Draftversion January6,2015 PreprinttypesetusingLATEXstyleemulateapjv.5/2/11 A SERENDIPITOUS ALL SKY SURVEY FOR BRIGHT OBJECTS IN THE OUTER SOLAR SYSTEM M.E. Brown1,M.E. Bannister2,3, B.P. Schmidt3, A.J. Drake1,S.G. Djorgovski1, M.J. Graham1,A. Mahabal1, C. Donalek1,S. Larson4, E. Christensen4, E. Beshore4, R. McNaught3 Draft version January 6, 2015 ABSTRACT We use seven year’s worth of observations from the Catalina Sky Survey and the Siding Spring 5 Survey covering most of the northern and southern hemisphere at galactic latitudes higher than 20 1 degrees to search for serendipitously imaged moving objects in the outer solar system. These slowly 0 moving objects would appear as stationary transients in these fast cadence asteroids surveys, so we 2 develop methods to discover objects in the outer solar system using individual observations spaced by months, rather than spaced by hours, as is typically done. While we independently discover 8 n knownbrightobjectsinthe outersolarsystem,the faintesthavingV =19.8±0.1,nonew objectsare a J discovered. We find that the survey is nearly 100% efficient at detecting objects beyond 25 AU for V . 19.1 (V . 18.6 in the southern hemisphere) and that the probability that there is one or more 5 remaining outer solar system object of this brightness left to be discoveredin the unsurveyedregions of the galactic plane is approximately 32%. ] P Subject headings: solar system: Kuiper belt — solar system: formation — astrochemistry E . h 1. INTRODUCTION images with archival images of the same location taken p at a different time is it recognized that the bright outer - The last decade has seen the discovery of most o of the brightest objects in the outer solar system solar system object appears as a one-time transient. r (Trujillo & Brown 2003; Brown 2008; Schwamb et al. The large number of serendipitous images of bright t outer solar system objects raises the possibility that a s 2009). The wide-field surveys for these brightest objects a appear moderately complete in both the northern and fully serendipitous archival survey could be attempted [ to find new objects. An ideal data set for such a survey southern skies, with only the generally avoided galactic would be one which covers large areas of the sky with 1 plane and ecliptic poles left to survey completely. All high enough temporal coverage that multiple detection v surveys miss some fraction of objects ostensibly in their of an object in the Kuiper belt could be made. 1 survey regions, however, due to temporal gaps, detec- Over the past decade, surveys for NEOs have come 4 tor gaps, stellar blending, and numerous other difficul- closest to achieving this ideal. In these surveys multiple 9 ties. Survey efficiencies have been estimated to be be- images are obtained to search for moving objects, but 0 tween70and90%(Schwamb et al.2010;Sheppard et al. 0 2011;Rabinowitz et al.2012),leavingthepossibilitythat these images are obtained over a time period too short 1. brightobjectsintheoutersolarsystemhaveescapedde- todetectthemotionofslowlymovingobjectsintheouter solar system. Objects in the outer solar system appear 0 tection. simply as stationary transients. 5 While each of the bright objects in the Kuiper belt Here we describe a serendipitous all sky survey for 1 has yielded an important boon of scientific information bright objects in the outer solar system using archival : about the origin and evolution of the Kuiper belt and v data from the Catalina Sky Survey (CSS) in the north- its objects (Brown 2008), mounting a dedicated survey i ern hemisphere and its sister survey, the Siding Spring X to find this small number of remaining bright objects Survey (SSS), in the southern hemisphere. wouldbeprohibitive. Wenote,however,thateachofthe r a bright objects discovered over the past decade serendip- 