Page History

Introduction

These release notes describe the schema and content of the eighth version of the TESS Input Catalog (TICv8) produced entirely by the Target Selection Working Group (TSWG).  TICv8 was finalized and prepared for delivery to the TESS Science Office (TSO) on 2019 April 15.  Updates to TIC v8 were posted in July  2020 (TIC v8.1) and September 2021 (TIC v8.2):  details are given below.

TICv8 has a number of minor issues (see below) that have not been fixed in this version due to time constraints during preparation. Specific details of the method of production, and the contents of TICv8, is described on the arXiv (Stassun et al. (2019); https://arxiv.org/pdf/1905.10694.pdf). The full documentation for TICv7, and earlier TIC versions, can currently be found in Stassun et al. (2018, AJ, 156, 102), and on the arXiv at (https://arxiv.org/abs/1706.00495).

The design of TICv8 is nearly identical to TICv7, in that the original columns and their formats are the same, but there have been updates to various parameter calculations, and additional columns have been included  to allow for more user flexibility. The TIC IDs have not been changed, and all future deliveries of the TIC will use the same TIC IDs for specific objects, preserving backwards compatibility. New objects added to the TIC will always receive new IDs. Objects found to be spurious will be marked with a disposition of ARTIFACT, but will retain their TIC ID and remain in the catalog. TIC IDs will always be unique, and a new TIC object will never receive the ID of an existing TIC object.

Document Update History

This document should be considered a ‘living document’, which will be updated as issues are identified, or as descriptions are made more accurate. The update history for this document is:

• 2018 August: original release
• 2020 July: update for TIC v8.1, which identified ~700 targets with new dispositions, and CTL v8.1
• 2021 September: update for TIC v8.2

Major changes in TICv8 compared to TICv7

TICv8 contains major changes in computed quantities compared to TICv7 (cf. Stassun et al (2019)). It should be noted that the methods used to estimate a variety of stellar parameters are still under active development, and can be affected by poor catalog photometry when there is no acceptable alternative photometry for a given star.

The major changes in TIC v8 compared to TIC v7 are:

1. The base catalog for TICv8 is the Gaia DR2 catalog. This differs from the previous versions of the TIC, which were based on 2MASS. 2MASS and Gaia DR2 objects were matched using the Gaia-provided 2MASS-to-Gaia look-up table (tmass_best_neighbour). All 2MASS objects retained their original TICv7 TIC ID. All Gaia objects not found in 2MASS, or earlier versions of the TIC, were given new TIC IDs.
2. Stars identified in 2MASS as a single source, but identified in Gaia as multiple sources, are identified with the disposition = SPLIT (column 86). In this case, the 2MASS object remains in the TIC with the original TIC ID, and two or more new sources have been added to the TIC with new TIC IDs. In these cases, the new TIC objects will have the TIC ID of the original object in the duplicate_id column (column 87).
3. In all cases, physical parameters have been calculated using the reported effective temperature (column 65), radius (column 71), and mass (column 73) in the TIC. This includes stars that may have measured physical parameters, such as log(g) (column 67), in spectroscopic catalogs matched to TICv8. The only stellar parameters that have been adopted directly from spectroscopic catalogs are effective temperatures (column 65) and metallicities (column 69), provided the error on the effective temperature (column 66) is less than 300 K.
4. When multiple spectroscopic parameters exist for a single star, in a single catalog, these values have been combined using a weighted mean.
5. The coordinates (columns 14,15, 25-28) and their uncertainties (columns 117-118) are provided in epoch 2000 due to mission requirements. Positional errors (columns 117-118) and proper motion errors (columns 18, 20) have been propagated. It should be noted that this leads to much larger errors than those in the nominal Gaia DR2 positions. Especially for Gaia DR2 stars, users should not try to propagate forward the TIC coordinates (columns 14, 15) using the proper motions (columns 17, 19) listed. Instead, users should use the original Gaia DR2 positions (columns 119, 120), proper motions (columns 14, 15), and corresponding errors for propagation (columns 121, 122). We provide RA and Dec, with errors, as given in the source catalog (Gaia DR2, 2MASS etc) in additional columns (119 - 122).
6. Many stars in the TIC have stellar parameters provided, regardless of their inclusion in the CTL. However, stars have their masses (column 73) calculated only if they have radii (column 71) smaller than the line shown in figure 10 of Stassun et al. (2019). This was to avoid providing masses for stars that may be giants using an empirical relation valid only for dwarf stars. If a star is identified as a subgiant, the mass (column 71) and log(g) (column 67) should not be considered to be reliable; however, the luminosity (column 78) and effective temperature (column 65) for these stars are considered to be reliable.
7. A number of additional columns have been provided for users on MAST; see columns 89-122 below.
8. The Tmag (column 61) is now calculated using colors that have been corrected for reddening. Gaia colors are preferred if they are given and match the Gaia quality criteria (gaiaqflag = 1, column 112). As a result, some stars may have different Tmags compared to previous versions of the TIC.
9. The cool dwarf list was updated in TICv8 to only include cool dwarfs with reliable parallaxes (< 20% error). Stars in the previous version of the cool dwarf catalog retain their TICv7 stellar parameters, but no longer have a priority calculated.
10. The only stars from the specially curated cool dwarf list which are included in CTLv8.01 are those with T<16. This was done to limit the size of CTLv8.01 to less than 10 million stars. Future versions of the CTL may include fainter cool dwarf stars.

