04.06.2013 Views

Gravity and Strings

Gravity and Strings

Gravity and Strings

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

172 Conserved charges in general relativity<br />

Apart from the problem of the gravitational energy–momentum tensor, the definition<br />

of conserved quantities in GR has many interesting points. Several approaches have been<br />

proposed <strong>and</strong> here we are going to study two: the construction of an energy–momentum<br />

pseudotensor for the gravitational field <strong>and</strong> the Noether approach. In both approaches there<br />

is a great deal of arbitrariness <strong>and</strong> in Section 6 we will study <strong>and</strong> compare several different<br />

results given in the literature in the weak field limit, finding complete agreement <strong>and</strong> a deep<br />

relation to the massless spin-2 relativistic field theory studied in Chapter 3.<br />

6.1 The traditional approach<br />

As we have stressed several times, the metric (or Rosenfeld) energy–momentum tensor of<br />

any general-covariant Lagrangian always satisfies (on-shell) the equation<br />

∇µTmatter µν = 0, (6.1)<br />

as a direct consequence of general covariance. This equation is crucial for the consistency of<br />

the theory. Furthermore, it is the covariantization of the Minkowskian energy–momentumconservation<br />

equation<br />

∂µTmatter µν = 0, (6.2)<br />

which is discussed at length in Chapter 2, <strong>and</strong> from which we can derive local conservation<br />

laws of the mass, momentum, <strong>and</strong> angular momentum <strong>and</strong>, in general, of those charges<br />

related to the invariance of a theory under certain coordinate transformations.<br />

In curved spacetime, however, Eq. (6.1) is not equivalent to a continuity equation for the<br />

tensor density √ |g| Tmatter µν that holds in Minkowski spacetime. Actually, we can rewrite<br />

Eq. (6.1) in the form <br />

|g| Tmatter µν<br />

=−Ɣρσ ν Tmatter ρσ , (6.3)<br />

∂µ<br />

<strong>and</strong>, in general, the r.h.s. of this equation does not vanish. From this equation we cannot<br />

derive any local conservation law.<br />

In a sense this was to be expected: only the total (matter plus gravity) energy <strong>and</strong> momentum<br />

should be conserved 3 <strong>and</strong>, therefore, we can only hope to be able to find local<br />

conservation laws for the total energy–momentum tensor. Now, how is the gravity energy–<br />

momentum tensor defined in GR? This is an old problem of GR. 4 It is clear that we cannot<br />

use the same definition (Rosenfeld’s) as for the matter energy–momentum tensor because<br />

that leads to a total energy–momentum tensor that vanishes identically on-shell. On the<br />

other h<strong>and</strong>, if we found a covariantly divergenceless gravitational energy–momentum tensor,<br />

the total energy–momentum tensor would have the same problem as the matter one.<br />

In fact, it can be argued, on the basis of the PEGI, that it is impossible to define a fully<br />

general-covariant gravitational energy–momentum tensor: according to the PEGI we can<br />

remove all the physical effects of a gravitational field locally, at any given point, by using an<br />

appropriate (free-falling) reference frame. This means that we could make the gravitational<br />

3 Actually, the coupling of gravity to the total, conserved, energy–momentum tensor was the main principle<br />

leading in Chapter 3 to GR.<br />

4 Some early references on the energy–momentum tensor of the gravitational field are [90, 355–627, 660,<br />

839].

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!