1 The Optical Gravitational Lensing Experiment. Planetary and Low-Luminosity Object Transits in the Fields of Galactic Disk. Results of the ∗ 2003 OGLE Observing Campaigns A. U d a l s k i1, M. K. S z y m a n´ s k i1, 5 0 M. K u b i a k1, G. P i e t r z y n´ s k i1,2, 0 I. S o s z y n´ s k i1,2, K. Z˙ e b r u n´1, O. S z e w c z y k1, 2 and L . W y r z y k o w s k i1 n 1Warsaw University Observatory,Al. Ujazdowskie 4, 00-478 Warszawa,Poland a J e-mail: 7 (udalski,msz,mk,pietrzyn,soszynsk,zebrun,szewczyk,wyrzykow)@astrouw.edu.pl 2 Universidad de Concepci´on, Departamento de Fisica, Casilla 160–C, 2 Concepci´on, Chile v 3 ABSTRACT 4 5 We present results of two observing campaigns conducted by the OGLE-III survey in 1 the 2003 observing season aiming at the detection of new objects with planetary transiting 1 companions. Sixfieldsof35′×35′eachlocatedintheGalacticdiskweremonitoredwithhigh 4 frequencyforseveralweeks inFebruary–July2003. Additionalobservations ofthreeofthese 0 fields were also collected in the 2004 season. Altogether about 800 and 1500 epochs were / collectedforthefieldsofbothcampaigns,respectively. h Thesearchforlowdepthtransitswasconductedonabout230000starswithphotometry p better than 15mmag. It was focused on detection of planetary companions, thus clear non- - o planetarycaseswerenotincludedinthefinallistofselectedobjects. Altogetherwediscovered r 40 stars with shallow (≤0.05 mag) flat-bottomed transits. In each case several individual t transits were observed allowing determination of photometric elements. Additionally, the s lowerlimitsonradiioftheprimaryandcompanionwerecalculated. a : From the photometric point of view the new OGLE sample contains many very good v candidatesforextrasolartransitingplanets. However,onlythefuturespectroscopicfollow-up i observations of the OGLE sample – determination of the amplitude of radial velocity and X exclusion of blending possibilities– may allow to confirm their planetary status. In general, r thetransitingobjectsmaybeextrasolarplanets,browndwarfs,M-typedwarfsorfaketransits a causedbyblending. All photometric data of objects with transiting companions discovered during the 2003 campaignsareavailabletotheastronomicalcommunityfromtheOGLEInternetarchive. 1. Introduction Detection of planetary transits in the spectroscopically discovered extrasolar planetary system HD209458 (Henry et al. 1999, Charbonneau et al. 1999) em- piricallyprovedthatthe photometricmethodofdetectionofextrasolarplanets, called the “transit method”, can be a potentially important tool in planetary ∗Based on observations obtained with the 1.3 m Warsaw telescope at the Las Campanas ObservatoryoftheCarnegieInstitutionofWashington. 2 searches. However,inspite ofmanyeffortsno extrasolarplanetwasdetectedin this manner during the next few years. The transit method offers many advan- tagescomparedtothetraditionalspectroscopicmethod. Whenthephotometric observations are combined with precise spectroscopy all important parameters of a planet like its size and mass can be derived providing extremely important data to studying in detail the structure of planets and evolution of extrasolar planetary systems. In2002theOpticalGravitationalLensingExperiment(OGLE)surveyannoun- cedresultsoftwophotometriccampaignsaimedatphotometricdiscoveryoflow depth transits caused by extrasolar planets or other sort of “low-luminosity” objects (Udalski et al. 2002abc, 2003). About 140 stars such stars were found amongabout150000starsintwolines-of-sightintheGalacticdisk–towardthe Galactic center and the constellation of Carina. It was obvious from the very beginning that the significant part of the discovered transits may be caused by stellar size companions, partcanbe aneffect of blending ofeclipsing stars with bright unresolvable neighbors while in the small number