Calculation of velocity dispersion of the nearby galaxies using different stellar template libraries

Abstract

We present the central velocity dispersion measurements of the nearby galaxies from the Sloan Digital Sky Survey (SDSS). Using the sample from the paper by Ho et al. 2009, we have selected 23 galaxies for which we calculate the velocity dispersion. We have used the Penalized Pixel-Fitting code (Cappellari and Emsellem2004) to measure the velocity dispersion throughout the four chosen spectral regions: (3800,4568)?A, (4568,5336)?A, (5336,6104) and (6104,6872)?A. In all these regions, we have separately calculated dispersions and corresponding errors. We found that the measured values may vary with the change of spectral region, but, if weighted properly with the measure of the goodness of the fit, the final results will be shifted coloser to those for the best fitting regions. We have also tested how the use of different spectral libraries (Miles, Valdes and Elodie databases) influences measurements and we showed that they do not affect measurements much. However, Elodie stellar library introduces the smallest errors in the velocity dispersion and it is the most stable throughout all four spectral regions. For these reasons it should be used preferentially when dealing with the SDSS spectra. We compare the results with the above mentioned paper and find a reasonable agreement. The agreement with the dispersions available in the HyperLeda database is very poor. The best agreement is obtained with SDSS measurements. We believe that our measurements are useful since SDSS velocity dispersions measurements are not available for many galaxies and the method of calculation of the velocity dispersion outlined in this work enables calculation of velocity dispersion for any galaxy. Of course, spectra with signal-to-noise ratio below 20 should be taken with caution.