The Surface Brightness Fluctuation survey of galaxy distances. II. Local and large-scale flows

We present results from the Surface Brightness Fluctuation (SBF) Survey for the distances to 300 early-type galaxies, of which approximately half are ellipticals. A modest change in the zero point of the SBF relation, derived by using Cepheid distances to spirals with SBF measurements, yields a Hubb le constant H-0 = 77 +/- 4 +/- 7 km s(-1) Mpc(-1), somewhat larger than the HST Key Project result. We discuss how this difference arises from a diffe rent choice of zero point, a larger sample of galaxies, and a different mod el for large-scale flows. Our result is 4% larger than found in a recent co mparison of the SBF Survey peculiar velocities with predictions derived fro m the galaxy density field measured by redshift surveys (Blakeslee et al. 1 999b). The zero point of the SBF relation is the largest source of uncertai nty, and our value for H-0 is subject to all the systematic uncertainties o f the Key Project zero point, including a 5% decrease if a metallicity corr ection for the Cepheids is adopted. To analyze local and large-scale flows- departures from smooth Hubble how-we use a parametric model for the distrib ution function of mean velocity and velocity dispersion at each point in sp ace. These models include a uniform thermal velocity dispersion and spheric al attractors whose position, amplitude, and radial shape are free to vary. Our modeling procedure performs a maximum likelihood fit of the model to t he observations. Our models rule out a uniform Hubble flow as an acceptable fit to the data. Inclusion of two attractors, one of which having a best-f it location coincident with the Virgo cluster and the other having a fit lo cation slightly beyond the Centaurus clusters (which we refer to by convent ion as the Great Attractor), reduces chi(2)/N from 2.1 to 1.1. The fits to these attractors both have radial profiles such that v approximate to r(-1) (i.e., isothermal) over a range of overdensity between about 10 and 1, but fall off more steeply at larger radius. The best-fit value for the small-s cale, cosmic thermal velocity is 180 +/- 14 km s(-1). The quality of the fi t can be further improved by the addition of a quadrupole correction to the Hubble flow. The dipole velocity offset from the CMB frame for the volume we survey (amplitude similar to 150 km s(-1)) and the quadrupole may be gen uine (though weak) manifestations of more distant density fluctuations, but we find evidence that they are more likely due to the inadequacy of spheri cal models to describe the density profile of the attractors. The residual dipole we find is comparable to the systematic error in these simple, param etrized models; in other words, our survey volume of R < 3000 km s(-1) is, in a mass averaged sense, essentially at rest with respect to the CMB. This contradicts claims of large amplitude flows in much larger volumes that in clude our sample. Our best-fitting model, which uses attenuated power-law m ass distributions for the two attractors, has enclosed mass overdensities a t the Local Group of 7 x 10(14) M. for the Virgo Attractor and 9 x 10(15) M . for the Great Attractor. Without recourse to information about the overde nsities of these attractors with respect to the cosmic mean we cannot provi de a good constraint on Omega(M), but our data do give us accurate measurem ents in terms of delta, the overdensities of the enclosed masses with respe ct to the background: delta Omega(M)(2/3) = 0.33 for the Virgo Attractor an d delta Omega(M)(2/3) = 0.27 for the Great Attractor.