Astron. Nachr./AN0000,No.00,1–13(0000)/DOIpleasesetDOI! Finding the most variable stars in the Orion Belt with the All Sky Auto- mated Survey Jose´ A.Caballero⋆,M.Cornide,andE.deCastro Departamento de Astrof´ısica y Ciencias de la Atmo´sfera, Facultad de F´ısica, Universidad Complutense de Madrid, E- 28040Madrid,Spain 0 Received05Sep2009,accepted04Jan2010 1 Publishedonlinelater 0 2 Keywords stars:variables:general—stars:pre-mainsequence—stars:oscillations—astronomicaldatabases:mis- n cellaneous—Galaxy:openclustersandassociations:OriOB1b a J Welookforhigh-amplitudevariableyoungstarsintheopenclustersandassociationsoftheOrionBelt.Weusepublic 5 datafromtheASAS-3PhotometricV-bandCatalogueoftheAllSkyAutomatedSurvey,infraredphotometryfromthe 2MASSandIRAScatalogues,propermotions,andtheAladinskyatlastoobtainalistofthemostvariablestarsinasurvey ] areaofside5degcentredonthebrightstarAlnilam(ǫOri)inthecentreoftheOrionBelt.Weidentify32highly-variable R stars,ofwhich16hadnotbeenreportedtovarybefore.Theyaremostlyvariableyoungstarsandcandidates(16)and S backgroundgiants(8),buttherearealsofieldcataclysmicvariables,contactbinaries,andeclipsingbinarycandidates.Of . theyoungstars,whichtypicallyareactiveHerbigAe/BeandTTauristarswithHαemissionandinfraredfluxexcess, h wediscoverfournewvariablesandconfirmthevariabilitystatusofanothertwo.Someofthembelongtothewellknown p σOrioniscluster.Besides,sixoftheeightgiantsarenewvariables,andthreearenewperiodicvariables. - o r (cid:13)c 0000WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim t s a [ 1 Introduction denemissionlinesareothersignpostsusedtoconfirmvery young ages in stars. Photometric variability in young stars 1 TheOriOB1bsubgroupassociation,mostlycoincidentwith ismostlyduetodark(cool)andbright(hot)spotsandflare v 2 theasterismoftheOrionBelt,containsarichpopulationof activity,whicharerelatedtothestrengthenedmagneticac- 6 veryyoungstars(e.g.Haro&Moreno1953;Walker1969; tivyduetofastrotation,andinteractionswithcircumstellar 6 Warren & Hesser 1978; Brown et al. 1994; Sterzik et al. discs,suchasintense,variableaccretionepisodesoroccul- 0 1995;deZeeuwetal.1999;Bricen˜oetal.2001).Theyoungest tations(e.g.classicalworks:Joy1945;Walker1956;Herbig . 1 regions, including Alnitak (ζ Ori), which illuminates the 1977 – theoretical works: Koenigl 1991; Collier Comeron 0 Flame Nebula (NGC 2024), and σ Ori, which illuminates &Campbell1993;Shuetal.1994–modernobservational 0 the Horsehead Nebula (Barnard 33 and IC 434), are to the works: Bouvier et al. 1993; Herbst et al. 1994; The´ et al. 1 : east of the Belt and have ages of 1 to 5Ma. Here lies the 1994;Hartmann&Kenyon1996;Hamiltonetal.2001;Eiroa v σ Orionis cluster, one of the most suitable sites for dis- etal.2002). i X covering and characterising substellar objects (Be´jar et al. TheOrionNebulaClusterintheOriOB1dsubgroupas- r 1999; Zapatero Osorio et al. 2002; Sherry et al. 2004; Ca- a balleroetal.2007).ThepopulationofstarssurroundingAl- sociationhasbeenatraditionaltargetforstudyingthepho- tometric variability of young stars (Jones & Walker 1988; nilam (ǫ Ori) and Mintaka (δ Ori), to the west of the Belt, Carpenteretal.2001;Herbstetal.2002;Stassunetal.2006; are older (5–10Ma) and spread over a wider area than the Parihar et al. 2009), but the surveys for variability in the muchdenserσ Orioniscluster(Caballero&Solano2008). OrionBeltarerelativelyscarce.Themostthoroughvariabil- SurroundingtheBelt,thereisanotherdispersedpopulation ity surveys in the Ori OB1 b association have been aimed of 10–20Ma-old stars, mostly associated to the Ori OB1a atthelow-masspre-mainsequencestarsandbrowndwarfs subgroup association (Jeffries et al. 2006; Caballero 2007) (Bailer-Jones & Mundt 2001; Zapatero Osorio et al. 2003; and its agglomerates, such as the 25 Orionis group or the Caballeroetal.2004,2006;Scholz&Eislo¨ffel2004,2005; ηOrionisoverdensity(Bricen˜oetal.2007;Caballero&Di- Bricen˜oetal.2005;Scholzetal.2009),whilethebrightest nis2008). variable stars were first recognised as early as the begin- HαandX-rayemissionandinfraredexcessarethepri- ningofthe20thcenturyandhavenotreceivedquiteatten- mary features for the identification of very young stars as tion.Onlyafew(multiple)young starsofimportantastro- those in the Orion Belt. Optical and near-infrared variabil- physical interest had been extensively monitored, such as ity,cosmiclithiumabundances,surroundingjets,orforbid- VV Ori (Miller Barr 1904; Struve & Luyten 1949; Terrell ⋆ Correspondingauthor:e-mail:[email protected] et al. 2007), σ Ori E (Walborn & Hesser 1976; Landstreet & Borra 1978), or Mintaka AE–D (Stebbins 1915; Koch (cid:13)c 0000WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim 2 Caballeroetal.:ThemostvariableASASstarsintheOrionBelt 1 0.6 0.5 0.5 0 g] ma 0.4 0) −0.5 3 [ 0 − 20 −1 AS 0.3 δ (J −1.5 V) AS00.2 σ( −2 0.1 −2.5 0 86.5 86 85.5 85 84.5 84 83.5 83 82.5 82 7 8 9 10 11 12 13 14 15 α (J2000) V ASAS−3 [mag] 0 Fig.1 Spatiallocationofthe10029ASASsourcesinthe Fig.2 Diagram showing σ(V ) vs. V . ASAS sources 0 0 studiedareawithlightcurveswithmorethan40datapoints, above the dashed line are variable star candidates in