Full proposal - Exoplanet Exploration Program - NASA

I. Executive Summary100GL 229BWe propose to use SIM’s unprecedentedastrometric capabilities to addresscritical problems relating to the formationand early evolution of both planetsand young, Sun-like stars. SIM willdetect unseen stellar, brown dwarf, orplanetary companions, determine precisedistances to faint, young stars, andimage disk-jet systems, allowing us toaddress the following questions:• What is the incidence of gasgiant planets around young, solar-massstars in the orbital range 1 AU to 5 AU?When and where do gas giant planetsform? By searching for planets around~150 pre-main sequence stars carefullyselected to span an age range from 0.5to 100 million years, we will learn atwhat epoch and with what frequencygiant planets are found at the water-ice“snowline” where they are expected toform. This will provide insight into thephysical mechanisms by which planetsform and migrate from their place ofbirth, and about their survival rate. Asshown in Figure 1, astrometry with SIMis the only technique available to determinethe distribution of planets shortly aftertheir formation in this critical interval ofmass and orbital distance.• What are the masses of young solartype stars? The current uncertainty of afactor of two or more in the masses of YoungStellar Objects (YSOs) results in comparableor larger uncertainties in determinations ofmass functions, ages, circumstellar disklifetimes, and star forming histories ofyoung clusters. We will determine thedistances, masses and luminosities of ~100pre-main sequence stars in binary systems,covering a range of 3 in mass and 100million years in age. With accurate distancesand companion information from SIM wewill be able to determine precise stellarCompanion Mass (M Jup )1010.151 Peg55 CncB70 Vir16 CygBJupiterSaturnUranusSolar Planets0.01Planet-Mass CandidatesPSR B1257+12 PlanetsBD-Mass Candidates5-yr FAME 250 µas0.8 M Sun at 140 pcEarth10-yr KI 30 µ as5-yr SIM 4 µ as10-yr RV 30 m / s0.0010.01 0.1 1 10 100Semi-Major Axis (AU)Figure 1. Comparison of our SIM survey, FAME, andthe Keck Interferometer for the detection of planetsaround young stellar objects at a distance of 140 pc.We adopt single measurement accuracies of 4 µas forSIM, 250 µas for FAME (both for 5 yr missions), and30 µas for the Keck-Interferometer (10 yr mission).We also show the sensitivity to planets for 30 m s -1radial velocity measurements appropriate for YSOs(10 yr survey).masses and thus provide the most reliablecalibrations of pre- main-sequence evolutionarytracks.• What is the origin of the apparentdearth of companion objects betweenplanets and brown dwarfs seen in maturestars? Our survey for young planets andobservations of young binaries will determinewhether the “brown dwarf desert” isthe expression of two different kinds offormation mechanisms, or the result ofdynamical evolution acting on bodies madevia a common process.• To what extent do jets, winds, andoutflows affect the planet forming regions ofYSO disks? We will demonstrate the powerof interferometric synthesis imaging by4

of giant planets is a crucial intermediatestep in this program because the formation,survival, and ultimately the habitability ofterrestrial planets are tied to the propertiesof giant planets, particularly in the criticalorbital range from 1-5 AU.The formation of gas giants like Jupiterrequires the presence of a substantial gaseousdisk around the protostar and thereforemust occur before that disk is dissipated,i.e. in 10 6 -10 7 years (Skrutskie et al. 1990).Once formed, giant planets profoundlyaffect the long-term habitability of terrestrialplanets through direct inward migration andthrough the scattering of solid planetesimals.In our own solar system this scatteringled to a water-rich asteroid belt and theformation of large planetary embryos actingas sources of Earth’s volatiles and perhapsof Earth’s Moon as well (Morbidelli et al.2000).Is the situation in our own solar systemtypical, or just one of a large range ofpossible outcomes of disk evolution?Around 5% of sun-like stars, radial velocitysurveys have discovered Jupiters and perhapseven Saturns (depending upon theactual—unknown—system inclinations) atorbital radii much smaller than those of ourown giant planets, favoring the latter idea(Cumming, Marcy, Butler 1999). Thediverse orbital properties of planets discoveredto date imply that dynamical evolutiondetermines the survival of many planetarysystems (Lin and Papaloizou 1986; Lin etal. 2000). However, we cannot determinefrom existing data whether the radial velocitydiscoveries are the tail of a greaterpopulation of giant planets distributed overlarger semi-major axes, or the preferredprimordial arrangement (making our systemthe exception). Neither radial velocitysurveys nor SIM’s astrometric observationsof mature planetary systems can answer thisquestion if the majority of giant planetsformed with initial Jupiter-Saturn type orbits(4-40 AU; Levison et al. 1998) and subsequentlymoved inward through dynamicalprocesses. It is therefore a high scientificpriority to establish the early distribution ofplanet masses and orbital properties bylooking for planetary companions around asample of young, pre-main-sequence (PMS)stars with ages from 0.5-100 Myr.The future prospects for detecting planetarycompanions around young stars are summarizedin Figure 1 which shows the sensitivityof radial velocity (RV) surveys and astrometricobservations (SIM, FAME, and theKeck Interferometer) to planets around a 0.8M ostar at the140 pc distance typical ofnearby star-forming regions. Since lowermass YSOs are often rapid rotators withactive chromospheres and strong sunspotactivity, they are not well suited for planetsearches via radial velocity methods becauseof the breadth and variability of theirspectral lines. The dispersion of radialvelocity measurements is typically ~10times worse for YSOs than for solar-type,main sequence stars (30 m s -1 vs. 3 m s -1 )with corresponding increases in the massesof planets that the RV technique can detect(Saar et al. 1998). Astrometric measurements,with SIM’s level of precision, areessential to detect Jupiter mass planetsaround YSOs. FAME’s single measurementaccuracy on faint (10

