Candidate nearby dwarfs in the NLTT: spectroscopic follow-up
1. Northern stars
The initial observations for this project were obtained between sept 29 and October 2 2000, using the KPNO 2.1-metre telescope, GoldCam spectrograph and 400 l/mm grating. The target list of 127 NLTT stars was selected using preliminary colour-magnitude selection criteria, which result in the inclusion of stars up to 30-40 parsecs from the Sun. Of the 127 targets, 28 are known late-type dwarfs and 29 have suspect 2MASS magnitudes (bright star diffraction spikes, unresolved companions etc.). The remaining 70 were observed spectroscopically. We summarise the main results on this pages; full results are given here .
One of the main tasks undertaken in these initial stages was defining calibrating relations between the measured narrowband indices (TiO5, CaH2, VO etc - see the PMSU pages for definitions) and both spectral type and absolute magnitude. Figure B2.1 plots the spectral type calibrations for both TiO5 and the Kirkpatrick et al. (1999) VO-a index:
Figure B2.2: Calibrating relations for computing MJ from TiO5, CaH2 and CaOH indices. All of the relations reverse at MJ > 11, as the molecular bands weaken in strength. Datapoints for individual ultracool dwarfs are labelled.
Our prime goal is identifying candidate members of the immediate Solar Neighbourhood (d < 20 pc.). That requires the derivation of spectroscopic parallax estimates, and hence an absolute-magnitude/spectral-type calibration. Since many of these stars lack accurate optical photometry, but all have 2MASS data, we have derived mean relations in the (MJ, sp. index) plane for the TiO5, CaH2 and CaOH indices. All of these calibration show the discontinuity at 8 < MJ < 9 which is evident in the (MJ, (I-J) plane (see the colour-magnitude diagrams ). Rather than fit a mean relation through that region, we fit the upper and lower sequences separately; for stars with indices in the region of overlap, we adopt a democratic solution - choose the value of MJ which provides the closer match to the estimate(s) based on unambiguous index or indices. The formal calibrations are, for TiO5:
Figure B2.3: Abundance calibration for M dwarfs. The blue points plot data for nearby disk dwarfs (from PMSU); open squares are sdMs and crosses esdMs from Gizis (1997); white points plot data for the NLTT sample. The two labelled outleirs are discussed further below.
The relative strengths of CaH and TiO5 bands allow moderately imprecise estimates of metallicity for individual stars. Figure B2.3 plots the calibrating relations, from Gizis (1997), where the sdM intermediate subdwarfs are likely to have <[M/H]> ~ -1 dex, and the esdM extreme subdwarfs probably have <[M/H]> < -1.5 dex. The majority of the NLTT dwarfs have TiO5 < 0.4, indicating spectral types later than M5. Three outliers, potential subdwarfs, are discussed further below.
The two latest-type dwarfs in the current NLTT sample are LP 762-38 and LP 647-13. The former is very similar in appearance to the M7 spectral type standard, VB 8. Based on the observed magnitude of K=11.21, we estimate a distance of 20.0+/-3 parsecs. LP 647-13 is later, with spectral type M9 and a likely distance of just over 10 parsecs, making it one of the four closest known M9 dwarfs.
Figure B2.5: LP 410-38, a candidate mild subdwarf. LP 702-1 appears to be a near solar-abundance disk dwarf.
Closer inspection of spectra of the three outliers plotted in the (CaH/TiO5) diagrams (Figure B2.3) suggests that only LP 410-38 is likely to be significantly metal poor. LP 702-1 and LP 824-383 have spectra of relatively low signal-to-noise, and that appears to account for the non-standard CaH indices. LP 410-38, however, appears to have significantly enhanced CaH absorption, characteristic of the intermediate-metallicity sdM subdwarfs.
It is clear from Figure B2,3 that the majority of the NLTT dwarfs are later than spectral type M5. As such, they provide an excellent complement to the predominantly earlier-type dwarfs in the PMSU sample in statistical analyses of the variation of properties with spectral type. Some of those studies require estimates of the bolometric magnitude, and Figure B2.6 plots mean calibrations of the J-band bolometric correction as a function of spectral type and TiO5 index:
Figure B2.7: Activity in M dwarfs: magenta points plot data from the PMSU sample; green points mark detections and red triangles upper limits for stars in the present sample.
Figure B2.7 plots the distribution of chromospheric activity amongst the NLTT dwarfs, where activity is measured as the ratio between the flux in the H-alpha line and the total bolometric flux. In general, the average level of activity is approximately constant amongst dMe dwarfs with spectral types earlier than M6 (note that only 20% of the PMSU dwarfs are dMe dwarfs, with H-alpha emission exceeding 1 Angstrom equivalent width). Observations of M7 and cooler dwarfs shows a distinct turndown in the mean level of activity (Gizis et al, 2000). These observations confirm that trend, filling in the distribution between M5 and M7, and showing a clear trend towards lower activity at the later types.
Observations with the Mayall 4-metre of NLTT dwarfs and stars from the ultracool sample. Reductions and analysis in progress.
Observations with the 2.1-metre and GoldCam of NLTT dwarfs and a few stars from the ultracool sample. Reductions and analysis in progress.
Observations with the Mayall 4-metre of a few NLTT dwarfs and stars from the ultracool sample. Reductions and analysis in progress.
Observations with the Blanco 4-metre and CTIO 1.5-metre of southern NLTT dwarfs and ultracool stars. Reductions and analysis in progress.
NStars proper motions index
NStars home page
INR home page