ebook img

The dynamic atmospheres of Mira stars: comparing the CODEX models to PTI time series of TU And PDF

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

Preview The dynamic atmospheres of Mira stars: comparing the CODEX models to PTI time series of TU And

Astronomy&Astrophysicsmanuscriptno.aa˙TUAnd5˙adapted (cid:13)c ESO2012 January30,2012 The dynamic atmospheres of Mira stars: comparing the CODEX models to PTI time series of TUAnd M.Hillen1,T.Verhoelst1,2,P.Degroote1⋆,B.Acke1⋆,andH.vanWinckel1 1 InstituutvoorSterrenkunde(IvS),K.U.Leuven,Celestijnenlaan200D,B-3001Leuven,Belgium e-mail:[email protected] 2 BelgianInstituteforSpaceAeronomy,Brussels,Belgium Received??,2011;accepted??,2011 2 1 ABSTRACT 0 2 Context.OurcomprehensionofstellarevolutionontheAGBstillfacesmanydifficulties.Toimproveonthis,aquantifiedunderstand- ingoflarge-amplitudepulsatoratmospheresandinterpretationintermsoftheirfundamentalstellarparametersareessential. n Aims.WewishtoevaluatetheeffectivenessoftherecentlyreleasedCODEXdynamicalmodelatmospheresinrepresentingM-type a Miravariablesthroughaconfrontationwiththetime-resolvedspectro-photometricandinterferometricPTIdatasetofTUAnd. J Methods.WecalibratedtheinterferometricK-bandtimeseriestohighprecision.Thisresultsin50nightsofobservations,covering 7 8subsequentpulsationcycles.Ateachphase,thefluxat2.2µmisobtained,alongwiththespectralshapeandvisibilitypointsin5 2 channelsacrosstheK-band.Wecomparedthedatasettotherelevantdynamical,self-excitedCODEXmodels. Results.Bothspectrumandvisibilitiesareconsistentlyreproducedatvisualminimumphases.Nearmaximum,ourobservationsshow ] thatthecurrentmodelspredictaphotospherethatistoocompactandhot,andwefindthattheextendedatmospherelacksH Oopacity. R 2 Sincecoverageinmodelparameter spaceiscurrentlypoor,moremodelsareneededtomakefirmconclusions onthecauseofthe S discrepancies.WearguethatforTUAnd,thediscrepancycouldbeliftedbyadoptingalowervalueofthemixinglengthparameter . combinedwithanincreaseinthestellarmassand/oradecreaseinmetallicity,butthisrequiresthereleaseofanextendedmodelgrid. h p Key words. Stars: AGB and post-AGB – Stars: oscillations – Stars: atmospheres – Stars: fundamental parameters – Techniques: - interferometric–Infrared:stars o r t s a 1. Introduction tionofthesourceofthisdiscrepancyishamperedbythelimited [ coverageinparameterspaceofthemodels,butwediscusspos- In spite of their great astrophysical importance (Muzzinetal. sible avenuesof furtherresearch,and stress the need forlarger 1 2009),manyaspectsofasymptoticgiantbranch(AGB)starsre- publishedmodelgridsandforobservationalcampaignscovering v main poorly understood (e.g. Kerschbaumetal. 2011, and ref- thevisualmaximumphaseindetail. 5 erences therein). Among the AGB stars, the luminous Mira 1 variables might be the most enigmatic with their long-period, 8 large-amplitudepulsationsandhighmass-lossrates.Inparticu- 2. Observations 5 larourunderstandingoftheiratmosphericstructure,whichisthe 1. linkbetweenenrichedstellar interiorandevolution-dominating TUAndwasextensivelyobservedaspartofthelargePTIMira 0 wind,isataturningpoint,thankstoadvancesonboththeobser- programme (Thompson 2002) during 60 nights from 1999 to 2 2005, over eight pulsation cycles. The publicly available PTI vationalside(IRinterferometry),andthemodellingsidewitha 1 data archive provides both squared visibility amplitudes (here- newgenerationofself-excitedatmospheremodels. : aftervisibilities)andintegratedphotoncounts.Wedescribethe v ThepublicreleaseoffourCODEXmodelseriesforM-type datareductionandcalibrationoftheseobservablesbelow. i MiravariablesbyIrelandetal.