1. The Formation and DynamicalEvolution of Planetsa) When and Where do Planets Form?Determining when and where giant gasplanets form is a key goal of our project.Theoretical timescales for planet formationare very uncertain and depend sensitively ondisk properties (Wuchterl et al. 2000). Sincethe formation timescale is closely related tothe amount of disk dissipation which in turnaffects the amount of planetary migration(Lin and Papaloizou 1986), the “where” and“when” of planetary formation are closelycoupled properties.Current theory indicates that gas giantplanets should form most readily in thecooler, outer reaches of a circumstellar diskwhere the ices of volatile species such aswater are stable, i.e. beyond a few AU(Pollack et al. 1996). These expectationsseem contrary to the recent discoveries ofplanets at locations much closer to theirparent stars (Marcy and Butler 2000 andrefs. therein), but current observationaltechniques cannot detect Jupiters at radii >5AU. SIM will be able to locate 5 AU from young stars,limited only by SIM’s ability to observe overan appreciable part of an orbit during itslifetime (Figure 1). The SIM observationsproposed here will make a definitive test ofthe theory that giant planes form at andbeyond the radius of water-ice condensation.If indeed giant planets are found to beabundant at and beyond the water ice “snowline”,there are two important implications.First, such planetary systems (if devoid ofclose-in Jupiters) could accommodateterrestrial planets in habitable zones, andcould potentially resemble our own solarsystem. Second, the abundance of giantplanets beyond the snowline, coupled withthe RV results for close-in giant planets, willallow an estimate of the rate of migrationand destruction of planets and providecrucial constraints on models of postformationplanetary migration.The combination of SIM’s detections ofplanets and the improved ages of youngstars, made possible by the combination ofSIM and ground-based observations (§II.B),will allow us to test directly the hypothesisthat giant planets form beyond the icecondensation radius in a few million yearsor less.b) What is Initial Mass Distribution OfPlanetary Systems around Young Stars? Theopening of gaps in a protostellar disk mayprovide a natural upper limit to the massesof planets formed by the accretion of solidsand gases in a stable disk. However, radialvelocity surveys cannot detect planetarymass objects around young stars to ascertainwhether the mass function of planets at thetime of their birth is consistent with this orany other theory. SIM can establish a massfunction for the young planets over a decadein mass (from 1 M Jto the putative 10-20 M Jplanetary cutoff) prior to the end of theepoch of dramatic orbital evolution.Although our main focus is on findingplanetary mass objects, our survey will alsodefine the ab initio incidence of browndwarf companions by finding >10M Jbrown dwarfs in orbits ranging from below1 AU to well beyond 20 AU (Figure 1).Radial velocity measurements towardmature stars suggest a “brown dwarf desert”between 10 and 50 M J(Marcy and Butler2000; Mazeh et al. 1998; Mazeh 1999). Ifthis effect is present as we look at progressivelyyounger stars, then formation processes,not dynamical mechanisms, arelikely to be the cause.c) How Might Planets Be Destroyed? OtherSIM programs will target the closest starswhich are, by accident of our solar neigh-7

of 5 yrs to ensure that Jupiter mass planetscan be detected with high reliably andcompleteness over at least the 1-5 AU rangeof orbital distances. This accuracy level isalso robust against the expected level ofastrometric jitter produced by starspots, byvariable scattered light from a disk (§II.A.4),and/or possible degradations in SIM’sastrometric accuracy. In many cases we willbe able to detect single planets with signaturesx2 smaller than this value, i.e. 0.5 M Jat 1 AU or 1 M Jat 0.5 AU as well as sub-Jupiter mass planets beyond 5 AU. Anillustrative sample of clusters, observingtime estimates, and specific stars are givenin Tables 2 and 3 (§II.A.6, 7).Our sampling strategy (§II.A.5) is consistentwith finding up to 3 planets orbiting astar with unknown periods between 100and >2000 days, although we will be able toderive only limited information on planetsRegionTaurus,CrA,Lupus,OphiuchusChamaeleonSco-Cen 2-20 160 Cluster mode.Narrow angle**TW Hya,and othernearbyyoungstarsTable 2. A Survey For Jupiter's at 1-5 AU*Total Time for 40Age Dist Observing# samples(Myr) (pc) ModeTime per star Stars and 2 axes1-2140 Single star, 10 stars (

with periods longer than about twice theduration of the SIM mission.4. Astrophysical Limits To SensitivitySIM’s astrometric accuracy for YSOsmay ultimately be limited by astronomicalnoise sources that induce shifts in theobserved photocenter. These noise sourcescan include starspots, starlight scattered bycircumstellar disks, disk hot spots, andvariable disk extinction. Since many ofthese attributes are characteristics of theyoungest stars with the densest disks, thereis an unavoidable tension between minimizingthese effects by rejecting sources andkeeping the most interesting objects. In thissection we try to quantify the effect of someof these processes on the astrometric dataand show how, by prudent source selectionand careful data analysis, we can makeobservations at the precision needed to findplanets at the desired levels.a) Astrometric Shifts due to StarspotsStarspots can cause changes in the positionof the photocenter of a star that canmask the astrometric signal from a planet.In the past decade a number of groupshave used photometric variability andDoppler imaging to investigate the surfacestructures on both cTTs and wTTs. Thedata can be interpreted in the followingways: relatively few, large, cool, longlivedstarspots in both cTTs and wTTs (~10-40% of a projected hemisphere vs. fewmillionths for the Sun); cool polar caps inrapidly rotating wTTs; hot spots in cTTsthat may represent regions of accretionimpact at the stellar surface; and a markeddecrease in spot activity for stars approaching1 M o(Shevchenko and Herbst,1998; Bouvier et al. 1995; Bouvier andBertout 1989; Schussler et al. 1996; andrefs. therein).We have developed a numerical model tocalculate the photometric and astrometricAstrometric Error (uas)effects of a distribution of spots over thesurface of rotating stars (Figure 2). Themodel incorporates a broad range of spotand star parameters as derived from thereferences cited above. A variety of runswere made for 100 evenly spaced observationsover a 5 year period at wavelengthsfrom 0.4-0.95 µm. The key result of thenumerical analysis (Figure 2) is that a V-band photometric variation of ∆F/F=10%(rms) corresponds to an astrometric variationof ~3 µas (rms) in the position of a 2 R opre-main sequence star at the distance ofTaurus.32.521.510.500 0.02 0.04 0.06 0.08 0.1Photometric Variation (1 sigma)Figure 2. The relationship between thephotometric variation (∆F/F) in V-band andphotocenter shift due to sunspots on a T Tauristar at 140 pc.These values are to be compared with the 8µas signal of a Jupiter in a 1 AU orbitaround a 0.8 M ostar at 140 pc. Because bothphotocenter excursions and the astrometricamplitude scale inversely with stellar distance,this noise source may ultimately set adistance-independent limit to our ability tofind planets smaller than ~0.1 M Jaroundyoung stars. The fact that the photocenterexcursions are well correlated with theintensity excursions will be critical formasking or modeling, and removing, theeffects of starspots. We will compensate forthe photocenter variation due to starspots in10

a number of ways:1. Minimize this effect by selectingstars with relatively small photometricvariations attributable to starspots (∆V

