"Topology and Relativity"

OK, so for this Friday's Theoretical Physics Seminar we had Maulik Parikh from IUCAA in India talking about two of his current projects paired for the occasion under the title "Topology and Relativity".

Topology and special relativity

This is a collaboration with Brian Greene and Jana Levin, to appear.

The starting point is the twin paradox on a cylinder. For memory, the twin paradox in flat space-time concerns the age difference of two twins, one of which has been sent to the outer space and back. There is in fact no paradox, since the Principle of Relativity states that all inertial observers are equivalent, whereas the space twin accelerated and decelerated on his Odyssey.

In the case where space is a circle and time a line, and so on a space-time cylinder, both twins, gracefully christened A and B, could be inertial and still go their different ways and meet again. Who, then, asketh Maulik, is younger?
Well, the catch is that the periodic identification of the space coordinate picks a globally preferred frame (the one with winding number zero I suppose). So the first lesson is that

A nontrivial topology breaks global Lorentz symmetry.

The preferred frame could be determined by experiments by sending photons in different directions. In particular, Einstein clock synchronization would only be possible for preferred observers. There is also a discontinuity in the time coordinate, as the inhabitants of Kiribati, lying (not anymore since 1995) on the International Date Line, know very well (after all, as far as time zones are concerned, the Earth's worldvolume is essentially a cylinder).

Suppose there is a compact extra dimension (of mm size according to the ADD scenario). Can we tell the velocity of our (3+1)-brane around it? Obama says "Yes we can!", for which Maulik thinks he should get the Physics Nobel Prize as well...

If the LHC fires gravitons in the extra dimensions, then measuring their return time could make it possible to determine the motion of the brane. This is in fact not realistic, since graviton interactions are suppressed by the Planck mass.

Another effect of an compact extra dimension would be a modification of Newton's law. Since a source placed at a point along the extra coordinate would be repeated infinitely at interval $L$, the standard potential
\[
V(r) = -\frac{GM}{\pi r^2}
\]would be replaced by
\[
V(r) = -\frac{GM}{\pi } \sum_{n=-\infty}^{\infty} \frac{1}{r^2 +(nL)^2}\\
\simeq -\frac{GM}{r} (1+ 2e^{-2\pi r/L})~.
\]For a moving brane, one would have to replace $L$ by $\gamma L$ where $\gamma$ is the relativistic factor. This opens the amusing possiblity of a magnified extra dimension, if our brane were to move ultra-relativistically.

Topology and the Fate of the Universe

The second part of Maulik's talk was on a disconnected topic, base on his paper Enhanced Instability of de Sitter Space in Einstein-Gauss-Bonnet Gravity. The geometry of the early universe is well-described by de Sitter space, which is perturbatively stable. However, Bousso and Hawking showed that dS space can be destabilized by non-perturbative effects (such as instantons, black hole tunneling, etc.). The probability of a gravitational instanton is
\[
\Gamma \sim \frac{\exp(-I_E[\text{instanton}])}{\exp(-I_E[\text{background}])}~.
\]where $I_E$ is the Euclidean action. Now instead of taking the usual Hilbert-Einstein action, Maulik considered the Einstein-Gauss Bonnet action (which appears for instance in the low energy effective action of heterotic string theory). The novelty is the in our dimension the Gauss-Bonnet action is a topological invariant
\[
I_{GB} = -\frac{\Lambda V_4}{8\pi G} - \frac{2\pi \alpha}{G} \chi~,
\]where $\Lambda$ is the cosmological constant, $V_4$ is the volume, $\alpha$ the coupling constant, and last but not least $\chi$ is the Euler number.

The Gauss-Bonnet topological term can enhance the instability of primordial de Sitter space.

In the case of the extremal Nariai black hole with topology $S^2\times S^2$, using the Hilbert-Einstein action as Bousso and Hawking did leads to an instanton probability $\Gamma = \exp[(-\pi L^2/3G)$, which means this is only relevant for a length scale $L$ close to the Planck scale. Disappointing.

In contrast, with its additional topological term, the EGB action leads to an enhanced production of Nariai black holes (with $\chi =2+2=4$) by a factor of $\exp(4\pi\alpha/G)$.

Maulik also mentioned a bound on the maximal curvature of empty dS space.

Other applications include the effect of the topological term on the probabilities of Calabi-Yau manifolds in the string landscape. For example, the quintic has $\chi=-200$, which means it could be suppressed by roughly $\exp(-2000)$.