(2011, hereafterISW11),greatly X increasestheopportunitiesforthegeneralcommunitytoassess ar theperformanceofsuchdynamicmodels.Here,wepresentthe 2.1.Interferometry firstconfrontationoftheCODEXmodels(seeSect.3)toafully phase-resolved 2 µm spectro-photometric and interferometric The now dismounted Palomar Testbed Interferometer (PTI) dataset covering eight pulsation cycles, as obtained with the was a Y-shaped, long-baseline near-IR interferometer located Palomar Testbed Interferometeron the Mira TUAnd. Thisstar at Palomar Observatory and operated by JPL/Caltech. It con- hasaspectraltypeM5IIIe,aprogenitormassM =1.15−1.4M , sisted of three 40 cm apertures with separations of 85-110m ⊙ metallicity Z = 0.004 − 0.02 (Mennessieretal. 2001), and a and used in pairwise combination(see Colavitaetal. 1999, for GCVSperiodP∼317d(Samusetal.2009). a detailed description). PTI thus measured only one interfero- metricobservableperspectralchannel:thesquareofthe“fringe The most intriguing result of this comparison is that there contrast” or visibility amplitude. Visibilities were recorded in seems to be a systematic discrepancy between the models and fivechannelswithintheK-band(2.0-2.4µm).Processingofthe our observations around visual maximum: the star appears too rawdata(correctingfordetectorbiasesandread-outnoise,and hot,anditsouteratmospheretoodevoidofwater.Anidentifica- averagingthe125sscansintofringeamplitudes)wasautomati- ⋆ Postdoctoral Fellow, Fund for Scientific Research of Flanders cally done on site, following procedures described in Colavita (FWO) (1999). Calibrating for the system visibility, i.e. the response 1 Hillenetal.:ComparingtheCODEXmodelstoPTItimeseriesofTUAnd of the measurement system (interferometer + atmosphere) to phases(hereaftersimplyphases)werecomputedbyfittingasine a point source, was done using the nbCalib tool (Bodenetal. totencyclesaroundthetimesofthePTIobservations. 1998) provided by the NExScI which uses a weighted-mean formalism to estimate the system visibility at the time and po- 3. TheCODEXmodels sition of the science observation from all calibrator measure- ments that fit the user-defined temporal and spatial constraints The CODEX atmosphere models (Irelandetal. 2008) are the (see vanBelleetal. 1999, for more details). To ensure calibra- mostadvancedpubliclyavailabledynamicalmodelsforM-type tion stability and consistency, essential for making full use of Mira variables. The pulsationsin these models are self-excited thisdatasetasaninterferometrictimeseries,severalcalibrators and limited in amplitude by the non-lineareffects of shocks in wereobservedeachnightandwereintercomparedbynbCalib. theouteratmosphereandbyturbulentviscosity.Foreachmodel Thestellarparametersofourcalibratorsweredeterminedby series, a few typical cycles were selected for detailed radiative fitting a reddened Kurucz model atmosphere (Kurucz 1993) to transfer calculations with an opacity-sampling method in LTE high-qualityarchivalphotometry.Theerrorswereestimatedus- with a phase-resolution of ∼ 0.1. This results in the centre-to- ingMonteCarlosimulations.Table 1(onlyavailableelectroni- limb variation (CLV) and spectrum, with ∆λ = 0.0002µmbe- cally)liststhecalibratorsandtheiradoptedparameters.Allcal- tween