The choice of N sampleis driven by the need tohave enough independent (2-axis) measurementsto detect and characterize multipleplanets. A minimum of 5+7*N planetobservationsis needed to characterize a multiplesystem in a least squares sense. The choiceof N sample=40 is consistent with an ability tocharacterize 3 planets with a factor of ~1.5redundancy in the data. A critical part of ourwork in the years before the launch of SIMwill be to carry out higher fidelity simulationsof multiple planet systems to fine-tunethe SIM observing strategy (§IV.A).6. Observing Time RequirementsSince the majority of the YSOs are faint,V~14 mag, we must be as efficient as possiblein observing them. We have used thescientific requirement of the reliable detectionof 1 M Jplanet at 1 AU to set a measurementaccuracy that varies with the mass ofthe parent star and its distance from Earth.We will tailor our observations for maximumefficiency using the following modes:a) Wide angle Observations. A Jupiter at 1AU from a 0.8 M ostar at 50 pc has an astrometricamplitude of 25 µas. A 2σ detectionrequires only a single measurement accuracyof 12 µas which can be accomplished efficientlywith multiple observations in wideangle mode.b) Narrow angle Observations. The sameplanet at Taurus (140 pc) has an astrometricsignal of 8 µas, requiring a narrow anglemeasurement of 4 µas, comfortably removedfrom SIM’s ultimate performance limit.To estimate the necessary duration of thenarrow angle observations, we have derivedan error budget from the information on theSIM web-pages that incorporates bothbrightness dependent and systematic errors.Assuming that the narrow angle systematicerror scales in a power-law fashion between1 o (1.7 µas) and 15 o (7.7 µas), we canachieve a 4 µas accuracy relative to areference star as far away as 4 o from thetarget. The advantage of a reference star thisfar away is that we have an excellent chanceof being able to use a SIM grid star as areference. Grid stars are bright (V

2 that meet all our criteria of mass, distance,age, lack of strong disk emission or pronouncedvariability (∆V

B. From the fact that the RV resultscover only a small part of the SIM discoveryspace around any given star, and using theresults of Trilling et al. (2000) as a roughguide, one might assume the incidence ofgiant planets around stars of any age isabout 10%.C. Use the Trilling et al. (2000) migrationcalculations as a guide to the loss ofplanets early in the history of sun-like stars.Since migration timescales, including gasand post-gas dust, would plausibly be 1-10Myr, we estimate that >30% of starsyounger than 10 million years have giantplanets, but only 10% of stars older thanthat maintain giant planet systems (i.e., atleast one giant planet). Nor, of course, canwe exclude the possibility that all youngstars have planets.Under these three different assumptions, weexpect to find roughly 7 planets (Case A),15 planets (Case B) or >35 planets (Case C)around star stars younger than 100 Myr anda few (2-5) planets around the olderPleaides sample. Thus the total number ofplanets we expect to detect ranges from 7 to>35, and possibly up to as many as >100planets.What do the above results imply for derivinginformation on giant planet formation?In Case A, the paucity of objects foundwould be a significant scientific resultimplying that giant planet formation may berarer than presently thought. It is also clearthat Cases B and C are distinguishable fromeach other; i.e., if indeed giant planet formationis a not-uncommon process, we shouldbe able to see the effects of migratory loss inour discovery statistics around youngerversus older stars.B. Fundamental Properties ofYoung Stellar ObjectsUnderstanding the formation and evolutionof stars is a second major objective of theOrigins Program. Since planet formation isa part of the phenomenon of star formationand evolution to the main sequence, understandingthe properties of YSOs is essentialto understanding the environment of planetformation. SIM offers unique capabilities toimprove our understanding of how starslike our Sun and smaller evolve toward themain sequence.Mass is the most fundamental property of astar and determines its early evolution andultimate destiny. Yet we don’t know themasses of most young stars, particularlythose of a solar mass and less, to within afactor of 2. Accurate knowledge of masseswill help calibrate early YSO evolution; theidentification of large numbers of accuratemasses and mass ratios will enable theoriststo determine definitively the formationmechanisms of binary star (Clarke 2000).Because the masses of only a few, young,low mass stars are well known, the premain-sequence(PMS) evolutionary tracksused for estimating YSO masses and agesare not accurately calibrated. Currentmodels for the tracks give masses and agesfor a 1L oK7 PMS star which vary by afactor of ~2 (Table 4). The discrepancy iseven greater for later type stars. As a result,parameters essential to theories of starformation, such as stellar ages, star formingregion history, initial mass function and thedistribution of masses in binaries are poorlyknown. The SIM program proposed herewill, by calibrating the calculations of PMSevolution over a decade in stellar mass,improve our knowledge of these astronomicalparameters and, equally importantly,resolve our uncertainties of the physicalinputs to the calculations, e.g. atmospheric14

Table 4. Theoretical Mass and AgeEstimates for a K7, L=1.0 L PMS StarCalculation M/M o (Myr)AgeCohen and Kuhi (1979) 0.85 2.0Swenson et al. (1994) 0.65 1.0D’Antona and0.45 0.8Mazzitelli (1997)Baraffe et al. (1998) 0.80 1.0Palla and Stahler (1999) 0.80 2.0opacities, convection mechanism, and theequations of state (cf. White et al. 1999).While FAME will make major improvementsin reducing uncertainties in mass,that mission cannot provide data of precisionsufficient to calibrate evolutionarytracks so that they can become predictiveover the full range of ages and masses —from ages when accretion ceases (1-10Myr) to the ZAMS (10-30 Myr later). Withonly a small investment in observing time(5% of our total), we can take advantage ofthe power of SIM for ultra-high precisiondistance and mass estimatesfor faint objects toproduce a unique databasethat theoreticians can useto calibrate PMS tracks to1% precision.Observations of visual andspectroscopic binaries(Mathieu 1994; Ghez et al.1995; Thiebaut et al. 1995;Simon et al. 1996; Caseyet al. 1998; Prato 1998)and the mapping of circumstellar(CS) diskrotation (Dutrey et al.1994, 1998; Koerner 1997;Mannings & Sargent 1997;Guilloteau & Dutrey 1998)provide the only reliabledeterminations of stellarTable 5. DynamicMasses Derived fromDisk Rotation(∝D/140 pc)Star M ✷ /M oSingles:MWC 1.65 ± 0.07480LkCa15 0.97 ± 0.03DL Tau 0.72 ± 0.11GM Aur 0.84 ± 0.05DM Tau 0.55 ± 0.03CY Tau 0.55 ± 0.33BP Tau 1.24±0.32Binaries:GG TauA 1.28 ± 0.07UZ Tau E 1.31 ± 0.08masses. We propose to use SIM to measuredistances and inclination angles of a sampleof ~100 stars for which sufficient dynamicaldata exist to enable us to measure the stellarmasses to high precision. Figure 4 comparesthe masses measured by CS disk rotation(Table 5) with PMS tracks calculated byBaraffe et al. (1998, BCAH98) andD’Antona and Mazzitelli (1997, DM97).Since these masses scale with distance, thedistance-independent parameter L/M 2 isplotted versus effective temperature. In thisformat, if a star does not lie on the trackcorresponding to its nominal mass, either itsdistance is not the reference value (140 pc),or the theoretical track is wrong. The distancesto these stars are not well enoughknown now to distinguish between thesetwo possibilities. The figure indicates that ameaningful test of the tracks requiresabsolute uncertainty of the measured massesof less than about 5%. Since the internalprecision of the dynamical mass measurementsis better than a few percent (Table 5),we require that the distance and inclinationmeasurements not compromisethe overall uncertainty.In the following sections, wedescribe three approaches toobtaining dynamical masses: 1)visual binaries (VBs), 2) spectroscopicbinaries (SBs), and 3)mapping of circumstellar diskswith CO interferometric millimeterobservations. Approaches1) and 3) require precise SIMdistances to determine themasses; approach 2) requiresSIM measurements of thebinary orbit parameters (particularlyinclination angle). Table 6gives the dependence of massestimates on distance and showsthat 1% distance measurementsare well within SIM’s design15

