Zusammenfassung
The ATLAS3D Survey has reported strong evidence for a non-universal stellar
initial mass function (IMF) and for a trend of the IMF with the effective
stellar velocity dispersion of early type galaxies (ETGs) (Cappellari et al.
2012). The ATLAS3D Survey consists of detailed Integral Field Spectroscopy of
260 ETGs closer than 42 Mpc. Cappellari et al. infer the IMF by comparing mass
measurements derived from fitting the kinematic data with measurements derived
from fitting the spectral energy distribution. Here we investigate possible
systematic errors and biases that could affect their conclusions. We show that
part of the reported trend between IMF and velocity dispersion is caused by a
selection effect on Heta absorption. Apart from an IMF trend with velocity
dispersion, we also find an IMF trend with distance, but no correlation between
nearest neighbour ETGs as would be expected if the dependence on distance would
reflect an environmental dependence. Part of the IMF trend with velocity
dispersion can be traced back to colour-dependent calibration issues with the
surface brightness fluctuation distances of Tonry et al. (2001) and Mei et al.
(2007). There is no IMF-dispersion trend for galaxies with distances larger
than 25 Mpc. The ATLAS3D Survey appears to be incomplete for masses lower than
10^10.3 solar masses. Restricting the analysis to the kinematic mass range M >
10^10.3 solar masses completely removes the IMF-dispersion trend. Our findings
strongly suggest that some of the trends between IMF and galaxy properties
reported by ATLAS3D are unphysical and caused by systematics.
Nutzer