2. THECSS,SSS,ANDCRTSSURVEYS itously appearedin multiple other survey images. These The CSS (Larson et al. 2003) operates on the 0.7 m serendipitous detections have been reported to the Mi- Catalina Schmidt telescope at the Catalina Observatory nor Planet Center in sources from the Skymorph data of CCD images from the Palomar 48-inch Schmidt 5 back in Arizona and covers 8.1 square degrees per field to a limiting magnitude of V ∼ 19.5. The CSS has covered to the original POSS I photographic plates of the 1950s approximately 19700 square degrees of sky between de- andmanyinbetween. The objects wereunrecognizedat clinations of -25 and +70 at galactic latitudes greater the time of their original imaging owing to the fact that than10degrees. TheSSSoperatesonthe0.5mUppsala even at opposition these objects move at speeds of only Schmidt telescope at Siding Spring Observatory in Aus- a few arcseconds per hour, so they appear identical to traliaandcovers4.2squaredegreesperfieldtoalimiting stationary stars in the images. Only by comparing the magnitude of approximately 19.0. The SSS has covered approximately14100squaredegreesofsky fromdeclina- 1CaliforniaInstitute ofTechnology, Pasadena, CA,USA tion -80 to 0. Accounting for overlapof the two surveys, 2UniversityofVictoria,Victoria,BC,Canada 3TheAustralianNationalUniversity,Canberra,Australia the total amount of sky covered is approximately 29700 4TheUniversityofArizona,LunarandPlanetaryLaboratory, square degrees. Tucson, AZ,USA In both surveys, most fields have been observed mul- 5 http://skyview.gsfc.nasa.gov/skymorph/ tiple times per season over many years. Figure 1 shows 2 beyond. An objectin acircularprogradeorbitat25 AU has a maximum retrograde motion at opposition of 4.9 arcseconds hr−1. Motions higher than this would would likely be detected in the NEO surveys. In typicaloperations,both the CSS and SSS take four imagesperfieldpernightoveratimeintervalofabout30 minutes. As ourfirststep inouranalysis,werequire the detection of four transient candidates on a single night within a diameter defined by 4.9 arcseconds times the maximum time interval. Note that we perform no other filtering here. The four detections are not required to show linear motion (as they would not for the slowest moving objects) or have similar measured magnitudes (whichtheymightnotatthe magnitudelimitofthe sur- vey). When four detections within anappropriatediam- eter are found, they are collected as a single transient with the average position, magnitude, and observation time of the four individual candidate transients. Some fraction of the CSS and SSS transients recur at thesamelocation. Thesetransientsarepresumablysome type of astrophysical source which has brightness varia- tions sufficiently large that the source does not appear in the deep catalog but the object occasional becomes bright enough to appear in individual images. To re- Fig.1.— (a) Coverage of the Catalina Sky Survey (CSS). The shadingshowsthenumberofoppositionseasonsinwhicheachfield move these clear non-solarsystem sources,we searchfor hasbeencovered4ormoretimes. Theequalareaplotiscentered all transients that have a transient detected on a differ- at an RA of 180 degrees and a declination of 0 degrees with grid entnightwithin4arcsecsofthesamelocation. Ourfinal marks placed every 30 degrees of RA and declination. Contours transient list, 1.2 million sources in the CSS fields and of galactic latitude of 10 and 20 degrees are shown as well as a lineshowingtheecliptic. (b)CoverageoftheSidingSpringSurvey 2.3millionsourcesintheSSSfields,willcontainallouter (SSS).Allparametersareasin(a)exceptthattheplotiscentered solarsystemobjectswithinthegeometricandbrightness atanRAof0degrees. limits of the survey, true astrophysical transients which appear only once at their location, and image artifacts, thefieldcoveragewiththegrayscaleindicatingthenum- which