limits. This value is chosen so that thedistance of a star will not compromise theoverall value of its mass and therefore willallow the best possible comparison with thetheoretical PMS tracks. Through AssociateRed DashedBlue SolidInvestigator Baraffe and her collaborators,our team has dedicated access to some ofthe best numerical codes available for thefuture development of evolutionary trackscalibrated using SIM.1. Visual BinariesTable 7 lists some of the visual binaries(VBs) for which orbital motions are nowbeing mapped from the ground and withHST (Ghez et al. 1995, Thiébaut et al. 1995,Simon et al. 1996). These have a > 20 milliarcsecso their primaries and secondaries areresolvable by SIM. SIM observations willmeasure not only the parallax but also, forsystems with detectable orbital motion ofthe primary and secondary, the ratio of thesemi-major axes of the primaryand secondary, a 1/a 2,and hence M 1/M 2permittingsolution for both M 1and M 2.Detection of any wobble inthe motion of the primary orsecondary would revealadditional companions.Figure 4. PMS tracks for stars of mass 0.1 to 1.2 M ocalculated by DM97 and BCAH98 and masses ofsingle cTTs measured by disk rotation. Theuncertainties displayed are ± 1 spectral type subclassand the propagated internal uncertainties of themass measurement and an assumed ± 10%uncertainty in luminosity. The small filled circlesmark ages 1, 2, 3, 5, 7, and 10 Myr for DM97, and 2,3, 5, 7, and 10 Myr for BCAH98.Table 7 shows that the precisionneeded for accurate massdeterminations falls wellwithin SIM’s wide anglecapabilities and can be obtainedwith a modest investmentin observing time. Thetable also estimates theprecision that FAME willobtain for a given brightness.For stars fainter than V=12.5mag, a FAME parallacticuncertainty of ~250 µas willyield an overall uncertainty inthe absolute mass of 10%,which is insufficient for ameaningful test of the evolutionarytracks.2. SpectroscopicBinariesAn active effort is in progress to identifyspectroscopic binaries (SBs) among theyoung stars and, since their periods tend tobe typically less than a year, to derive theirorbital parameters (Mathieu 1994;Neuhäuser 1999). Also underway is aproject to identify additional double-linedSBs by complementing visible light spectroscopyof SB1s with high spectral resolutionobservations in the near-IR where thesecondary star is brighter than it is in opticallight (Prato 1998; Mazeh et al. 2000;16

Steffen et al. 2000).The SIM observationswill convert SBs toVBs by measuring theinclination angle, i. Fordouble-lined SBs, thiswill yield each of thecomponent masses. We will include the fewknown (Corino et al. 2000), eclipsingdouble-lined SBs in our program as they areidentified to provide an independent determinationof their masses.3. Mass Measurements of Stars withResolved CO DisksMaps of disk rotation measured by mmwaveinterferometry provide a unique meansto measure masses of single stars, and anumber of precision values are alreadyavailable (Table 5). The derived massscales linearly with distance to the starbecause it depends on the physicalradius at which a given velocity ismeasured. Table 5 lists the massderived from recent 12 CO J=2-1observations of cTTs in Taurus(Simon et al. 2000). The precision ofthe masses in Table 5 is limited byuncertainties in the disk inclinationand in the distance. SIM will improvethe precision of the mass measurementsfor the brighter stars relative toFAME, and enable a precise measurementof the fainter stars (V>14 mag).4. The Total ProgramThe total masses program will consistof observations of the visual andspectroscopic binary stars and thestars with CO disks (Tables 5 and 7).Over the next 3-5 years, we willidentify new targets, with a particularemphasis on late K-M spectral typestars, M350 0.6DL Tau 13.1 71 270 0.6GM Aur 12.9 71 250 0.6LkCa15 12.1 71 180 0.6Visual BinariesFW Tau 16.4 24 >500 1.0F0 Tau 15.4 24 >500 0.6HV Tau 14.0 24 >350 0.6DI Tau 12.9 24 250 0.6Spectroscopic binariesHaro 1-14c 12.3 21 180 1.2(P=591d)0425+3016 11.6 24 250 1.2(P=2530d)1559-2233 11.2 21 120 1.2(P=2.4d)GW Ori(P=242d)9.8 7 70 1.217

time by SIM, will enable abreakthrough in our understandingof the early evolutionof solar type and lower massstars. Further, the greatlyimproved evolutionary trackswill give us a chronometerwith which to order the sequenceof events that leads tothe formation of planetarysystems.C. The Structure andOrigin of YSO Disks andJetsJets and outflows play a keyrole in the evolution of accretiondisks and thus may be animportant factor in the formationof and survivability ofgiant planets close to centralstars.We propose a small program to imageHerbig-Haro jets driven by YSOs in Taurus.Little is known about the size scale onwhich jet acceleration and collimation occurwhich would allow us to distinguish betweena “disk” wind (Konigl 1989) or a“stellar” wind (Shu et al. 1995). Extrapolatingthe widths of jets from HST imagesinwards suggests widths of ~15 AU, easilyresolvable by SIM, that favor a disk windinterpretation (Burrows et al. 1996;Reipurth et al. 2000). However, stellar windmodels cannot be ruled out and wouldpredict much narrower jets that would beunresolvable by SIM.Jets radiate strongly in optical wavelengthemission lines such as [S II] 671.7/673.1 nmand the 656.3 nm Hα line. SIM will resolvestructures down to 0.01" (1.5 AU) in nearbyYSO jets that could constrain certain jetmodels in a way impossible even withmajor advances in ground-based adaptiveFigure 5. Left panel: Models of the central region of DG Tau(0.28'’x0.14'’) assuming (top) the jet is launched from acompact region (r ~10 R o) at a narrow opening angle, or(bottom) it is launched from the disk within a region ~5 AU.Right panel: The corresponding simulated Hα images for SIM.The star’s emission has been subtracted in both the model andSIM images (using data from adjacent wavelength channels).optics. The result will be a scientificallyimportant and visually compelling demonstrationof space interferometer imaging, acritical precursor for the observationsplanned with the Terrestrial Planet Finder.We propose SIM visibility measurements tomap the brightness distribution of five YSOjets in Taurus (Table 8). In the single SIMspectral channel that includes the Hα line, asignificant fraction (~20%) of the total fluxis expected to come from the extended jetso that a lack of contrast will not be an issuein the data reduction.We have simulated aperture synthesisimages for the DG Tau (V=12 mag) system,assuming 7 sidereostats providing 13independent baselines from 0.8 m to 10 msampled at 0.8 m intervals and by rotatingthe baseline in steps of 10 o . The simulationsinclude appropriate levels of phase, amplitude,and photon noise and show that SIMshould be able to resolve a jet driven by a18