of objects the transits could be of planetary origin. Unfortunately, the photometry, which allows de- terminationofsize,cannotunambiguouslydistinguishplanetaryorbrowndwarf from late M-type dwarfstellar companions as all of them may have the radii of the order of 0.1–0.2 R⊙ (1–2 RJup). The OGLE transiting objects were subject of many follow up studies in the past two years. The most important follow up observations of the OGLE ob- jectswerehighresolutionspectroscopicobservationsforradialvelocitychanges. In the past two yearsthe planetary status of four of the OGLE transiting com- panions was confirmed by measuring small amplitude radial velocity variation with the same period and in appropriate phase as resulting from the OGLE photometric orbits: OGLE-TR-56 (Konacki et al. 2003a, Torres et al. 2004a, Bouchyetal.2004b),OGLE-TR-113(Bouchyetal.2004a,Konackietal. 2004), OGLE-TR-132(Bouchyetal.2004a)andOGLE-TR-111(Pontetal.2004). An- otherOGLEobject–OGLE-TR-10–alsoverylikelyhostsanextrasolarplanet although additional radial velocity observations are necessary to fully confirm its planetary status (Konacki et al. 2003b, Bouchy et al. 2004b). Additionally, the high resolutionspectroscopyallowedto rule out possible blending scenarios makingtheconfirmationsveryreliableandsound. Itisworthnotingthatexcept forOGLE-TR-132,whichwasattheedgeofdetection,alltheremainingOGLE planets were on the OGLE short lists of the most likely planetary candidates based on their photometric properties. The sixth known extrasolar transiting planet, TrES-1, has recently been found by the small telescope wide field TrES survey (Alonso et al. 2004). Detection of the first transiting planets allowed to significantly increase the sample of extrasolar planets with precisely known masses and radii, thus pro- viding crucial empirical data for testing the planetary models etc. Surprisingly the first three OGLE planets (OGLE-TR-56, OGLE-TR-113 and OGLE-TR- 132) turned out to form a new class of “very hot Jupiters” – giant planets orbiting their hosts stars with periods shorter than 2 days. Such systems have been unknown so far in the solar neighborhood. It is clear that the detection 3 of additional transiting planets is of high importance for extrasolar planetary science. In2003theOGLEsurveyconductedtwoadditionalobservationalcampaigns aimed at the detection of additional transiting exoplanets. Altogether six new fieldslocatedintheGalacticdiskintheconstellationsofCarina,Centaurusand Muscaweremonitoredregularlywithhightime resolutionforafewweekseach. The search for objects with transiting companions was performed in similar manner as in the past campaigns. However, it was more focused on planetary transits. Hence,manycleardetectionsoflowmassstellartransitingcompanions were removed from the final sample of objects with transiting companions. Similarly to our previous transit samples (Udalski et al. 2002abc, 2003) the photometric data of our new objects with transiting companions from the 2003 fields can be found in the OGLE Internet archive. Thus the follow-up spectroscopicobservationsandconfirmationoftheirplanetaryornon-planetary status can be made in short time scale. Details and pointers to the OGLE Internet archive can be found at the end of this paper. 