Solar System Objects Observed with TESS--First Data Release: Bright Main-Belt and Trojan Asteroids from the Southern Survey

arXiv:2001.05822v1 [astro-ph.EP] 16 Jan 2020 1 2 3 5 4 6 okl bevtr,Rsac etefrAtooyadEar and Astronomy for Centre Research Observatory, Konkoly Eötvös Pázmány H-1117 Loránd University, Péter sét edo-iw(o)o 24 of having (FoV) camera field-of-view each cameras, a wide-field four by attained ont h otenelpi pole ecliptic southern the to down oeigtesyfo h citclttd of latitude ecliptic the sector from observed Hemisphere, sky 13th TESS Ecliptic the Southern the sectors, the covering on 13 completing fields these primary after ended the Throughout observations 2019, of year 18, (S13). resonance first Earth, July mean-motion The around 1:2 on Moon). in TESS the orbit of with spacecraft (in orbits a observations two with continuous to nearly with corresponds of sector accordance TESS days each mission, 27 sectors” where “TESS roughly primary sectors) of its simply, Dur- terms of (or in 2018. years scheduled 25, two are commissioning, observations July first after on the and operations ing routine 2018 its 18, started April on launched I al nttt o srpyisadSaeRsac,70 Research, Space and Astrophysics for Institute Kavli MIT LEEovo oń nvriy ohr srpyia O EötvösELTE Astrophysical Group Gothard Loránd Research University, Cosmology Lendület Near-Field CSFK MTA ( Szen Szombathely, 9700 Group, Research Exoplanet MTA-ELTE OA YTMOJCSOSRE IHTS IS AARELEASE DATA FIRST – TESS WITH OBSERVED OBJECTS SYSTEM SOLAR ikre l 2015 al. et Ricker apal@szofi.net 1 rpittpstuigL using 2020 typeset 17, Preprint January version Draft h rniigEolntSre Satellite Survey Exoplanet Transiting The Andr Kiss https://tess.mit.edu/observations/ 1 l rvosgon-ae uvy,b tlata re fmagnit of ar order periods an least rotation at long Keywords: by with surveys, bodies ar ground-based of unamb amplitudes previous number with all and the bodies that periods of shows accurate cur number clearly where light aster the 9912 namely triples main-belt contains and characteristics, observed TSSYS-DR1, manner, the named homogeneous of release, o a end data TESS bright This of analysis the th tions. photometric in on of populations focusing sca release body data the small Bodies, various first degrees, place the six to present approximately enough we by large objects. plane is System disk ecliptic Solar debris the observe to of approach vicinity new the and Exo unique Transiting a the surveys, vides space-borne previous with Compared sP as ´ G , brMarton abor ´ al ´ 1,2,3 1. R , srnmcldtbss aaous–Atooia databases Astronomical – catalogues databases: Astronomical ehd bevtoa ehius htmti io planets Minor – photometric Techniques: – observational Method: ES a ucsflybeen successfully has TESS) , INTRODUCTION A br Szak obert ´ T E tl ATX .1.0 v. AASTeX6 style X 1 L , ◦ RJNATRISFO H OTENSURVEY SOUTHERN THE FROM ASTEROIDS TROJAN × aszl ´ 24 ◦ ats ´ Moln o ´ n h rs o is FoV gross the and 1 1 hscvrg is coverage This . sb Kiss Csaba , ar ´ 1,4 a y1A uaet Hungary Budapest, ány 1/A, ms Plachy Emese , β R sraoy zmahl,Hungary Szombathely, bservatory, 1 ≈ tiaB Attila , br Szab obert ´ ABSTRACT -6 asrSre,Cmrde A019 USA 02109, MA Cambridge, Street, Vassar hSine,H12 uaet okl hg Miklós út 15- Thege Konkoly Budapest, H-1121 Sciences, th ◦ meh .12 Hungary 112, u. h. Imre t , w oe r lorfre oas to referred These also mode). are stamp” “postage modes called (hence, two also minutes 2 is of mode cadence this a with the pre-selected observed in while are minutes sources mode 30 (FFI) image is plane. full-frame observations ecliptic so-called TESS the of scanned from cadence #1 south The Camera just while fields starred subsequent pole continuously the ecliptic #4 according southern Camera and, the design, at numbers survey camera the the to by sky, the identified in also 96 rectangle of size contiguous a nearly with a to equivalent atta h citcpaei vie by avoided is plane ecliptic the the considering that even fact mission, primary the during objects retrieved. is frame CCD whole the that implies cadence short locnieigternnzr cetiiis thousands eccentricities, non-zero their considering also xsrneo 2 of range axis than higher ∼ inclination an with ∼ objects glance, first At odi ´ 1,4 o ´ Udsac fTS oteSnadtesemi-major the and Sun the to TESS of distance AU 1 hsmsindsg losu oosreSlrSystem Solar observe to us allows design mission This ere r xetdt eosre,btdet the to due but observed, be to expected are degrees 6 1,4 Kriszti , 1,4 Zs , faBogn ofia ´ nS an ´ ude. lntSre aelt TS)pro- (TESS) Satellite Survey planet bevtos o ES ogcdnealso cadence long TESS, for observations: srain fsalSlrSystem Solar small of bservations ohrpre.Orcatalogue Our reported. both e hl t rmr iso avoids mission primary its While . ◦ 1 i n oinToa popula- Trojan Jovian and oid arneczky ´ × entl neetmtdby underestimated definitely e e bandadetatdin extracted and obtained ves − ehih fteSlrSystem Solar the of height le gosfnaetlrotation fundamental iguous 24 edo-iw nti paper this In field-of-view. e ar ´ surveys : 3 . ◦ Ufrtemi-etasteroids, main-belt the for AU 3 1,4 h niiulcmr osare FoVs camera individual The . slaKalup Csilla , RGTMI-ETAND MAIN-BELT BRIGHT : 1 seod:gnrl– general asteroids: , yl .Szab M. Gyula , ogcadence long 2,1 L , ∼ 7 Hungary 17, o ´ aszl ´ 5,6 degrees. 6 and , L. o ´ and 2

