I'm reading the paper "Area, Entanglement Entropy and Supertranslations at Null Infinity" and there is a point that I have a doubt on what the authors are actually doing. First, let me summarize the context. They are considering asymptotically flat spacetimes on which we introduce retarded coordinates $(u,r,z,\bar{z})$ near $\mathcal{I}^+$ so that the metric becomes $$ds^2=-du^2-2dudr+2r^2\gamma_{z\bar{z}}dzd\bar{z}+\frac{2m_B}{r}du^2+rC_{zz}dz^2+rC_{\bar{z}\bar{z}}d\bar{z}^2+D^zC_{zz}dudz+D^{\bar{z}}C_{\bar{z}\bar{z}}dud\bar{z}+\dots\tag{3}$$ In the above notation $D_A$ is the covariant derivative on the unit sphere and $\gamma_{AB}$ its corresponding round metric. The coordinates $z,\bar{z}$ are holomorphic coordinates on the sphere.

Now, in section 3, page 5, the authors say:

We wish to study the area of a cut $\Sigma$ of $\mathcal{I}^+$ defined by $$u=u_\Sigma(z,\bar{z})\tag{9}$$ in the geometry (3) and also to study its variation under supertranslations of $\Sigma$ $$u_\Sigma\to u_\Sigma+f\tag{10}$$ with the geometry held fixed. Of course this area is infinite so we must introduce both a regulator and a subtraction. We regulate the area by the replacement of $\mathcal{I}^+$ wth the past lightcone of a point which approaches $i^+$. For the flat Minkowski metric the null hypersurface $$r=-\frac{1}{2}(u-u_0)\tag{11}$$ approaches $\mathcal{I}^+$ for $u_0\to +\infty$ with $u$ held fixed. More generally we solve the ODE $$\left(1-\frac{2m_B}{r}\right)du+2dr=0\tag{12}$$ which guarantees that the surface is null, and choose the integration constants at each $(z,\bar{z})$ so that the surface lies at large radius, approaching infinity, for finite $u$.

If I understood, they wish to define a familly of surfaces at finite radius such approach $\mathcal{I}^+$ in the large $r$ limit.

In Minkowski spacetime they do it directly by defining, in Eq. (11) the surfaces as level sets of a function.

My issue is with Eq. (12). Where it comes from? How do we derive it? I believe what the authors want is to define the normal covector $n_a$ of the surface. In that case is Eq. (12) just the null covector condition for $n_a$ written in Bondi coordinates dropping corrections proportional to $1/r$? If not, where the equation comes from?

  • 1
    $\begingroup$ Minor comment to the post (v1): In the future please link to abstract pages rather than pdf files. $\endgroup$ – Qmechanic Jun 29 '19 at 5:16

(12) is an equation for a null hypersurface, just like (11) is, this hypersurface is a past lightcone of a very late-time event, serving as an approximation of future null infinity $\mathcal{I}^{+}$. Only (11) is for a Minkowski spacetime while (12) is for asymptotically flat spacetime in Bondi gauge and so it is written as a differential equation. This equation should be understood as following: if a point $p$ with coordinates $(u,r,z,\bar{z})$ is on a hypersurface, then a nearby point $p'$ with coordinates $(u+du,r+dr,z+dz,\bar{z}+d\bar{z})$ is also on a hypersurface only if (12) is satisfied for the coordinate differentials.

This equation may be interpreted as $n_{\mu} d x^\mu=0$, where $n^\mu$ is a normal to this hypersurface. Since this is a null hypersurface this normal is a null vector, $n_\mu n^\mu = 0$ and is also tangent to the hypersurface.

For more familiarity with the Bondi–Sachs coordinates, I would recommend reading the Scholarpedia page on Bondi–Sachs formalism, as well as lectures by Compère, section “Gravity in Bondi gauge” onward.

  • $\begingroup$ So, we are describing the hypersurface using a one-form $\omega$ such that $\omega(X)=0$ for all $X\in T\Sigma$. For level sets $f = c$ as in (11), the appropriate form is $\omega = df$. Now in the general case, one is defining the form - hence the distribution of tangent spaces to the hypersurface - and getting the surface out of it, which by Frobenius' theorem works out when $\omega \wedge d\omega = 0$. Is this the overall idea? $\endgroup$ – user1620696 Jun 30 '19 at 15:30
  • $\begingroup$ No, the “idea” is to write equation for a past lightcone of a very late-time event in a Bondi gauge. By trying to find coordinate-free version of equations you are missing the point of gauge fixing. There is a residual freedom left in this description: it is precisely BMS group, which is the main feature of the B–S formalism. $\endgroup$ – A.V.S. Jun 30 '19 at 15:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.