λ = 0.45−2.50µm.The currentmodelseries are made ibrators have angular diameters well below the PTI resolution to match o Cet, R Leo, and R Cas, although even these well- limitof1mas(vanBelle&vanBelle2005)andare,onaverage, observedMiras have large uncertaintieson their basic parame- 10◦awayfromthesciencetarget. ters(e.g.the2verydifferentmodelsforRCas).Theywerepre- The calibrated visibilities of TUAnd from 50 nights were sented to the communityand comparedto photometricand in- retained,havingamedianrelativeprecisionrangingfrom5%at terferometricobservationsinISW11.Themodelparametersare 2.2µm,7.5%at2.4µmand10%at2.0µm. mass M, luminosity L, metallicity Z, mixing length parameter α , andturbulentviscosity parameterα . For a detailed justifi- m ν cationofthecurrentsetofparametervalues,werefertoISW11. 2.2.Spectro-photometry With a slightly meandering (± 5d) pulsation period of Togetherwiththeuncalibratedvisibilities,PTIdeliveredtheraw ∼317 d, TU And is compared to the R52 (M = 1.1M , L = ⊙ photoncountsofeachobservation,typicallyinblocksoffiveob- 5200L ,Z = 0.02,α = 3.5,α = 0.25)ando54(M = 1.1M , ⊙ m ν ⊙ servationstakenwithinminutesandinterleavingsciencetargets L = 5400L , Z = 0.02, α = 3.5, α = 0.25) model series, ⊙ m ν withcalibrators.WederivedthespectralshapeovertheK-band whichhavederivedperiodsof307and330d,respectively. andflux-calibratedthe2.2µmchannelwithabootstrapmethod. When comparing a model series to observations, the only The spectral shape calibration procedure consists of divid- free parameter is the distance, or equivalently, the phase- ing the science-target normalised photon counts by the ones averagedRosselandangulardiameterθ .Consideringtheangu- R of the calibrator and multiplying the result with the intrinsic larresolutionofPTI,weconvertedthemodelintensitiestosyn- spectral shape of the calibrator. To obtain an absolute flux in theticobservablesforθ ∈ [1.5,4.5]mas,instepsof0.05mas. R the 2.2 µm channel, one calibrator measurementon either side Convertingthemodelintensitiestothespectro-photometricob- of the science observation with respect to airmass is required servableswasdonewithEq.1.Thesquaredvisibilitiesarecom- to solve for the atmospheric extinction per unit airmass A = putedfollowingWittkowskietal.(2004): v (log[I /I ]+log[S /S ])/(m −m )andtheinstrumentspectral respon2se1R =S /(1I ×2exp(−2A m1)),withI theintrinsicandS λuk(ν(B/λ)·F)⊗Gk2λdλ tchaelibmreaatesdurflveudxflius1xt,hea1nndcamlciuthlaeteadviram1saIss,o=fSthie/iRthc×aleixbpra(Atorm.Th)ei, V2(B/λeff)= Rλl λuk(ν(0)·F)⊗Gk2λdλ (2) whereS denotesthemeasuredphostcoincouscnitsavndm thvesaciir- Rλl sci sci withF,G,λ,andλ asinEq.1,and massofthesciencetarget.Foreachobservingblock,theabove l u procedure is applied 5000 times to a randomly picked science ν(B/λ)= IµJ (πθ qB/λ)qdq (3) observation,correspondingrandomcalibrator observationsand Z λ 0 R calibratorintrinsicfluxes.Finally,thedistributionsofallobserv- ing blocks within a single night are co-added, and the average themonochromaticHankeltransformoftheintensityprofile Iµ λ andstandarddeviationaretakenfromtheresultingnormaldis- atthebaselineBandwavelengthλ. tribution,asthefinalcalibratedfluxanditserror. Toobtainthecalibratorintrinsicfluxes,thePTIopticalcon- 4. Results figuration was taken into account:after beam combinationand dispersionbyaprism,thewavefrontwassentthrougha broad- Figure 1 shows the resulting photometric time