will be the overwhelming majority of the list. ber of seasons each field has been observed at least 4 Figure2showsthelocationsofeachofthesetransients. times. The overall cadence of the surveys is extremely Significantstructurecanbeseeinthetransientlocations. non-uniform; the four (or more) observations could be In the SSS, in particular, transients occur frequently on within a single lunation or could be spread over a six the field edges, suggesting inconsistent astrometric solu- monthoppositionseason. Aswewillseebelow,however, tions in these areas (which will lead stationary stars to thesedetailshavenoeffectonourdetectionscheme. The occasionallybeclassifiedastransientsinlargenumbers). practicallimitthatwefindisthatanobjectneedstohave Similar effects can be seen for the CSS data in the far been detected at least4 times in a single seasonin order north. Other regions of clear artifact can be seen. In for us to extract it from the data set. addition, the higher density of transients in the SSS is In addition to being searched for NEOs and other clear. TheSSScontains163transientspersquaredegree moving objects, the CSS and SSS data are analyzed comparedto61intheCSS.Thislargernumberofsouth- by the Catalina Real-time Transient Survey (CRTS) ern transients will make the SSS moving object search (Drake et al. 2009) to search for transients. The CRTS comparatively more difficult. compares source catalogs from each image with source Nonetheless, real known Kuiper belt objects are also catalogs generated from deep coadds of multiple images present in the data. As an example, Figure 3 shows the in addition to comparing with other available deep cat- orbit of Makemake – the brightest known KBO after alogs; objects which do not appear in these catalogs are Pluto – as well as the transients that are detected in deemedcandidatetransients. CRTSperformsadditional this region of the sky. Of these 789 transients, 53 are filteringinanattempttocullthetrueastrophysicaltran- detections of Makemake itself. The other known bright sientsfromthelargenumberofimageartifacts. Webegin KBOs likewise have many detections. our analysis, however, with the full catalog of candidate transients. The full CSS catalog has nearly 1.8 billion 3.2. The Keplerian filter transient candidates from 2005 Dec 6 to 2012 April 21, All objects in the outer solar system move on well- while the SSS catalog has 2.1 billion candidates from defined Keplerian orbits. We use this fact to look for 2005 Feb 20 to 2012 April 29. collections oftransients which define any physicallypos- 3. THESLOWLYMOVINGOBJECTSEARCH sible Keplerian orbit. Three points in the sky at differ- ent times are required to define an orbit, but because 3.1. Creation of the transient list the three points contain 9 parameters while an orbit is The cadence of the NEO surveys has allowed them to defined by only 7 parameters, the three points overcon- detect objects out to the orbit of Uranus. We will thus strain the orbit. Thus arbitrary sets of three points will defineourheliocentricradiusofinteresttobe25AUand not be able to be fit to a Keplerian orbit. 3 have a 5 degree parallax over a six month season. We conservativelyallow motions up to 5 degrees for our 120 daymaximumseparation. Thissingleseasonspatialand temporal filtering brings the number of potential orbital triplets that need to be checked down to 160 billion for CSS and 18 trillion for SSS. While this initial filter cuts thenumberoforbitstobefitsignificantly,eventhisnum- berwouldbecomputationalprohibitiveforfullKeplerian orbital fitting. Weapplyonemorefilterwhichisappropriateforthese distant objects observed over a modestly short time in- terval. AsshowninBernstein & Khushalani(2000),mo- tions of distant objects in the solar system can be ap- proximated over a short time period as moving linearly through the solar system perpendicular to the earth- object vector. This sky plane approximation is defined by only 6 parameters: the motion vector in the plane of the sky, the distance to the object, and the position of the object at a single point in time. For each triplet of transients, we perform a least-squares fit for these 6 parameters. We then use the fitted parametersto calcu- late the predicted positions of the object at the times of observation and the residuals from this prediction. To test the accuracy of this approximation, we exam- Fig.2.