey. Aknowledgements:to FAPESP and CNPq for finnancial support. References [1] Barbá,R.H., Niemela,V.S.,Baume,G., Vazquez,R.A.: 1995(preprint). [2] Caputo,F. & Viotti,R.: 1970 A&A 7, 266. [3] Conti, P.S.: 1994 Sp. Sc.Rev. 66, 37. [4] Conti,P.S. & Vacca,W.D.: 1990 AJ 100, 431. [5] Crowter,P.A. & Willis, A.J.: 1994,Sp. Sc.Rev. 66,85. [6] Damineli,A.: 1996 ApJL 460,L49. [7] Damineli,A. & deFreitas-Pacheco,J.A.: 1982MNRAS 198, 659. [8] Fullerton,A.W.: 1990Ph. D. ThesisUniv. Toronto [9] Hammann,W-R., Koesterke,L.,& Wesselowsky,U.: 1993A&A 274,397. [10]Hoffmann,K-H., Saeggewiss,W. & Weigelt,G.: 1995 (preprint) [11]Iglesias,C.A., Rogers,F.J. & Wilson,B.G.: 1992 ApJ 397,717. [12]Kiriakidis, M., Fricke,K.J. & Glatzel W.: 1993 MNRAS 264, 50. [13]Koesterke,L. & Hammann,W-R.: 1995 (preprint). [14]Lamers,H.J.G.L.M.: 1995,IAU Coll 155 (emimpressão). [15]Langer,N.,Hammann,W-R., Lennon,M., Najarro, F., Pauldrach,A.W.A. & Puls, J.: 1994 A&A 290, 819. [16] van den Bergh, S.: 1991 Phys. Rev. 204, 385. Section 6 Solar Systemand Astrometry Proceedingsof the XXIth SAB 171 NONGRAVITATION AL FORCES IN PRESENT AND PAST SOLAR SYSTEM Rodney S. Gomes Observatório Nacional Abstract The association of nongravitational forceswith resonant dynamics is a rich source of dynamical behavior. Some observed Solar System featuresare explained by this coupling phenomenon, like the Earth resonant dust ring, due to the lock of dust particles affected by Poynting-Robertson drag in external resonances with the Earth. Jupiter 1:1 resonant arcs are also supposed to exist by a similar process. Yarkovsky effectmay be responsible to the delivery of asteroid fragments to the Earth’s neighborhood, afterthese bodies goes through resonant regions with Jupiter. The evolution of the early Solar System must have been remarkably affectedby mutual interaction of gas drag forcesand resonances with protoplanets. This interaction must have had an important influencein the formation of Solar System bodies,including planets and asteroids. I. Introduction We aim at pointing out the important dynamical aspectsof a trapping into a mean motion resonance when a nonconservativeforce is present. We divide this work into the following three sections. In section 2, we study Yarkovsky force and the consequences of its effect at mean motion resonances with Jupiter for a lm diameter body, like an asteroid fragment. In section 3, the formation of dust clouds in the Solar System is presented, as a consequence of the interaction of Poynting-Robertson drag with resonant dynamics. Finally, in chapter 4, we show some relevant aspects of the action of gas drag in a primordial solar nebula associatedwith the gravitational effects of a protoplanet. II. Yarkovsky effect Supposea rotating spherical lm diameter body with its rotational axis normal to its orbital plane. Its rotational status implies in a hotter afternoon hemisphere,where from solar radiation will be mostly reradiatecl. This in turn yields a radiation force with a transverse component either in the direction of the body’s velocity or opposite to it, dependingon its rotation sense. In this case we may have either a dissipative or an antidissipative force acting on the body. The acceleration causedby this force at 1 AU onto that lm diameter body, supposingthat it has an uniform density equal to 2.5g /cm3, is 0.65 x 10-lo,m/.s2(Afonso et al, 1995). This is a small force,but its effect may be important as it is nonconservative. The secular variation imposed on 172 Proceedingsof the XXIth SAB 30 ’T-1" 0.8 g20 3 0.6 I y o.4 K 0.2 oUH w»iliÉÍ>li.: 11*0 5x10° 107 1.5x10’ O 5x10® 107 1.5X107 3 1 fe 2.9 I % 2.8 2.7 O 5x10® 107 1.5xl07 TIME (years) Figure 1: Evolution of orbital elements for a body affected by the Yarkovsky effect in the gravitational field of the Sun,Jupiter and Saturn. the semimajor axis and eccentricity by the Yarkovsky effect up to secondorder in the eccentricities is given by (Gomes,1995): K 55 s‘/° = -ÿ2 + I\ 47 sÿ = -ÿ<2+¥ei) where a is the body’s orbit semimajor axis, e its eccentricity, is a constant, and K is another constant related to the absolute value of Yarkovsky force,being positive or negative dependingon its dissipative or antidissipative character. Now suppose the body is in the gravitational field of Solar System planets. As its semimajor axis decaysor increasesthrough the above expressions,it will reach many resonant regions with the planets. In particular let us suppose that the body starts at 3AU in the gravitational field of the Sun and the perturbers Jupiter and Saturn,for the dissipative case. Figure 1 showsthe evolution of 3 orbital elements,the semimajor axis, the eccentricity and the orbital inclination. At first, the semimajor axis starts decaying expected from the Yarkovsky dissipative effect,but it eventually gets trapped in the 5:2 resonance with Jupiter. From this time, the semimajor axis stops its secular decay, although it gains a high amplitude variation around the resonant semimajor axis. The Proceedingsof the XXP'1SAB 173 2.0 1 Mars I H - o.s Oxl O® time (years) lO’ X.Ox 10T Figure 2: The same orbit as in figure 1, now showingthe pericentric distance and the average distance of the inner planets to the Sun. 2.5 II 2 I-I Venus o.s 6X10* timelO’ (years)I.BKIO' 2x1 O* 2.6X10’ Figure 3: Pericentric distance or a body starting at 2.3 AU, with an antidis- sipative Yarkovsky effect. After capture into the 3:1 resonance with Jupiter, the body may get inside the inner planets’ orbits. eccentricity starts a somewhatchaotic variation after trapping, but in averageit takes high valuesuntil the body gets into a hyperbolic orbit. The effect of this resonance on the pericentric distanceis shownin figure 2. From this figure we notice that the body’s orbit is Mars crossing, Earth crossing and even Venus crossing for a relatively long time during its trapped status. This phenomenonmay cause the delivery of asteroids fragmentsinto the Earth’s atmosphere(Afonsoet al, 1995). 