Photometric searches (II): 2MASS near-infrared data and optical counterparts

Main investigators: Cruz, Reid, Liebert
Late-type M, L and T dwarfs span a wide range of photometric properties . This variety means that no single set of search criteria is suitable for finding all nearby ultracool dwarfs. We therefore use a combination of techniques to identify candidate nearby stars and brown dwarfs, each technique targeting a somewhat different regime in (absolute magnitude, colour/spectral type) space.

1. Ultracool Dwarfs: Near-infrared Photometric Selection

1.1 Defining the sample

The (M_J, (J-K)) colour-magnitude diagram for nearby stars shows that (J-K) is essentially degenerate for early- and mid-type M dwarfs, but can provide a means of identifying cooler dwarfs, with distinctively redder (J-K) colours. The second incremental release of 2MASS data provides accurate JHK_S photometry for hundreds of millions of point sources covering 47% of the sky. That cxatalogue provides a homogeneous dataset to search for candidate ultracool dwarfs.

Figure A1.1 illustrates the search criteria adopted for this phase of our NStars survey. As in Figure 1.11, we plot photometric data for nearby stars and L dwarfs with known parallaxes, but in this case the magnitudes are scaled to a distance of 20 parsecs. The lines superimposed on the figure delineate our selection criteria: (J-K) > 1.0, and J < J_lim, where

The simple colour-magnitude limits outlined in Figure A1.2 encompass a wide range of astronomical sources: red giants, carbon stars, reddened Galactic stars, T Tauris and dusty QSOs, besides the targeted ultracool dwarfs. Those contaminants can be eliminated to some extent by imposing selection limits in the (J-H)/(H-K) plane. Our chosen limits are

As an indication of the stellar sample segregated by these criteria, consider the two archetypical ultracool dwarfs, VB8 and VB 10
VB8, spectral type M7, has M_J=10.71, or J=12.21 at 20 parsecs, and (J-K)=0.95
VB10, spectral type M8, has M_J=11.05, or J=12.55 at 20 parsecs, and (J-K)=1.10
The (J-H)/(H-K) colours of the two stars are
VB8: (J-H)=0.58, (H-K)=0.47; VB 10: (J-H)=0.66, (H-K)=0.44
VB8 fails to meet our criteria; VB10 is well within the sample limits. Thus, our sample should isolate ultracool dwarfs with spectral types later than ~M8.

Figure A1.3 illustrates the (J, (J-K)) distribution for point-source selection based on these colour criteria. The data are drawn from approximately one-third of the sky (although the 2MASS second incremental covers only half of that area, so effectively ~17% sky coverage). We also exclude all sources within 10 degrees of the Galactic Plane.

The colour-coding of the sources in the (J-H)/(H-K) colour-magnitude diagram provides a hint as to the likely make-up of the sample. The uncoloured open squares predominantly represent the brightest sources in J. It is clear that most of those sources lie near the upper bound of the JHK selection box, close to the locus of red giants and AGB stars. Indeed, follow-up observations have generally confirmed most of those objects as evolved stars. The fainter `good quality' sources (blue points) lie mainly in the vicinity of the dwarf sequence in the JHK plane, with some extending upwards along the L dwarf sequence. Many of these have indeed proven to be ultracool dwarfs. Note, however, the increased number density in a band across the L dwarf region at (J-H)~ 1.4: most of these stars prove to be reddened sources.

1.2 Refining the sample

Section 1.1 outlined the basic technmiques employed in our survey. This section describes in more detail the application of those techniques to analysis of data from the 2MASS Second Incremental Release.

There are a total of 2224 sources in the Second Incremental Release which meet the colour/magnitude criteria of the ultracool sample, and which have Galactic latitudes |b| > 10^o. Visual inspection of those sources allows the immediate elimination of 430 sources as confused, near bright stars, parts of large galaxies, etc. The celestial distribution and near-infrared properties of the remaining 1794 targets are shown in Figure A1.5. Clear concentrations remain at low latitudes, particularly south of the equator around 6 hours (Orion, Monoceros), 16 hours (Ophiuchus, Sco-Cen, Lupus) and 20 hours (Capricorn, Aquarius)

We have divided the ultracool sample into three subsets based on their apparent magnitudes, as illustrated in Figure A1.6. These are defined as follows, with each sub-sample described in more detail on the linked page:

• The faint sample: sources with K_S > 9.5; 1051 sources.
• The intermediate sample: sources with K_S < 9.5 and J > 9; 100 sources.
• The bright sample: sources with J < 9; 643 stars.

NStars Photometric surveys index

NStars home page

INR home page