Spectral Typing Using TiO Bands

Reid et al. 1995:

• compare spectra to template spectra of known spectral type
• MMK System (Morgan et al. 1943)
• no late M-type spectral templates in MMK system
• KHM system (Kirkpatrick 1992)
• uses spectral features in the wavelength range 6300-9000 A
• Reid et al. 1995
• TiO bandheads at 6500A and 7050 A
• CaOH and CaH bands at 6250 and 6400/6800
• TiO band is temperature dependent
• can be used as a distance indicator (spectrophotometric parallax)
• "index" is defined as : R=F(W)/F(cont)
• F(cont) is the pseudocontinuum flux (e.g. mean flux in a sideband)
• Figure 1 from Reid et al. shows the spectrum of a typical mid-M star with the bands used for spectral typing indicated:

• They derive a direct (linear) relationship between the spectral number (M #), and the TiO index: Sp = -10.775*TiO5 + 8.2

THE PROBLEM:

Since my Kast spectra do not include the features near the TiO bandhead (due to the maximum grating tilt allowed by the telescope), I need to find a way to define spectral types based on the TiO1 band head. Hopefully these will agree with the spectral types quoted by Rodriguez et al. 2011, which were derived using broadband photometry. 

The Write-Up

I'm just starting to work on the write-up (which will hopefully turn into a publishable paper) on my observations of low mass stars over 5 months with the Lick 3-m telescope and a combination of the Hamilton Echelle Spectrometer and the Kast Dual Spectrometer.

First comes The Big Table, for which I need measured equivalent widths for H-alpha and Li. So I need to go back over all my old data and re-calculate my EWs to make sure I still agree with myself. One of our undergrads (Nicole Arulanantham) has already done this for our August data. Our method is to plot the spectra using IRAF's splot package, then measure the equivalent width using a Voigt profile fit (hit k, then v on either side of the line, making sure to include the flux in the wings). We do this a few times (~5) and take the average. This is because there is some intrinsic error in the calculation - the EW will depend on where you place the continuum level (where you cut off the wings).