174 Proceedingsof the XXIth SAB .2 1.15 £ 5:6 ext. res. wit hi thie Earth .!# I M 1 .05 lO* 2x10* 3x10* £ 0.15 ~3 o.i Ia> 0.05 i' IO* time (year's)2x10* 3x10* Figure 4: Evolution of a dust particle affected by Poynting-Robertsondrag, showinga trapping in the 5:6 resonance with the Earth. The capture into the 5:2 resonance with Jupiter shown in figure 1 doesnot nec¬ essarily occurs. Other examplesof numerical integrations would show a very unstable trapping in the 4:1 resonance with Jupiter at ~ 2AU which would place the body into a hyperbolic orbit in a relatively short time. Other exampleswould showa sharpincrease of the eccentricity through the secularresonance alsoat about 2AU, giving rise to inner planet’s crossingorbits for a long time. A final exampleis shownin figure 3. The peri¬ centric distanceis plotted against time for a body starting at 2.3AU submitted to an antidissipative Yarkovsky effect. In this case, the body is trapped in the 3:1 resonance with Jupiter and its eccentricity increasesuntil it takes a hyperbolic orbit. For a long time the body’s orbit will be Mars crossing,Earth crossingand for a shorter time, but still non negligible, it will be Venus crossing. Ill - Poynting-Robertson drag The effect of radiation pressure on a. micron sizedparticle induces an important dissipative effect onto the particle (Burns et al, 1979), the so calledPoynting-Robertson drag. The secular variations in the semimajor axis and the eccentricity due to this dissipative effect are given by: C 2 + 3e2 Sa= ~ ° (1 + e2)t 57= ,e - 2 a2y/1 - e2 Proceedingsof the XXIth SAB 175 10 i I S 0 10 Figure 5: Numerical simulation showing dust arcs around Jupiter’s orbit formed by dust particles coming from Trojan asteroids. The decayin the semimajor axis may be couterbalancedby resonant effectscoming from perturbing planets thus giving rise to resonance trappings. For this case of fast dissipative evolution (as compared to Yarkovsky effect), only external resonances are important. Here they are associatedto converging orbits and trapping is a rather probable event (Henrard, 1982). Figure 4 showsa typical example of a trapping into the 5:6 resonance with the Earth. After trapping, the eccentricity will tend to increase to an equilibrium value (Gomes, 1995) if there is no close approach to the perturber that leads to resonance rupture, like this example. Trapping induced by Poynting- Robertson drag has always an unstable character,even when no close approach takes place (Gomes, 1995). This means that a particle will not stay ’for ever’ in a resonance lock, just consideringdynamical effects(of course it could be ejected from resonance by pure collision with another particle). Even so, the temporary lock of dust particles in external resonances with the Earth causes a densernumber of particles just around the Earth’s orbit (this must happen also with Venus and Mars, in a lesseffective way). This dust resonant ring, which was first proposedby Jacksonand Zook (1989),has now observational evidence (Dermott et aí, 1994). Formation of dust rings may also occur at the 1:1 resonance with Jupiter (Liou and Zook, 1994). In this case particles may already start at the very resonant region. Thesewouldbe particles releasedfrom the Trojan asteroidsby continuous comminution. Figure 5 showsthe results of a numerical simulation, wheredust particles are released from the Trojan asteroids and their rectangular planar coordinates are plotted with respect to a referential frame rotating with Jupiter. This is the case with j3 = 0.012, where /3is the ratio of the radiation pressure force to the gravitational force. For larger 176 Proceedingsof the XXIth SAB 10 m ’SSSR 10 Figure 6: Numerical simulation showingdust arcs around jupiter’s orbit formed by dust particles coming from the Eos asteroid family. /?(smallerparticles), the radiation pressure on the just releasedparticle will make it assume a larger orbit as it ’sees’the Sun with a smaller mass. In this case the particle may start out of the resonance region and posterior trapping is very difficult, so in respect with Trojan particles, the bigger ones are responsible for the formation of a possiblecloud. On the other hand,due to the same radiation pressure1 effect,particles releasedin the main belt can be instantly placedin the Trojan region. This now will be possible for small particles (high f3), as the gain in semimajor axis is higher for theseparticles. Figure 6 showsthe possibleformation of a dust ring in the (displaced) 1:1 resonant region with Jupiter, coming from particles that would be releasedfrom asteroidsbelonging to the Eos family. In this case we take /3= 0.25. Although here the dust ring doesnot look so evident as that of figure 5, this is a more real case in the sense that dust is really formed in the