series of TU bandKBarrfilterontothedetectorpixels(Colavitaetal.1999). And asa functionof phase,with two consecutivecyclesof the EachoftheKuruczmodelsintheMonteCarlodistributionsob- R52 model series repeatedly overplotted. The two other avail- tained with the spectralenergydistributionfitting, is converted able cycles of this series give qualitatively identical results, so intotheeffectivefluxIineachofthePTIchannelsas: are not shown. The Rosseland angular diameter is chosen to λu[(M·F)⊗G]λdλ match the absolute flux at 2.2 µm, with a resulting value of I = Rλl (1) θR = 2.20± 0.15 mas or d = 1.13± 0.07 kpc. Although the λu[F⊗G]λdλ error on the absolute calibration in this channelis rather large, Rλl the contamination from upper atmospheric molecular layers at with M theKuruczmodel, F the filter profile,G a Gaussianof these wavelengths is smallest. This channel thus provides the FWHM =0.7µm(M.Colavita,privatecommunication),andλ, best view at the continuum photosphere in the spectral region l λ theedgesofthespectralchannelunderconsideration. wherethebulkoftheluminosityisemitted.Moreover,fixingthe u Finally,weconstructedavisuallightcurvebyretrievingob- angulardiameterlike thisgivesa consistentpicturein boththe servedV-bandphotometryfromtheAAVSOandAFOEV.Visual spectro-photometryandtheinterferometry. 2 Hillenetal.:ComparingtheCODEXmodelstoPTItimeseriesofTUAnd 1.1 Flux/Flux2.20001....8970 0.6 2.0µm 2.1µm 1.1 Flux/Flux2.200001.....68970 2.3µm 2.4µm Fig.1.Spectro-photometry.Thetwoupperpan- Flux (100 Jy)2.201121.....64028 ersloehlswsopsweehrscotpwtahtontehetealhbvetshao2reil.au2vttiieµosumnflaiulncxhflthauinnexnsteiphnlee,cmttlhraaaetgltetnshrih,itaruadpdneepdswa.tnihItenhel 6 8 each panel two cycles of the R52 model se- mag10 ries are consecutively overplotted as the solid V-12 line.Ontheleft,thecyclesareshownasafunc- 14 0 1 2 3 4 5 6 7 0.0 0.2 0.4 0.6 0.8 1.0 tionofvisualphasewhileontherighttheyare φV φV phase-folded. TheR52modelseriesshowsreasonableagreementwiththe observedvisuallightcurvein the lowerpanel,consideringthat 2.4µm 0.6 (1) the amplitude was not tuned to match this light curve and 2.3µm 2.2µm (2)non-LTEeffectsplayanimportantroleatthesewavelengths 0.4 2.1µm 2.0µm (Irelandetal.2008),causinganincreaseinvisualbrightnessof 0.5to1.0magnitude.Incasethehighervalueofthiscorrection 0.2 is valid, the perceived difference between model and observa- φV=0.0 φV=0.1 0.0 tionscould pointto the presenceof an excessin visuallightat 0.6 maximumphasestogetherwithatoolargemodelamplitude. 0.4 Thetwoupperpanelsshowthevariationinthespectralshape with time. A clear discrepancy occurs around times of visual 0.2 maximum, then the model predicts too much flux. Moreover, φV=0.2 φV=0.3 the amplitude of this excess shows an interesting wavelength- 0.0 0.6 dependence:thestrongesteffectisobservedat2.0µm,asmaller discrepancyat2.1and2.4µmandagoodagreementat2.3µm. 0.4 2V The visibility time series is presentedin Fig. 2, overplotted 0.2 with the same cycles of the R52 model series. Each panel dis- plays the visibilities at a given phase, using different symbols 0.0 φV=0.4 φV=0.5 for observationscoming from differentcycles (at least two per 0.6 panel). The model reproduces the observations well at phases 0.4 0.3-0.6,butnotaroundvisualmaximum.Comparisonwithother cycles does not give better results. In fact, this reveals that 0.2 the model has larger cycle-to-cyclevariationsat pre-maximum φV=0.6 