Table 8. Potential Sources InTaurus For Jet ImagingStarV(mag)Hα(EW)[SII](EW)DG Tau 12.0 113 9CW Tau 12.4 135 2DM Tau 14.0 139 ---DP Tau 14.285 -DO Tau 14 100 1disk wind with just a few hours of integrationtime (Figure 5). We anticipate spending40 hours (8 hours ×5 objects) to observe 5stars plus an additional 10 hours to measure1 or 2 bright unresolved point sources forcalibration purposes.Without invoking full synthesis imaging wewill carry out a second imaging experimentin direct support of the astrometric observations.As discussed in §II.A.4, time variableillumination of disk structures may induceshifts in the photocenter measured by SIM.We will use the visibilities obtained inmulti-wavelength single-baseline u-vsampling at several orientations to characterizethe disk photocenter and any asymmetries.These data will allow us to developtechniques to assess and mitigate, if necessary,the influence of disk structures on theastrometry of the entire sample. We allocate25 hours to carry out these experimentstoward 5 stars with a variety of disk characteristics.D. Precursor and SupportingObservationsA variety of precursor observations areessential for defining the best possible listsof target and reference stars (Table 9). Thehighest priority use for SIM funds (§IV.B)will be to pay for the travel and postdoctoralassistance needed to identify and characterizethe stars for planet search part of program.Co-I Ghez (UCLA) will coordinatethis program which will take advantage ofnational observatories as well as facilities towhich relevant team members have access(Table 12).1. VariabilityWe will use a combination of FAME andground-based photometry using smalltelescopes to reject highly variable stars orto identify variable stars that might besimply modeled. Associate-InvestigatorHerbst has considerable experience inmaking these observations. Beichman andShao are members of the FAME team andwill have access to FAME photometry priorto its release for this purpose.2. BinarityWe will use AO imaging from large groundbased telescopes and HST snapshots tosearch for close binary companions to targetstars, rejecting stars with companions closerthan ~100 AU. We will also use spectroscopyto identify spectroscopic binariesusing 2-3 spectra to look for radial velocityvariability. Objects removed from the planetsearch sample could be added to the fundamentalproperties sample. FAME data willbe useful in weeding out target and referencestars with stellar or brown dwarfcompanions.3. Reference StarsIn addition to using SIM grid stars wheneverpossible, we will use 1-2 bright referencestars (V~10-12 mag) within 4 o forplanet searches in clusters. We recognizethe difficulty of finding reference stars instar forming regions where extinction canblock the distant giants we would like touse. However, since we have selectedrelatively nearby, high latitude, star-formingregions, it is possible to see through theseregions to distant stars over a 4 o scale. Wewill select reference stars by using 2MASSand visual photometry to identify candidate19

K0-M0 giants brighter than V~10-12 mag(Bessell and Brett 1988). We would thenuse spectroscopy to verify that the stars aretruly giants free of perturbing companions.The clustering of YSO targets reduces thenumber of reference stars we will have toidentify and characterize.4. Observations Of DisksWe will use the following techniques todetermine whether disks or other nebulositywill be a problem for stars in our sample:use of 2MASS and other near-IR data toreject stars with massive accretion disks(Lada and Adams 1992; Meyer et al. 1998);expansion of the HST snapshot imagingsample (Stapelfeldt et al. 1998) to investigatenebulosity or disk structures at the 0.1″scale; continued observing programs atKeck and Palomar to assess the presence ofdisks and binary companions using adaptiveoptics imaging.5. Other ObservationsWe will augment and refine the target list forthe stellar masses part of the program withprecursor observations using SIM funds plusother research support. These observationswill include identification and characterizationof visual (Ghez, Simon, Mathieu) andspectroscopic binaries (SB1s, Mathieu;SB2s, Prato) as well as continued millimetermapping of disks around single stars(Simon).E. Data Analysis and DataProducts1. Companion SearchesWe assume that the Interferometry ScienceCenter (ISC) will provide time-tagged delayline values, observed visibilities and photometricamplitudes, instantaneous baselineTable 9. Precursor Observations In Support Of Planet Search# of Sources/ ParticipantsProgramFacilityNightsVariabilityFAME; 0.6 m WeslyanTelescope and Yale 0.6 mtelescope (CTIO)200 stars; 10-20nights/yr for 3 yrHillenbrandHerbstSearch for unseencompanions: Radialvelocity variability orshift relative tomolecular cloudSearch for unseencompanions(imaging)Identifying newYSOs within 50 pcand with 5

