Testing adiabatic contraction with SDSS elliptical galaxies
We study the profiles of 75 086 elliptical galaxies from the Sloan Digital Sky Survey (SDSS) at both large (50-500 kpc/h) and small (~3 kpc/h) scales. Weak lensing observations in the outskirts of the halo are combined with measurements of the stellar velocity dispersion in the interior regions of the galaxy for stacked galaxy samples. The weak lensing measurements are well characterized by a Navarro, Frenk and White (NFW) profile. The dynamical mass measurements exceed the extrapolated NFW profile even after the estimated stellar masses are subtracted, providing evidence for the modification of the dark matter profile by the baryons. This excess mass is quantitatively consistent with the predictions of the adiabatic contraction (AC) hypothesis. Our finding suggests that the effects of AC during galaxy formation are stable to subsequent bombardment from major and minor mergers. We explore several theoretical and observational systematics and conclude that they cannot account for the inferred mass excess. The most significant source of systematic error is in the IMF, which would have to increase the stellar mass estimates by a factor of two relative to masses from the Kroupa IMF to fully explain the mass excess without AC. Such an increase would create tension with results from SAURON (Cappellari et al. 2006). We demonstrate a connection between the level of contraction of the dark matter halo profile and scatter in the size-luminosity relation, which is a projection of the fundamental plane. Whether or not AC is the mechanism supplying the excess mass, models of galaxy formation and evolution must reconcile the observed halo masses from weak lensing with the comparatively large dynamical masses at the half light radii of the galaxies.