main belt, whereasit is not sure that a significant amount of dust can be produced by the Trojan asteroids. In any case, there is not observational evidence of this Trojan ring. Particles are cold to be observedfrom the Earth, and surveyors sent to the outer Solar System always come near Jupiter and so not near the Trojan region. As a final comment, another promising candidate to have a dust ring around its orbit is Neptune. In this case dust would come from the Kuiper belt objects that seem to be numerous. The weak effect of Poynting-Robertson drag would also facilitate resonance trapping into external resonances with Neptune. Proceedings of the XXIth SAB 177 IV - Gas drag Gas drag acted in the primordial SolarSystemnebula. The general expressionfor gas drag can be expressedas: F = -CVÿV where fj, = 0 or /.t= 1 v Vga8= 0.995vkev ?—i—r 0.15 IT o.i i s 0.05 O O 10® 2x10® 3x10® 4x10® 5x10° 11 1 1° •I* 9 & 8 o 10® 2x10° 3x10® 4x10® 5x10® TIME (years) Figure 7: Trapping into the 1:2 resonance with Jupiter, for a body subject to gas drag, with /.<= 1 The case /.i= 0 is related to a lessdensegas and smallerbodies,the opposite valid for the case /i = 1. The interactive dynamics of both forms of drag with resonance are similar anyway. Figure 7 showsan exampleof resonance trapping for a body subject to gas drag (case/J,= 1), where Jupiter is the only perturber. Two interesting features are noticed here, differently from the Poynting-Robertsondrag case. First the eccentricity stabilizesat a low value, thus preventing close approachwith the perturber. Second, the semimajor axis oscillates with a constant amplitude, thus preventing resonance 178 Proceedingsof the XXIth SAB rupture. Thus gas drag differs essentially from Poynting-Robertson drag (also from Yarkovskyeffect and solar wind; see Gomes,1995,for details) in the sense that trapping is stable under a dynamical point of view. Resonance rupture will be causedsolely by intercollisions of planetesimals(Malhotra,1993). This stable aspect of resonance trapping inducedby gas drag must havehad an important contribution to Solar System formation. The planetsthemselvesmay havebeenformedin near resonance regionsfrom one or two primordial ones (Patterson,1987;Beaugéand Ferraz-Mello,1994). Gasdrag may also have contributed to an early formation of the Trojan asteroids(Peale,1993). IV - Conclusions The interaction of nongravitational forceswith resonant gravitational forcesinduces important phenomenain present and past Solar System. Due to Yarkovsky effect aster¬ oids fragments may be trapped in resonances with Jupiter or pass by secularresonances induced by the pair Jupiter-Saturn. Oneof the effectsof theseresonances is the increase of orbital eccentricity, which can induce Earth- crossing orbits and possibledelivery of these bodies into the Earth’s atmosphere. For the case of Poynting-Robertson drag, resonance trappings of dust particles into exterior resonances with the inner planets are responsiblefor the formation of resonant rings around the inner planets orbits (the Earth’s ring has been observationally proved to exist). Also a 1:1 resonant dust ring around Jupiter’s orbit may exist. Finally, for the early Solar System, the association of gas drag forces with resonant gravitational effectscoming from a protoplanet must have induced stable resonant locks of planetesimals, stability only broken by mutual colli¬ sions. Many aspects of today’s Solar System may be a consequence of this early Solar System phenomenon,including the formation of trans-Jovian planets and the Trojan asteroids. References Afonso,G.B., Gomes,R.S. and Florkzac,M.A1995. Asteroid fragmentsin Earth-crossing orbits. Planetary and Space Science 43:6 787-795. Beaugé,C., and S.Ferraz-Mello 1994. Resonancecapture and the formation of the outer planets. Mon. Not. R. Astron. Soc. 270, 21-34. Burns,J.A., Lamy,P.L., Soter,S.1979. Radiation Forces on Small Particles in the Solar System. Icarus 40 1-48. Dermott,S.F.,Jayaraman,S., Xu,Y.L., Gustafson,B. A.S.,Liou,J. C. 1994. A circumsolar ring of asteroidal dust in resonant lock with the Earth. Nature 369 719-723. Gomes,R.S. 1995. The effect of nonconservative forces on resonance lock: stability and instability. Icarus 115 47-59. Henrard,J.1982. Captureinto resonance: an extension of the use of adiabatic invariants. Celestial Mechanics 27, 3-22. Proceedingsof the XXI1h SAB 179 Jackson,A.A. and H. A.Zook 1989. A Solar System dust ring with the Earth as its shepherd. Nature 337, 629-631. Liou,J.C. and Zook,H-A. and Jackson A.A. 1995. Radiation pressure, Poynting- Robertson drag, and solar wind drag in the restricted three-body problem. Icarus 116, 186-201. Malhotra.,R. 1993. Orbital resonances in the Solar Nebula: strengthsand weaknesses. Icarus 106, 264-273. Patterson,C.W. 1987. Resonancecapture and the evolution of the planets. Icarus 70, 319-333. Peale,S.J.1993. The effect of the Nebula on the Trojan precursors. Icarus 106, 308-322. Proceedingsof the XXIth SAB 181 VOYAGERS IMAGES REVEALING THE STRUCTURES OF THE PLANETARY RINGS S. Giuliatti Winter Grupo de Dinâmica Orbital e Planetologia Campus de Guaratinguetá-U NESP Abstract The great tour of the Voyagersspacecraftreveleacla large amount of information about the ring systems which encircle the four giants planets: Jupiter, Saturn,Uranus and Neptune. This paper describesthe relevant structures discoveredby the Voyager images. 