vectors, corrected for differentialstellar aberration andgravitational deflections bysolar system bodies. We willthen use the photometric data tolook for variability and thevisibility data to look forresolved structures or luminouscompanions that might producespurious astrometric signatures.We will then correct the datafor relative parallax and propermotion to produce a time seriesof relative two-dimensionalastrometric measurements between thetarget and the reference star(s). The corrected,relative two-dimensional astrometricmeasurements for each star constitute afundamental deliverable of our project.Table 10. Analysis of A Three Planet SimulationPlanet 1 Planet 2Planet 3Input ParametersSemi-major axis (µas) 3216 8Period (day) 869 548 345Mass (M J ) 2.2 1.5 1Inclination(deg) 75 75 75Derived ParametersSemi-major axis (µas) 35.4±0.5 15.1± 0.3 8.1± 0.3Period (day) 881±3 562±2 349±1Mass (M J ) 2.4±0.1 1.4±0.1 1.0±0.04Inclination (deg) 78.6±0.7 70.4±1.3 83.0± 1.8Simulations of multiple signal extractionsuggest that detection of multiple companionsshould proceed iteratively by systematicallyand successively identifying, esti-mating, and removing statistically significantperiodicities from the data set. Inparticular, this means the identification ofperiodicities from LS periodograms, andthen the fit of a Keplerian orbit model to theastrometry data. From the inclination andsemi-major axis of the model the companionmass (formally the companion/star massratio) is directly inferred.We will use the standard method of Lomb-Scargle (LS) periodograms (see Black andScargle 1982) which should be reliable androbust for periods up to the duration of theSIM mission (although Cumming et al.(1999) suggest alternatives to simple LSperiodogram analysis that may be relevantto our SIM analysis). Because the astrometricmeasurements are relative, there is aformal degeneracy as to which object hasthe periodicity; this degeneracy can easilybe broken by identifying commonperiodicities among pairwise periodogramsbetween multiple target and reference stars.This degeneracy can also be broken if anindependent observation method (e.g. radialvelocity or photometry) indicates commonperiods among the target or reference stars.Provisional detections of some of the largerand higher-frequency periodicities are likelyto be available at the halfway point of themission. However as the extraction ofsmaller-amplitude signals are dependent onthe correct removal of larger signals thesecompanions will largely go undetected untilnear the end of the mission. As a final step,the optimal signal extraction will utilize asimultaneous estimation of companion orbitmodels initialized with the results of theserial extractions. We will use an extensiveprogram of Monte Carlo simulations basedon our exact SIM observing strategy toassess the completeness and reliability ofour survey so that we understand quantitativelythe significance of the absence ofplanets around a particular star.We have developed planet detection algorithmsbased on astrometric work carried outusing the Palomar Interferometer Testbed(co-Investigator Boden; e.g. Boden et al.1999; Boden and Lane 2000). We have usedthese algorithms on simulated data forsystems of up to three planets orbiting a 0.8M ostar at the distance of Taurus. Consider a21

co-planar system of three planets inclined by75 o . We simulated our proposed 2-dimensionalobservational sequences with 40measurements, each with a precision of 4µas with samples distributed through a 5-yrperiod with a power law α=0.98 (§II.A.5).Successive examinations of the LSperiodogram revealed the three planets withhigh significance and led to the parametersgiven in Table 10. Even in the 3-planetsystem, we were able to derive the mass of a1 M Jplanet at 1 AU with 4% accuracy.We will build on our existing planet searchalgorithms to develop a working, prototypepipeline as early as possible during theprogram. This prototype will be used toimprove the algorithms, understand interfaceswith the ISC, and optimize our observationstrategy.2. Luminous CompanionsThe above analysis will be carried out tohelp characterize the fundamental propertiesof binaries in the T Tauri star sample. Onemodification will be to examine the visibilitydata for direct evidence of light from thecompanion itself. We estimate that we willable to measure the brightness and separationof a companion as faint as ∆V=5 magat 10 mas separation and ∆V=4 mag at 5mas. In these cases, we will be able tocharacterize the properties of a coeval pairof stars with well determined mass, temperatureand luminosity differences.3. Imaging AnalysisThe basic data for this mode are the calibratedvisibilities at all wavelengths on thevarious baselines and rotation angles. Wewill make images with these data usingstandard Fourier-based image processingtechniques (AIPS). Modifications to thistechnique might include using adjacentcontinuum channel images to improve thephase accuracy in the line channels andusing images from HST or other facilities tofill in short spacings not well covered bySIM itself (

and observation to advances in our understandingof the Universe. Explanation ofthese connections is fundamental to enhancingthe public’s understanding of the benefitsof scientific research and is in keepingwith the goals of national education reformefforts such as those outlined in the ‘Benchmarks’publication of Project 2061.Members of this scientific team have experiencepartnering with the formal and informaleducational community. Team membersYorke, Stapelfeldt, Hillenbrand, andNorman, have past and present partnershipsbetween elementary and secondary schoolsthrough participation in programs such asProject ASTRO. In addition, Stapelfeldt andGhez have experience in presenting at theNational Science Teachers convention,developing courses for college professors tohelp bring research into the classroomsetting, and developing exhibits and materialfor museums. A few members (D. Norman,L. Hillenbrand) have participated in programsdesigned to encourage groups historicallyunderrepresented in science, e.g.science workshops for girls through programslike the Girl Scouts, Rural Girls inScience Camp and Expanding Your Horizons.One team member (Mathieu) is involvedwith the National Institute for ScienceEducation in Madison, Wisconsin developingcurriculum modules that use leadingedgeresearch as a tool for introducingscience concepts into introductory collegeclasses. The EPO team can use the searchfor planetary systems to infuse K-14 curriculummodules with new ways to presentelementary concepts in math, astronomy,physics, chemistry and biology.As shown in the organizational chart (Figure6) for the YSO science team, Educationaland Public Outreach (EPO) efforts will besupervised and coordinated by the DeputyPI, ensuring that EPO will be a high priorityof our team. The budget presented belowshows that ~4% of the project funds will beused for EPO activities with the majority ofthe funds to be expended after launch toprovide partial support for a post-doc withinterests in education.The EPO team has established contacts withOSS Forums and Broker/Facilitators at JPLand DePaul University and will workclosely with the Office of Space Sciences(OSS) to be sure that the programs andmaterials developed reflect the nationalgoals and standards for science, mathematicsand technology education. The EPO teamwill also establish additional new contactswith other minority institutions and teachingcenters through NASA’s EducationalDivision, Minority University Research andEducation Division and the RegionalTeacher Resource Centers. The EPO teamrecognizes the need to evaluate the impactof our efforts. Therefore, we will ask external,professional evaluators to review ourprogram periodically.IV.Cost and BudgetA. Science Team StructureThe overall SIM-YSO program is led by thePrincipal Investigator, C. Beichman (JPL),who has overall responsibility for the successof project and for the execution of thetasks listed in the Work Breakdown Structure(WBS) described below. The scienceteam is organized both by scientific interests(Table 12) and into six working groups thatreflect the major tasks that must be accomplishedfor a successful program (Figure 6).This WBS was used to develop the costestimate for this proposal. It should benoted that funds are allocated purely on thebasis of the WBS and are not sent to team23

