The Growth Rate of Supermassive Black Holes and Its Dependence on the Stellar Mass of Galaxies at the Present Epoch

We study the distribution of accretion rates onto supermassive black holes in AGNs of the local Universe (z < 0.15) based on near-infrared and hard X-ray surveys (2MASS and Swift/BAT). Using sufficiently accurate black holemass estimates has allowed the Eddington ratio.Edd to be reliably estimated approximately for half of the objects in the AGN sample; a rougher estimate from the correlation of M-BH with the galaxy stellarmassM* has been used for the remaining ones. As a result, for a wide range of galaxy masses, 9.28 < log(M-*/M-circle dot) < 12.28, including the most massive galaxies in the local Universe, we have shown that the distribution f(lambda(Edd)) above log lambda(Edd) = - 3 is described by a power law with M*-independent parameters and falls off with a characteristic slope approximate to 0.7 up to the Eddington limit (log lambda(Edd) similar to 0), where there is evidence for a break. In addition, there is evidence that at log lambda(Edd) < -3 the dependence f(lambda(Edd)) has a lower slope or flattens out. The mean characteristic growth time of supermassive black holes at the present epoch has been estimated. It has turned out to depend weakly on the galaxy stellar mass and to exceed the lifetime of the Universe, but by no more than one order of magnitude. The mean duty cycle of supermassive black holes (the fraction of objects with lambda(Edd) > 0.01) in the local Universe has been estimated. It also depends weakly on M-* and is 0.2-1%. On the whole, these results obtained for the present epoch confirm the trends pointed out in previous studies for the earlier Universe, refining the parameters of the dependence f(lambda(Edd)vertical bar M-*) at z < 0.15. The revealed universal (weakly dependent on the galaxy stellar mass) pattern of the dependence f(lambda(Edd)) probably stems from the fact that, at present, the episodes of mass accretion onto supermassive black holes are associated mainly with stochastic processes in galactic nuclei rather than with global galaxy evolution processes.