this markedwith(blue)smalldots.Northisup,eastistotheleft. work. X Ori is out of limits (V ∼ 12.7mag, σ(V ) ∼ 0 0 The100ASASsourceswithpossiblevariablelightcurves, 1.4mag). thethreeOrionBeltsupergiantstars(Alnitak,Alnilam,and Mintaka, from east to west), and σ Ori are marked with fornextphasesoftheanalysis.TheASASdatacoveredal- (red) filled circles, big stars, and a small star, respectively. mostsixyearsduringthefirstdecadeofthe21stcentury. Note the blank strip where ASAS did not survey (between Of the five available apertures (0, 1, 2, 3, 4), the “0” Alnitak and σ Orionis’ declinations) and the overlapping aperturephotometryseemedtoprovidebetterdata(i.e.less between different ASAS pointings (at about the Mintaka’s dispersion and number of grade “C” and “D” data points declination). –probablyuselessdataorwithoutphotometricvalueatall duetobadpixels,frame-edgeandbrightstar/planetproxim- & Hrivnak 1981). Many of the intermediate-mass T Tauri ity,orexcessivefaintness).AccordingtoPojman´ski(2002), starsintheOrionBeltwithvariabilitydenominations (e.g. thefiveaperturescorrespondtodiametersizesoftheaper- BG Ori, V505 Ori) were compiled by Fedorovich (1960) ture photometry from 2 to 6pixels; the “0” aperture corre- and have never been monitored again. After a century of spondstothenarrowerphotometry(Pojman´ski,priv.comm.). photometricsearchesintheregion,therearestillserendip- Fromthispointon,weusedonlytheJohnsonV magnitude itous discoveries of bright, UX Orionis-, A-type, variable withtheASAS“0”aperturephotometry,whichweindicate stars(Caballeroetal.2008). withthesymbolV . 0 Our aim is to find the most variable stars in the Orion Fig. 2 illustrates the selection of variable stars. It is a Belt, giving especial emphasis to the young Herbig Ae/Be σ(V )vs.V diagraminwhichthevariablestarcandidates 0 0 (HAeBe) and T Tauri stars. To accomplish that, we have havelargervaluesofstandarddeviationσ(V )withrespect 0 usedtheASAS-3PhotometricV-bandCatalogueoftheAll to non-variable stars of the same average magnitude V . 0 SkyAutomatedSurvey(ASAS)1,whichis“alowcostproject ASASstarsbrighterthanV ≈8.5magareaffectedbysat- 0 dedicatedtoconstantphotometricmonitoringofthewhole urationinthedetectorandwerenotconsidered.Thelower availablesky”(Pojman´ski1997,2002). envelopeatthenon-saturated,brightestendofthecloudof data points in the diagram indicates the best photometric accuracy of the system, which is of about 0.02mag. The 2 Analysis obvious rise of the lower envelope at faint magnitudes is a combinationofthestandarddeviations(root-mean-squares) First,wecompiled63276ASASlightcurvesofobjectsin from Poisson noise in the targets and from sky noise in asquareofside5.0degcentredonAlnilam;seeFig.1.We the photometric aperture (e.g. Irwin et al. 2007). The line used the default parameters N > 4 data points and r < usedastheselectioncriterionisthelowerenvelopeshifted 15arcsec (N is the number of data points within a circle tolargerσ(V )sbymagnitude-dependentamounts(roughly ofradiusrcentredonthecoordinatesofanASASsource). 0 0.06magatV ∼8.5–11.5mag,0.12magatV ∼12.5mag, Ifnotforcertainblankstrip,thesurveyedareawouldhave 0 0 and0.18magatV ∼13.5mag).Thischoicewasrelatively been 25deg2. To maximise the scientific return and min- 0 arbitrary: more or less restrictive selection criteria would imise possible further troubles with light curves of poor have given significantly less or more variable star candi- quality, we discarded all ASAS light curves with N ≤ 40 dates,respectively.Thus,theusedselectioncriterionwasa datapoints.Itleft10029lightcurves(16%)withN >40 compromisetominimisethenumberoffalsepositives(vari- 1 http://www.astrouw.edu.pl/asas/. ablestarcandidatesthatactuallydonotvary)andmaximise (cid:13)c 0000WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim www.an-journal.org Astron.Nachr./AN(0000) 3 Table1 IdentifiedvariablesintheOrionBelt. No. α2MASS δ2MASS V0 σ(V0) δV0 No⋆bs Spectral Name Class Variable (J2000) (J2000) [mag] [mag] [mag] type 01 052859.57 –033352.3 13.621 0.412 0.055 190 ... V1159Ori Dwarfnova Known 03 053101.46 –032324.3 12.492 0.341 0.056 294 ... IRAS05285–0325 Giant New 06 053322.45 –030955.7 11.185 0.085 0.032 207 ... No.06 Unknown New 07 053748.79 –030748.7 10.780 0.101 0.021 286 ... IRAS05353–0309 Giant New 11 053209.94 –024946.8 11.581 0.524 0.032 191 F5–8Vpe RYOri HAeBe Known 12 053833.68 –024414.2 12.885 0.368 0.054 187 K4e TXOri TTauri Known 13 053825.87 –024351.2 13.059 0.356 0.054 142 K3e TYOri TTauri Known 14 054354.26 –024335.4 11.628 0.135 0.032 222 ... 