2. Observational Data Observations presented in this paper were collected with the 1.3-m Warsaw telescope at the Las Campanas Observatory, Chile (operated by the Carnegie Institution of Washington), equipped with a wide field CCD mosaic camera. The camera consists of eight 2048×4096 pixel SITe ST002A detectors. The pixel size of each of the detectors is 15 µm giving the 0.26 arcsec/pixel scale at the focus of the Warsaw telescope. Full field of view of the camera is about 35′×35′. The gain of each chip is adjusted to be about 1.3 e−/ADU with the readout noise of about 6 to 9 e−, depending on the chip. More details on the instrumental setup can be found in Udalski (2003). Twoobservingcampaignswereconductedin2003. Thefirstofthem–OGLE campaign #3 – started on February 12, 2003 and lasted up to March 26, 2003. The photometric data were collected during 39 nights spanning 43 days. Three fields of the Galactic disk were observed continuously with the time resolution of about 15 minutes. Acronyms and equatorial coordinates of these fields are provided in Table 1. The second campaign – OGLE campaign #4 – started on March 25, 2003. While the main observing material was collected to the middle of May 2003, the fields were also observed from time to time until July, 25, 2003. Moreover, additionalbut less extensive photometric monitoring of the campaign#4 fields was conducted during the 2004 observing season – from March 14, 2004 up to July30,2004. AsinthepreviouscampaignsthreefieldsoftheGalacticdiskwere continuously monitored with the time resolution of about 15 minutes. Details on the location of these fields and their acronyms can also be found in Table 1. All observations were made through the I-band filter. The exposure time of eachimage was set to 180seconds. Altogether about 820images were collected foreachfield duringourOGLEcampaign#3 and1050imagesin2003plus 460 4 Table 1 Equatorialcoordinates oftheobservedfields Field RA(J2000) DEC(J2000) Campaign#3 CAR106 11h03m00s −61◦50′00′′ CEN106 11h32m30s −60◦50′00′′ CEN107 11h54m00s −62◦00′00′′ Campaign#4 CEN108 13h33m00s −64◦15′00′′ MUS100 13h15m00s −64◦51′00′′ MUS101 13h25m00s −64◦58′00′′ imagesin2004forthecampaign#4fields. Themedianseeingofimagesinboth campaigns was about 1′.′2. 3. Data Reductions Allcollectedimageswerepreprocessed(de-biasingandflat-fielding)inrealtime with the standard OGLE-III data pipeline (Udalski 2003). Photometricreductionsofthe2003seasonimageswereperformedoff-lineaf- ter the end of both campaigns. Additional 2004season data were reduced with the same software setup in real time at the telescope. Similarly to our previ- ous campaignsthe OGLE-III photometric data pipeline basedonthe difference image analysis (DIA) method was used – see details in Udalski (2003). Becausenostandardstarswereobservedduring2003campaignsprecisecal- ibrationofthe OGLEphotometryoftransitfields tothe standardsystemcould not be obtained. However, part of the CAR106 field overlaps with CAR SC1 field observedand well calibrated during the OGLE-II phase of the OGLE sur- vey. Based on the mean magnitudes of several bright stars observed in both – OGLE-II and OGLE-III – the mean shift of the magnitude scale between the OGLE-III magnitudes and OGLE-II calibrated data was derived. Other fields were calibrated based on magnitudes of obtained in this manner “local” stan- dards in the CAR106field. The errorof the magnitude scale shouldnot exceed 0.1–0.15 mag. Astrometric solution for the observed fields was performed similarly to the previous campaigns (Udalski et al. 2002ac),i.e.,by cross-identificationof a few thousand brightest stars in each chip image with the Digitized Sky Survey im- ages of the same part of the sky. Then the transformation between OGLE-III pixelgridandequatorialcoordinatesofthe DSS(GSC) astrometricsystemwas calculated. 