1. INTRODUCTION The rings of Saturn were first observed by Galileo in 1610 as an unusual phe¬ nomenon around the planet. Huygens in 1656 found the solution to this puzzle de¬ scribing the ring as a thick, solid structure. In spite of many attempts in describing the ring as composedof small particles only in 1859 Maxwell proved that the stability of the ring required it to be formed by a large number of small moonlets. Later, this hypothesis was confirmed by Keeler’s and Campbell’s spectroscopic observations(Van Helclen,1984). Andrew Common obtained the first photograph of Saturn in 1883. Since then a great expansion of observational capability in photometry, polarimetry and especially with the advent of spacecraft, revolutionized the field of planetary ring science. Saturn presents the most complex planetary ring system in the Solar System. The main ring system is composedof two bright rings, A and B, the crepe C ring and the D ring. In 1966,during the ring plane crossing,the E ring was detectedin photographs by Feibelman (1967). It is a faint ring which encompasses the orbits of the satellites Mimas, Enceladus,Tethys, and Dione. It has a brigth peak in the Enceladusorbit suggestingthat this satellite could be the source for the ring particles. The F ring was discoveredby Pioneer 11 in 1979. It is a narrow ring located about ~ 4000 km from the main ring system. The nine “classical” rings of Uranus were detectedby Elliot et al. in 1977during a star occultation through the Kuiper Airborne Observatory’s 91-cm telescope. The rings were named in order of increasingdistancefrom Uranus: 6, 5, 4, a , /3,rj, 7, 6 and e. The rings are very narrow varying between 2 and 12 km in width. There are two exceptions: the 77ring (W = 55 km) and the e ring (W = 20 to 96 km). The e ring shows a linear relation betweenits width and orbital radius,it has a width of 20 km at pericentre and 96 km at apocentre of its orbit. Nicholson et a1. 1978 analysedthe results obtained on April 11, 1978 observationsand modeled the ring as a keplerian ellipsewith different outer and inner eccentricities. Althought this model is in agreeement with the observations,differential apsiclalprecessiondueto the oblatenessof 182 Proceedings of the XXI' h SAB Uranus would changethis configuration. Derrnott and Murray (1980) have shown how self-gravity, collisions and differential precessioncan act together in order to stabilize the e ring. The author is analysing this model by numerically simulating the ring particles using the tree codemethod developedby Richardson(1994). After the discovery of Uranus rings many attempts have been madein order to detect rings around Neptune. During an occupation on July 22, 1984 observersled by A. Brahic and W. Hubbard found some evidenceof materialaround Neptune. However, this material do not encircle the planet, implying the existence of •‘arcs” around Nep¬ tune. Becauseof theseobservations,Voyager2 was reprogramedto collect more data about this recent discovery. 2. THE STRUCTURE OF THE RINGS 2.1 Voyagers Cameras All the information describedbelowhasbeen extracted from the VoyagerImaging and ProcessingUserguide.The Voyagerspacecraftare an advancedthree-axis stabilized ‘Mariner’ classof spacecraftwhich have been designedfor the exploration of the outer planets. They carried equipment for a total of eleven scientific investigations, including imaging science. The cameras and the instruments necessary for imaging scienceare mounted on the scan platform, which incorporates sensors for the imaging, infra-red radiation (IRIS), UVS and PPS experiments. The scan platform can rotate about two angles,the azimuth and elevation angles,and the sensors are all ‘boresightecl’in order to point in a common direction. There are three on board computers, inter-connected,which control the orienta¬ tion, instrument operation and scan platform movement of the spacecraft. Oneof these computers, the Flight Data System (FDS) is the one that supports the imagingscience. It is responsiblefor the control of the Imaging ScienceSubsystem(ISS) throughout the mission. Theseinstruments consist of: • a narrow angle(NA) camera with a focal length of 1500 mm and a square field of view of 7.4 mraclsor 0°.42. • a wide angle(WA) camera with a focal length of 200 mm and a square field of view of 56 mracls or 3°.21. The optical axes of the two cameras are ‘boresighted’ to be concentric and point in the same direction. Therefore,the image taken by the NA camera is contained in the centre of the image taken by the WA camera. All imageswere taken with the clear filter. The duration of the exposure time is controlledby the shutter assemblycontrols, giving durations normally between 0.05 to 15 sec. The exposure time of each image dependson the feature that has to be analysed.The image formed by the camera, lens is exposedonto the face plate called the vidicon tube. The output of each vidicon is a. square array of picture elements (pixels), with 800 samplesby 800 lines in a raw image. The light measuredby the vidicon’s photosensitivesurface within eachpixel is assigned a data number (DN) from 0 to 255. Proceedingsof the XXIth SAB 183 Theseimagescan either be transmitted directly to Earth in real time, or they can be stored in digital form on the on-board spacecrafttape recorder to be transmitted in the future. I •' * , V \ Figure 1. This