of objects with a few degrees of inclination are also pos- and what kind of object selection principles are available sible to be observed with the aforementioned spacecraft for a mission like TESS. In Sec. 3 we discuss the main attitude configuration. This limit of 6◦ . i is more strict steps of the data reduction and photometry, emphasizing for distant objects, such as Centaurs or trans-Neptunian the importance of differential image analysis. Sec. 4 sum- objects. marizes the structure of the available data products while According to earlier simulations (Pál, Molnár & Kiss in Sec. 5 we make a series of comparisons with existing 2018), one can expect good quality photometry of mov- databases aiming to collect photometric data series for ing targets down to V . 19 mag with a time resolution small Solar System bodies. Our findings are summarized of 30 minutes corresponding to the data acquisition cy- in Sec. 6. cle of the TESS cameras in full-frame mode. Although the cadence for the postage stamp mode frames would 2. OBJECT SELECTION allow a similar precision down to the brighter objects Regarding to the identification and querying Solar Sys- (i.e. V . 16 mag), the corresponding pixel allocation tem objects on TESS FFIs, one can ask two types of would be too expensive. In this aspect, TESS short questions: cadence observations are analogous the Kepler/K2 mission (Borucki et al. 2010; Howell et al. 2014) and simi- • When and by which Camera/CCD was my target larly, only pre-selected objects could be observed in this of interest observed? mode (Szabóet al. 2015; Pál et al. 2015). Specifically, • Which objects were observed by a certain Cam- one should allocate roughly a thousand pixel-wise stamp era/CCD during a given sector? if observations for a certain object are required. The rule-of-thumb for the apparent tracks of main-belt aster- We can also connect these questions to the K2 Solar Sys- oids on long cadence TESS images is the movement of tem observations. Namely, the first question is related ≈ 1 pixel/cadence (see also Fig. 2 in Pál, Molnár & Kiss to the computation of the pixel coverage of an asteroid 2018). Of course, NEOs and trans-Neptunian objects track, as it was done in the case of K2 mission while could have apparent speeds which are larger and smaller, observing pre-selected objects (see e.g. Pál et al. 2015; respectively. Kiss et al. 2016; Pál et al. 2016) and the second question The yield of such a survey performed by TESS is a se- is related to the observations of serendipitous asteroids ries of (nearly) uninterrupted, long-coverage light curves crossing large, contiguous K2 superstamps (Szabóet al. of Solar System objects – like in the case of previous 2016; Molnár et al. 2018). space-borne studies mentioned below. From these light In order to identify the objects which were observed curves, one can obtain fundamental physical character- by a certain Camera/CCD during a given sector, we istics of the bodies such as rotation periods, shape con- followed a similar approach as it was done in our K2 straints and signs of rotating on a non-principal axis - asteroid studies (Molnár et al. 2018; Szabóet al. 2016) with a much lesser ambiguity than in the case of ground- and in the case of simulations of TESS observations based surveys. This ambiguity is mainly due to the fact (Pál, Molnár & Kiss 2018). Our solutions are based on that ground-based photometric data acquisition is in- an off-line tool called EPHEMD, providing a server-side terrupted by diurnal variations – which yield not just backend for massive queries optimized for defining longer stronger frequency aliasing but higher fraction of long- time intervals and larger field-of-views within the same term instrumental systematics. In addition, the knowl- call (see Pál, Molnár & Kiss 2018, for more details). In edge of rotation period helps to resolve the ambigu- fact, the catalogue presented in this paper is retrieved by ity of rotation and thermal inertia (see e.g. Delbo et al. simply executing EPHEMD queries on per-CCD basis for 2015) in thermal emission measurements of small bod- each sectors. Due to the dramatic decrease of the aster- ies. Further combination of spin information with ther- oid density at higher ecliptic latitudes, in this catalogue mal data (see e.g. Müller et al. 2009; Szakáts et al. 2017; (DR1) we included only the observations from Camera Kiss et al. 2019)2 can therefore be an important initia- #1. tive. This paper describes the first data release, TSSYS- 3. DATA REDUCTION AND PHOTOMETRY DR1 of the TESS minor planet observations, based on As it was mentioned above, the whole data process- the publicly available TESS FFI data for the first full ing of this catalogue was based on the observations per- year of operations on the Southern Hemisphere. The formed by Camera #1 while surveying TESS sectors structure of this paper goes as follows. In the next sec- ranging from 1 up to 13. The processing has been car- tion, Sec. 2 we describe how the objects were identified ried out on a per-CCD basis, executing the same set of routines on the 13 × 4 = 52 blocks of images correspond- 2 https://ird.konkoly.hu/data/SBNAF_IRDB_public_release_note_2019March29.pdfing to a single-sector-single-CCD acquisition run. The TSSYS-DR1: Solar System objects as observed by TESS: first data release 3

Table 1. Quality flags and bits for the individual light curve data points. The data point flags are interpreted in a bitwise logical-or combination of these individual flags. The bit positions between 0 and 11 (values from 1 to 2048) are inherited from the FITS headers of the calibrated FFI data products, in accordance with the TESS Science Data Products Description Document(Tenenbaum & Jenkins 2018). The bit positions from 12 to 14 (mask values from 4096 to 16384) are specific for this particular data release and might be altered in the future. Note that bits at the position 1, 6, 8 and 9 (having a description in parentheses) are not used in the TESS FFI data products.