5 4. Search for Transits The search for objects with transiting companions was performed in almost identical way as in the past campaigns. Before the transit search algorithm was applied to the collected data, two steps were performed. First a preselection procedure was run. We limited our search for transiting objects to stars with very precise photometry, i.e.,those with the rms from the entire time series collected in the 2003 season smaller than 15 mmag. Similarly to the 2002 Carina campaign we did not make any preselection based on colors of stars. About 100 000 and 130 000 stars passed our “good photometry” cut in the campaign #3 and #4 fields, respectively. Next the photometric data of all objects were corrected for small scale sys- tematic effects using the data pipeline developed by Kruszewski and Semeniuk (2003). Udalski et al. (2003) showed that this algorithm efficiently removes various small scale systematic errors present in the original data and makes it possible to detect lower depth transits. Finally, all stars were subject to the transit search algorithm – the BLS algorithmof Kov´acs,Zuckerand Mazeh (2002). Similar parameters of the BLS algorithm as in the previous campaigns were used. The search for transits was limited to periods from the range of 1.05 days to 10 days. The final list was prepared after careful visual inspection of all light curves which passed the BLS algorithm. The experience from our previous campaigns and spectroscopic follow up observations indicates that the size of exoplanets hardlyexceeds the size ofJupiter. None ofmany largersize transiting compan- ions discoveredby OGLE turned out to be a planet. Therefore we set a tighter limit on the depth of transits than in the previous campaigns and removed all objects that revealed transits deeper than 0.05 mag. It was also realized from the very beginning that when only two transits had been observed during the campaign the derived periods may be uncertain. Usually, the shortest period allowed by the distribution of epochs was selected in such cases, but longer periods could not be excluded. Spectroscopic follow- up observations of Bouchy et al. (2004b) showed that indeed in a few cases so selected periods were inconsistent with spectroscopy. The problem practically disappears when three transits are observed. Therefore we left on our list only those objects for which three or more transits were covered during our cam- paigns(this constraintwaslifted in two exceptions,i.e.,inthe case oftwo good candidates in campaign #3 where only two transits were observed. The span of data collected for campaign #3 fields was much shorter than for the second campaign making the detection of many transits more difficult). To minimize the number of objects contaminating our final sample we re- movedfromtheoriginallistofdetectedstarsalargenumberofsmallamplitude events caused by grazing eclipses of regular stars of similar size and brightness that produce V-shaped eclipses and large number of objects revealing small deptheclipsesbutsimultaneouslyasmallamplitude sinusoidalvariationcaused by distortion of the primary – a clear sign of a relatively massive companion (Udalski et al. 2002a, Drake 2003, Sirko and Paczyn´ski 2003). Also evident 6 blendsoffainteclipsingsystemswithabrighterstarwereremoved. However,it shouldbestressedthatinthecaseofmorenoisylightcurvesoffainterstarsitis not easy to distinguish between grazing eclipses and very non-central transits. Therefore,some ofthe starson ourlist might still be doublestars,whatcan be easily verified by the future spectroscopy. The final periods of our candidates were found after careful examination of theeclipselightcurve–byminimizingdispersionduringtheeclipsephasesthat are very sensitive to period changes. The formal accuracy of periods depends on the number of individual transits observed and the total duration of obser- vations. It is of the order of 10−4P for the only six week long campaign #3 and10−5P forthe twoseasonlongcampaign#4(whenadditionaltransitswere observed in the 2004 season data). 