— (a) The locations of transients in the CSS. To allow viewability onlyevery 20th transient is shown. Geometry is as in ined the positions of real KBOs over single oppositions. Fig. 1(a). (b) The locations of transients in the SSS. Every 20th As an example, if 3 observations are made of Makemake transientisshown. GeometryisasinFig. 1(b). – at opposition, one month before, and one month after –theresultinglinearapproximationtotheorbitpredicts thepositionofMakemaketowithin25arcsecondsforthe fourmonthssurroundingopposition. Theapproximation is the worst for the the nearest and most eccentric ob- jects. But similar observations of the positions of 2005 31 s) EB299, for example – with an eccentricity of 0.51 and a ee 30 semimajor axis of 52 AU, and currently near perihelion degr 29 at25.7AU–predictthepositiontowithin45arcseconds n ( for the same period. For our Keplerian filter we conser- atio 28 vatively require that the maximum residualbetween the n linearsky-plane fit and the data is 50 arcsecondsorless. cli 27 e In addition, we require that the heliocentric distance re- d 26 trievedfromthe sky-planeapproximationbe largerthan 192 190 188 186 184 10 AU and that the orbital energy be within a factor RA (degrees) of 1.5 of the maximum for a bound object at that dis- tance. Experimentation with synthetic orbits suggests Fig.3.— All transients detected in the vicinity of Makemake. that these limits will detect real objects beyond 25 AU The path of the orbit of Makemake over the 8 year period of the survey is shown as a solid line. The 53 times that Makemake innearlyallcombinationsofthreeobservationsinanop- wasidentifiedasasingle-nighttransientareshownasfilledcircles. position season. After applying this filter, we have 140 Significantstructureisoftenseeninthetransientlocations,usually million triplets remaining in the CSS data and 3 billion associated with the edge of a field or with artifacts from bright in the SSS. stars. The final step in the Keplerian filter is full orbital Fitting of orbits is a complex non-linear problem. At- fitting. We use the code of Bernstein & Khushalani tempting to fit all combinations of three points in the (2000)whichefficientlycalculatesorbitsincludingplane- transient list (∼1018 attempted fits for CSS and ∼1019 tary perturbations for outer solarsystem objects. When forSSS)iscomputationallyprohibitive,soweseekmeth- the orbits of real KBOs are fit by this code, we obtain ods to minimize the number of orbital computations re- sub-arcsecond residuals. Again, to be conservative, we quired. First,wewillconsideronlycombinationsoftran- require that the orbit of a transient triplet, when fit by sientswithpairsoftransientsseparatedbynomorethan this code,yieldresidualssmallerthanonly5 arcseconds. 