Updates to the TICv8 CTL

This section of the release notes highlights major changes to how stars are selected for the TICv8 CTL as new versions of the CTL are generated.

The current version of the CTLv8.01, and is the original CTL for TICv8.

1. Stars are selected to be in the CTL if they meet the following requirements: Tmag < 13 (column 61), the Gaia quality flag is set to 1 (column 112), the star has a dwarf-like radius (the star’s radius is smaller than the line in figure 10 of the full documentation), and the star’s placement on the absolute Gmag (based on column 59) vs. gaiabp (column 108) - gaiarp (column 110) color-magnitude diagram suggests the star is not a white dwarf. Please see section 3.1 of the Stassun et al. (2019) for more details.
2. There is no longer a specially curated ‘Bright Star’ list, since there is now a separate brighter star CTL. As a result, stars no longer have their priorities set to be 1 if their TESS magnitude is brighter than Tmag = 6  (column 61). Thus ‘Bright stars’ are no longer automatically added to the exoCTL.
3. There is no longer a specially curated ‘Known-Planets’ list. The stellar parameters for such objects in existing catalogs were found to be too inhomogeneous to properly calculate a priority. Thus known planet hosts are no longer automatically added to the CTL.
4. CTL stars with calculated masses and radii that result in an unphysical value of log(g) > 5 (column 67) have had their CTL priorities set to 0. This was done to avoid prioritizing stars that may have poorly measured effective temperatures (column 65), extinction values (column 82) , and/or parallaxes (column 22), but that still pass the quality assurance cuts. Stars in specially curated lists have been excluded from this criterion.
5. Stars in the CTL have their priorities (column 88) multiplied by a factor of 0.1 if they are within 10 degrees of the Galactic plane (column 26). This was done to avoid prioritizing stars for two-minute cadence which may be affected by large amounts of extinction. This is a change from previous version of the TIC that de-boosted priorities if |b| < 15 degrees.
6. The only stars from the specially curated cool dwarf list which are included in CTLv8.01 are those with T<16. This was done to limit the size of CTLv8.01 to be less than 10 million stars. Future versions of the CTL may include fainter cool dwarf stars.

Currently known quirks and issues with TICv8, versions 8.1 and 8.2

General Issues

There are a number of minor issues, and quirks, which have been identified by the TSWG. Users should read this section carefully before accepting TIC quantities at face value.

1. Some stars may have reported stellar parameter errors that are larger than the parameter quantity. Users are cautioned against using these parameters in calculations for publications.
2. Users are reminded that stars that have ecliptic latitudes between -6 and 6 degrees have their priorities set to zero. This “gap” in priority is meant to mimic the expected gap in camera coverage for the two year primary TESS mission.
3. Some stars in the hot subdwarf list do not have errors for their stellar parameters. These were adopted ‘as-is’ for consistency.
4. The disposition column identifies objects that are included in the TIC, but are likely spurious, or related to other TIC objects in a non-trivial, non-astrophysical way. Currently, this column is populated as a NULL, DUPLICATE (6), ARTIFACT (7), or SPLIT (8). See section on Updates, below. 
5. Users should note the metallicity (column 69) has been adopted when available from a spectroscopic catalog, but it is not directly used in any calculation. This may result in physical parameters that are not strictly consistent with the reported metallicity.
6. The listed RA (column 14) and Dec (column 15) are not measured positions, but positions which have been propagated to EPOCH J2000.0. Users should not try to propagate forward the TIC coordinates using the proper motions listed. Instead, users should use the original catalog position (column 119 & 120), proper motions (columns 17 & 19), and corresponding errors for propagation.
7. If a star is identified as a subgiant, and a mass and log(g) are provided, users should not consider these parameters to be reliable, because the empirical relations designed to calculate the mass are valid for dwarf stars. However the luminosity and effective temperature should be accurate.
8. Asymmetric errors are computed using a Monte Carlo method and are only provided for CTL stars which are not in a specially curated list. There are borderline cases where the peak of the property-distribution was too close to the validity-limit of the property-relation in order to compute reliable errors. These stars will have no or asymmetric errors for just some properties.
9. Negative errors for some Bmag and Vmag uncertainties: there are 40354 stars with negative Vmag uncertainties and 4395461 stars with negative Bmag uncertainties.  The negative error values are all for stars from APASS DR9, and are magnitudes for which only one observation is used, and therefore are Poisson errors rather than standard deviation.  When constructing the TIC, these negative values were not caught and converted into positive values, and thus are recorded as negative values in TIC8.  Users should feel free to convert them to positive values and use them as a valid estimate of the uncertainty.
10. Negative E(B-V) values: there are two different origins for negative uncertainties in the "eneg_EBV" and "epos_EBV" columns.  In the first case, if "EBVFlag" = 'schlegel' they are simple numerical and conversion errors between double and single precision values. These uncertainty values actually should be assumed to be 0.0. There are approximately 2.5 million stars in this group. If "EBVFlag" = 'panstarrs' the negative values originate from former CTL targets that got a negative value from the Monte Carlo method that was not restored properly. In these cases the value is assumed to be NULL, e.g., the negative value does not have any useful meaning. 