Voyager2 image (FDS 4410750)showsthe main ring system of Saturn and the narrow F ring. 2.2 New Rings Besides giving a detailed view of all the ring systems (for example, Figure 1) , Voyagers1 and 2 also discoverednew rings. The G ring of Saturn is a faint ring visible in two Voyagerimages.It lies between the F and E rings. There was some evidenceabout this ring from the Pioneer 11 instruments before Voyagersencounter with Saturn. The G ring is mainly composedof G 7 dust (cr = 10“ —10~ ) and has a width of 7000km. Pioneer 11 instruments detected some signals which can be larger bodies,maybe the parent bodies of the G ring, of about 0.1-1 km. It has been proposed that the G ring could be the remanescent of a disrupted satellite. Voyager 2 confirmed the existence of the nine rings of Uranus (Figure 2) and detected two new rings: the A ring orbiting betweem the e and 6 rings and the 1986U2Rring orbiting interior to the 6 ring and extending to ~ 3000 km in the di¬ rection of the planet. Due to the results from the radio and imaging observationsthe nine “classical” rings (Figure 2) are mainly composedof large particles of the order of meters. In 1989 Voyager 2 revealedthat Neptune has a ring system (Figure 3) composed of three rings named from the planet increasing distance: the Galle, Le Verrier and 184 Proceedingsof the XXIth SAB •/ •/ / / Á i Figure 2. The nine classicalrings of Uranus can be seen in this image (FDS 2681416)takenin back scatteredlight. 1 I i Figure 3. This image was taken by Voyager2 (FDS 1141251)and the ring system of Neptune. Adams rings. The four arcs, Courage, Liberte, Egalité and Fraternité, are the dens¬ est parts of the Adams ring. All these rings are brigther in forward-scattered ligth indicating a substantial population composedof dust. The ring system of Jupiter was discoveredby Voyager 1 in 1979 and additional Proceedingsof the XXIth SAB 185 mm Figure 4. A Voyager2 image (FDS 2069302)of the ring system of Jupiter. images were obtained by Voyager 2. After analysing a sampleof 25 Voyager 2 images Showalter (1985) concluded that the ring system is composedof three components (Figure 4): - interior halo: W = 30000km and vertical extension of 10000km aboveand below to the equatorial plane of the ring; - main ring: W —7000km, encircling the orbits of Adrasteaand Metis; - gossamer ring: W = 81000km, near to Thebe’s orbit. The particles present in Jupiter ring system are brigther in forward-scatteredlight (a ~10“ 6 ) implying a population of dust. Lorentz resonance can explain the structure presentedin the Jupiter ring system. A study by Schafferand Burns (1987) showed that there are three locations of this kind of resonance: the limit betweenthe halo and the main ring is in a 3:2 resonance,and the inner edgeof the halo and the outer edge of the gossamerring are in 2:1 and 2:3 Lorentz resonances,respectively. 2.3 Resonance The resonant interaction betweenthe ring particles and the satellite givesrise to density waves on the ring material. For the case of a satellite on an inclined orbit its resonant perturbation on the ring material will create a bending wave (Shu1984). There are density waves on the A and B rings causedby Janus,Prometheus and Pandora. Mimas causes density and bending waves on the A ring (Figure 5). The shepherdingmechanismproposedby Goldreich and Tremaine (1979),to avoid the spreadingof a narrow ring due to dissipativesforces,was confirmedby the discovered of two satellites,Cordeliaand Ophelia,by Voyager2 encounter with Uranus. The narrow e ring is shepherdedby thesetwo satellites. As studied by Porco and Goldreich (1987) Cordelia is in a 24:25Linclblad resonance with the inner edgeof the ring and Ophelia 186 Proceedingsof the XXIth SAB m 1'V Figure 5. This imahe (FDS 4399350)showshe bendingand the density waves on the A ring due to the resonant perturbations of the satellite Mimas. in a 14:15 Lindblad resonance with the outer edgeof the ring. Galatea, the satellite discoveredby Voyager2, is the responsiblefor the radial and longitudinal stability of the arcs of Neptune. Galatea is about 900 km interior to the Adams ring and it is in a 42:43 Corotation resonance with the ring particles. This resonance induce the formation of sites of equilibrium with a maximum extension of 4° and a radial width of 0.6 km (Porco,1991). These values are in good agreement with the observedfrom the Voyager2 images. 2.4 New Satellites Waveson the rings are the result of the satellite perturbation on the particles. A techniqueto find new satellitesis to analysewaves on the ring. Voyager1 and 2 imagesof the EnckeDivision showedan azimuthal wave pattern on either or both inner and outer edges(Cuzzi and Scargle,1985). The satellite Pan, responsiblefor the formation of the waves on the ring, was detectedin some Voyager images by Showalter (1991). The Encke Division has also an incomplete ring in the same orbit of Pan suggestingthat this satellite can be keepingthe ring on a horseshoe orbit (Showalter,1991). The two closesatellites of the F ring of Saturn, Prometheusand Pandora (discov¬ ered by Voyager 1), and hypothetical moonlets are claimed to be responsible for the strange morphology present in this ring such as the clumps, waves, kinks and braids. There are some evidencefrom the Voyagerimagesthat there are small moonletsorbiting in the neighbourhoodof the F ring. Kolvoord et a1.