5. Results of the 2003 Campaigns Seventeen stars with small transiting companions from the fields of 2003 cam- paign#3havemetourcriteria. Inthe fields of2003campaign#4twentythree objects have been found. ) g .06 a m ( t i .04 s n a r t f .02 o h t p e 0 D 14 15 16 17 I (magnitude) Fig.1. Depthof transitvs. I-bandmagnitude ofthe hoststar. Horizontal dotted lineatthe depthof0.05marksthelimitofOGLEsearch. Similarly to our pastcampaignswe list the basic dataon selectedobjects in Table2. WepreserveournotationstartedinOGLE-III2001campaign: thefirst objectin Table2 is designatedasOGLE-TR-138. InTable 2the followingdata are tabulated: Identification, equatorial coordinates (J2000), orbital period, epoch of mid-eclipse, I-band magnitude outside transit, the depth of transit, number of transits observed (N ) and remarks. Accuracy of the magnitude tr scale is of about 0.1–0.15 mag. Although our search was limited to periods 7 ) g .06 a m ( t i .04 s n a r t f .02 o h t p e 0 D 0 .2 .4 .6 .8 log P Fig.2. Depthoftransitplottedagainsttheorbitalperiod(logP). longerthan1.05days,one objectwith shorterperiod (detected withthe period of 2P) also entered the list. In the Appendix we present the light curves and finding charts of selected objects. For each object the full light curve and a close-up around the transit areshown. It shouldbe notedthat the magnitude scale changesin the close-up windows, depending on brightness, noise and transit depth. The finding chart isa 60′′×60′′ subframe ofthe I-bandreferenceimagecenteredonthe star. The starismarkedbyawhitecross. NorthisupandEasttotheleftintheseimages. Fig. 1 presents the depth of transitplotted againstI-bandmagnitude of the hoststar. AsexpectedthedetectionlimitisroughlyflatuptoI≈15.7magand then rises slowly for noisier light curves of fainter stars. In Fig. 2 the depth of transit is shown against logP. No relation or trends between these parameters are seen. 6. Discussion Fortynewobjectswithsmalltransitingcompanionswerediscoveredinsixfields in the Galactic disk during two 2003 season OGLE campaigns. Although the number of detected objects is smaller than in the past campaigns our search in the2003seasondatawasfocusedmainlyonthesmallestcompanionswhichhave larger probability to be extrasolar planets. Therefore tighter selection criteria were applied and all uncertain cases, objects with longer lasting transits – evi- dently of stellar origin, objects revealing significant variability between eclipses suggesting more massive, stellar companions and other evident contaminators wereremovedfromthe listofdetectedobjects. Specialcarewastakentodetect transits as shallow as possible, as the probability of planetary origin in such cases is larger. All existing detections of transiting exoplanets indicate that the size of the extrasolar planets practically does not exceed that of Jupiter so 8 Table 2 Objectswithplanetaryorlowluminositytransitingcompanions Name RA(J2000) DEC(J2000) P T0 I ∆I Ntr Rem. [days] –2452000 [mag] [mag] OGLE-TR-138 11h04m23.s77 −61◦43′10.′′7 2.64550 685.20281 16.04 0.016 3 OGLE-TR-139 11h02m43.s79 −61◦37′46.′′7 2.53420 684.64467 16.41 0.030 4 OGLE-TR-140 11h01m13.s91 −61◦44′21.′′6 3.39330 684.68281 16.45 0.043 4 OGLE-TR-141 11h01m52.s09 −61◦58′01.′′4 5.67860 683.75001 14.80 0.019 2 OGLE-TR-142 11h00m34.s75 −62◦04′09.′′0 3.06280 684.10926 15.68 0.034 4 OGLE-TR-143 11h34m03.s83 −60◦54′13.′′0 3.34980 682.92526 13.69 0.006 3 OGLE-TR-144 11h34m52.s78 −60◦52′35.′′6 2.44560 685.19335 16.74 0.034 5 ST OGLE-TR-145 11h33m04.s34 −60◦34′04.′′2 2.74120 685.48763 16.33 0.019 4 ST OGLE-TR-146 11h34m05.s52 −60◦39′38.′′4 2.94460 685.79072 16.57 0.043 5 OGLE-TR-147 11h32m30.s45 −60◦38′41.′′6 3.84170 686.11365 15.67 0.030 3 OGLE-TR-148 11h31m31.s66 −60◦36′52.′′6 1.43290 683.45961 16.83 0.022 7 ST OGLE-TR-149 11h30m45.s10 −60◦43′57.′′9 4.55170 684.68689 16.35 0.026 3 OGLE-TR-150 11h32m29.s92 −60◦58′25.′′5 2.07380 683.36717 15.03 0.005 6 OGLE-TR-151 11h32m21.s84 −60◦50′45.′′5 1.48350 683.46444 15.94 0.022 9 OGLE-TR-152 11h54m05.s59 −62◦04′38.′′2 3.73000 685.88809 16.30 0.019 4 ST OGLE-TR-153 11h55m21.s60 −62◦03′55.