2M054354–0243.6 Contactbinary Known 15 053939.99 –024309.7 13.246 0.246 0.054 197 ... RWOri TTauri Known 18 053848.04 –022714.2 13.340 0.295 0.053 224 K7e+M: Mayrit528005AB TTauri New 21 053620.91 –021057.6 12.607 0.192 0.056 252 F3e PQOri HAeBe Known 29 053738.67 –014616.6 12.752 1.441 0.054 179 M8–9III: XOri Giant Known 30 053756.59 –014050.3 10.734 0.250 0.021 273 ... IRAS05354–0142 Giant New 33 053002.89 –013005.3 12.786 0.263 0.056 176 ... No.33 Unknown New 36 053425.82 –012106.5 12.667 0.325 0.055 170 ... V469Ori TTauri Known 41 052653.53 –010902.3 13.478 0.480 0.057 144 ... KisoA–0903135 TTauri? New 43 053759.04 –010552.8 13.554 0.371 0.054 200 ... V993Ori TTauri? Known 46 053126.41 –005833.1 10.454 0.076 0.021 192 F0 HD290509 Unknown New 48 053945.65 –005550.9 10.934 0.075 0.021 203 ... GSC04767–00071 TTauri? Known 51 053624.29 –004212.0 13.399 0.388 0.056 227 K3e PUOri TTauri Known 53 053448.91 –003716.7 13.384 0.298 0.056 178 ... V472Ori TTauri? Known 54 052724.67 –003511.2 11.031 0.134 0.033 184 ... IRAS05248–0037 Giant New 55 053543.27 –003436.7 12.411 0.278 0.057 246 K4e StHA48 TTauri New 60 052634.28 –001927.2 12.353 0.137 0.056 181 ... No.60 Giant New 65 053628.55 +000445.7 11.987 0.131 0.056 257 ... No.65 Unknown New 66 054329.25 +000458.9 11.459 0.513 0.032 202 F0? GTOri Giant Known 70 054418.80 +000840.4 8.906 0.061 0.019 204 A7IIIe V351Ori HAeBe Known 72 054611.86 +003225.9 13.390 0.439 0.053 167 ... VSSVI–32 Unknown New 73 054159.81 +003527.1 12.698 0.435 0.053 185 ... No.73 Unknown New 74 053642.64 +003834.3 11.010 0.101 0.033 254 A2 HD290625 HAeBe New 75 053347.85 +005536.3 10.433 0.166 0.021 181 ... IRAS05312+0053 Giant New 78 052817.85 +011006.1 11.219 0.165 0.033 173 F8–K0e StHA40 TTauri Known the number of actual variable stars, but always keeping a correspond to the best and mean data, respectively2. Be- maneagablenumberofobjectstobefollowedup. cause of the large ASAS pixel size (of about 15µm) and typicalfullwidthathalfmaximum,ofnolessthan1.4pixels Therewere100ASASlightcurvesabovetheselection (Pojman´ski2002),wediscardedallvisualbinariesofroughly criterion in Fig. 2. We were able to identify the 2MASS thesamebrightnessseparatedbylessthanabout20arcsec. (Skrutskie et al. 2006) and USNO-B1 (Monet et al. 2003) Besides, we also discarded relatively faint variable candi- possiblestellarcounterparts of99ofthem.Theunidenfied datesatlessthan1arcmintoverybrightstars,includingAl- ASASsourcewas054200–0154.7(V =13.167mag,σ(V ) 0 0 nilam,Mintaka,andσOri,whichareseenbythenakedeye =0.245mag,N =49),whichisactuallyabrightspotin obs and lead to a high sky background. After these removals, NGC2024.Forthevisualidentification,weusedtheAladin wekept32variablestarswhoselightcurveswerelikelynot interactiveskyatlas(Bonnareletal.2000).Theactualnum- contaminatedbyotherstars.The32lightcurvesareshown ber of stars with light curves was 78, because there were inFigs.A1toA4. 21 variable star candidates with two corresponding light curves.Theseparateddatasets(withidenticalassociatedco- The identification number, 2MASS coordinates, aver- ordinates and magnitudes within error bars) correspond to ageV0 magnitude,standarddeviationσ(V0),meanerrorof thesamestarindifferentobservedfields(Pojman´ski2002). magnitudeδV0,andnumberoffinallyusedlightcurvedata points, N⋆ , of the 32 identified variable stars are listed WerevisitedtheASASdataforthe78starsbyretrieving obs in Table 1. The ratios between the standard deviations and alldatapointswithqualitygrades“A”and“B”inacircleof mean errors vary between ∼2.5 and 27; the expected ratio radiusr=15arcseccentredonthestars,andmergingthem intoauniquelightcurve.“A”and“B”qualitygrades,which 2 Thequalitygrades“A”and“B”ofanewASASframedependson aretabulated by the ASAS catalogue and derived fromthe theaveragedispersionbetween‘old’and‘new’magnitudesforstarsinthe averagephotometricqualityoftheframeforeachaperture, range8–11mag(Pojman´ski,priv.comm.). www.an-journal.org (cid:13)c 0000WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim 4 Caballeroetal.:ThemostvariableASASstarsintheOrionBelt Table2 Identifiedvariableswithpowerspectrummaxima 1 largerthan500mag2. 0.9 No. Power P Name Remarks 0.8 [mag2] [d] s0.7 ar 01 767.19 44.345 V1159Ori P =44.2–46.8d st0.6 of 03 1785.4 78.226 IRAS05285–0325 Newperiodic on 0.5 11 632.39 600.38 RYOri Noclearpeak acti0.4 29 8779.7 416.05 XOri P =424.2±1.8d Fr0.3 30 913.68 373.31 IRAS05354–0142 Newperiodic 66 1671.2 920.32 GTOri Pulsating 0.2 0.1 0 2 101 102 103 104 Power spectrum maximum [mag2] 1.8 Fig.3 Cumulative distribution function of power spec- 1.6 trum maxima of the light curves of the 32 identified vari- 1.4 ablesintheOrionBelt.Starstotherightofthedashedver- g] 1.2 ticallineatpowerspectrumof500mag2 areclassifiedhere ma asprobableperiodicvariablestars. K [s 1 J− 0.8 0.6 for a non-variable star should be lower than 2 (e.g. Bailer- 0.4 Jones&Mundt2001).InTable1,wealsoprovidethespec- 0.2 traltypefromtheliterature(ifithaseverbeengiven),name, variabilityclass(seeSection3),andifthevariabilitystatus 0 −2 0 2 4 6 8 10 12 V −J [mag] wasknownornot.Forthosestarsthathaveneverbeenmen- 0 tioned in the literature, we use the nomenclature “No. X”, Fig.4 Optical-infraredcolour-colourdiagram(J−K vs. s whereXistheidentificationnumber. V −J).Filled(red)stars:newvariableyoungstars;filled 0 ForatypicalnumberoflightcurvedatapointsofN⋆ ∼ obs (red) circles: previously known variable young stars; filled 200 and a temporal coverage of almost six years, one may (red) squares: variable young star candidates; open (blue) derivethattherewasanASASvisitevery∼11dforeachof circles:variablegiants;open(blue)up-triangle:dwarfnova the32stars.Actually,thetypicaltimeintervalbetweencon- (V1159 Ori); open (blue) down-triangle: contact binary secutive ASAS visits was shorter because there were long (2M054354–0243.6); open (blue) squares: other new vari- non-observinggapswithoutdata.Themediantimeinterval ablestars. betweenconsecutiveASASvisitswas2.0–3.0d,depending on the star. Approximately 15–20, 20–33, 65–85, and 85– 95%ofthetimeintervalswereshorterthan1,2,5,and10d, average non-colour corrected flux densities with qualities respectively,andonlyinfourcases,coincidentwiththefour “3”or“2”(highandmoderatequalities,respectively).While Australsummers,thetimeintervalswerelongerthan100d. allthe32starshaveanear-infraredcounterpartinthe2MASS Weperformedarutinarytime-seriesanalysisonthe32 catalogue, only 12 stars were identified in the IRAS cata- lightcurvesbyapplyingtheScargle(1982)procedure.Fig.3 logues (e.g. IRAS catalogue of point sources, version 2.0; showsthecumulativedistributionfunctionofthemeasured IPAC 1986). Furthermore, only one star, the A7IIIe Her- maximaofpowerspectrum.Roughly85%ofthestarshave big Ae/Be star V351 Ori, has high/moderate quality mea- powerspectrummaximalowerthan500mag2,wherethere surements at the four IRAS passbands at 12, 25, 60, and seems to be an inflection point in the cumulative distribu- 100µm.Acolour-colourdiagramcombiningASASoptical tion function. The other 15% (6) of the stars with larger and2MASSnear-infraredmagnitudesisshowninFig.4. powerspectrummaximaaregiveninTable2andwereclas- Besides,wealsocollectedpropermotionsfromTycho-2 sifiedbyusasprobableperiodicvariablestars.Weprovide (Høgetal.2000),PPMX(Ro¨seretal.2008),andUSNO-B1 periods for two previously known periodic variable stars forthe32stars.Giventhatonlyafewofthemhaveappre- (V1159 Ori and X Ori) consistent with those in the liter- ciable proper motions, we do not list them, but are men- ature, and found probable periods for two previously un- tionedinSection3whenappropriated. knwon variable giant stars (with IRAS denomination), but Finally, only 13 stars have variable star designations, failedtofindanyreliableperiodictyforRYOriandGTOri. and there were suspects of variability for other three stars. ThesixofthemwillbestudiedindetailinSection3. Besides,animportantfractionofthetargetshavealsobeen Weperformedacompilationofadditionalinfrareddata tabulated as variable stars of the Southern Hemisphere by by collecting JHK magnitudes from 2MASS and IRAS Pojman´ski (2002), to which our work complements. Next, s (cid:13)c 0000WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim www.an-journal.org Astron.Nachr./AN(0000) 5 Table3 Infraredphotometryofidentifiedvariables. No. J H Ks F12 F25 F60 F100 [mag] [mag] [mag] [Jy] [Jy] [Jy] [Jy] 01 13.817±0.027 13.781±0.046 13.675±0.050 ... ... ... ... 03 6.097±0.019 5.083±0.027 4.639±0.021 1.08±0.08 0.39±0.04 <3 <20 06 8.295±0.034 7.552±0.024 7.368±0.026 ... ... ... ... 07 5.590±0.021 4.665±0.076 4.222±0.033 1.09±0.07 0.47±0.08 <0.9 <30 11 9.444±0.023 8.885±0.055 8.277±0.031 0.75±0.09 0.71±0.09 <2 <20 12 10.131±0.026 9.280±0.024 8.666±0.024 0.38±0.04 0.50±0.07 <0.7 <40 13 10.445±0.027 9.726±0.024 9.311±0.028 ... ... ... ... 14 10.240±0.026 9.943±0.026 9.816±0.021 ... ... ... ... 15 10.647±0.027 9.920±0.023 9.530±0.019 ... ... ... ... 18 10.156±0.023 9.463±0.026 9.187±0.019 ... ... ... ... 21 11.303±0.022 11.065±0.028 10.979±0.023 ... ... ... ... 29 2.322±0.292 1.379±0.306 0.861±0.346 96±4 49±7 6.0±0.9 <20 30 5.152±0.018 4.608±0.076 3.972±0.320 2.0±0.2 0.63±0.07 <2 <40 33 11.340±0.024 10.878±0.025 10.756±0.021 ... ... ... ... 36 10.655±0.023 9.647±0.030 8.759±0.021 ... ... ... ... 41 11.547±0.027 10.652±0.027 9.898±0.019 ... ... ... ... 43 10.846±0.027 9.955±0.024 9.355±0.027 ... ... ... ... 46 9.457±0.023 9.247±0.023 9.133±0.021 ... ... ... ... 48 8.589±0.023 7.933±0.055 7.797±0.027 ... ... ... ... 51 10.536±0.023 9.551±0.033 8.764±0.019 <0.4 0.38±0.05 <3 <20 53 10.860±0.022 10.043±0.026 9.538±0.019 ... ... ... ... 54 5.206±0.017 4.225±0.202 3.911±0.258 1.48±0.09 0.49±0.04 <5 1.7±0.3 55 10.243±0.021 9.509±0.033 9.061±0.019 ... ... ... ... 60 7.508±0.023 6.596±0.029 6.280±0.021 ... ... ... ... 65 10.762±0.024 10.447±0.025 10.054±0.019 ... ... ... ... 66 9.802±0.027 9.043±0.025 8.219±0.026 0.87±0.13 1.02±0.17 <5 <40 70 7.950±0.020 7.504±0.040 6.846±0.026 1.16±0.07 4.1±0.2 26±3 21±3 72 10.936±0.024 10.046±0.023 9.448±0.019 ... ... ... ... 73 10.448±0.023 9.818±0.023 9.351±0.019 ... ... ... ... 74 10.459±0.024 10.342±0.024 10.317±0.023 ... ... ... ... 75 6.304±0.026 5.375±0.027 5.091±0.020 0.47±0.03 <0.2 <0.4 <20 78 9.422±0.023 8.825±0.040 8.366±0.018 0.41±0.06 0.67±0.06 0.47±0.05 <1.1 weclassifythe32variablestarsbasedonspectroscopic,as- Besides,sixofthestarsinthisclassdisplayX-rayemission, trometric,andphotometriccriteria. andothersix,infraredfluxexcess.OnlythethreeσOrionis members are known to simultaneously display Hα and X- rayemissionsandinfraredfluxexcess. 