φV=0.7 phases than are seen in the observations. A closer look at the 0.0 0.6 wavelengthdependenceshowsthatthediscrepancyisleastpro- nouncedat2.3and2.4µm.Between2.0and2.2µm,themodel 0.4 shows very little wavelength-dependence, unlike the observa- tionswhichshowacleardecreaseinvisibilitytowards2.0µm. 0.2 This observation is supported by the more quantitative as- 0.0 φV=0.8 φV=0.9 200 300 200 300 sessment presented in Figs. 3 and 4 (available electronically Spatial frequency (cycles/arcsec) only). Epochswith observationson at least two differentbase- Fig.2.Visibilitytimeseries.Eachpanelshowsallsquaredvisi- lineswereselected,toresolvethedegeneracybetweenCLVba- bilitiesasafunctionofspatialfrequency,obtainedatthevisual sicsizeandshape.Eachofthesedatasetswithgooduv-coverage phasegiveninthelowerleftcornerofthepanel.Differentsym- wascomparedtoeachmodel,independentlyofmodelphaseand forarangeofθ .Thefiguresshowtheresultingχ2-maps.Inall bols are used for observationsfrom different cycles. The same R two model cycles as in Fig. 1 are overplotted as the full and panelsacurvedvalleyshowsup:everydatasetcanbefittedwith dashedlinesineachpanel. amodelofanyphase,albeitwithasignificantlylargerdiameter for visual maximum models. However, within these valleys a better χ2 is typically foundwith a modelat the correct(photo- 5. Discussion metric)diameterbutwithaphasenearertovisualminimumthan theactualvalueoftheobservations.Inconclusion,wefindthat Irelandetal.(2008)demonstratethattheCODEX modelshave themodelsaroundvisualmaximumshouldresemblethoseofvi- reachedalevelofmaturity,concerningcomputationalissuesand sualminimummoreinCLVshapeandwavelength-dependence. inputphysics,that allowsquantitativecomparisonswith obser- Finally,wealsocomparedtheobservationstothesimilaro54 vations.Pulsationsareself-excitedandmolecularlayersarepro- model series, except for its luminosity. The results for o54 are ducednaturallyandinaphysicallyself-consistentway. alike and thus not shown here. The visibilities seem to match Nevertheless,ourobservationsofTUAndshowacleardis- slightlybetter,butthelightcurvediscrepanciesbecomelarger. crepancy with the models around phases of visual maximum, 3 Hillenetal.:ComparingtheCODEXmodelstoPTItimeseriesofTUAnd ter values that fits all observables. Accepting the LMC period- 1.8 3.5 luminosityrelationandassumingthattheeffectofmetallicityis φV=6.6 φV=7.8 3.0 insufficient(seeISW11),theonlyfreeparametersareM,αm,and 1.4 2.5 α .Sincethelastmainlyinfluencesthepulsationamplitude,but ν odel φV 1.0 12..50 2logχ ncloutdtehethpaetrtihoedporrotbhleem‘paartenhtanstdari’sroandeiuosf,tchaeliabbraotvinegauththeomrsixcoinng- M 0.6 1.0 length parameter as a function of mass. Decreasing the rather 0.5 0.2 0.0 high αm = 3.5 to a more ‘realistic’ value requires an increase −0.5 in mass or a decrease in metallicity (albeit with a smaller de- −0.21.6 2.0 2.4 2.8 3.2 3.6 4.0θR4. 4(m1a.6s)2.0 2.4 2.8 3.2 3.6 4.0 4.4 pendency)topreservetheperiod,whichwouldincaseofoCet be inconsistentwith its kinematicallydeterminedage. TU And Fig.3. The χ2 maps of two data sets with a good uv-coverage though, was classified as a member of a population with ini- (see text). The log χ2 is shown as a function of Rosseland an- tial masses in the 1.15 − 1.4M range and Z ∈ [0.004,0.02] ⊙ gular diameter and phase, of the same two model cycles as in (Mennessieretal. 