′′3 4.39560 685.15132 15.39 0.030 2 OGLE-TR-154 11h52m55.s80 −62◦16′53.′′6 3.66946 687.41198 14.96 0.013 4 OGLE-TR-155 13h33m26.s00 −64◦16′38.′′4 5.27700 701.09815 15.62 0.008 5 OGLE-TR-156 13h33m54.s81 −64◦09′43.′′9 3.58341 699.47628 15.26 0.034 8 OGLE-TR-157 13h34m22.s90 −64◦07′17.′′5 5.86821 698.99295 17.05 0.043 7 OGLE-TR-158 13h33m38.s08 −64◦05′23.′′6 6.38410 698.60150 15.89 0.019 5 OGLE-TR-159 13h31m54.s49 −64◦02′39.′′6 2.12676 697.54746 16.35 0.038 16 ST OGLE-TR-160 13h30m42.s95 −63◦57′30.′′6 4.90185 698.32754 14.34 0.008 3 OGLE-TR-161 13h30m55.s71 −63◦58′38.′′2 2.74730 696.73100 15.69 0.006 4 OGLE-TR-162 13h31m01.s09 −63◦58′34.′′9 3.75819 731.37770 14.87 0.011 4 OGLE-TR-163 13h32m52.s59 −63◦57′44.′′1 0.94621 696.24111 15.78 0.034 23 OGLE-TR-164 13h30m58.s60 −64◦11′53.′′0 2.68153 697.11292 16.15 0.019 10 OGLE-TR-165 13h31m31.s55 −64◦10′51.′′5 2.89185 698.14260 16.81 0.043 9 OGLE-TR-166 13h32m23.s86 −64◦10′35.′′2 5.21920 699.83984 15.31 0.011 5 OGLE-TR-167 13h30m26.s30 −64◦10′09.′′2 5.26100 698.58614 15.88 0.022 4 OGLE-TR-168 13h32m12.s07 −64◦32′56.′′2 3.65080 696.94702 14.13 0.013 3 OGLE-TR-169 13h17m23.s43 −64◦54′54.′′8 2.76877 698.04287 16.78 0.019 13 OGLE-TR-170 13h15m14.s50 −64◦49′28.′′8 4.13680 699.13261 16.62 0.026 7 OGLE-TR-171 13h13m52.s00 −64◦41′30.′′0 2.09180 696.49729 17.07 0.038 9 OGLE-TR-172 13h13m12.s81 −64◦47′13.′′8 1.79323 697.28151 15.33 0.006 8 OGLE-TR-173 13h14m56.s11 −65◦02′00.′′4 2.60590 698.87378 14.89 0.019 4 OGLE-TR-174 13h25m58.s75 −65◦14′50.′′6 3.11012 697.52261 16.50 0.016 7 OGLE-TR-175 13h27m47.s31 −65◦14′34.′′8 1.48830 697.01551 16.87 0.019 11 ST OGLE-TR-176 13h25m33.s73 −64◦55′34.′′1 3.40478 700.66215 15.15 0.022 5 OGLE-TR-177 13h24m48.s41 −64◦51′37.′′8 5.64386 699.53686 16.79 0.038 3 9 the depth of transits caused by planets transiting FG-type stars must be of the order of 10 mmag or less. Photometric observations of the transit shape allow to derive the size of transitingcompanions. Unfortunately,thesizealonecannotunambiguouslyde- fine the type of the transiting objects. Transits can be caused by extrasolar planets or brown dwarfs or small late M-type dwarfs as all these objects can have sizes of about R . To distinguish between these possibilities, radial ve- Jup locity follow-upmeasurementsare necessaryto estimate masses ofcompanions. Another important contaminator of photometrically selected transiting objects mightbeblendingofaregulartotallyeclipsingstarwithacloseopticalorphys- ically related (wide multiple system) unresolvable neighbor that can produce transit-like light curve. Therefore some of our candidates listed in Table 2 can actually be fake transits caused by blending effects. Unfortunately, even in the clean unblended case additional information on the size of the host star is necessary to derive the size of transiting body, be- causethedepthoftransitprovidesonlytheratioofthosesizes. Suchadditional information can come from moderate resolution spectroscopy suitable for spec- tralclassification,IRphotometry (Dreizler et al. 2002,Gallardoet al. 2004)or, in principle, from optical colors of host stars. Unfortunately, in the case of the pencil beam survey of the Galactic disk, like OGLE, significant and unknown interstellar extinction makes dereddening of optical colors of individual stars very uncertain and unreliable and thus of no use for estimation of stellar sizes. When only single band photometry is available it is not possible to obtain actual size of the companion when the errors of individual observations are comparabletothetransitdepth,duetowellknowndegeneracybetweenradiiof the host star and the companion, R , R , inclination, i, and limb darkening,u. s c Photometricsolutionsofsimilarqualitycanbeobtainedfordifferentinclinations of the orbit and radii of components (in the I-band the transit