3 Identifiedvariablestars RYOri(No.11). Itisawellinvestigated,emission-line 3.1 HAeBeandTTaurivariablestars staroftheOrionpopulationattheboundarybetweenHAeBe 3.1.1 Previouslyknownvariableyoungstarsand andTTauristars(e.g.Haro&Moreno1953;Herbig&Bell candidates 1988; Yudin & Evans 1998; Eiroa et al. 2001; Mora et al. 2001;Herna´ndezetal.2004;Acke&vandenAncker2006). This class is mostly made of probable accreting T Tauri Itsphotometricvariabilitywasdetectedveryearly(Picker- stars with intense Hα emission. Of the nine Hα emitter ing1904).AccordingtoBibo&The´(1991),itisaUXOri- stars,threebelongtotheσOrioniscluster(TXOri,TYOri, typestarwithacolourreversalinthecolour-magnitudedi- RW Ori), three have been associated to the Alnilam and agram, indicative of an edge-on disc. The disc of RY Ori Mintakayoungstarpopulations(V469Ori,PUOri,V472Ori),isalsodetectedbecauseofitslong-time-knowninfraredex- andthreehavenotbeenlinkedtoanyyoungregioninpartic- cess(Cohen1973;Glass&Penston1974;Weintraub1990; ularbutarebrightstarsattheboundarybetweenHAeBeand Weaver & Gordon 1992). Its photometric variability in the TTauristars(RYOri,V351Ori,StHa40–twoofthemmay ASAS lightcurve is obvious, with small mean photometric haveedge-ondiscs).Thetruenatureoftheotherthreenon- errorswithrespecttothepeak-to-peakamplitudefrom10.8 Hα emitter stars (PQ Ori, V993 Ori, GSC 04767–00071) to 13.0mag (δV = 0.032mag). RY Ori is one of the six 0 shouldbeinvestigatedwithfurtherspectroscopicanalyses. identified variable stars with power spectra maxima larger www.an-journal.org (cid:13)c 0000WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim 6 Caballeroetal.:ThemostvariableASASstarsintheOrionBelt ima at about V ∼ 12.5 and 14.0mag. Remarkably, the 0 700 TX Ori light curve displayed an apparent dimming for he- liocentricJulianday3 HJD>3500. 600 500 RW Ori (Mayrit 931117, No. 15). It is another mem- 2g] berintheσ Orioniscluster,butitismuchlessknownthan a400 m thestarsabove.ItsphotometricvariabilityandHαemission er [ werefirstdetectedbyPickering(1906)andHaro&Moreno w300 o P (1953), respectively. Afterwards, it has been only investi- 200 gatedbyFranciosinietal.(2006–withXMM-Newton,Her- na´ndezetal.(2007–withSpitzer),andCaballero(2008).Its 100 ASASlightcurvevariedbetweenV ∼12.8and14.0mag 0 (theoutlierdatapointisprobablyunreliable). 0 10−1 100 101 102 103 104 Period [d] PQOri(No.21). Inthediscoverypaper,McKnelly(1949) 10.5 classified the star as a rapid, irregular variable with an ob- servedrangeofvariabilityof12.5–13.5mag(photographic 11 magnitude).Althoughitslightcurvedisplayedmanymax- ima and minima, he was unable to determine “any sem- 11.5 blance of periodicity”. Herbig & Kameswara Rao (1975), g] Cohen & Kuhi (1979), and Herna´ndez et al. (2004) mea- a m 12 suredF0,F5:,andF3.0±1.5spectraltypesandobservedno [0 V emissionlines.However,whilethefirstauthoursclassified 12.5 itasastarintheOrionpopulation,Herna´ndezetal.(2004) didnotfindsignificantdifferencesfromF3main-sequence 13 stars nor near infrared excess, and categorized it as an ob- jectwithuncertainevolutionarystatus.Thevariabilitypat- 13.50 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 ternsintheASASlightcurveofPQOri,althoughranging Phase betweenV ∼12.3and13.3mag,donotresemblethoseof 0 otheryoungstarsinthiswork,anditsJ−K colourisquite Fig.5 Periodogram of the ASAS light curve (top) and s blue with respect to other early F stars in Ori OB1 b. As a phase-folded light curve of RY Ori to the period P = result,itsfielddwarfnatureseemstousmoreplausible. 600.38d(bottom). V469Ori(No.36). Itisanemission-linestarwithX-ray than 500mag2 (Table 2 – actually, it is the star with the emission(Haro&Moreno1953;Wiramihardjaetal.1989; lowestpowerspectrummaximum).Theperiodogramofthe Nakanoetal.1999).AccordingtoCaballero&Solano(2008), ASASlightcurve,displayedinFig.5,showsnoclearpeak, the star is a member in the young Ori OB1 b stellar popu- withanumberofpowermaximaincreasingtowardslonger lationsurroundingAlnilam,andisproablyassociatedtothe periods. The phase-folded light curve at the period of the [OS98]29J,[OS98]29H,and[OS98]29Kremnantmolec- heighestpeak,atP ∼600d,doesnotshowanyevidenceof ularclouds(Ogura&Sugitani1998).BasedonitsV −K 0 s realperiodicity.Actually,fromthefull-lengthlighcurvein colour and the empirical scale of effective temperatures of Fig.A1,thetimescaleoftheirregularvariabilityofRYOri main-sequence stars of Alonso et al. (1996), and extrapo- isoftheorderofafewdays,justlikemanyconfirmedyoung latingittothepre-mainsequence,V469Oriislikelyalate starsinthiswork. GeorearlyKestar.ItsASASlightcurvelookslikethoseof otherTTauristars. TXOri(Mayrit521199,No.12)andTYOri(Mayrit489196, No.13). Theyaretwoclose(ρ∼40arcsec)pre-mainse- V993Ori(No.43). Itisalong-time-knownvariablestar quencestarsintheσOrionisclusterthathaveappearedina discovered by Luyten (1932). Caballero & Solano (2008) numberofyoungandemission-linestarcatalogues(Herbig classifieditasaprobablememberoftheOriOB1byoung & Kameswara Rao 1972; Cohen & Kuhi 1979; Herbig & stellar population surrounding Alnilam. It has a very red Bell 1988; Ducourant et al. 2005). As in the case of many colour J − K = 1.26±0.04mag for its brightness (H = s variablestarsinthiswork,theywerediscoveredasvariable stars by Pickering (1904). Both TX Ori and TY Ori have 3 The values of heliocentric Julian day used in the Figures are those been subject of comprehensive studies and data