Updates to TIC 8 (v8.1 and v8.2)

TIC v8 was updated twice (2020 April (v8.1) and 2021 July (v8.2)) to identify objects that are included in the TIC, but are actually spurious objects or related to other TIC stars in a non-trivial, non-astrophysical way.  The general categories for these objects are “split”, “join”, and “artifact”, which are described below. Changes to the affected objects are reflected in the “disposition” (column 86) and/or “duplicate ID” (column 87) fields of their TIC entries.

The TIC 8.1 release changed the dispositions of ~700 stars that were targets in TESS Year 1 observations.  The TIC 8.2 release was generated from a systematic analysis of TIC 8.1.  The search for artifacts around target stars was limited to stars brighter than Tmag = 13. The search for duplicates and splits was not limited by magnitude, but by the origin of target stars: only 2MASS stars unmatched to Gaia DR2 were searched for duplicates and splits. In total about 34 million stars were updated.

The description below is an overview of the updates for TIC 8.2.  For a complete discussion of the TIC 8.1 and 8.2 updates, please refer to Paegert et al. (2021) (https://arxiv.org/pdf/2108.04778.pdf):  Section 3.1.1 deals with artifacts, Section 3.1.2 covers joins, and Section 3.1.3 describes splits.

SPLIT stars

In the case of a SPLIT, a star originating from TIC7 is actually an unresolved double- or multiple-star in 2MASS and has at least two entries in Gaia DR2 that did not get matched up properly in TIC8.  In this case, the TIC 7 star is not a real star, and thus all magnitudes (Tmag, Vmag, J, H, K, etc.) and stellar properties (Teff, radius, mass, etc.) of the TIC 7 star are most likely corrupted  in TIC 8 and should NOT be used. TIC 8.2 fixes the problem by copying the coordinates, TESS magnitude, and stellar characteristics of the brightest Gaia star over to the TIC ID of the corrupted star and then setting the disposition of the brightest Gaia star to DUPLICATE to indicate that it not real and should be ignored. (It is important to note that observed magnitudes, such as the 2MASS JHK, are not corrected and care should be taken before using them in any calculations). The disposition field of the corrected star is set to SPLIT, and the duplicate ID field of the brightest Gaia star is set to the TIC ID of the corrected star.  The duplicate ID field of any other stars resolved by Gaia DR2 and associated with the corrected star is set to the TIC ID of the corrected star:  the stellar parameters of these stars are valid, as they come from Gaia DR2, and remain unchanged. 

The impact on photometry of stars misidentified in TIC 7 depends on the fluxes of the Gaia DR2 stars into which it was resolved, as well as the fluxes of nearby stars.

JOINs

Joins are stars originating from TIC 7 that did not get matched up with a Gaia DR2 star in TIC 8, meaning there are two entries in the TIC referring to the same star.  In these cases, one of the stars is given the disposition DUPLICATE to indicate that it is not a real star, and the duplicate ID field is set to the TIC ID of the other star.  To preserve backward-compatibility, the ID of the TIC 7 star is preserved (i.e. its disposition is unchanged from NULL) and the Gaia star is given the disposition DUPLICATE.  The stellar parameters of the TIC 7 star are updated using the improved, Gaia-derived values, unless it is a star from a specially curated list, in which case the parameters are taken from those curated lists.

ARTIFACTs

Artifacts are generally spurious “stars” found by extracting sources from the original catalog source images and are often the result of diffraction spikes near bright stars, or other instrumental artifacts mistakenly interpreted as stars. The 2MASS entries of TIC 8 were searched for such spurious images.  The disposition of TIC entries that have been identified as ARTIFACTs is set to ARTIFACT, indicating that the entry is not a real star.

TIC v8.2 Column Schema - Executive Summary

...