light curve is practically insensitive to the limb darkening parameter u). Thus, the selection of the proper solution is practically impossible. Only the lower limit of the size of the companion can be calculated assuming that the transit is central, i.e.,i=90◦. The corresponding radius of the primary is in this case also the lower limit. In Table 3 we provide lower limits for the components radii calculated in the same manner as in Udalski et al. (2002abc) assuming additionally that the host star follows the mass-radius relation for main sequence stars (R/R⊙= (M/M⊙)0.8). The resulting mass of the primary is also listed in Table 3. In practice the transits might be non-central, thus the size of the star and com- panion can be larger than given in Table 3. Also when the host star is evolved the estimation may be inaccurate. The estimations of the minimum sizes of transiting companions cannot be consideredastheiractualdimensionsassometimesmistakenlyinterpreted. The final parameters of the system can only be derived when the spectroscopic ob- servationsareavailable. TheaimofpresentingTable3istoprovideinformation necessary for preselection of the most promising planetary candidates from our 2003 transit sample. Objects with the estimation of the minimum radius of 10 transiting companion larger than 0.15–0.18 R/R⊙ are almost certainly stellar binary systems and are marked by “ST” in the remarks column of Table 2. Table 3 Lowerlimitsofradiiofstarsandcompanions (centralpassage: i=90◦). Name Rs Rc Ms Name Rs Rc Ms [R⊙] [R⊙] [M⊙] OGLE-TR-138 0.74 0.082 0.69 OGLE-TR-158 0.29 0.035 0.21 OGLE-TR-139 1.00 0.150 1.00 OGLE-TR-159 1.22 0.207 1.28 OGLE-TR-140 0.56 0.101 0.48 OGLE-TR-160 1.02 0.081 1.02 OGLE-TR-141 0.79 0.095 0.75 OGLE-TR-161 2.53 0.177 3.19 OGLE-TR-142 0.46 0.074 0.38 OGLE-TR-162 1.02 0.092 1.03 OGLE-TR-143 0.49 0.034 0.41 OGLE-TR-163 0.62 0.098 0.55 OGLE-TR-144 1.43 0.228 1.56 OGLE-TR-164 1.12 0.134 1.15 OGLE-TR-145 1.82 0.219 2.12 OGLE-TR-165 0.92 0.165 0.90 OGLE-TR-146 0.71 0.127 0.65 OGLE-TR-166 1.31 0.118 1.40 OGLE-TR-147 1.05 0.158 1.06 OGLE-TR-167 0.62 0.081 0.56 OGLE-TR-148 1.64 0.213 1.86 OGLE-TR-168 0.77 0.077 0.73 OGLE-TR-149 0.91 0.128 0.89 OGLE-TR-169 1.48 0.177 1.63 OGLE-TR-150 1.83 0.110 2.13 OGLE-TR-170 1.23 0.172 1.29 OGLE-TR-151 1.02 0.132 1.02 OGLE-TR-171 0.81 0.138 0.77 OGLE-TR-152 1.70 0.205 1.95 OGLE-TR-172 1.19 0.083 1.24 OGLE-TR-153 0.79 0.119 0.75 OGLE-TR-173 0.38 0.046 0.30 OGLE-TR-154 1.24 0.124 1.31 OGLE-TR-174 0.75 0.083 0.70 OGLE-TR-155 1.42 0.113 1.54 OGLE-TR-175 1.71 0.205 1.96 OGLE-TR-156 0.49 0.079 0.41 OGLE-TR-176 1.02 0.132 1.02 OGLE-TR-157 0.88 0.158 0.85 OGLE-TR-177 0.68 0.115 0.61 Solid line in the close-up windows in the Appendix shows the transit model lightcurvecalculatedfor the centralpassage. Inmostcases the centralpassage fit is practically indistinguishable from non-central so at this stage it is impos- sible to derive other values than the lower limits of radii provided in Table 3. Because the search for transiting objects in the 2003 campaigns data was focusedonplanetarysystems the finallist containsmany objectswith compan- ions for which the lower size limit is well within the planetary range. Based on the shape of transits and photometric properties one can select many very promising planetary transit candidates in our sample, for instance OGLE-TR- 158, OGLE-TR-162, OGLE-TR-164, OGLE-TR-167, OGLE-TR-172, OGLE- TR-174,OGLE-TR-176. However,it cannotbe excluded thatother more noisy or low transit depth systems also host a planet – in such cases it is difficult to assess the real transit shape. Results of the past OGLE transit campaigns have shown how severe a non- planetary background in the planetary transit searches can be. Very often the OGLEsampleoftransitstarswasmisinterpretedaspure“planetary”candidates