Bit position Value Description 0 1 Attitude Tweak. 1 2 (Safe Mode.) 2 4 Spacecraft is in Coarse Point. 3 8 Spacecraft is in Earth Point. 4 16 Argabrightening event. 5 32 Reaction Wheel desaturation Event. 6 64 (Cosmic Ray in Optimal Aperture pixel). 7 128 Manual Exclude. The cadence was excluded because of an anomaly. 8 256 (Discontinuity corrected between this cadence and the following one.) 9 512 (Impulsive outlier removed before cotrending.) 10 1024 Cosmic ray detected on collateral pixel row or column. 11 2048 Stray light from Earth or Moon in camera FOV. 12 4096 Formal photometric noise exceeds the threshold of 0.5 magnitude. 13 8192 Point rejected due to the presence of unexpected histogram region. 14 16384 Manual removal of an outlier point. pipeline providing the light curves is exclusively based dian image, employed as a median differential background on the FITSH package (Pál 2012). In this section we reference image (MDBRI). This MDBRI was then sub- summarize the main steps of the photometric process- tracted from all of the images acquired by the same ing. CCD in the same sector and the resulting differences were smoothed using a median window filtering com- 3.1. CCD-level steps bined with spline interpolation with a grid size of 64×64 pixels. This step allowed the derivation of large-scale Each of the CCD image series is processed as follows. background variations and nicely helped to minimize and Based on the available orbital and pointing data, we model the variations inducted by scattered light and zo- selected nearly a dozen of frames called individual me- diacal light. The derived background variations were dian reference frames (IMRFs) spanning a ∼ 2-day pe- then subtracted from all of the images and image con- riod long interval close to the center of the observations volution were applied between the MDBRI and these evenly. These frames coincide for all of the four CCDs background-subtracted images. Note that this step does for a given sector, i.e., these correspond to the same ca- not subtract the intrinsic background since such a back- dence and usually have a time step of 4 hours between ground practically does not exist for TESS images due each frame. Another set of criteria was based on the con- to the very strong confusion and large pixel size. The straint that both the Sun and the Moon should have been image convolution steps correct not only for the PSF below the sun-shade of the spacecraft, meaning that both variations but for the offsets inducted by the differen- the Sun-TESS-boresight and the Moon-TESS-boresight tial velocity aberration as well. The latter one can be as angle should have been larger than 90◦. This combined large as one tenth of a pixel throughout a sector and it selection criteria ensured the lack of stray light in all of is the most prominent further away from the spacecraft the cameras at the same time while the duration ensured boresight (which includes Camera #1 CCDs #3 and #4, an expected coverage of several tens of pixels of a main- which are the closest to the ecliptic plane). Once the con- belt asteroid while still keeping the differential velocity volved MDBRIs are derived, the resulting residual image aberration at a considerably low level. In addition to the was processed by a spline-smoothed median window fil- aforementioned selection criteria, if a prospective frame tering with a block size of 1 × 64 pixels. This filtering was flagged with a “reaction wheel desaturation event” removed the vertical stripes exposed in the TESS CCDs (see Tenenbaum & Jenkins 2018), the next or previous in parallel with the increased stray light. The steps of frame was selected instead. the aforementioned processing are displayed in Fig. 1 via In the next step, IMRFs were used to create a me- 4 Original difference Background Background variations removed Simple difference difference Convolution & Stripes removed

Figure 1. Panels showing the various stages of the image-level data processing using asteroid (2429) Schurer as an example, observed during Sector 2, by Camera #1, CCD #3. The left column shows the 10′ ×10′ vicinity of the target, the middle column shows the neighbourhood (3.7◦ × 1.4◦) area while the right column is the full CCD frame, all at the TESS FFI cadence 2018247095941 (JD 2458365.92767). Images in the first row show the original unprocessed data. The second row is the difference in the background structure with respect to a frame where the Earth and Moon were below the sun-shade of TESS. The large-scale variations due to the stray light are clearly visible. The third row shows the difference between the first two rows. The fourth row shows the naive difference between the target image and the median differential-background reference image. The residuals due to the uncorrected differential velocity aberration are clearly visible. The fifth row shows the results of the image convolution followed by subtraction. This step also makes the TESS-specific, but otherwise comparatively faint vertical CCD stripes visible. In addition, the left stamp in this row shows that the sources, even ones brighter than the target objects are completely removed, with some residual structure only visible at a much brighter star at the upper-right corner of this stamp. Images in the sixth row show the results of the stripe removal process. The target at the center is clearly visible. TSSYS-DR1: Solar System objects as observed by TESS: first data release 5 the example of (2429) Schurer. on fitting a sinusoidal variation in parallel with the decor- relation of the phase angle variations up to the second 13.4 order (see also Sec. 4.2 later on). The most dominant 13.5 frequency was computed by interpolating in the vicinity

13.6 of the frequency spectrum were the root mean square of

13.7 the aforementioned ﬁt residual was found to be the small- est (see Section 4.2). The light curves were then folded 13.8

TESS Magnitude and binned after phase angle correction. Folding was per- 13.9 formed with two periods, one corresponding to the domi- 14 nant frequency while the other period we used was twice 14.1 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 the dominant period, assuming a double-peaked light- Rotation phase (period: 8.99931 hours) curve generated by the rotation of an elongated body. In total, 9912 objects are included in the present data Figure 2. Folded light curve of (692) Hippodamia, hav- release, for which accurate light curve information were ing a rotation period P = 8.9993 h. While this rotation derived with a reasonable signiﬁcance. Out of these 9912 period satisﬁes the Nyquist criterion, the phase coverage objects, 125 have only provisional designations and there- is not uniform due to the P/C ratio of ∼ 18. fore are not numbered minor planets.

