Abstract
In this paper we prove a positive energy theorem related to fourth-order gravitational theories, which is a higher-order analogue of the classical ADM positive energy theorem of general relativity. We will also show that, in parallel to the corresponding situation in general relativity, this result intersects several important problems in geometric analysis. For instance, it underlies positive mass theorems associated to the Paneitz operator, playing a similar role in the positive Q-curvature conformal prescription problem as the Schoen–Yau positive energy theorem does for the Yamabe problem. Several other links to Q-curvature analysis and rigidity phenomena are established.
Similar content being viewed by others
Notes
See Appendix B for details on our conventions on Q-curvature.
In this cases we will say that \(Q_g\) is semi-positive.
In the case \(\tau >n-2\), we can achieve \(\delta =\sigma \).
Notice that, actually, \(R_{g}=O_2(r^{-\tau -2})\) which is stronger than the decay we are using, but for our purposes keeping just one weight parameter given by \(\delta \) is enough to achieve the required estimates.
For an explicit construction of this see Example 1 in [38]
In fact, from Theorem 1.17 in [9], it follows that
$$\begin{aligned} u_{k}=1+\frac{a_k}{r^{n-2}} + o(r^{-(n-2)}), \end{aligned}$$where \(a_k\) are constants.
References
Avalos, R., Lira, J.H.M.: Nicolas: energy in fourth order gravity. arXiv e-prints, page arXiv:2102.00545 (2021)
Burgess, C.P.: Quantum gravity in everyday life: general relativity as an effective field theory. Living Rev. Relat. 7, 5 (2004)
Donoghue, J.F.: General relativity as an effective field theory: the leading quantum cerrections. Phys. Rev. D 50(6), 3874–3888 (1994)
Hooft, G.T., Veltman, M.: One loop divergencies in the theory of gravitation. Ann. Inst. Henri Poincare A 20, 69–94 (1974)
Starobinsky, A.: A new type of isotropic cosmological models without singularity. Phys. Lett. B 91(1), 99–102 (1980)
Maldacena, J.: Einstein gravity from conformal gravity. arXiv:1105.5632v2
Mannheim, P.D.: Making the case for conformal gravity. Found. Phys. 42, 388–420 (1980)
Arnowitt, R., Deser, S., Misner, C.W.: Coordinate invariance and energy expressions in general relativity. Phys. Rev. 122(3), 9971006 (2006)
Bartnik, R.: The mass of an asymptotically flat manifold. Commun. Pure Appl. Math. 39(5), 661–693 (1986)
Schoen, R.: Conformal deformation of a riemannian metric to constant scalar curvature. J. Diff. Geom. 20, 479–595 (1984)
Lee, J.M., Parker, T.H.: The Yamabe problem. Bull. Am. Math. Soc. (N.S.) 17(1), 37–91 (1987)
Carlotto, A.: Rigidity of stable minimal hypersurfaces in asymptotically flat spaces. Calc. Var. Partial. Differ. Equ. 55(3), 1–20 (1988)
Carlotto, A., Chodosh, O., Eichmair, M.: Effective versions of the positive mass theorem. Invent. Math. 206(3), 1–20 (2016)
Carlotto, A., Schoen, R.: Localizing solutions of the einstein constraint equations. Invent. Math. 205(3), 1–20 (2016)
Brendle, S., Eichmair, M.: Large outlying stable constant mean curvature spheres in initial data sets. Invent. Math. 107(3), 663–682 (2014)
Eichmair, M., Metzger, J.: Large isoperimetric surfaces in initial data sets. J. Differ. Geom. 94(1), 159–186 (2013)
Huisken, G., Yau, S.T.: Definition of center of mass for isolated physical systems and unique foliations by stable spheres with constant mean curvature. Invent. Math. 124, 281–311 (1996)
Huang, L.H.: Foliations by stable spheres with constant mean curvature for isolated systems with general asymptotics. Commun. Math. Phys. 300(2), 331–373 (2010)
Nerz, C.: Foliations by stable spheres with constant mean curvature for isolated systems without asymptotic symmetry. Calc. Var. Part. Differ. Equ. 54(2), 1911–1946 (2015)