120days. Thisconstraintisnearly,thoughnotprecisely, Inadditionwerequirethattheretrievedheliocentricdis- equivalent to requiring three transients observed over a tanceoftheobjectbelargerthan15AU.Weputnoother single opposition season. As a short hand, however, we orbital constraints on the fit. willrefer to this constraintasrequiring3 transientsover FullKeplerianfilteringyields4.8milliontripletsinthe a single opposition. CSS data and 235million triplets in the SSS data which If we confine ourselves to a single opposition season, can be fit to Keplerian orbits. we can also confine ourselves to a significantly smaller area of the sky. A stationary object at 25 AU would 4 3.3. Further filtering Keplerian filtering yields impressive results. For the CSS, for example, out of 1019 possible combinations of Makemake three transients observed across the whole sky over a 7 2002 TX300 yearperiod,only one out ofeverytwo trillioncouldpos- Haumea sibly be fit to single-season Keplerian orbits for objects 2002 VE95 in the solar system beyond 25 AU. Nonetheless, these Nereid Huya Orcus Eris triplets are predominantly not real objects, but rather chance alignments of astrophysical transients or noise, so further filtering is needed. Therearemanypotentialmethodswecouldusetofur- ther filter the transients. We could, for example, require that all transients in a triplet have the same magnitude Fig.4.— The 8 objects detected in our survey. Each of the within limits, or we could tighten the constraints on the objects is found to be a previously discovered object in the outer finalorbitfitting. Butbecauseofourlargedatasetwith solarsystem. ThegeometryisasinFig. 1(a) the potential that anyrealobject will be observedmany times, we chose instead to require that there be addi- available in an opposition. tional observed transients which also fit the same orbit. All known outer solar system objects brighter than The simplest requirement to implement is that an or- V = 19.1 were detected with only 2 exceptions. Pluto bit be required to have 4, rather than 3, observations is in a region of the galactic plane which was never ob- during an opposition season. The large number of fields served. Quaoar was in the survey region for only the in which observations have been made 4 or more times first season when only 2 images were obtained. In later in a season (Figure 1) suggests that this method could seasons Quaoar moved into the galactic avoidance zone. be quite efficient at detecting objects. Nereid, an irregular satellite of Neptune which does not Tofindquadruplets,ratherthanjusttriplets,whichfit follow a heliocentric Keplerian orbit, was detected 14 a Keplerian orbit, we simply note that any real quadru- times and recovered during an opposition season when pletswillbemadeupofcombinationsofthetripletsthat its orbit was indistinguishable from a heliocentric orbit. wehavealreadycollected. We examineallofthe triplets The bright outer solar system objects with V . 19.1 that already passed our initial Keplerian filter and we wereallnotjustdetected,theywereobservedandwould look for pairs of triplets in in which two of the three have been detected independently multiple times. The transientsinthetripletarecommonacrossthepair,thus least well-observed object of this brightness was Orcus. defining 4 unique transients. We pass these quadruplets It was independently detected 381 times as a Keple- oftransientsthroughtheBernstein & Khushalani(2000) rian quadruplet. It was detected at each opposition in orbitfittingroutineandretainanyquadrupletsforwhich the calculated χ2 of the fit is below 10. We find 1192 which 4 or more observations were made of its region. Any one of these detections would have sufficed. The goodquadrupletfits toKeplerianorbitsinthe CSSdata other bright objects were independently detected com- and5515intheSSSdata. Thesenumbersaresufficiently parable or greater numbers of times. It appears that for small that we now examine the results in detail. bright objects, where detection is not limited by signal- 4. RESULTS to-noise, our algorithm is essentially 100% efficient at detecting objects when a sufficient number of observa- 4.1. CSS tions is obtained. Because of the nature of the data, Examining the locations of CSS quadruplets we see which are simply reported detections of transients, we that the majority are tightly clumped into a few dis- cannot