3.3. Sampling characteristics 3.2. Target astrometry and photometry The observing strategy of TESS is highly deterministic These cleared images were then used as the input compared to many of the surveys and ground-based ob- of the aperture photometry where the centroids are servations. Namely, the cadence is strictly C =0.5h for computed by the EPHEMD tool with TESS set as the a nearly uninterrupted observing period of L . 25−28 d. observer’s location. Absolute astrometric plate solu- This property implies the Nyquist criterion which does tions have been derived using the Gaia DR2 catalogue not allow the unambiguous rotation characterization for (Gaia Collaboration et al. 2016, 2018) while the pro- objects having a period of P ≤ 2C = 1 h. This is inter- jection function was obtained by a third-order Brown- esting for small objects, having a size of approximately Conrady model on the top of tangential projection with or smaller than the spin barrier limit of ∼ 100 m: such additional refinements using a third-order polynomial ex- objects can rotate faster than ∼ 2.2 h (Pravec & Harris pansion. The fluxes are extracted using the proper way 2000). needed to interpret convolved differential images (see Eq. The strict cadence also yields sampling artefacts of ob- 83 in Pál 2009). The zero-point of the light curves were jects having a rotation period which is close to the integer obtained using a global fit against the GAIA DR2 RP multiple of the cadence C. For instance, (692) Hippo- magnitudes. Due to the almost perfect overlap of the damia has a rotation period of P = 8.9993 hours, which TESS and GAIA RP passbands – see also Fig. 1 in is almost exactly 18 times longer than the TESS FFI ca- Ricker et al. (2015) and Fig. 3 in Jordi et al. (2010)– dence (see Fig. 2). In order to characterize the strength this yields a good and accurate match of the zero point. of this sampling effect, let us assume that the period of However, offsets can be presented due to the PSF vari- the object is P = nC + ε where ε represent a short time ations across the field-of-view of the fast TESS optics. difference and n is an integer number (e.g. n = 18 and We note here that the formal uncertainties does not in- ε = −0.0007 h for (692) Hippodamia). In order to fully clude the respective uncertainty of this offset. Individ- sample the rotational phase domain, one should expect ual light curve files were then generated by transposing that the second instance (t = C) has the same phase the photometric results and flagged afterwards accord- as the last phase after at or around the kth rotation ing to the quality flags presented in the TESS FFI head- where for the total observation timespan is L ≈ kP . ers (Tenenbaum & Jenkins 2018). Light curves with in- Here k is also an integer, the total number of rotations sufficient number of data points were removed from the covered during the observations. The phases are equal database and the post-filtering of these remaining light if (knC/P ) − (C/P ) = k, from which we can compute curves also added additional types of quality flags (see that CP/L should be smaller than |ε|. This limit for Table 1). This post-filtering process includes exclusion (692) Hippodamia is |ε|692 = (1/2 h)·(8.9993 h)/(25 d) ≈ of the points with high formal photometric uncertainty, 0.0075 h, definitely larger than |ε| = 0.0007h, we ob- outlier detection based on histogram clipping and man- tained above for this object, resulting in a stroboscopic ual removal of points in the most prominent cases. effect. This stroboscopic effect is also present in K2 The filtered light curves were then analyzed by per- observations, see e.g. the case of (14791) Atreus in forming a period search. This period search was based Szabóet al. (2017). 6

354 220281 9.2 18.1

18.2

18.3 9.3 18.4

18.5 9.4 18.6

TESS Magnitude TESS Magnitude 18.7 9.5 18.8

18.9

9.6 19 -12 -11 -10 -9 -8 -7 -6 -5 -15-14-13-12-11-10-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 Time (JD, t-2458480) Time (JD, t-2458370)

9.25 18.1

18.2 9.3 18.3 9.35 18.4

9.4 18.5

9.45 18.6

TESS Magnitude TESS Magnitude 18.7 9.5 18.8 9.55 18.9

9.6 19 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 Rotation phase (period: 4.27735 hours) Rotation phase (period: 6.34017 hours)

0.09 0.174

0.08 0.172

0.07 0.17 0.06 0.168

Residual 0.05 Residual 0.166 0.04

0.03 0.164

0.02 0.162 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Frequency (c/d) Frequency (c/d)

Figure 3. Object light curve plots for (354) Eleonora (left column of 3 individualplots) and (220281) 2003BA47 (right column of 3 individual plots). These plots are available for all of the 9912 objects presented in this study.

4. DATABASE PRODUCTS AND STRUCTURES In the following, we describe these data prod- Per-object data products were saved and stored in ac- ucts in more detail. The full data release is cordance with the aforementioned steps. The primary going to be available from the web address of data products include four ﬁles per object, namely: http://archive.konkoly.hu/pub/tssys/dr1/.