(1990)examined a seriesof Voyager images to look for periodicities in the azimuthal brightness that might be associated with nearby satellites. They detected the signature of Prometheus and provided some Proceedings of the XXIth SAB 187 V \ Figure 6. This image (FDS4390511) was taken in back scatteredlight. It showsthe spokeson the B ring. evidence for an undiscovered satellite at a separation in semi major axis of 1180 km from the F ring. By analysing a sequence of five Voyager2 images which showsthe ring divided into separate strands Giuliati- Winter et al. (1995) concluded that this struc¬ ture can be preservedover ~ 45° in longitude and it can be causedby small moonlets orbiting in the gap between the strands. However, these satellites with radii typically ~ 5 km are not ruled out on the basis of the limited Voyagerspacecraft coverage of the F ring. 2.5 Dust The B ring of Saturn presents a strange feature called spokes(Figure 6). They are located in the most optically thick region of the ring. Many studies revealedthat the spokesare composedof microscopicparticles and they are levitating above the orbital plane of the ring. The electromagneticeffectscould be acting on theseparticles in order to make them chargedand thus levitating off the ring (Burns, 1990). A long exposure Voyager 2 image revealed a large number of dust rings around Uranus. The relation between this dust region and the rings are not clear yet. There is a model which suggest that the classicalrings and some hypothetical satellites could be the source of this dust (Esposito and Cowell, 1989,Cowell and Esposito, 1990). Neptune ring system also presents a dust region which extends from between the Adams and Le Verrie rings in the direction to the planet (W = 38000 km). 3. FUTURE PROSPECTS During the period of 1995-96 the rings of Saturn will be seen on edge-on from Earth. At this time the faint rings of Saturn become visible, especially the E ring. 188 Proceedingsof the XXIth SAB It has been organized an international campaign in order to improve the observations made during this period. The Planetology Group composedby members from the Observatório Nacional, UNESP-Campus Guaratinguetá and Universidade Federal de Curitiba is participating on this campaign,whereour principal objectives are to analyse the radial distribution of the F, G and E rings, verify the temporal variations in radial distribution and the E-W variations,and the periodicity in the F ring structure. As has beenanalysedby Giuliatti- Winter and Murray (1995)there is the formation of a gap in the F ring during the closestapproachbetweenPrometheus and the ring particles each ~ 18 years. The last closest approachwas in May, 1993 and this event can span up to two years. This configuration can be observedduring the ring plane crossingunless there is any kind of mechanismwhich can avoid the formation of the gap. The Galileo spacecraftwill improve the resolution on the rings of Jupiter and will allow to analysethe internal ring structure and possibleadditional moonlets. It will be important to know accurately the position of the known satellites,especiallythose one closeto the ring system in order to verify their relevanceto the ring structure. Because of the low inclination of the spacecraft’sorbit it will not be possible to see the ring in three dimension,however the results from Galileo will be as good or better than the Voyagerobservations. Cassini,a combinedSaturn Orbiter and Titan Probe mission,will be launchedin 1997 reaching its destination,Saturn, in 2004. Cassini will orbit Saturn during a four year mission. The primary objectives of the CassiniImaging ScienceSubsystem (ISS) related to the ring system is to study the structure of the rings and the interactions be¬ tween the rings and the satellitesin an attempt to analysethe evolution of the Saturnian system. The author is grateful to the Brazilian agency CNPq for providing financial support (Proc. 30110/94-9)and to Mitchell Gordon for kindly providing the Voyagerimages. REFERENCES Burns, J. A. (1990).Planetary Rings. In The New Solar System (J. Kelly Beatty and A. Chaikin, Eds.), 153-170. CambridgeUniversity Press & Sky Publishing Corporation, Cambridge. Colwell, J. E. and L. W. Esposito (1990). A numerical model of the Uranian dust rings. Icarus 86, 530-560. Cuzzi, J. N. and J. D. Scargle(1985). Wavy edgessuggest moonlet in Saturn’sEncke Gap. Astrophys. J. 292, 276-290. Dermott, S. F. and C. D. Murray (1980). Origin of the eccentricity gradient and apse alignment of the e ring of Uranus. Icarus 43, 338-349. Elliot, J. L., Dunham,E. W., and R. L. Millis (1977).Discovering the rings of Uranus. Sky Telescope. 