members just by virtue oftheir being on the scienceteam.The Theory Group is led byJ. Lunine (LPL) for planetformation issues and by LeeHartmann (CfA) for YSOphenomena. During the prelaunchphases, the group willensure that the proposedmeasurements areappropriate to developing thebest understanding of thephysical processes of planetformation and stellar evolution.This will be donethrough participation in targetstar selection (ensuring anappropriate spread of ages,metallicity, etc.,) developmentof software tools forprompt analysis of detectedplanets, and interface with theHartmann-led group on YSOphenomena to optimizeselection of disk observationsfor information on planetaryformation processes. Thetheory and data analysisgroups will conduct simulationsof planet-detection withthe SIM-YSO target list toquantify the uncertainties inderived parameters such asplanetary initial mass function,and optimize the list tominimize these uncertaintieswhere possible.Precursor Observations. Thisgroup is led by A. Ghez ofUCLA and has the critical role during theyears before launch of gathering the dataneeded to enable final target selection. Teammembers and postdocs will make and reduceTable 12. Science Team Affiliations and InterestsName Institution(Observatory)Primary ScientificInterestCo-Investigators (Core Team)Beichman, C JPL (Palomar) Planet SearchesBoden, A. JPL/IPAC Planet Searches(Palomar, Mt.Wilson)Ghez, A. UCLA (Keck, YSO BinariesLick)Hartmann, L. CfA YSO PropertiesHillenbrand, L. CIT (Palomar,Keck)YSO Properties,Planet SearchesLunine, J. LPL Planet FormationAnd Migration,Disk StructureSimon, M. SUNY YSO PropertiesStauffer, J. CfA (MMT,Magellan, Mt.YSO PropertiesPlanet SearchesHopkins 48", 60")Velusamy, T. JPL Disk/Jet ImagingAssociate InvestigatorsBaraffe, I. U. Nancy YSO EvolutionCarpenter, J. CIT (Palomar, YSO PropertiesHerbst, W.Keck)Weslyan(Van Vleck 0.6mtelescope)Planet SearchesT Tauri StarVariabilityKulkarni, S. CIT (Palomar, Planet SearchesKeck)Lin, D. UCSC (Lick,Keck)Planet FormationAnd MigrationMathieu, R. U. Wisc.YSO Properties(WIYN)Norman, D. SUNY EPOPrato, L. UCLA (Lick, YSO PropertiesKeck)Shao, M. JPL Planet SearchesStrom, S. NOAO YSO PropertiesStapelfeldt, K. JPL (Palomar) Disk/Jet ImagingYorke, H. JPL Jet Theoryobservations from Northern and SouthernObservatories as well as with HST to assessphotometric and spectroscopic variability,and identify the presence of disks, companions,and nebulosity (Table 9).24

C. BeichmanPrincipal InvestigatorSimonDeputy PILunineTheoryGroupHartmannGhezPrecursorObservationsBodenObs.Planning& DataAnalysisStaufferTargetSelectionHillenbrandVelusamyImagingScienceSimonEducation& PublicOutreachBaraffeKulkarniLinMathieuYorkeCarpenterHerbstMathieuPratoSimonStromBeichmanCarpenterKulkamiShaoSimonStapelfeldtVelusamyPratoStromGhezHartmannStapelfeldtYorkeGhezMathieuNormanStapelfeldtStromYorkeFigure 6. A simplified Organization Chart and Work Breakdown Strucutre for the SIMYSO project shows the basic tasks that must be accomplished prior to launch andduring the mission.Observation Planning. This group is led byAndy Boden (JPL and ISC) and will addressthe key issues of sampling unknownperiods and the choice of most efficientdata acquisition modes. This group will helpto develop an integrated simulation capabilityto minimize the number of observationsneeded for planet detection, as well as tooptimize their scheduling.Target Selection. This team is led by JohnStauffer (CfA) and takes input from theteams responsible for Theory, precursorobservations, and observation planning tocome up with an integrated list of targets.This team works intensively during the yearor so before launch to define and refine theobserving list.SIM Data Analysis. The largest part of ourbudget goes to the team members,postdocs, and programmers working toreduce the SIM data. The available budgetlimits the scope of this activity, so we mustrely heavily on receiving reliable, usefulproducts from the ISC in a timely manner.The Data Analysis team (led by A. Boden,JPL and ISC) will interface with the ISC todefine appropriate products and to understandkey ISC algorithms. The group willalso develop algorithms for planet searchingand characterization. In conjunction withthe entire science team, the data analysisgroup will reduce and analyze the data toderive relevant astrophysical quantitiessuitable for archiving by the ISC and publicationin professional articles. We show amodest software development from theonset of the project to support the planetsearch prototyping.25

0.2 0.2 0.20.2Table 13. SIM Science Team Workforce (FTE)2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011Theory 0.7 0.7 0.7 0.7 0.7 1 1 1.1 1.1 1.1 1.1Postdoc 0.5 0.5 0.5 0.5 0.5 0.6 0.6 0.6 0.6 0.6 0.6University 0.1 0.1 0.1 0.1 0.1 0.3 0.3 0.3 0.3 0.3 0.3JPL 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.20.20.20.2Precursor Obs 2.45 2.45 2.45 2.45 0.85 0 0 0 0 0 0Postdoc 2 2 2 2 0.5 0 0 0 0 0 0University 0.2 5 5 0.2 5 5 5 0 0 0 0 0 0JPL 0.2 0.2 0.2 0.2 0.1 0 0 0 0 0 0Obs Planning 0.2 0.2 0.2 0.3 0.6 0.4 0.2 0 0 0 0Postdoc 0 0 0 0.1 0.5 0.20.1 0 0 0 0University 0 0 0 0 0 0 0 0 0 0 0JPL 0.20.20.20.20.1 0.20.1 0 0 0 0SIM Data Analysis 0.35 0.35 0.35 0.85 1.2 1.7 1.7 1.7 2.5 2.5 2.5Postdoc 0 0 0 0.5 0.5 1 1 1 1.5 1.5 1.5University 0.1 0.1 0.1 0.1 0.20.20.20.20.25 0.25 0.25JPL 0.25 0.25 0.25 0.25 0.5 0.5 0.5 0.5 0.75 0.75 0.75Target Selection 0.3 0.3 0.3 0.4 0.8 0.6 0 0 0 0 0Postdoc 0 0 0 0.1 0.5 0.5 0 0 0 0 0University 0.1 0.1 0.1 0.1 0.2 0 0 0 0 0 0JPL 0.20.20.20.20.1 0.1 0 0 0 0 0Imaging Science 0 0 0 0 0.1 0.1 0.75 0.75 0 0 0Postdoc 0 0 0 0 0 0 0.5 0.5 0 0 0University 0 0 0 0 0 0 0 0 0 0 0JPL 0 0 0 0 0.1 0.1 0.2 5 5 0 0 0EPO 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.3 0.3 0.3 0.3Postdoc 0 0 0 0 0 0 0 0.2 0.2 0.2 0.2University 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1JPL 0 0 0 0 0 0 0 0 0 0 0Management 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1Postdoc 0 0 0 0 0 0 0 0 0 0 0University 0 0 0 0 0 0 0 0 0 0 0JPL 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1Totals 4.2 4.2 4.2 4.9 4.5 4.0 3.9 4.0 4.0 4.0 4.0Postdoc 2.5 2.5 2.5 3.2 2.5 2.3 2.2 2.3 2.3 2.3 2.3University 0.7 0.7 0.7 0.7 0.9 0.6 0.6 0.6 0.7 0.7 0.7JPL 1.1 1.1 1.1 1.1 1.1 1.1 1.1 1.1 1.1 1.1 1.1Imaging Science. A small group of scientistsled by T. Velusamy (JPL) will carry out thejet imaging experiment and support theanalysis of the disk imaging relevant to theastrometric observations.Education and Public Outreach. We haveallocated approximately 4% of the teambudget to enable team members to carry outthe efforts described in §III and to supportthe SIM project’s EPO team as requested.To ensure high visibility for this task, theeffort will be led by Deputy-PI M. Simon(SUNY) who has worked on a variety ofAAS-led activities in education.Management. There will be a small administrativeeffort at JPL to coordinate funding,meetings, reports to the project and toNASA, etc. This activity will include a parttimecontract monitor/financial analyst (~0.1FTE) to track funding and milestones onspecific work areas.26