2001), which makes it a better candidate to Figs. 1 and 2. The phase of the data set is given in the upper testtheabovehypothesis. rightcornerof the paneland is plottedas the horizontaldotted We can at this point not gauge the impact of using an α ν line. The vertical dotted line marks the photometric diameter. whichisnottunedtotheamplitudeofTUAnd.Asaresult,we The intersection of the dotted lines shows where the minimum feelthatthemodelgridshouldspansomescopeinthisparameter χ2shouldbeifthemodeliscorrect. toallowmorecomparisonswith‘small-amplitude’pulsators. Could either change in the model parametersalso induce a stabilizingeffectonthe H O layer,asrequiredbyourobserva- 2 bothinthespectrumandinthevisibilities.Themostnaturalin- tions,oristhereaneedtoaddmorephysicstothemodels? terpretation of the wavelength-dependence of this discrepancy seems to be in terms of the shape of the H O opacity curve, 2 whichshowsanincreasefrom2.2µm towards2.0and2.4µm. 6. Conclusions TheformationofH Olayersdependssensitivelyonthedensity, 2 The combination of a spectro-photometric and interferometric temperature,andpressurestratificationintheupperatmosphere. time series is a powerfultool for testing the validity of pulsat- In the models, the conditions in the upper atmosphere are set ing AGB model atmospheres, even at low spectral resolution. bythedynamics.Shockfrontsthatemergefromthecontinuum PTIobservedalargesampleofM-typeMirasandsemi-regulars, photosphere between phases -0.3 and -0.1 and then propagate severalwith a similar samplingto the one presentedhere,cov- outwards lead to the formation of an H O layer around phase 2 ering periods between 150 and 470d. To fully exploit this rich 0.2-0.3,whichthenstronglydecreasesinstrengthatphase0.7- database, better coverage in the model parameter space is ur- 0.8.Atφ = 0.5bothvisibilitiesandspectrashowgoodagree- V gentlyneeded,eventhoughitrequiresalargecomputationalef- ment.However,as demonstratedby Figs. 1, 2, and3, thenear- fort. Only in this way can the model parameters be calibrated maximumobservationsare representedbetter by modelphases and(simply)linkedtoMiraobservableproperties.Fromtheob- nearertominimum,andthuswiththepresenceofanH Olayer 2 servationalpointofview,interferometricstudieswithbetteruv- intheatmosphere.Thisinterpretationissupportedbythebetter coverageand spectral resolution should focus on phases 0.0 to agreementat2.3and 2.4µm. ThereCO is animportantsource 0.15, since none of the current model series predicts any ex- of opacity,in the photosphereand in the extendedatmosphere, tendedmolecularatmosphereinthatphaserange. evenatvisualmaximum.Asamuchmorestablemolecule,CO dependslessontheexactconditionsandpartlyhidesthelackof Acknowledgements. ThePalomarTestbed Interferometer was operated bythe anH2Omolecularlayerinthespectrumandvisibilities. NASA Exoplanet Science Institute and the PTI collaboration. It was devel- In retrospect, it is interesting to note that the image recon- opedbytheJetPropoulsionLaboratory,CaliforniaInstituteofTechnologywith structionofTLep(P∼380d, LeBouquinetal.2009),madeata funding provided from the National Aeronautics and Space Administration. Weacknowledge with thanks the variable starobservations fromthe AAVSO phaseof0.8-0.9,showstheclearsignatureofanextendedmolec- InternationalDatabasecontributedbyobserversworldwideandusedinthisre- ular layer with a brightness ratio of ∼10%, unlike its closest search. model series o54, which supports our findings. Woodruffetal. (2009) found fair agreement between the wavelength depen- dence of their measured 1.1-3.8 µm UDs and the predictions References of the CODEX models, for several Miras at phases 0.5 to 0.8, Boden, A. F., Colavita, M. M., van Belle, G. T., & Shao, M. 1998, in butneededslightlylargermodelcontinuumdiametersthanpre- PresentedattheSocietyofPhoto-OpticalInstrumentationEngineers(SPIE) dicted.Ontheotherhand,Wittkowskietal.