rigorously prove this assertion of 100% algorith- tinct locations in the sky (Fig. 4). For each quadru- mic efficiency through population modeling, for exam- plet we examine if other quadruplets fit the same or- ple, as is often done in surveys. We instead point out bit by again running the full orbit fit through the that the real known objects in the sky provide, essen- (Bernstein & Khushalani 2000) routine. In this manner tially,amodelpopulationwiththousandsofindependent we find that the 1192 CSS quadruplets define 8 distinct quadruplet test cases. Our algorithm correctly recovers objects (Table 1). Each of these recovered objects is, in eachoftheseindependentcombinationsofrealobjectob- fact, a known bright object in the outer solar system. servationswith100%efficiencywhenthereare4ormore Requiring only 4 detections within an opposition season detections in a single opposition season. We conclude reduces the false positive rate in this dataset to zero. that for sufficiently bright objects in regions with good While wehavenorigorousmethodofassessingthe de- coverage,we will lose zero objects. tectionefficiencyofthissurvey,weestimatetheefficiency The The next faintest objects detected, 2002 TX300 by examining the detections of the brightest known ob- and Huya, with predicted V ∼ 19.6 and measured jects. Table 2 shows the brightest known objects in the V ∼ 19.4, each have more than a dozen individual de- solarsystembeyond25AU(withtheexceptionofUranus tections, but are only detected 4 times in an opposition and Neptune, which saturate in our survey). Looking at season twice and once, respectively. It is clear that ob- the initial unfiltered list of transients,we determine how jects of this magnitude will occasionally be missed not many times each object was detected as a transient in a because of algorithmic inefficiencies but simply because survey,and,furthermore,howmanyseparateoppositons they will not always be detected even when they are in were linked. Our algorithm was 100% efficient at re- the observed field. covering known objects when 4 or more detections were Between 19.5< V < 20.1 many objects were detected 5 threeorfewertimes per oppositionseasonbut only2002 extreme ecliptic latitudes or close to the galactic plane. VE95 (with a measured magnitude of V = 19.8) has a No bright objects in the outer solar system have been single 4-detection opposition. Even with just this single discovered with inclinations higher than the 44 degree opposition season in which 2002 VE95 is detected, the inclinationofEris,sowedonotanticipateanyundiscov- algorithm correctly identifies 2002 VE95 in the data. ered objects at the ecliptic poles. We conclude that the detection efficiency must be We estimate the probability that any bright objects nearly 100% for V . 19.1 for regions well covered in remain to be discovered. The sky density of bright ob- the survey and it must begin dropping around V ∼19.4 jects appears approximately uniform within 30 degrees until it reaches zero by V ∼20. While the precise shape of the ecliptic, while no bright objects have been found oftheefficiencycannotbedefined,thegeneralshapeand at higher latitudes (Schwamb et al. 2014). Our survey behavior appears clear. covered80% of the sky within 30 degrees of the ecliptic. The uncovered 20% is located within 20 degrees of the 4.2. SSS galacticplane. ThesurveyofSheppard et al.