How does TeVe S measure up to the task of describing gravitational lensing?
n TeVeS an extragalactic system lenses light, or radio waves, just as would GR, were the latter supplemented by DM in the amount and with the distribution necessary to reproduce the observed galactic dynamics. . . the TeVe S scheme is falsifiable—by comparison of the calculated potential with that inferred from the lensing —to a larger extent than is the DM paradigm for which any discrepancy can be tucked away into the invisible component. . . .
Zhao et al. (2006) . . . compare TeVeS predictions with a large sample of quasars doubly imaged by intervening galaxies. . . . The corresponding mass-to-light ratios are found to be in the normal range for stellar populations, with some exceptions. This result clashes with the claim by Ferreras et al. (2008) that lensing by galaxies from the very same sample can only be explained in MOND by including a lot of DM apart from neutrinos. But the last authors use a mixture of MOND and GR instead of TeVeS.
What should the probability distribution by angular separation of the two images in a sample of lensed quasars?
This important question has proved troublesome for the DM paradigm. In TeVeS it has been investigated by Chen and Zhao (2006) and lately by Chen (2008). . . . these workers compare predictions of both TeVeS for a purely baryonic universe with cosmological constant and of GR with DM and baryons with the CLASS/JVAS quasar survey. After the preliminary work the later paper reports that TeVeS comes out on top. All the above is accomplished with spherical mass models of the galaxies; a step towards the modeling of asymmetric lenses within TeVeS has been taken by Shan, Feix et al. (2008).
When it comes to weak lensing (distorted but unsplit images) by clusters of galaxies, a pure MOND account is less than satisfactory. The case of spherically symmetric clusters is fairly summarized by Takahashi and Chiba (2007). . . . These authors . . fail to get a fit with observations unless they add a neutrino component a la Sanders (2003, 2007); the required neutrino mass is unrealistically large, so it seems that a DM component is needed to buttress the MOND effect. . . .
Nonspherical cluster systems are also problematic. In the massive colliding clusters systems MACSJ0025.4–1222 (Bradaˇc et al. 2008) and 1E0657–56 (Clowe et al. 2004) the galaxy components have been rudely separated from the hot gas concentrations. Weak lensing mapping using background galaxies shows the gravitating mass to be preponderately located in the regions containing the galaxies, rather than in the gas which accounts for the bulk of the visible baryonic mass (Clowe et al 2006, Bradaˇc et al 2008). Collisionless DM would indeed be expected to move together with the galaxies and get separated from the collisionless gas; hence the widespread inference that much DM exists in these systems. However, this view conflicts with the finding (Mahdavi et al. 2007) that in the merging clusters A520 the lensing center is in the hot gas which is separate from the galaxy concentration. Angus, Famaey et al. (2006) considered it possible to explain the lensing seen in 1E0657–56 by TeVeS with a reasonable purely baryonic matter distribution, but later concluded (Angus, Shan et al. 2007) that a collisionless component is needed after all, with neutrinos just barely supplying a resolution. This conclusion is confirmed by a careful study of Feix, Fedeli et al. (2008) who . . . conclude that the source of gravity in 1E0657–56 must include an invisible component.
The weak lensing by cluster Cl0024+17 provides another relevant case study. Jee et al. (2007) find its deduced mass surface density to exhibit a ring which does not coincide either with the galaxy distribution, or the hot gas. Again this has been hailed as graphic proof of DM. But Milgrom and Sanders (2008) argue that such feature is actually expected in MOND, lying as it does at the transition between the Newtonian and the MOND regime. Famaey et al. (2007) conclude that the lensing in Cl0024+17 can be modeled in MOND by including 2 eV neutrinos. A truly TeVe S model of Cl0024+17 is still outstanding.
Turn now to cosmology.
Critics of MOND used to argue that the complex power spectrum of cosmological perturbations of the background radiation, which is said to be well fit by the “concordance” DM model of the universe, proves that DM is essential to any rational picture of the cosmos. . . . Skordis et al. (2006) have shown that, without invoking DM, TeVeS can largely be made consistent with the observed spectrum of the spatial distribution of galaxies, and of the cosmic microwave radiation, if one allows for contributions to the energy density of massive neutrinos, and of a cosmological constant. . . . Thus although the elimination of galaxy bound DM was the original motivation for MOND, and thus for TeVe S, the later may potentially provide a way to eliminate cosmological (homogeneously distributed) DM!
Apart from the DM mystery, cosmology furnishes us with a “dark energy” mystery. Dark energy is the agent responsible for the observed acceleration of the Hubble expansion in the context of GR cosmological models. . . . Diaz-Rivera et al. (2006) find an exact deSitter solution of TeVeS cosmology which can represent either early time inflation epochs or the late time acceleration era. Hao and Akhoury (2005) conclude that with a suitable choice of the TeVeS function F, the scalar field can play the role of dark energy. According to Zhao (2006) the choice of F implicit in the work of Zhao and Famaey (2006) leads to cosmological models that evolve at early times like those of standard cold DM cosmology, and display late time acceleration with the correct present Hubble scale, all this without needing DM or dark energy. . . .
The advent of large lensing surveys may open a way to distinguish between these two theories, as well as between GR and other modified gravities, by exposing correlations between galaxy number density and weak lensing shear (Schmidt 2008, Zhang et al. 2007). The effect of the dark energy on the expansion can be separated out by comparing cosmological models with the same expansion history in two theories. And the ultimate confrontation between GR and TeVeS cosmology may be accomplished by cross-correlating galaxy number density with cosmic microwave background antenna temperature (Schmidt et al. 2007).