compila- tabulatedbytheASAScatalogue.Actually,theyarenotheliocentricJulian days HJD, but the HJD minus the constant 2450000.0 (do not confuse tionsbyOliveira&vanLoon(2004)andCaballero(2008). withtheconstantusedtotransformtheJuliandayintothemodifiedJulian Theirlightcurvesarequitesimilar,withminimaandmax- dayMJD,whichis2400000.5). (cid:13)c 0000WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim www.an-journal.org Astron.Nachr./AN(0000) 7 9.82±0.02mag),whichmayindicatethepresenceofacir- (van den Ancker et al. 1998; Marconi et al. 2001; Ripepi cumstellar disc or, alternatively, that it is a background gi- et al. 2003). The pre-main sequence nature of the star is ant star. It may be the optical/near infrared counterpart of supportedbythepresenceofHαinemission,apronounced IRASZ05354–0107,asourceintheIRASFaintSourceRe- inversePCygniprofile,SIλ1296A˚ andOIλ1304A˚ lines jectCatalog(Moshir1992).Thereisnospectroscopyavail- inemission,andinfraredexcess(Valentietal.2000andref- able to confirm or discard its possible T Tauri status. The erencesabove).WithV ≈8.91magand20–30Jyatthe60 0 largepeak-to-peakamplitudeofitslightcurveofV ∼13.0 and 100µm IRAS passbands, it is by far the brightest star 0 tofainterthan14.5mag,althoughshedslightonthephoto- in our sample at all wavelengths, but also the less variable metricvariabilityofV993Ori,isnotconclusive. one. However, the amplitude of variability, between V ∼ 0 8.80 and 9.15mag, is slightly larger than previously mea- sured.Itmaybeduetoourlong(ASAS)temporalcoverage, GSC04767–00071(No.48). Schirmeretal.(2009)clas- probablylongerthananyotherpreviousmonitoring. sifieditasachromosphericallyactivestarsuspectedofvari- abilityintheROTSE-1database.ThestarGSC04767–00071 isprobablyassociatedtotheX-raysource1RXSJ053944.7– StHa40(No.78). ItwasfirstlistedintheHα-emission 005612(Vogesetal.1999)andhasaverylowPPMXproper starcataloguesofMacConnell(1982)andStephenson(1986) motion,of(µ cosδ,µ )=(–6±2,–9±2)masa−1.Theop- and followed-up afterwards by Downes & Keyes (1988), α δ ticalandnear-infraredphotometrysupportsmembershipin Torresetal.(1995),Gregorio-Hetem&Hetem(2002),and Ori OB1 b. It may be a very young solar analog. The am- Maheswaretal.(2003).StHa40isaTTauristarwithvari- plitudeofphotometricvariabilityofitsASASlightcurve,of able spectral type (from F8, through G2 and G5, to K0), V ∼10.8to11.1mag,isrelativelysmall. variableHαemission(frompEW(Hα)=–14.5A˚ toafaint 0 absorption),lithiumabsorption(frompEW(LiI)=+0.13to +0.33A˚),andradialvelocity(indicativeofpossiblespectral PUOri(No.51). Itisapre-mainsequencestarnearthe binarity). Besides, Gregorio-Hetem & Hetem (2002) asso- Alnilam supergiant with Hα in strong two-lobe emission ciated StHa 40 to a ROSAT X-ray source and derived ba- (Haro & Moreno 1953; Herbig & Kameswara Rao 1972; sic stellar and circumstellar parameters. Maheswar et al. Cohen&Kuhi1979;Wiramihardjaetal.1989),photomet- (2003) estimated a magnitude difference ∆V ∼ 1.6mag ricvariability(Fedorovich1960;Bricen˜oetal.2005),mid- between two optical spectra taken almost three years apart infraredfluxexcessattheIRASpassbands(Weintraub1990; and after comparison with spectro-photometric standards. Weaver&Jones1992),andforbiddenemissionlines([OI] TheHαlinewasinemission[absorption]whenthestarwas λ6300.3A˚; Hirth et al. 1997). It has a very red colour of brighter,V ∼10.9mag[fainter,V ∼12.5mag].However, J−K =1.77±0.03mag(Caballero&Solano2008).From s whilethebrightestMaheswaretal.(2003)’sestimationfor our light curve, PU Ori is an irregular variable with mini- V agreesverywellwithourlightcurve(fromwerewemea- mumandmaximummagnitudesV ∼12.5and14.5mag. 0 sure V ≈ 11.0mag), none of the 173 StHa 40 data 0,bright pointscoveringalmostsixyearswerefainterthanV ≈ 0,faint V472 Ori (No. 53). It is an Hα emitter star (Haro & 11.8mag. Moreno 1953; Wiramihardja et al. 1989) of known photo- metricvariability(Fedorovich1960;Cieslinskietal.1997). 3.1.2 Previouslyunknownvariableyoungstars Itmaybetheoptical/nearinfraredcounterpartofIRASZ05322– 0039.V472OriisaprobablememberintheMintakaclus- Ofthefourstarsinthisclass,onlytwo(Mayrit528005AB ter(Caballero&Solano2008).ItsASASlightcurvehasa andStHa48)haveuncontrovertiblesignpostsofyouth.The peak-to-magnitudelargerthan1mag. othertwostars(KisoA–0903135andHD290625)havenot beenwellinvestigatedyetandmaynotbelongtotheyoung OrionBeltstellarpopulation. V351Ori(No.70). ItisaHAeBeA7IIIevariablestarof δ ScutitypediscoveredbyHoffmeister(1934).Ithasbeen investigatedindetail,especiallyafterKovalchuk(1984)ob- Mayrit 528005 AB (No. 18). It is a hierarchical triple served a flare-like event on its light curve at the UBVR systemintheσOrioniscluster,madeofaclosebinary(ρA−B = bands (Wouterloot & Walmsley 1986; Sterken et al. 1993; 0.40±0.08arcsec)resolvedwithadaptiveopticsatρAB−C ∼ Rosenbush 1995; Marconi et al. 2000; Balona et al. 2002; 7.6arcsectothelow-massstarMayrit530005(SOriJ053847.5– Vieria et al. 2003). It has been proposed that a temporar- 022711; Caballero 2005, 2009). The binary primary has a ilystrongaccretionofmatterontothestarorextinctionby cosmic lithium abundance (Zapatero Osorio et al. 