(2011)haverecently Conference,Vol.3350,SocietyofPhoto-OpticalInstrumentationEngineers comparedAMBERmediumspectralresolutionobservationsof (SPIE)ConferenceSeries,ed.R.D.Reasenberg,872–880 Colavita,M.M.1999,PASP,111,111 four Miras to the CODEX models and found good agreement, Colavita,M.M.,Wallace,J.K.,Hines,B.E.,etal.1999,ApJ,510,505 althoughtheycomparedthemodelstostarsofadifferentperiod, Ireland,M.J.,Scholz,M.,&Wood,P.R.2008,MNRAS,391,1994 henceluminosity. Ireland,M.J.,Scholz,M.,&Wood,P.R.2011,MNRAS,1633 In summary,atvisualmaximumthe currentmodelspredict Kerschbaum,F.,Lebzelter,T.,&Wing,R.F.,eds.2011,AstronomicalSociety ofthePacificConferenceSeries,Vol.445,WhyGalaxies CareaboutAGB acontinuumphotospherethatistoocompactandhotasalready StarsII:ShiningExamplesandCommonInhabitants suggestedbyISW11,incombinationwithtoosmallanH Oop- 2 Kurucz,R.L.1993,VizieROnlineDataCatalog,6039,0 ticaldepthintheouteratmosphere,asfoundhere. LeBouquin,J.-B.,Lacour,S.,Renard,S.,etal.2009,A&A,496,L1 The basic question now is, can this be salvaged within the Mennessier,M.O.,Mowlavi,N.,Alvarez,R.,&Luri,X.2001,A&A,374,968 Muzzin,A.,Marchesini,D.,vanDokkum,P.G.,etal.2009,ApJ,701,1839 currentmodellingframework?Isitjustaquestionoftuningthe Samus,N.N.,Durlevich,O.V.,&etal.2009,VizieROnlineDataCatalog,1, parameterstomorerealisticvalues?Followingthediscussionof 2025 ISW11, it seems difficult to find any combination of parame- Thompson,R.R.2002,PhDthesis,UNIVERSITYOFWYOMING 4 Hillenetal.:ComparingtheCODEXmodelstoPTItimeseriesofTUAnd vanBelle,G.T.,Lane,B.F.,Thompson,R.R.,etal.1999,AJ,117,521 vanBelle,G.T.&vanBelle,G.2005,PASP,117,1263 Wittkowski,M.,Aufdenberg,J.P.,&Kervella,P.2004,A&A,413,711 Wittkowski,M.,Boboltz,D.A.,Ireland,M.,etal.2011,A&A,532,L7 Woodruff,H.C.,Ireland,M.J.,Tuthill,P.G.,etal.2009,ApJ,691,1328 5 Hillenetal.:ComparingtheCODEXmodelstoPTItimeseriesofTUAnd,OnlineMaterialp1 Table1.Thederivedcalibratorparameters. Calibrator SpT T (K) UDDiameter(mas) eff HD166 K0V 5540±90 0.61±0.05 HD1404 A2V 8840±175 0.47±0.02 HD2628 A7III 7200±50 0.48±0.05 HD3651 K0V 5250±50 0.74±0.06 HD6920 F8V 6350±200 0.61±0.04 HD7034 F0V 7600±450 0.49±0.04 HD7229 G9III 5550±150 0.80±0.12 HD9712 K1III 4750±240 0.78±0.20 1.8 3.5 φV=1.8 φV=3.1 1.4 3.0 1.0 2.5 0.6 2.0 0.2 φV Model −10..82 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 1.5 2logχ φV=3.2 φV=3.3 1.0 1.4 1.0 0.5 0.6 0.0 0.2 −0.5 −0.21.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 θR (mas) Fig.4.SimilartoFig.3,butforthefourotherdatasetswithgood uv-coverage. The log χ2 is shown as a function of Rosseland angulardiameterandphase,ofthesametwomodelcyclesasin Figs. 1 and 2. The phase of the data set is given in the upper rightcornerof the paneland is plottedas the horizontaldotted line. The vertical dotted line marks the photometric diameter. The intersection of the dotted lines shows where the minimum χ2shouldbeifthemodeliscorrect.

See more

The list of books you might like

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