• the light curve file, containing the time series of the 4.1. Light curve files brightness measurements for a particular object; The light curve files basically represent the post- transposition stage of the photometric output. Since • the residual r.m.s. frequency spectrum; photometry is performed on a per-frame basis and a sin- fiphot • a metadata file (best-fit rotation frequency, peak- gle call to the photometric task (FITSH/ ) per- to-peak amplitude, light curve type); and forms the flux extraction for all of the minor planets associated with that particular frame, light curve files also • validation sheets, including the plots of the afore- include the target name, the timestamp, the (x, y) pixel mentioned data products, coordinates and estimations for the background structure. Although differential imaging analysis and the • and per-object summary plots and slides, including subsequent photometry yields zero local background on the folded light curve with the most likely rotation subtracted images in theory, some artefacts – such as period. stray light spikes, unmasked blooming, prominent resid- TSSYS-DR1: Solar System objects as observed by TESS: first data release 7 ual structures around bright but unsaturated stars – Version 3 of Nesvorný, Broˇz& Carruba (2015) and the cause deviations from the zero level. Such information is most recent version of the Asteroid Lightcurve Database therefore useful for further filtering of outliers and asso- (LCDB, Warner, Harris & Pravec 2009). Of course, the ciate quality flags to the photometric data points. In ad- overlap with neither of the aforementioned databases are dition to the aforementioned data, light curve files are ex- complete and there are only 1563 objects for which both tended with three additional columns showing the phase proper orbital elements and LCDB data are available. angle values, observer-centric distances and heliocentric distances. 4.4. Validation plots For a quick manual vetting of the results of the pho- 4.2. Residual spectra tometric analysis, we create a four-panel summary plot Residual spectra are generated by frequency scanning for each object. The four plots are the unfolded light with a step size and coverage in accordance with the curve, the residual spectrum, the folded light curve with TESS sector time-span and the TESS FFI cadence, re- the dominant period and the folded light curve with the spectively. Namely, the total time-span of ∼ 27 days double of the dominant period. on average imply a stepsize of ∆f = 0.01c/d while the Nyquist criterion maximizes the scanning interval in 4.5. Object light curve plots and slides f = 24c/d. The residual spectrum is then computed max These plots contain the same information as the valida- for a certain input frequency f by minimizing the param- tion plots, but in a bit different arrangement and these eters A, B, and ki (i =0, 1, 2) for the model function display only a single folded light curve with the most 2 i likely rotation period. The plots also show this rotation m(t)= X ki[α(t) − α0] + (1) period in the units of hours. We note here that the time i=0 instances for both the plots and all of the light curve data +A cos[2πf(t − T0)] + B sin[2πf(t − T0)]. products are given in in Julian Days (JDs). As an ex- where m(t) is the observed magnitude (corrected for the ample, two of such object light curve plots are displayed variations in the solar and observer distances) at the in- in Fig. 3 for the objects (354) Eleonora and (220281) stance t, α is the phase angle, α0 is the mean phase angle 2003BA47. These objects represent the bright end and throughout the observations, and T0 is an approximate the faint end of our catalogue. mid-time of the observations. The actual values of α0 5. COMPARISON WITH EXISTING DATABASES and T0 do not alter the residuals (hence the spectra), however, setting the aforementioned values helps to min- 5.1. Asteroid Lightcurve Database – LCDB imize the numerical round-off errors and k0 can also be The most comprehensive database available in the lit- interpreted as a mean brightness magnitude throughout erature is the Asteroid Lightcurve Database4 (LCDB, see the observations. Warner, Harris & Pravec 2009). The most recent (Au- 4.3. Metadata gust 2019) release of this database contains 4842 objects for which a valid rotation period and brightness In the case of the light curve and residual spectrum variation amplitude is associated5. While this amount analysis, metadata represents the rotation frequency of data is nearly half of the entries available in the (and/or equivalently, the rotation period), the charac- TESS minor planet data, the LCDB cites 2788 biblio- teristics of the light curve shape and the peak-to-peak graphic sources (concerning the entire database), there- amplitude as well as any associated external database. fore one should consider the inhomogeneity while inter- While the processing scripts store metadata in separate preting LCDB statistics. However, we expect that the files in a form of key-value pairs, the final data prod- aforementioned quality constraints of selecting 4842 ob- uct includes a list of concatenated metadata in a tabular jects ensure the robustness of the data products. form. In total, we identified 624 objects which are avail- In addition, this metadata table is extended with var- able both in TSSYS-DR1 and LCDB (with sufficiently ious large asteroid database information for convenience strong qualification). We note here that there are 1535 and further analysis. This information can be used to objects available both in TSSYS-DR1 and LCDB if we create additional types of statistics and have estimations do not consider the amplitude quality criteria mentioned for biases (see Sec. 5 for examples). In our published above. In Figs. 4 and 5 we displayed the rotation fre- database, we included the most recent version of the quency and amplitude correlations, respectively, between synthetic proper orbital elements of Kneˇzević& Milani (2000), as available online3, the asteroid family catalog 4 http://www.minorplanet.info/lightcurvedatabase.html 5 We note here that incomplete amplitude information but set- 3 https://newton.spacedys.com/astdys2/ tled rotation periods are available for 20462 objects. 8

2h 3h 6h 12h 24h 48h 120h 240h 480h

1000 480h

0.1 240h

h 120 100

48h

1 24h 10 12h

3h LCDB rotation frequency (c/day) and period (hours) 10 1 2h 0.1 0.2 0.5 1 2 5 10 10 1 0.1 Rotation frequency ratio: fLCDB/fTESS TESS rotation frequency (c/day) and period (hours)

Figure 4. The left panel shows the rotation periods of the 624 objects for which reliable rotation characteristics (i.e. periods and amplitudes) are available both in the LCDB catalogue and the TESS observations presented in this paper. The thick line and the two dashed lines correspond to the same rotation frequencies as well as the 1 : 2 and 2 : 1 ratios, respectively. The right panel displays the histogram of the frequency ratios of the objects available both in the LCDB catalogue and the presented TESS minor planet catalog. In total, ∼ 80% of the matched objects have the same derived rotation periods while in the case of ∼ 8% of the objects, the newly derived preferred periods are either the double or half of the periods available in the LCDB. TESS measurements clearly identiﬁed longer rotation periods for the majority of the remaining ∼ 60 objects.