2002), circumstellardustcloudstakesplacetoexplaintheremark- double-peak Hα (Caballero 2006), and strong X-ray emis- ableV351Orichangeofbehaviourfrom“thatofa[HAeBe] sions (Wolk 1996; Fraciosini et al. 2006; Caballero et al. starwithstrongphotometricvariations[...]tothatofanal- 2009),andanspectralenergydistributiontypicalofclassII most non-variable star” (van den Ancker et al. 1996). The objects(Herna´ndezetal.2007).TheASASlightcurveshows main period of pulsations of V351 Oriis only of the order variability between V ∼ 13 and 14mag, and an apparent 0 of 0.06d with a peak-to-peak amplitude of about 0.1mag seriesofbrighteningsatHJD∼3000. www.an-journal.org (cid:13)c 0000WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim 8 Caballeroetal.:ThemostvariableASASstarsintheOrionBelt Kiso A–0903 135 (No. 41). It is an emission-line star 9000 only referenced once by Kogure et al. (1989), who mea- sured a weak Hα emission intensity and placed it in the 8000 Orionregion.Itsoptical(DENIS;Epchteinetal.1997)and 7000 near-infrared(2MASS)photometryandnullpropermotion 6000 (withinerrorbars;USNO-B1)areconsistentwithmember- 2g] isthiips oinnethoefOthrieOmBo1stbvaarsisaobclieatsitoanr.sWinitthheσVst0ud=y0..W48h0ilmeaigts, wer [ma45000000 brightest magnitude is V ∼ 12.8mag, the faintest one is o 0 P V ∼15.0magorfainter. 3000 0 2000 StHa 48 (No. 55). It is a K4-type T Tauri star with Hα 1000 in emission and Li I in absorption (Stephenson 1986; Ma- 0 heswar et al. 2003; Caballero & Solano 2008). The star, 10−1 100 101 102 103 104 Period [d] whichislocatedinthesurroundingsofAlnilam,hasawide amplitudeofvariability,withmaximumandminimumbright- 9 eness at V ≈ 11.9 and 13.3mag, respectively, and time 0 scalesofvariationofafewdays. 10 11 HD290625(No.74). ItisanA2starclassifiedasaprob- able member in the young stellar population of Ori OB1 b ag] m 12 (Guetter1979;Nesterovetal.1995).TheASASlightcurve [0 displayed variations between V ≈ 10.8 and 11.3mag, ap- V 0 13 proximately.TheTycho-2and2MASSphotometryandthe lowpropermotion,oflessthan4masa−1(Tycho-2,PPMX, 14 USNO-B1),areconsistentwiththespectraltypeandalong heliocentricdistance.Besides,lifesofA2starsarerelatively 15 short, which confirm the HD 290625 youth. However, the 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Phase starisacoupleofmagnitudesfainterthanexpectedfor“nor- mal”A2Orionstarsatd ∼0.4kpc.Insteadofresortingto Fig.6 Same as Fig. 5, but for X Ori and a period P = another young population in the far background of Orion, 416.05d. ofwhichthereisnoevidence,weproposethatHD290625 is a variable HAeBe star with a scattering edge-on disc in MidcourseSpaceExperiment(8.3–21.3µm;Kraemeretal. the Orion complex, just like UX Ori or the σ Orionis star 2003).Therewerealsoonlythreestarspreviouslyclassified StHa50(Mayrit459340–The´ etal.1994;Caballeroetal. asgiants:XOri,GTOri,andIRAS05354–0142(forwhich 2008)or,alternatively,abright,blue,subdwarf(therehave variability had not ever been proposed). Therefore, in this been previous detections of such objects in the area: Ca- workwepresentsixnewvariablegiants.Besides,wemea- ballero&Solano2007;Vennesetal.2007). sureperiodsofphotometricvariability(withmaximaofthe powerspectrumabove500mag2)between78and920dfor 3.2 Variablebackgroundgiants four of the eight giants, while only one, X Ori, had been reportedbeforetodisplayperiodicity. Unfortunately,becauseoftherelativeresemblancebetween the spectral energy distributions of distant giant stars with extendedcoolenvelopesandofyoungstarswithdeveloped 3.2.1 Knownbackgroundgiants discs (due to the common strong infrared flux excess), gi- antstarssometimescontaminatelistsofmembercandidates X Ori (No. 29). This giant or superginat star was the in star-forming regions. For example, the bright (in the in- reddest object in the Alnilam-Mintaka area studied by Ca- frared)giantstarIRAS05358–0238wasonceclassifiedasa ballero & Solano (2008), but is located at about twice the ClassIobjectintheσOrioniscluster(Oliveira&vanLoon heliocentricdistancetotheOriOB1association(Juraetal. 2004;Herna´ndezetal.2007;Caballero2008).Sincethesur- 1993).ItisanM8–9-typeMiraCetivariablefoundbyWolf veyed area in this work is rich in star-forming regions (in- (1904)withP =424.75±1.77d(Templetonetal.2005)and cludingσOrionis),wehadtobecarefulwithdisentangling silicate dust emission (Sloan & Price 1998; Speck et al. bothpre-mainsequenceandgiantstarpopulations. 