E. ReferencesBaraffe, I. et al.1998 A&A, 337, 403.Bessell, M. S. and Brett, J. M., PASP, 1988, 1134.Black, D.C. and Scargle, J.D. 1982, ApJ, 263, 854.Boden, A. and Lane, B.F. 2000, preprint.Boden, A. et al. 1999, ApJ, 515, 356.Bouvier et al. 1995, A&A, 299, 89.Bouvier, J. and Bertout, C. 1989, A&A, 211, 99.Brown, R., Sozzetti, A., Casertano, S., 1999, “Simulations of Planet Searches with SIM,” STScI Internal Report,preprint.Burrows, C.J. et al. 1996 ApJ, 473, 437.Casey, B.W. et al. 1998, AJ, 115, 1617.Clarke, C. 2000, in Birth and Evolution of Binary Stars, Proc. of IAU 200, eds. B. Reipurth and H. Zinnecker, inpress.Cohen, M. and Kuhi, L. 1979, ApJ(S), 41, 743.Corino et al. 2000, in Birth and Evolution of Binary Stars, Proc. of IAU 200, eds. B. Reipurth and H. Zinnecker, inpress..Cumming, A, Marcy, G. W., Butler, R. P., 1999, ApJ, 526, 890.D’Alessio et al. 1999, ApJ, 527, 893.D’Antona F. and Mazzitelli, I. 1997, in Cool Stars in Clusters and Associations, eds. G. Micela and R. Pallavicini,Mem. S. A. It., 68, 807Dutrey, A. et al. 1998, A&A, 338, L63.Dutrey, A., Guilloteau, S., & Simon, M. 1994, A&A 286, 149.Fischer, D. 1998, Ph.D. Thesis.Freedman, W. et al. 1994, ApJ, 427, 628.Ghez, A.M. et al. 1995, AJ, 10, 753.Gomez et al. 1993, AJ, 105, 1927.Guilloteau, S. and Dutrey, A. 1998, A&A, 339, 467.Herbst et al. 1994, AJ, 108, 1906.Koerner, D.W. 1997, Origins of Life and Evolution of the Biosphere, 27, 157.Konigl, A., 1989, ApJ, 342, 208.Lada, C. J., and Adams, F. C. 1992, ApJ, 393,. 278.Levison, H.F., Lissauer, J.J., Duncan, M.J., 1998, AJ, 116, 1998.Lin, D. N. C. and Papaloizou, J. 1986, ApJ, 309, 846.Lin, D.N.C., J. C. B. Papaloizou, C. Terquem, G. Bryden, and S. Ida 2000, in Protostars and Planets IV, (ed. VinceMannings, A. P. Boss, and S.S. Russell), 1111.Mannings, V. and Sargent, A.I. 1997, ApJ, 490, 792.Madore, B. and Freedman, W. 2000, in preparation.Marcy, G. and Butler, P. 2000, PASP, 112, 137.Mathieu, R.D. 1994, ARA&A, 32, 465.Mazeh, T. 1999, Precise Stellar Radial Velocities, eds J.B. Hearnshaw and C.D. Scarfe, IAU Colloq. 170, 131.Mazeh, T. Goldberg, D. and Latham, D 1998, ApJ, 501,L199.Mazeh, T., Prato, L. and Simon, M. 2000, in Birth and Evolution of Binary Stars, Proc. of IAU 200, eds. B.Reipurth and H. Zinnecker, p. 22.Meyer, M. R, Calvet, N., Hillenbrand, L. A. 1998, AJ, 114, 288.Morbidelli, A. et al. 2000, preprint.Neuhäuser, R. 1999, priv. comm.Palla, F. and Stahler, S. 1999, ApJ., 525,772.Pollack, J., Hubickyj, O., Bodenheimer, P., Lissauer, J. J., Podolak, M. Greenzweig, Y. 1996, Icarus, 124, 62.Prato, L.A. 1998, PhD Thesis, SUNY-SB.Swenson et al. 1994, ApJ, 425, 286.Reipurth, B. et al. 2000, ApJ, in press.Saar, S. H., Butler, R. P., Marcy, G. W. 1998, ApJ, 498, L153.Schussler, P. Caligari, A. Ferriz-Mas, Solanki, S.K. and Stix, M., 1996, A&A, 314, 503.Shevchenko, V. S., and Herbst, W., AJ, 116, 1419.33

Shu, F., Najita, J., Ostriker, E., Shang, H., 1995, ApJ, 455, L155.Simon, M. et al. 1996, ApJ, 469, 890.Simon, M., Dutrey, A., and Guilloteau, S. 2000, ApJ, submitted.Skrutskie, M. F., Dutkevitch, D., Strom, S. E., Edwards, S., Strom, K. M. and Shure, M. A. 1990, AJ, 99, 1187.Stapelfeldt, K. et al. 1998; in Proceedings of Exo-Zodiacal Dust Workshop, eds. D. E. Backman, L. J. Caroff, S.A., Sandford, D. H. Wooden, NASA CP, p 289.Steffen et al. 2000, in Birth and Evolution of Binary Stars, Proc. of IAU 200, eds. B. Reipurth and H. Zinnecker,p.19.Thiébaut, E. et al. 1995, A&A, 304, L17.Trilling, D. E., Benz, W., Guillot, T., Lunine, J. I., Hubbard, W. B., Burrows, A. 1998, ApJ, 500, 428Trilling, D. E., Lunine, J., Benz, W. 2000, ApJ, in press.White, R.J. et al. 1999, ApJ, 520, 811.Wuchterl, G., Guillot, T., & Lissauer, J. 2000 in PPIV, eds. Mannings, Boss, & Russell, (Tucson: UofA press),1081.34