Carlotto, A.: Four Lectures on Asymptotically Flat Riemannian Manifolds. In: Cacciatori, S., Güneysu, B., Pigola, S. (eds.) Einstein Equations: Physical and Mathematical Aspects of General Relativity. Springer, Switzerland (2019)
Lee, D.A.: Geometric relativity. Graduate Studies in Mathematics, vol. 201. American Mathematical Society, Providence, RI (2019)
Schoen, R., Yau, S.T.: On the proof of the positive mass conjecture in general relativity. Commun. Math. Phys. 65, 45–76 (1979)
Schoen, R., Yau, S.T.: Comm. Math. Phys. 79, 231–260 (1981)
Schoen, R., Yau, S.T.: Conformally flat manifolds, kleinian groups and scalar curvature. Invent. Math. 92, 47–71 (1988)
Djadli, Z., Malchiodi, A.: Existence of conformal metrics with constant \(q\)-curvature. Ann. Math. 168, 813–858 (2008)
Gursky, M.J.: The principal eigenvalue of a conformally invariant differential operator, with an application to semilinear elliptic PDE. Commun. Math. Phys. 207(1), 131–143 (1999)
Chang, S.-Y.A., Yang, P.C.: Extremal metrics of zeta functional determinants on 4-manifolds. Ann. Math. 142, 171–212 (1995)
Hang, F., Yang, P.C.: Q-curvature on a class of manifolds with dimension at least 5. Commun. Pure Appl. Math. LXIX, 1452–1491 (2016)
Gursky, M.J., Malchiodi, A.: A strong maximum principle for the Paneitz operator and a non-local flow for the \(Q\)-curvature. J. Eur. Math. Soc. (JEMS) 17(9), 2137–2173 (2015)
Qing, J., Raske, D.: On positive solutions to semilinear conformally invariant equations on locally conformally flat manifolds. Int. Math. Res. Not. 2006(9), 94172 (2006)
Hang, F., Yang, P.C.: Sign of Greens function of Paneitz operators and the \(Q\) curvature. Int. Math. Res. Not. IMRN 19, 9775–9791 (2015)
Hang, F., Yang, P.C.: \(Q\) curvature on a class of 3-manifolds. Commun. Pure Appl. Math. 69(4), 734–744 (2016)
Humbert, E., Raulot, S.: Positive mass theorem for the Paneitz–Branson operator. Calc. Var. Part. Differ. Equ. 36(4), 525–531 (2009)
Michel, B.: Masse des opérateurs GJMS. arXiv e-prints, page arXiv:1012.4414, December (2010)
Michel, B.: Geometric invariance of mass-like asymptotic invariants. J. Math. Phys. 52(5), 052504 (2011)
Maxwell, D.: Solutions of the Einstein constraint equations with apparent horizon boundaries. Commun. Math. Phys. 253(3), 561–583 (2005)
Dilts, J., Maxwell, D.: Yamabe classification and prescribed scalar curvature in the asymptotically Euclidean setting. Commun. Anal. Geom. 26(5), 1127–1168 (2018)
Avalos, R., Lira, J.H.: Reduced thin-sandwich equations on manifolds euclidean at infinity and on closed manifolds: existence and multiplicity. J. Math. Phys. 61, 122501 (2020)
Chen, S.S.: Conformal deformation on manifolds with boundary. Geom. Funct. Anal. 19(4), 1029–1064 (2009)
Gursky, M.J., Hang, F., Lin, Y.-J.: Riemannian manifolds with positive Yamabe invariant and Paneitz operator. Int. Math. Res. Not. IMRN 5, 1348–1367 (2016)
Li, J., Li, Y., Liu, P.: The \(Q\)-curvature on a 4-dimensional Riemannian manifold \((M, g)\) with \(\int _MQdV_g=8\pi ^2\). Adv. Math. 231(3–4), 2194–2223 (2012)
Lin, Y.-J., Yuan, W.: Deformations of Q-curvature I. Calc. Var. Part. Differ. Equ. 55(4), 29 (2016)
Ndiaye, C.B.: Constant q-curvature metrics in arbitrary dimension. J. Funct. Anal. 251, 1–58 (2007)
DeTurck, D.M., Kazdan, J.L.: Some regularity theorems in Riemannian geometry. Ann. Sci. École Norm. Sup. (4) 14(3), 249–260 (1981)
Acknowledgements
The authors would like to thank the CAPES-COFECUB, CAPES-PNPD and ANR (ANR- 18-CE40-002) for their financial support. Also, we would like thank the comments, suggestions and critics made by an anonymous referee to this paper, which have helped us improve the presentation and content of the paper.
