Suma topol´ ogica

RECX25 is intended to contain basic astrophysical information on every component of every star system ever reported to be within 25 pc of the Sun by trigonometric parallax, divided into three categories of quality:

• VETTED: Systems with weighted mean parallaxes greater than 40 mas (within 25 pc) and with weighted mean errors less than 10 mas. The error cutoff is not based on a fractional error because the reliability (though not necessarily usefulness) of a parallax is best described in terms of the actual observable angular measurement.

• FLIMSY: Systems with mean parallaxes greater than 40 mas but errors larger than 10 mas, which (as explained above) we deem to be unreliable. Some are genuinely nearby stars, many are not.

• BOOTED: Systems with mean parallaxes less than 40 mas, that were at least once reported to be greater than 40 mas. This is largely for posterity, so researchers can still find systems that were once considered 25 pc objects.

At this writing, like the 10 pc sample curated by Todd Henry, RECX25 is a one-entry-per- object flat text file, with columns for data. The concept here is that, apart from parallaxes, only one best value will be used.

At the moment RECX25 keeps track of the following:

• Coordinates

Decimal 2MASS ICRS coordinates and epoch (Julian Date) (Skrutskie et al. 2006) Proper Motions and Position Angles (best source, usually van Leeuwen 2007) Parallaxes (separated for bookkeeping purposes):

Weighed Mean Parallax and source key

individual parallaxes: YPC (van Altena et al. 1995), HIP97 (Perryman et al. 1997)6_{, vLe07 (van Leeuwen 2007), HIPPARCOS re-reductions (see § 3.3.2), RE-}

CONS/CTIOPI, Other (see _§3.3.1)

• Photometry Guide V SuperCOSMOS BJR1R2I JohnsonU BV Kron-CousinsRI 2MASSJHKs • Spectroscopy Spectral Type • Multiplicity Real Component ID Number of Components

Hierarchical Key

Separation and Position Angle Delta Magnitude with filter

• Names

Preferred Name

GJ, LHS, HIP, YPC, variable, HD, Durchmustrung (BD/CD/CPD), 2MASS, other

• Notes (not in the main table)

References are given explicitly for most data values. The only data for which we currently record multiple values are the parallaxes; the table currently supports up to 20 values. Parallaxes from any member of the system are applied to all members of the system7_{, and}

combined to form weighted mean system parallaxes. Care has been taken to make certain that datasets do not count more than once, as that would unfairly bias the results. For instance, RECONS has occasionally re-published parallaxes if the solution changes noticeably (e.g. GJ 633 in Riedel et al. 2010). The reduction in the older paper is flagged as obsolete and not included in the weighted mean, as that would unfairly bias the parallax toward the RECONS data. In the same way, the original HIPPARCOS parallaxes in Perryman et al. (1997) are included in the table, but they are uniformly superceded by the new van Leeuwen (2007) values.

7_{Currently, the only multiple system we donotcombine into a single weighted system mean parallax isα}

Centauri: Proxima Centauri (768.85±0.29 mas) is 17σcloser thanαCentauri AB (747.23±1.17 mas). Gaia will almost certainly resolve more systems.

Proper motion and position angle are taken from the source of the parallax (HIP, RE- CONS, YPC, other), LSPM (L´epine & Shara 2005), or UCAC2 (Zacharias et al. 2004) (in that general order). Proper motions for companions are assumed to be the same unless we have specific astrometry (from parallax sources) that says otherwise. Although there are multiple sources of proper motions, we will not be combining them as we do parallaxes; most proper motion catalogs are built from similar sources (e.g. HIPPARCOS, Palomar Optical Sky Survey (POSS) and Science and Engineering and Research Council (SERC) plates), and are thus not actually unique. Eventually it is hoped that we will standardize on UCAC4 for anything not in HIPPARCOS, and have the entire database in the ICRS ref- erence frame. Accurate coordinates (and their precise epoch in Julian Date) are taken from 2MASS (Skrutskie et al. 2006). Eventually the goal will be to precess each star to Epoch 2000, ICRS Equinox using this 2MASS position and a good proper motion.

U BV RI photometry is currently either unfiltered white Hpmag (a proxy for V) from

HIPPARCOS, some kind of V from YPC or U BV RI from a clear hierarchy of trusted sources starting with Bessell (1990a), various papers by Weis (1993, 1996) (which use the Kron-system RI), our own photometry, and papers by Koen and Kilkenny (e.g. Koen et al. 2002).