53, 412-416, 430. Esposito, L. W. and J. E. Colwell (1989). Creation of the Uranusrings and dust bands. Nature 340, 322. Feibelman,W. A. (1967).Concerningthe “D” ring of Saturn. Nature, 214, 793-794. Proceedingsof the XXIth SAB 189 Giuliatti Winter, S. M., Gordon, M. K. and C. D. Murray (1995). Multiple strands in the F ring: Additional evidence for embeddedmoonlets? Submitted to Icarus. Giuliatti Winter, and C. D. Murray (1995). The closest approach between Prometheus and Saturn’s F ring. Submitted to Icarus. Goldreich,P. and S. Tremaine (1979). Towardsa theory for the Uranian rings. Nature 277, 97-99. Kolvoord,R. A., Burns, J. A. and M. R. Showalter(1990). Periodic features in Saturn’s F-ring: Evidence for nearby moonlets. Nature 345, 695-697. Nicholson,P. et al. (1978). The rings of Uranus: Resultsof the 10 April 1978 Occupa¬ tion. Astron. J., 83, 1240-1248. Porco, C. C. and P. Goldreich(1987). Shepherdingof the Uranian rings. I. Kinematics. Astron . J. 93, 724-729. Porco, C. C. (1991). An explanation for Neptune’srings arcs. Science 253, 995-1001. Richardson,D. C. (1994). Tree code simulations of planetary rings. Mon. Not. R. Astron. Soc., 269, 493-511. Schaffer,L. E. and J. A. Burns (1987). The dynamicsof weakly chargeddust: Motion through Jupiter’s gravitational and magnetic fields. J. Geophys. Res. 92, 2264- 2280. Showalter,M. R. (1985). Jupiter’s ring system resolved: Physical properties inferred from the Voyager images. Ph.D. Dissertation, Cornell University, Ithaca. Showalter, M. R. (1991). Visual detection of 1981S13,Saturn’s eighteenthsatellite, and its role in the Encke gap. Nature 351, 709-713. Shu,F. H. (1984). Wavesin planetary rings. In Planetary Rings (R. J. Greenbergand A. Brahic, Eels.),513-561. Univ. of Arizona Press, Tucson. Van Helclen,A. (1984). Rings in astronomy and cosmology,1600-1900. In Planetary Rings (R. J. Greenbergand A. Brahic,Eds.),12-22. Univ. of Arizona Press,Tucson. Proceedingsof the XXIth SAB 191 IS THE SOLAR DIAMETER VARIABLE? N.V. Leister, M. Emilio, P.C.R.Poppe Instituto Astronómicoe Geofísico UniversidadedeSãoPaulo Abstract We discuss the possible variations of the solar diameter, and its connection among other indices of activity or global parameters such as the irradiance or the neutrino production. The dependenceof this variation with respect to the helio¬ graphic latitude is compatible with the non-radialpulsation with 1=6. key-words - Sun(the): oscillations of, Astrometry 1 Introduction The solar radius has been measuredmore than three centuries,and displaysconsider¬ able variation. The solar radius has beenmeasuredwith an accuracy of a few parts in 10-4. Traditional theory of stellar structure concerns itself with stars that are spherical symmetrical, nonrotating and devoid of magneticfields. It expects for a star with the caracteristicsof the Sun, a radius that increasesby at most a few 10-11of its own value per years. The variousmethods that are in use for measuringthe apparent solar semi-diameter, consistsin directly measuringthe anglebetweentwo oppositelimbs,besidesthe other type of methodsthat consistsin timing the duration betweensuccessivecontacts of opposite limbs with fiducial lines on the sky. Analysisof clifferentsdata sets including differentsobservationaltechniquesmade by Gilliland (Gilliland 1981) suggest of a cyclic variation of solar radius with periods be¬ tween 10 and 120 years. Two of data sets are derivedfrom independentmeridian circle observationsconducted at the Royal Observatory at Greenwichand the United States Naval Observatoryat WashingtonD.C. (Eddyand Boornazioan1979)was compiledby Shapiro (1980)and historical Mercury transit observationsby Parkinson,Morrison,and Stephenson(1980)(Fig.l). More recently, astrometrists have measuredthe apparent solar radius by means of a prismatic astrolabe(Laclare1983,Leister 1989,Leister and Benevides-Soares1990,Poppe 1994).Since1974,a regularprogram is under way at Calernand A.MoraesObservatories. Theseseriesconsistsof 3,000independentmeasurements distributed over all seasons and cover solar cycle 21 (Fig.2). The individual analysisof thesedata set showa variation of the apparent solar diameterwith periodsbetween1 and 10 years. The most particulary signaloccuringnear 1,000days,in phasein both set whenindependentlyanalyseis made. The Table 1 shows the individual values obtained in differents data sets with re- spectivesepoch. The considerablevariation display arises,in fact, becausethe various 192 Proceedingsof the XXIthSAP Shapiro'sMercury 0.5 Greenwich . - Transit Data MeridianCircle Í .** ( horizontal) 0 4 44 Table1: Valuesof the solar radius Reference Mean Epoch Radius(arcsec) Remarks Lalande 1764 945.5 From Heliometer measurements Delambre 1806 961.4 Airy 1876 961.82 From Greenwich Auwers 1895 959.63 From heliometer measurements Sofiaet al. 1978 959.77i0.06 From solar eclipsesdata Leister et al. 1986 959.40i0.05 From prismatic astrolabe Sofiaet al. 1992 959.53i0.06 solar disk sextant on ballon flights Laclare et al. 1985 959.42i0.02 From prismatic astrolabe observationmethodsdo not directly measure a true radius. The radius of a surfacegiven optical depth. Rather,they measure propertiesof the Sun’slimb darkeningfunction,and are affectedby many sources of degradationof the solarimage. However,the agreement among some valuesis evident,mainly long seriesof visual observationsmadewith a same kind of instruments. 