We assume that dark matter is comprised of axion-like particles (ALPs) generated by the realignment mechanism in the post-inflationary scenario. This leads to isocurvature fluctuations with an amplitude of order one for scales comparable to the horizon at the time when the ALP field starts oscillating. The power spectrum of these fluctuations is flat for small wave numbers, extending to scales relevant for cosmological observables. Denoting the relative isocurvature amplitude at k * = 0.05 Mpc −1 by f iso , Planck observations of the cosmic microwave background (CMB) yield f iso < 0.31 at the 2σ-level. This excludes the hypothesis of post-inflationary ALP dark matter with masses m a < 10 −20 -10 −16 eV, where the range is due to details of the ALP mass-temperature dependence. Future CMB stage IV and 21-cm intensity mapping experiments may improve these limits by 1-2 orders of magnitude in m a .

Spatial variations in the distribution of galaxy luminosities, estimated from redshifts as distance proxies, are correlated with the peculiar velocity field. Comparing these variations with the peculiar velocities inferred from galaxy redshift surveys is a powerful test of gravity and dark energy theories on cosmological scales. Using ∼ 2 × 10 5 galaxies from the SDSS Data Release 7, we perform this test in the framework of gravitational instability to estimate the normalized growth rate of density perturbations f σ8 = 0.37 ± 0.13 at z ∼ 0.1, which is in agreement with the ΛCDM scenario. This unique measurement is complementary to those obtained with more traditional methods, including clustering analysis. The estimated accuracy at z ∼ 0.1 is competitive with other methods when applied to similar datasets. Introduction.-Unraveling the origin of cosmic acceleration remains one of the biggest challenges in fundamental physics. Lacking a natural explanation within the standard paradigms of cosmology and particle physics, many theoretical models have been proposed, ranging from the inclusion of new scalar fields to genuine modifications of general relativity [1,2]. Models that provide the same expansion history of the universe can generally lead to a very different evolution of density fluctuations. In the linear regime, the latter is fully captured by the growth rate,

Aims. We explore the lensing properties of asymmetric matter density distributions in Bekenstein's tensor-vector-scalar theory (TeVeS). Methods. Using an iterative Fourier-based solver for the resulting non-linear scalar field equation, we numerically calculate the total gravitational potential and derive the corresponding TeVeS lensing maps. Results. Considering variations on rather small scales, we show that the lensing properties significantly depend on the lens' extent along the line of sight. Furthermore, all simulated TeVeS convergence maps strongly track the dominant baryonic components; nonlinear effects, being capable of counteracting this trend, turn out to be very small. Setting up a toy model for the cluster merger 1E0657 − 558, we infer that TeVeS cannot explain observations without assuming an additional dark mass component in both cluster centers, which is in accordance with previous work.

Strong gravitational lensing by galaxies in MOdified Newtonian Dynamics (MOND) has until now been restricted to spherically symmetric models. These models were able to account for the size of the Einstein ring of observed lenses, but were unable to account for doubleimaged systems with collinear images, as well as four-image lenses. Non-spherical models are generally cumbersome to compute numerically in MOND, but we present here a class of analytic non-spherical models that can be applied to fit double-imaged and quadruple-imaged systems. We use them to obtain a reasonable MOND fit to 10 double-imaged systems, as well as to the quadruple-imaged system Q2237+030 which is an isolated bulge-disc lens producing an Einstein cross. However, we also find five double-imaged systems and three quadruple-imaged systems for which no reasonable MOND fit can be obtained with our models. We argue that this is mostly due to the intrinsic limitation of the analytic models, even though the presence of small amounts of additional dark mass on galaxy scales in MOND is also plausible.