2000). In particular, the 9.7 and 18µm silicate dust fea- Oftheeightidentifiedvariablebackgroundgiants,only tureoriginatesfromthecircumstellarenvelopesofoxygen- one (No. 60) has no IRAS excesses or were not detected richasymptoticgiantbranchstars(Kwoketal.1997).From with the Spatial Infrared Imaging Telescope onboard the ourperiodogram(Fig.6),therearetwoclearpeaksatP ≈ (cid:13)c 0000WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim www.an-journal.org Astron.Nachr./AN(0000) 9 1800 1000 1600 900 800 1400 700 1200 2g] 2g] 600 ma1000 ma wer [ 800 wer [ 500 o o 400 P P 600 300 400 200 200 100 0 0 10−1 100 101 102 103 104 10−1 100 101 102 103 104 Period [d] Period [d] 10.5 10.2 11 10.4 11.5 10.6 g] g] a a m 12 m [0 [010.8 V V 12.5 11 13 11.2 13.5 11.4 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Phase Phase Fig.7 Same as Fig. 5, but for GT Ori and a period P = Fig.8 Same as Fig. 5, but for IRAS 05354–0142 and a 920.32d. periodP =373.31d. 416d,consistentwithliteraturevaluesassumingareasobale areremarkable,aswellasitsabnormallyblueV −J colour 0 errorof2%,anditsharmonicatP ≈832d.Thereisalsoa foritsredJ −K andIRASfluxexcessemission. s spurious(systematic)peakatP =1d.Thetriangularshape ofminimaandmaximainthephase-foldedlightcurvemay 3.2.2 Newvariablegiantcandidates be useful for modelling the astrophysical properties of the Mirastar. IRAS05354–0142(No.30). WithV −K =7.1±0.3mag, T s ithadthereddestoptical-infraredcolouramongthe∼1500 Tycho-2/2MASS stars investigated by Caballero & Solano GTOri(No.66). Itisasemi-regularpulsatingstar,orig- (2008). According to them, “the closeness of IRAS 0534– inally proposed to be a likely Mirastar with variations be- 0142 to the Ori I–2 globule may explain part of its red- tween11.5and12.5magbyHoffmeister(1934).Afterwards, dening, but not all. It might be an S-type or a C-type gi- it has been monitored by several authors (e.g. Kurochkin ant with a very late spectral type and very low effective 1949; Braune 2001). The F0 spectral type determined by temperature”. From our ASAS data, IRAS 05354–0142 is Nesterov et al. (1995) is in contradiction with the red op- thebrighteststarinTable2(V ≈10.2–11.2mag)andhas 0 tical and near-infrared colours, the detection of GT Ori by the fourth largest power spectrum. Besides, it is one the IRASatthe12and25µm,anditslowpropermotion,ofless four most clearly variable stars in this study, together with than 3masa−1. The photometric and astrometric informa- XOri,RYOri,andGTOri.Thelightcurvepowerspectrum tionandthestrongperiodogrampeakatP ∼920dandthe peaks at P = 373.31d, but there are also secondary max- phase-foldedlightcurveinFig.7aremoreconsistentwith ima at shorter periods (Fig. 8). The multi-pattern, phase- thesemi-regularpulsatingstatusandaprobableK-type,gi- folded light curve indicates that a harmonic study is nec- antnatureofGTOri.Thelargeamplitudeofvariabilityof essary to disentangle the different oscillation modes of the the star, of almost 3mag (V ≈ 10.6–13.5mag), and the giant.Atabout3arcmintothenorthliestheirregularvari- 0 overlappingofdifferent(shorter)periodsinthelightcurve ablestarV1299Ori,foundbyKaiser(1992).Itwastheonly www.an-journal.org (cid:13)c 0000WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim 10 Caballeroetal.:ThemostvariableASASstarsintheOrionBelt 1800 800 1600 700 1400 600 1200 2g] 2g]500 ma1000 ma wer [ 800 wer [400 Po Po300 600 200 400 200 100 0 0 10−1 100 101 102 103 104 10−1 100 101 102 103 104 Period [d] Period [d] 11.5 12.5 12 13 g] 12.5 g] 13.5 a a m m [0 [0 V 13 V 14 13.5 14.5 14 15 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Phase Phase Fig.9 Same as Fig. 5, but for IRAS 05285–0325 and a Fig.10 SameasFig.5,butforV1159OriandaperiodP periodP =78.226d. =44.345d. SIMBADobjectunidentifiedbyCaballero&Solano(2008) locus in Fig. 4 (V0 −J ∼ 4.8mag), and very low proper in the Alnilam/Mintaka region. Tabulated coordinates are motionsmakeNo.60tobeafairgiantcandidate. likelyincorrect.IRAS05354–0142andV1299Orimightbe the same star, but the variability type and amplitude given 3.3 Miscellanea by Kaiser (1992) are inconsistent with the data presented here. In this class, we include variable objects that are neither young stars nor background giants. Of the eight objects in this class, only two (the dwarf nova cataclysmic variable IRAS 05285–0325 (No. 03), IRAS 05353–0309 (No. 07), V1159Oriandthecontactbinary2M054354–0243.6)were IRAS05248–0037(No.54),andIRAS05312+0053(No.75). knowntovary.Atleastthreeoftheothersixstars,including TheyhadbeenonlyreportedbyKraemeretal.(2003)as thenearbyearlyF-typestarHD290509,mightbeeclipsing “infrared”sourcesandhave“highquality”fluxdensitiesat orinteractingbinaries. theIRAS12µm(and25µm)pass-bandsandpropermotion absolutevaluesofafewmilliarcsecondsperyear.Ofthem, only one had a power spectra peak larger than 500mag2 V1159Ori(No.01). Itisawell-knowndwarfnovacat- (Table 2). The periodogram of IRAS 05285–0325 shows a aclysmic variable of ER UMa (or RZ LMi) type, which is forest of peaks associated to the different pulsation modes a subgroup of SU UMa-type dwarf novae4 with short re- ofthegiant.Themainmode,whichisquiteclearinFig.9, currence time of super-outburst (τ ≤ 45d). Their super- correspondstoaperiodP =78.3±0.1d. 4 Adwarfnovaisaclosebinarystarsysteminwhichoneofthecom- ponentsisawhitedwarf,whichaccretesmatterfromitscompanion.In No. 60. It is reported here for the first time and has not contrasttoclassicalnovae,outburstsandsuper-outburstsindwarfnovae result from instabilities in the accretion disc. SU UMa dwarf novae are ameasuredIRASfluxdensity.However,itsveryredoptical characterisedbytwodistincttypesofoutburstsintimescalesofdays(more and near-infrared colours, which puts the star in the giant frequent)andweeks(lessfrequentandwithlargeramplitude). (cid:13)c 0000WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim www.an-journal.org