2 Solar Radius Variations Therealmotivation for observingthe solarradiusis the suspicionthat it might be variable. Theseinvestigation leaveopen the possibility that the apparent solar radius might vary by as much as a few tenths of an arcsecond. The (anti)correlationbetweensolar diameter and some indexes of solar activity has beennoticed (Laclare1996). Proceedingsof the XXIthSAB 193 959,9 < I 9595 1 -c r ;; |C959,1 ? - -- M 958.7 o 500 1000 1500 2000 2500 3000 3500 4000 4500 JD=2444500+t Figure 2. The date set of solar radius measure with prismatic astrolabe (Leister and Benevides-Soares1990 - modified). Table 2: Radius measurements spectralpeackcharacteristic:peakswith amplitudes A T(month) A(arcsec) Remarks 136.6 7.32E-03 0.10 Cerga 133.4 7.50E-03 0.09 Valinhos 61.6 1.62E-03 0.08 Cerga 33.3 3.00E-02 0.18 Valinhos 30.5 3.28E-02 0.13 Cerga 20.4 4.90E-02 0.07 Cerga 19.0 5.26E-02 0.04 Valinhos 10.4 9.61E-02 0.07 Valinhos 10.4 9.61E-02 0.06 Cerga 8.5 1.18E-01 0.09 Valinhos 8.3 1.21E-01 0.03 Cerga We have examinedtwo different data sets of solar radius measurements for evidence of cyclic behavior with periodsbetween1 and 11 years. Thesetwo data sets derivedfrom independentprismatic astrolabeconductedat the CERGA and A.MoraesObservatories. The prismatic astrolabemeasurements from A.Moraeshavebeendiscussedby the authors (Leister1989,Leister and Benevides-Soares1990)Asplotted in Figure 2, the mean value of the radius over 1980 to 1994is 949.40±0.05 arcsec,and this valueagree with CERGA mean value959.42±0.02see in Table 1. Theharmonic analysisover the CERGAandA.Moraessetsdatarevealsa aproximately 11 years periodic oscillation (anti)correlatedwith the Solar activity over the period to 1978-1994(Fig.3). The Table 2 givesthe radius spectralpeakcharacteristic. The anticorrelation between solar neutrino counting rate and some indexesof so¬ lar activity has been noticed long ago (Sakurai1980,Subramanian1979). Gavryusev (V.Gavryusevet al. 1994)evaluateand discusscorrelation of radius data of the Sun with severalcombinationsof harmonics obtained from neutrinos analysis. In spite of a large dispersion there is some significant power close to frequency (1 year)-1. Is not very easy decide whether this is due to the more or lessregular winter interruption of the data, or it is due to real variations of the solarradius observations,but 194 Proceedingsof the XXIth SAB 250.0 2000 g in V. Eco 150.0 2 .»*• • • * *.* O .* •% 100.0 * 2w 3&. (0 7*i 50.0 •• o 0.0 3000 5000 7000 9000 MODIFIEDJULIAN DATE Figure 3. Solar semi-diameteran opposing phaserelative to the solar activity given by the SunspotIndex Data Center (Laclareet al. 1996 - modified) there are a significantpeakin the neutrino flux power spectrum, closeto thesefrequency computedby Gavryusev & Gavryuseva(1993).The Figure 4 showsthe variations of the observeddiameter with the neutrinos flux for the A.Moraesdata set. Finaly, the heliographic inclination of the observeddiameter changesduring the year and also with a given zenith distance. The measurements being sorted into classesof heliographiclatitude. The results are shownin Figure 5, which seems to indicate that the solarradius varieswith period 1 year and semi-amplitude0.04 arcsecond.The exact forms of these dependenciesare difficult to calculate,especiallyfor visual observations; those uncertainties make interpretation of such observations difficult, but these results suggestthat this variation is compatiblewith non-radial pulsation with / = 6. 3 Conclusion The observationof p-modefrequencychangesover a solarcycletime scale(Woodardand Noyes1985)Using the ACRIM satellite solar irradiancedata,theseauthors arguedthat the low degree(l<2) p-modefrequencydecreaseby about 0.4nHzbetween1980and 1984. Subsequentobservationsby Fossatet al. (1987)using velocity data provided consistent evidencefor p-modefrequencyvariability. Although thereare conflictingreport concerning the evidencefor this variability, we considerthat the variability is a solar fenomenon. The possibility that the Earth’s atmosphereis also responsiblefor some long-term variations still remains. An intriguing correlationbetweenthe stratospheric circulation reversalwith the 1,000 day oscillation of the apparent radius has been noted by Ribes et al. (1988).Sincethe Quasi-biennialOscillation (QBO) are located within the tropics, it would be surprising that they affect astrometric station at different latitudes at the same time. The 1000-dayperiodicity is roughly in phasefor observational station such as CERGA 43.5°N,Belgrade43.5°N,and SaoPaulo 23.0°Sand cannot easily be explained Proceedingsof the XXIth SAB 195 rr a a o -EL •Q- X X X X X X XX X i _ | i _ 46600 47100 44100 44800 45100 45800 46100 MODIFIEDJULIAN DATE SEMI-DIAMETER SUNSPOT( ZELENKA) -Q- NEUTRINOS—X— Figure 4. Semi-diametervariations and neutrinos flux. T T 1 O 004 LU oC/5 Figure 5. Correction to semi-diameterand heliographiclatitude 196 Proceedingsof the XXIthSAB as an atmosphericeffect. References [1] Gilliland, R.L., 1981, The Astroph. Journal 248, 1144 [2] Eddy, J.A.,andBoornazion,A.A., 1979,Bull. A.A.S. 1, 437 [3] Shapiro,1.1.,1980,Science208, 51 [4] Parkinson,J.H., Morrison,L.V., and Stephenson,F.R., 1980,Nature 288, 548 [5] Laclare,F., 1983,Astron.Astrophys125, 200 [6] Leister,N.V., and Benevides-Soares,P., 1990,C.R.Acad.Sci.Paris 313 SII, 399 [7] Poppe,P.C.R.,1994,Msc Mestrado, 110 pages [8] Leister,N.V., 1989,PhDDoutorado, 150pages [9] Auwers,A.A.,1891,Astron.Nachr128, 361 [10]Sofia,S.,Heaps,W.,andTwigg, L.W., 1994, Astrophys.J.427, 1048 [11]F.Laclare,C.Delmas,J.P.Coin,and A.Irbah,1996SolarPhysicsin press, 0 [12] Sakurai,K., 1980,Pub.Astron.Soc.Japan32, 547 [13]Subramanian,A., 1979,Curr. Sci. 48, 705 [14]V.Gavryusev, E.Gavryuseva, Ph.Delache,and F.Laclare, 1994, Astron.Astrophys. 286, 305 [15] V.Gavryusev, E.Gavryuseva, 1993, Astron.Astrophys.accepted [16] Woodard,M.F., and Noyes,R.C., 1985,Nature 318, 449 [17]Fossat,E., Geliy, B., Grec, G., and Pomerantz,M., 1987, Astron.Astrophys.177, L47 [18]Ribes,E., Ribes,JJ.C., 1988,Nature 332, 689 Impressão e Acabamento: 5584-6023 ln' J I , I 2 3 4 tl(l06yr)