Right now, each star is indexed by its own rough coordinates and given at least one preferred name, generally the Gliese number or a Luyten name (L, LP, or LHS). The guide coordinates, rough magnitude, and preferred name are all usable placeholder of unknown provenance and uncertain quality, useful until we obtain better values (particularly in the

case of brown dwarfs and other stars not in the 2MASS catalog).

The Notes section is not part of the main table, but is a separate text file named with the HHMM+DDMM coordinates of the system; detailed notes and the notes in ARICNS will be included as needed. Already, several paragraphs of information have been laid out for the more complicated systems. This unrestricted method both reduces the clutter of the main database, and frees up the files for more useful details.

The Sun is included, as it was in CNS3p, ARICNS and NStars; this time it has a “par- allax” of 206265′′ _{(1 AU) to allow proper calculations of absolute magnitudes and space}

positions; there are also entries for each planet in the Solar System.

3.2.1.1 Multiples

Multiple star systems present additional organizational difficulties. To start with, there are several common ways of naming their components. Although the more pedantic may care, common usage has validated the idea of extending existing names- there is no LHS 1749B in Luyten (1979a), but it is the accepted name for the companion to LHS 1749 as of Jao et al. (2003). RECX25 extends names in the same fashion. As for the method of assigning letters to primaries and companions, there are multiple different schemes on the subject. RECX25 (and RECONS) use the following8_:

8_{The IAU and WDS (http://ad.usno.navy.mil/wds/wmc/wmc post191.html checked 2012 JUL 15.) rec-}

ommend Hierarchical Order_{, as used by (for example) Raghavan et al. (2010). If a currently-known}

companion is split into two, the name is split between the two stars (Aa, Ab), going byV magnitude bright- ness. If a new companion is found at a resolved separation from the first, it gets a new letter (B). With the example of the GJ 644 system, GJ 644 A=Aa, GJ 644 B=Ab1, GJ 644 D=Ab2, GJ 643=B, GJ 644 C=C.

The benefits are a sense of the system’s structure, and no ambiguity between resolved and unresolved components (it would be clear whether observations of GJ 644 B refer to GJ 644 B (Ab1) itself, or the combination (Ab) before it was split into B and D). Unfortunately, it comes with serious organizational problems: It can take up ton−1 letters to distinguish between ncomponents. The names themselves offer

Johnson V magnitude(Where available). Component ID is purely a function of John- son V magnitude in descending order. This does not always correspond to the component letter given by someone else- GJ 644C is the faintest component of the GJ 644/643 quintuple system, and is thus the E component by our reckoning. A slight variant would use bolometric luminosity, but given the way blackbody radiation works, the differences would be minor.

Discovery Order. WDS and CNS traditionally add letters by order of discovery. In this instance, GJ 644 C was the third component of GJ 644 discovered, and is the C component even though the later D component is brighter. When a new component is resolved, the brighter component assumes the original letter, and the fainter component takes a new letter. Stars that already had names are not renamed when they are discovered to be in a bound system (i.e., GJ 643). Exoplanets use this system, and the RECX25 database uses it when the V magnitudes of companions are not known.

RECX25 (and the RECONS 10 pc list) use another column that reports the number of components in the system- the primary is given a number, all other components get zeros. For instance,αCentauri A is listed as ‘3’, whileαCentauri B and Proxima Centauri are both listed with ‘0’. This enables a quick sort of the table to find out how many system primaries there are, and the multiplicity of each. Meanwhile, in a nod to hierarchical arrangement of the system, separations and delta magnitudes are given for the various components: αCen A and

no easy way to distinguish between stars and barycenters (Aa is a real star, Ab is a spectroscopic binary containing the stars Ab1 and Ab2) no indication that Aa is a real star and Ab is only a barycenter. The lowercase letters can be confused with planets (WDS recommends treating planets no differently from stars, but this has not caught on - perhaps forcing letters would solve this (AAA, ABA, ABB, BAA, CAA) but this is not the system advocated by the IAU). One-object-per-line tables are not conducive to separately storing information that applies to an unresolved multiple anyway.

B list the A-B separation; Proxima Centauri lists the A-C separation and delta magnitudes (relative to A). Brown dwarfs, white dwarfs, and stars are considered components, planets are counted separately.

Coordinates and magnitudes are contingent on a resolved multiple. As the resolution limit of 2MASS is roughly 3′′

, stars with smaller separations are generally unresolved (although individual U BV RI photometry may be resolved). In the unresolved case, the coordinates and magnitudes are applied to the more luminous of the pair (magnitudes are given a “J” to denote their joint, convolved status); the secondary is given “no data” entries in those fields. In the future, multiples may be deblended where adequate information exists, but this has not been done yet.