You are in: Instruments > NIR > Spectral Templates > Examples

nir logo

Spectral Templates - Examples

There are two instruments among the suite currently available at the Gemini telescopes which are particularly suitable for near-IR studies of stellar kinematics in external galaxies:
The CO overtone bands from evolved stars are the strongest absorption features in the 1-3 μm range of stellar systems. The features are sharp and deep, thus very sensitive to stellar motions,  and at least the first two are located in regions of the IR spectrum relatively free from telluric lines. The equivalent width of the CO bands is a function of the effective temperature and surface gravity of the star: decreasing the effective temperature or increasing the stellar radius increases the CO equivalent width.

Assuming that a galaxy spectrum is the convolution of an average stellar spectrum (combination of all the individual spectral types present in the local stellar population) with the line-of-sight velocity distribution, the most commonly method used to extract the velocity dispersion from the galaxy spectrum employs a cross-correlation technique using one or more star spectra as templates ((Tonry & Davis 1979). This technique is sensitive to the effect of template mismatch, since the velocity dispersion measured from the galaxy spectrum will decrease as the equivalenth width of the template star absorption lines increases. The template sample should therefore span a range of CO equivalent widths (more than simply of spectral types - Silge & Gebhardt (2003) have examined the issue and concluded that it is the EW of the template that affects the fitting, not the details of the spectral type).  The figure below shows the EW of the second CO overtone (at 2.32μm) in our sample, plotted as a function of the effective temperature. The data ranges from less than 5 to over 20Å. The figure also presents a sample of spectra in the first and second CO overtones, to ilustrate the different equivalent widths and profiles.


In order to ilustrate the use of the templates, we selected actual data from a representative Gemini programme, GS-2005B-Q-65, a GNIRS/IFU+111l/mm observation of the Seyfert galaxy NGC7582, centred at 2.24μm. The data have been reduced by R. Riffel as part of his PhD thesis, and a full description of the reduction procedure and analysis will be presented in the final publication. The sample spectra used here were extracted from a couple of spatial positions in the final datacube simply to ilustrate the use of the templates. The task stsdas.contrib.redshift.xcor was used to calculate the cross-correlation.

Two stars were used to ilustrate the effect of the template EW in the fit: HD20038, an F7IIIw star with EW(CO 2.322μm)~8Å; and HD2490, an M0III star with EW(CO 2.322μm)~22Å. The result is shown in the figure below. PLEASE NOTE THAT THE ABSOLUTE VALUES OF THE VELOCITY DISPERSION ARE UNCERTAIN - the purpose of the figure is to illustrate the effect of using different templates, a much more complete treatment of the errors is needed to give any further meaning to the results than the expected one: the larger the EW of the CO band in the template, the lower the resulting velocity dispersion required to reproduce the galaxy spectrum.


[Science Operations home] [NIR home]

Last update 2007 April 09; Claudia Winge