150

100

0.1 50 LCDB light curve amplitude (magnitudes)

0 0.1 0.2 0.5 1 2 5 10 0.1 1 Amplitude ratio: ALCDB/ATESS TESS light curve amplitude (magnitudes)

Figure 5. Left panel: light curve peak-to-peak amplitudes for the 624 objects where rotation characteristics are available both in the LCDB catalogue and the TESS observations presented here. Right panel: the histogram of the distribution of the amplitude ratios. The thick vertical line shows the unity ratio while the thin vertical dashed line at ∼ 1.076 shows the median value of the amplitude ratios. the two databases. Considering the rotation periods, we tion here that TESS clearly identiﬁes the objects with found that the agreement is perfect for ∼ 80% of the ob- longer periods better, suspecting an unclear origin of the jects while there are a few dozens of objects where the otherwise shorter reported periodicity in LCDB (see the double-peaked ambiguity yields a 1 : 2 or 2 : 1 ratio. The points above the 1 : 1 and 2 : 1 line on the left panel of amount of such ambiguities is roughly the the same (19 Fig. 4 or the histogram distribution at the right tail on vs. 28) for the two ratios. Otherwise, it is worth to men- the right panel of the same Figure). TSSYS-DR1: Solar System objects as observed by TESS: first data release 9

Regarding to the interpretation of the correlations be- jects. These classes include not only main-belt and tween amplitudes (see Fig. 5), the larger amplitudes Trojan asteroids but trans-Neptunian objects (Pál et al. present in the LCDB is a clear signature of the bias in 2015), irregular satellites of giant planets (Kiss et al. the TESS observations. Namely, TESS observes minor 2016; Farkas-Takcs et al. 2017), and the Pluto-Charon planets close to the opposition, i.e. at small phase an- system (Benecchi et al. 2018). K2 observations also im- gles while LCDB contains many kinds of observations plied the discovery of the satellite of (225088) 2007 OR10 (yielding better coverage in phase angles), not just ones (Kiss et al. 2017) when its slow rotation was detected close to the opposition. According to the expectations (Pál et al. 2016). (Zappala et al. 1990), higher phase angles would yield With the exception of the discovery and photome- higher amplitudes, which can explain the shift in the try of the trans-Neptunian object (506121) 2016BP81 correlation diagram and the corresponding histogram. (Barensten et al. 2017), all of these object classes were However, one should note that because of this TESS- measured as targeted observations, i.e. with pre- specific observing constraint as well as due to the fact allocated K2 target pixel files (arranged into special that the presented data release contains only a single boomerang-shaped pixel blocks). In the case of main- epoch while LCDB aggregates data from many observ- belt and Trojan asteroids, there are examples of tar- ing runs, such a statistical comparison between TESS geted observations (Marciniak et al. 2019; Szabóet al. and LCDB amplitudes needs to be considered tentative. 2017; Ryan, Sharkey & Woodward 2017) as well as While the presented TESS data series are highly homo- photometry on contiguous superstamps (Szabóet al. geneous, it shows an amplitude characteristics only for a 2016; Molnár et al. 2018) when asteroids serendipitously single observing geometry, leaving many aspects of shape crossed these celestial areas. However, the data reduction characteristics ambiguous. pattern does not differ significantly for pre-allocated re- ductions and the analysis of contiguous superstamps with 2h 3h 6h 12h 24h 48h 120h 120h the exception of the aforementioned querying of the objects (by tools like EPHEMD) in the latter case. See, e.g., K2/Trojans Szabóet al. (2017) for a detailed description about the K2/M35 field 48h data reduction for K2 minor planet observations obser- K2/Uranus field 65210 vations. 1 24h In order to compare the objects observed by any 24534 initiative of the K2 Solar System Surveys with this 12h 24537 recent TESS-based photometry, we identifies 6 main- belt and Trojan objects that were observed both by 6h 37750 K2 and TESS. These were (24534) 2001CX27, (24537) 45086 42573 2001CB35, (37750) 1997BZ, (42573) 1997AN1, (45086) 3h 10 1999XE46 and (65210) Stichius. We found that the de- 2h Kepler/K2 rotation frequency (c/d) and period (hours) rived rotation periods match perfectly in 5 of the 6 cases, see Fig. 6. There was only a slight offset for (65210) 10 1 Stichius, due to its faintness and long rotation period of TESS rotation frequency (c/d) and period (hours) ∼ 32 hours. Figure 6. The 6 minor planets for which both Kepler/K2 and TESS measurements are available. 3 out of these 6 5.3. Period statistics objects are Jupiter Trojans while the another two are In Fig. 7 we displayed the histograms of the detected main-belt asteroids. With the exception of the Tro- rotation periods for this TESS-based asteroid survey, jan asteroid (65210) Stichius, the periods match within the LCDB and the K2 serendipitous main-belt aster- 1.5%. The agreement for (24534) 2001CX27 and (42573) oid detections on the M35 and Neptune-Nereid fields 1997 AN1 are less than one tenth of a percent. (65210) (Szabóet al. 2016), as well as on the Uranus field Stichius show a difference of ∼ 8% between the derived (Molnár et al. 2018). A tentative fit in the long-period periods. part of these histograms clearly show that both ground- based and shorter duration but otherwise uninterrupted space-borne measurements underestimate the number of 5.2. K2 Solar System Studies – K2SSS objects in the population of slow rotators. Therefore, we While having scanned various fields close to the eclip- can safely conclude that the nearly one-month long con- tic plane, the K2 mission (Howell et al. 2014) also pro- tinuous data acquisition of TESS would provide us the vided a highly efficient way to provide uninterrupted most unbiased coverage and confirmation of slowly rotat- observations for various classes of Solar System ob- ing asteroids. However, it is still an interesting question 10

