Repeated studies of the CMB based on WMAP data have revealed an apparent asymmetry in the distribution of temperature fluctuations over the celestial sphere. The studies indicate that the amplitudes of temperature fluctuations are higher in one hemisphere than in the other. We consider whether this asymmetry could originate from a large scale inhomogeneity in the gravitational field enclosing the present Hubble volume. We examine what effect the presence of an inhomogeneity in the gravitational field of size larger than the present Hubble radius would have on the temperature distribution of the CMB and start eliciting its observational signature in the CMB power spectrum. The covariance function contains, in addition to the diagonal entries of the conventional CMB temperature anisostropy power spectrum, non-diagonal entries. We find that specific non-diagonal entries of the covariance function are sensitive to the strength of the inhomogeneity, while the diagonal entries are not. These non-diagonal entries, which are not present in the case of a homogeneous background geometry, are observational signatures of a large-scale inhomogeneity in the background geometry of the universe. Furthermore, we find that an inhomogeneity in the gravitational potential of super-Hubble size would yield a power asymmetry in the CMB with maximal asymmetry at an angle of 90 degrees to the CMB dipole axis. The axis of the CMB power asymmetry was recently estimated by Eriksen et. al. to be located at angles between 83 and 96 degrees to the CMB dipole axis, which is consistent with the prediction of our model. This implies that the location of the observed power asymmetry in the CMB sky could be accounted for by a large-scale inhomogeneity in the gravitational field enclosing the present Hubble volume.

If the Pioneer anomaly has a gravitational origin, it would, according to the equivalence principle, distort the motions of the planets in the Solar System. Since no anomalous motion of the planets has been detected, it is generally believed that the Pioneer anomaly can not originate from a gravitational source in the Solar System. However, this conclusion becomes less obvious when considering models that either imply modifications to gravity at long range or gravitational sources localized to the outer Solar System, given the uncertainty in the orbital parameters of the outer planets. Following the general assumption that the Pioneer spacecraft move geodesically in a spherically symmetric spacetime metric, we derive the metric disturbance that is needed in order to account for the Pioneer anomaly. We then analyze the residual effects on the astronomical observables of the three outer planets that would arise from this metric disturbance, given an arbitrary metric theory of gravity. Providing a method for comparing the computed residuals with actual residuals, our results imply that the presence of a perturbation to the gravitational field necessary to induce the Pioneer anomaly is in conflict with available data for the planets Uranus and Pluto, but not for Neptune. We therefore conclude that the motion of the Pioneer spacecraft must be non-geodesic. Since our results are model independent within the class of metric theories of gravity, they can be applied to rule out any model of the Pioneer anomaly that implies that the Pioneer spacecraft move geodesically in a perturbed spacetime metric, regardless of the origin of this metric disturbance.