(2011)cov- eredapproximatelyone thirdofthe galacticplanebelow Atotalof5515Keplerian-fittingquadrupletsarefound ecliptic latitudes of 30 degrees to a depth of approxi- in the SSS survey. None of these quadruplets can be mately R = 21 with a completeness of approximately linked to another quadruplet, thus it appears that these 75%. There are 6 known objects brighter than V∼19, 4 arelikelyallfalsepositives. Withthesignificantlyhigher of which were in our survey region and one recovered in number density of detected transients in the SSS fields, the galactic plane survey. requiring 4 detections is insufficient for removing all of From these surveys, we estimate that the probability thefalsepositives. Clearly,sincenoneofthequadruplets that there is one or more remaining objects in the outer canbelinked,addingthatrequirementthatanobjectbe solar system brighter than V ∼ 19. To do so we con- linked 4 times in each of two opposition seasons reduces struct Monte Carlo models of KBO populations with a the false positive rate to zero. But we also find that a varyingskydensities. Selectingthe simulationwhichare less stringent requirement – that we link the object 5 compatible with the detections of the two data sets, we times inone oppositionratherthanjust 4 times –is also find that the probability that there is one additional ob- sufficient to drop the false positive rate to zero. It is ject yet to be detected is 32%. The probability that possible,ofcourse,thatoneormoreofthe 5515quadru- there are two or more is 10%. For all regions except for pletsisarealobjectthatissufficientlyfainttoonlyhave thegalacticplane,theselimitsextendtoverydistantob- 4detections andwhichis removedby the morestringent jects;abodyat10,000AU,forexample,wouldstillmove filtering. Thispossibilitydemonstratesthattheaddition ∼10 arcseconds in a month at opposition and would be of a more stringent detection criterion lowers our true detected in our analysis. detection threshold in this part of the survey. With no Schwamb et al. (2014) estimate that surveys of the effective way to perform followup observations of candi- outer solar system have been approximately 70% com- dates, such a lowering of the detection threshold is nec- plete to R = 19.5. This survey suggests that at the essary in order to remove false positives. brightest end the surveys to date have been even more Determining an efficiency for the SSS is more difficult efficient andthat the most likely scenario is that no new with a lack ofdetections ofrealobjects,but we makean bright objects remain to be discovered. estimate based on the experience with the CSS. First, In addition to demonstrating the only modest proba- the SSS images are about 0.5 magnitudes less deep than bility of the existence of additional bright objects in the CSS images. Second, the requirement of 5 detections outersolarsystem,thissurveydemonstratesthe relative raises the detection threshold. While a 5 detection re- quirement would have detected the V ∼ 19.4 objects in ease of detecting slowly moving solar system objects in transient surveys. the CSS,all fainter objects wouldhavebeen missed. We thusestimatethatourdetectionefficiencyisnearly100% for V . 18.6 and begins to drop by V ∼ 18.9. We have ThissearchformovingobjectsintheCRTScataloghas no reliable method of determining where the efficiency beensupportedbygrantNNX09AB49Gfromthe NASA droptozerobutwesuspectithappensquicklyfaintward Planetary Astronomy program. The CRTS survey of V =18.9. was supported by the NSF grants AST-0909182, AST- 1313422, and AST-1413600.The CSS survey is funded 5. DISCUSSION by the National Aeronautics and Space Administration Nonewbrightoutersolarsystemobjectsweredetected under GrantNo. NNG05GF22Gissued through the Sci- inthisallskysurveytoanapproximatelimitofV =19.4 ence Mission Directorate Near-Earth Objects Observa- inthenorthernsurveyandanestimatedlimitofV =18.9 tions Program. This serendipitous survey was conceived in the southern survey. If any bright objects remain to duringaserendipitousconversationatLSST“AllHands” be discovered in the outer solar system they must be at meeting between MEB and MJG. REFERENCES Bernstein,G.