h h h h h h h h h 2 3 6 12 24 48 120 240 480 curate comparison with LCDB. We should also express our hope that the extended mission of TESS would include wide coverage of the ecliptic plane, further expand-

100 ing our collection of asteroid observations and increase the number of multi-epoch observations. Facilities: TESS (Ricker et al. 2015), Gaia DR2 10 (Gaia Collaboration et al. 2018) Software: FITSH (P´al 2012), EPHEMD

1 10 1 0.1 Rotation frequency (cycles/day) and period (hours) We would like to thank Brian D. Warner for the careful review of our paper and his highlights of many aspects of Figure 7. The number of objects as the function of their light curve interpretation and caveats. A.P. would like to periods, provided by various databases. The black curve thank Matt Holman, George Ricker, Roland Vanderspek, shows the period distribution for TSSYS-DR1 (9912 ob- Joel Villasenor and Deborah Woods for the fruitful dis- jects), the red curve shows the period distribution for the cussions about the astrometry of TESS full-frame images 4842 LCDB objects for which a valid rotation period and and Solar System topics in general. This paper includes brightness variation amplitude have been derived at the data collected by the TESS mission. Funding for the same time. The blue curve shows the period distribution TESS mission is provided by the NASA Explorer Pro- for 113 serendipitous main-belt asteroids provided by the gram. This project has been supported by the Lendület analysis of three K2 superstamps. The thin dashed lines Program of the Hungarian Academy of Sciences, project guide the eye to provide a tentative slope at the long No. LP2018-7/2019. Additional support is received from period (low frequency) parts of these distributions. the K-125015 and GINOP 2.3.2-15-2016-00003 grants of the National Research, Development and Innovation Of- where the cut-off of TESS is, above which the rotation fice (NKFIH, Hungary). Zs.B. acknowledges the sup- period statistics become significantly biased. The diver- port provided from the National Research, Development gence between the LCDB and TESS histograms stars and Innovation Fund of Hungary, financed under the at rotation periods of 8 − 10 hours. Below this period, PD17 funding scheme, project no. PD-123910. L.M. the two statistics nicely agree down to the periods of was supported by the Premium Postdoctoral Research ∼ 2hours range. Program of the Hungarian Academy of Sciences. Cs.K. was supported by the UNKP-19-2´ New National Ex- 6. SUMMARY cellence Program of the Ministry of Human Capacities. Partial funding of the computational infrastructure and In this paper we presented the first data release of the database servers are received from the grant KEP-7/2018 complete Southern Survey of the Transiting Exoplanet of the Hungarian Academy of Sciences. Gy.M.Sz. was Survey Satellite in terms of analysis of bright, main-belt supported by the Hungarian NKFI Grant K-119517 and and Trojan asteroids crossing the field-of-view of Cam- the City of Szombathely under Agreement No. 67.177- era #1. This survey triples the number of asteroids with 21/2016. The work of G.M. was supported by the PD- accurately determined rotation characteristics. Another 128360 project of the National Research, Development advantage of the presented catalogue is that it is fully and Innovation Office, Hungary. This work has made homogeneous considering both data acquisition and data use of data from the European Space Agency (ESA) mis- processing principles. Further fine-tuning in the pipeline sion Gaia, processed by the Gaia Data Processing and presented here is also possible, and we have the intention Analysis Consortium (DPAC). Funding for the DPAC to process and add further object classes, including Cen- has been provided by national institutions, in particu- taurs, trans-Neptunian objects and near-Earth objects lar the institutions participating in the Gaia Multilateral (see also Milam et al. 2019). Agreement. TESS is now observing the Northern Hemisphere, opening the possibilities to re-observe many of the ob- REFERENCES jects presented in this data release with a completely different observing geometry with respect to the spin-axis orientation of these bodies. Such further observations Barentsen, G.; Pál, A.; Sárneczky K., & Molnár, L. 2017, would help us to interpret the derived light curve char- MPC102428 Benecchi, S. D.; Lisse, C. M.; Ryan, E. L.; Binzel, R. P.; acteristics, specifically the amplitude in a more accurate Schwamb, M. E.; Young, L. A. & Verbiscer, A. J. 2018, Icarus, manner and therefore helping the analysis for a more ac- 314, 265 TSSYS-DR1: Solar System objects as observed by TESS: first data release 11

Borucki, W. J., Koch, D., Basri, G., et al. 2010, Science, 327, 977 Molnár, L.; Pál, A.; Sárneczky, K.; Szabó, R.; Vinkó, J.; Szabó, Delbo, M., Mueller, M., Emery, J. P., Rozitis, B., Capria, M. T., Gy. M.; Kiss, Cs.; Hanyecz, O.; Marton, G.; Kiss, L. L. 2018, 2015, Asteroid Thermophyiscal Modeling, in: Asteroids IV, eds. ApJS, 234, 37 P. Michel, F.E. DeMeo & W.F. Bottke, University of Arizona Müller, T. G., Lellouch, E, Böhnhardt, H. et al. 2009, EM&P, Press, Tucson 105, 209 Farkas-Takács, A.; Kiss, Cs.; Pál, A.; Molnár, L.; Szabó, Gy. M.; Nesvorný, D.; Broˇz, M. & Carruba, V. 2015, in Asteroids IV, Hanyecz, O.; Sárneczky, K.; Szabó, R.; Marton, G.; Mommert, Patrick Michel, Francesca E. DeMeo, and William F. Bottke M.; Szakáts, R.; Müller, T. & Kiss, L. L. 2017, AJ, 154, 119 (eds.), University of Arizona Press, Tucson, 895 pp. Gaia Collaboration: T. Prusti et al., 2016, A&A, 595, A1 Pál, A. 2009, PhD thesis (arXiv:0906.3486) Gaia Collaboration; Brown, A. G. A.; Vallenari, A.; Prusti, T.; de Pál, A. 2012, MNRAS, 421, 1825 Bruijne, J. H. J.; Babusiaux, C.; Bailer-Jones, C. A. L., 2018, Pál, A.; Szabó, R.; Szabó, Gy. M.; Kiss, L. L.; Molnár, L.; Sárneczky, K. & Kiss, Cs. 2015, ApJL, 804, 45 A&A, 616, A1 Pál, A.; Kiss, Cs.; Müller, Th. G.; Molnár, L.; Szabó, R.; Szabó, Howell, S. B., Sobeck, C., Haas, M., et al. 2014, PASP, 126, 398 Gy. M.; Sárneczky, K. & Kiss, L. L. 2016, AJ, 151, 117 Jordi, C.; Gebran, M.; Carrasco, J. M.; de Bruijne, J.; Voss, H.; Pál, A., Molnár, L. & Kiss, Cs. 2018, PASP, 130, 114503 Fabricius, C.; Knude, J.; Vallenari, A.; Kohley, R.; Mora, A, Pravec, P. & Harris, A. W. 2000, Icarus, 148, 12 2010, A&A, 523, A48 Ricker, G. R. et al. 2015, J. Astron. Telesc. Instrum. Syst. Vol. 1, Kiss, Cs.; Pál, A.; Farkas-Takács, A. I.; Szabó, Gy. M.; Szabó, R.; id. 014003 Kiss, L. L.; Molnár, L.; Sárneczky, K.; Müller, T. G.; Ryan, E. L.; Sharkey, B. N. L. & Woodward, Ch. E. 2017, AJ, Mommert, M.; Stansberry, J. 2016, MNRAS, 457, 2908 153, 116 Kiss, Cs.; Marton, G.; Farkas-Takács, A.; Stansberry, J.; Müller, Szabó, R., Sárneczky, K., Szabó, Gy. M., et al. 2015, AJ, 149, 112 Th.; Vinkó, J.; Balog, Z.; Ortiz, J.-L. & Pál, A. 2017, ApJL, Szabó, R.; Pál, A.; Sárneczky, K.; Szabó, Gy. M.; Molnár, L.; Kiss, 838, 1 L. L.; Hanyecz, O.; Plachy, E. & Kiss, Cs. 2016, A&A, 596, A40 Kiss, Cs.; Szakáts, R.; Marton, G.;Farkas-Takács, A.; Müller, Th.; Szabó, Gy. M.; Pál, A.; Kiss, Cs.; Kiss, L. L.; Molnár, L.; & .; Al´ı-Lagoa, V. 2019, Small Bodies: Near and Far database Hanyecz, O.; Plachy, E.; Sárneczky, K.; Szabó, R. 2017, A&A, of thermal infrared measurements of small Solar System bodies 599, A44 Release Note: Public Release 1.2, 2019 March 29 Szakáts, R.; Kiss, Cs.; Marton, G.; Varga-Verebélyi, E.; Müller, Kneˇzević, Z. & Milani, A. 2000, CeMDA, 78, 17 T. & Pál, A. 2017, European Planetary Science Congress, id. Marciniak, A.; Al´ı-Lagoa, V.; Müller, T. G.; Szakáts, R.; Molnár, EPSC2017-223 L.; Pál, A.; Podlewska-Gaca, E.; Parley, N. et al. 2019, A&A, Tenenbaum, P. & Jenkins, J. M., 2018: TESS Science Data 625, A139 Products Description Document, EXP-TESS-ARC-ICD-0014 Milam, S. N.; Hammel, H. B.; Bauer, J.; Brozovic, M.; Grav, T.; Rev D. Holler, B. J.; Lisse, C.; Mainzer, A.; Reddy, V.; Schwamb, M. Warner, B. D., Harris, A. W. & Pravec, P. Updated 2019 August E.; Spahr, T.; Thomas, C. A.; Woods, D 2019, Combined 14, 2009, Icarus, 202, 134 Emerging Capabilities for Near-Earth Objects (NEOs), White Zappala, V.; Cellino, A.; Barucci, A. M.; Fulchignoni, M.; Paper for NASA on astrophysics assets, e-print Lupishko, D. F. 1990, A&A, 231, 548 (arXiv:1907.08972)