&Khushalani,B.2000,AJ,120,3323 Drake,A.J.,Djorgovski,S.G.,Mahabal,A.,Beshore,E.,Larson, Brown,M.E.2008,inTheSolarSystemBeyondNeptune, ed. S.,Graham,M.J.,Williams,R.,Christensen,E.,Catelan, M., M.A.Barucci,H.Boehnhardt,D.P.Cruikshank,& Boattini,A.,Gibbs,A.,Hill,R.,&Kowalski,R.2009, ApJ, A.Morbidelli,335–344 696,870 6 Larson,S.,Beshore,E.,Hill,R.,Christensen,E.,McLean,D., Schwamb,M.E.,Brown,M.E.,Rabinowitz,D.L.,&Ragozzine, Kolar,S.,McNaught,R.,&Garradd,G.2003, inBulletinof D.2010,ApJ,720,1691 theAmericanAstronomicalSociety, Vol.35,AAS/Divisionfor Sheppard,S.S.,Udalski,A.,Trujillo,C.,Kubiak,M., PlanetarySciences MeetingAbstracts#35,982 Pietrzynski,G.,Poleski,R.,Soszynski,I.,Szyman´ski,M.K.,& Rabinowitz,D.,Schwamb, M.E.,Hadjiyska,E.,&Tourtellotte, Ulaczyk,K.2011,AJ,142,98 S.2012,AJ,144,140 Trujillo,C.A.&Brown,M.E.2003,EarthMoonandPlanets, Schwamb,M.E.,Brown,M.E.,&Fraser,W.C.2014,AJ,147,2 92,99 Schwamb,M.E.,Brown,M.E.,&Rabinowitz,D.L.2009,ApJ, 694,L45 7 TABLE 1 Measuredpropertiesof the detected objects linked average a e i distance identified detections Vmag. (AU) (deg) (AU) object 51 17.06±.02 45.501± 0.003 0.160509± 0.00007 29.002± 0.001 51.867±.001 Makemake 47 17.37±.01 43.102±.002 0.195129± 0.00009 28.205± 0.001 51.217±.001 Haumea 33 18.95±0.08 39.273±.002 0.22397±0.00005 20.567± 0.001 47.706±.001 Orcus 28 18.53±0.01 67.82±0.02 0.4384±.0002 43.992±0.001 96.895±.006 Eris 8 19.5±0.1 43.21±0.03 0.118±0.002 25.855± 0.001 41.186±0.004 2002TX300 4 19.2±.1 31±16 0.4±0.5 2.34±0.06 30.37±.06 Nereid 4 19.3±0.1 39.74±0.03 0.29±.01 15.49±0.01 28.7±.01 Huya 4 19.8±0.1 39.7±0.4 0.30±0.01 16.33±0.02 28.15±0.02 2002VE95 TABLE 2 The brightestknown objects in the outer solar system name Vmag distance a e inc notes (avg) (AU) (AU) (deg) 134340Pluto 14.0 31.6 39.4 0.25 17.2 notinsurveyregion 136472Makemake(2005FY9) 16.9 52.1 45.45 0.16 29.0 53 detections, 8 oppositions linked 136108Haumea(2003EL61) 17.3 51.1 43.17 0.19 28.2 49 detections, 6 oppositions linked 136199Eris(2003UB313) 18.7 96.7 67.71 0.44 44.2 39 detections, 6 oppositions linked Nereid 18.8 30.0 30.1 0.01 1.76 14 detections, 1 opposition linked 50000Quaoar(2002LM60) 18.9 43.2 43.58 0.04 8.0 notinsurveyregion 90482Orcus(2004DW) 19.1 47.8 39.41 0.22 20.6 37 detections, 5 oppositions linked 55636(2002TX300) 19.6 41.5 43.10 0.12 25.9 17 detections, 2 oppositions linked 28978Ixion(2001KX76) 19.6 41.7 39.59 0.24 19.6 nodetections; lowgalacticlatitude 230965(2004XA192) 19.7 35.9 46.90 0.24 38.1 notinsurveyarea 38628Huya(2000EB173) 19.7 28.8 39.77 0.28 15.5 19 detections, 1 opposition linked 120178(2003OP32) 19.9 41.4 43.03 0.11 27.2 1detection (2010EK139) 20.0 39.9 70.26 0.54 29.4 1detection inSSS 84922(2003VS2) 20.0 36.5 39.29 0.07 14.8 6detections, maxof3insingleopposition 145451(2005RM43) 20.0 35.3 90.35 0.61 28.7 7detections, maxof3insingleopposition 90568(2004GV9) 20.1 39.1 42.17 0.08 22.0 toofarsouthforCSS;nodetections inSSS 145453(2005RR43) 20.1 38.6 43.13 0.14 28.5 6detections, maxof3insingleopposition 20000Varuna(2000WR106) 20.1 43.4 42.91 0.05 17.2 2detections 47171(1999TC36) 20.1 30.8 39.31 0.22 8.4 5detections, maxof2insingleopposition 145452(2005RN43) 20.1 40.7 41.37 0.02 19.3 4detections, maxof2insingleopposition 229762(2007UK126) 20.1 45.5 73.06 0.49 23.4 4detections, maxof1insingleopposition 55637(2002UX25) 20.2 41.8 42.55 0.14 19.5 7detections, maxof3insingleopposition 278361(2007JJ43) 20.2 41.8 48.21 0.16 12.1 toofarsouthforCSS;nodetections inSSS 55565(2002AW197) 20.3 46.6 47.54 0.13 24.3 2detections 174567Varda(2003MW12) 20.3 47.9 45.85 0.14 21.5 1detection 55638(2002VE95) 20.4 28.4 39.22 0.29 16.4 12 detections, 1 opposition linked 303775(2005QU182) 20.4 47.9 110.28 0.67 14.0 0detections (2004NT33) 20.4 38.1 43.41 0.15 31.2 notinsurveyregion 202421(2005UQ513) 20.4 48.8 43.22 0.15 25.7 nodetections 144897(2004UX10) 20.5 38.9 39.08 0.04 9.5 3detections, maxof1insingleopposition 119951(2002KX14) 20.5 39.5 38.74 0.04 0.4 nodetections 208996(2003AZ84) 20.5 45.4 39.40 0.18 13.6 nodetections