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by A. Malchiodi.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix: Some analytic results concerning AE manifolds
In this appendix we will collect some results concerning AE manifolds which are used in the core of the paper. Most of these results are well-known for experts. We include them for the sake of completeness and to deliver a self-contained presentation. For detailed proofs and discussions on these topics, we refer the reader to Bartnik [9], Lee-Parker [11].
Let us start with the following fundamental theorem regarding the properties of the Laplacian on AE manifolds.
Theorem A.1
Let (M, g) an asymptotically Euclidean manifold with a structure at infinity \(\phi :M\setminus K \rightarrow {\mathbb {R}}^n\setminus B_1\) with decay rate \(\tau \). If \(\delta \not \in ({\mathbb {Z}} \setminus \{-1,\cdots ,3-n\})\) and \(q>1\) then
is Fredholm. Moreover
Remark A.1
The decay rates in the set \({\mathbb {Z}} \setminus \{-1,\cdots ,3-n\}\) are called exceptional and we say that \(\delta \) is non-exceptional if \(\delta \not \in {\mathbb {Z}} \setminus \{-1,\cdots ,3-n\}\).
Let us now analyse the relation between different potential structures of infinity. In particular, the following theorem concerns the existence of harmonic coordinates.
Theorem A.2
Let (M, g) be an AE manifold, with \(g\in W^{k,q}_{loc}\), \(k\ge 1\), \(q>n\), and \((\Phi ,x):M\backslash K\mapsto E_R\doteq {\mathbb {R}}^n\backslash \overline{B_R(0)}\) where \(K\subset \subset M\), \(R\ge 1\) is a structure of infinity of order \(\tau > 0\) with \(1-\tau \) non-exceptional, and fix \(1<\eta <2\). There are functions \(y^{i}\in W^{k+1,q}_{\eta }\), \(i=1,\cdots ,n\), such that \(\Delta _gy^i=0\) and \((x^i-y^i)\in W^{k+1,q}_{1-\tau ^{*}}(E_R)\) for \(\tau ^{*}\doteq \min \{\tau ,n-2\}\), which implies
Furthermore, the set of functions \(\{1,y^i\}\) is a basis for \(H_1=\{u\in W^{k,q}_{\eta }: \Delta _gu=0\}\).
Proof
Let us first extend the functions \(x^i\) smoothly to all of M. Then, near infinity, \(\Delta _gx^i=g^{kl}\Gamma ^i_{kl}\doteq \Gamma ^i\in W^{k-1,q}_{-1-\tau }\). Notice that if \(1-\tau >2-n\), then Theorem A.1 implies the existence of \(v^i\in W^{k+1,q}_{1-\tau }\) solving \(\Delta _gv^i=\Gamma ^i\). Now, if \(1-\tau < 2-n\) (the equality case is an exceptional case), we cannot, a priori, improve the decay given by \(r^{2-n}\). This actually follows from Theorem 1.17 in [9]. Therefore, in any case, we know that there are solutions \(v^i\in W^{k+1,q}_{1-\tau ^{*}}\) to \(\Delta _gv^i=\Gamma ^i\), where \(\tau ^{*}=\min \{\tau ,n-2\}\). Therefore, there is some \(v^i\in W^{k+1,q}_{1-\tau ^{*}}\) satisfying \(\Delta _g(x^i-v^i)=0\). Let us then define \(y^i\doteq x^i-v^i\), which implies the first estimate in (A.2), and furthermore, since \(\frac{\partial y^i}{\partial x^j}=\delta ^i_j + o(r^{-\tau ^{*}})\), we see that near infinity \(\{y^i\}\) are coordinates which are asymptotically Cartesian. This, in turn, implies the second estimate in (A.2). The final claim concerning \(H_1\) follows since \(\{1,y^i\}\) spans an \((n+1)\)-dimensional subspace of \(H_1\), but from Proposition 2.2 in [9] \(\mathrm {dim}(H_1)=n+1\). \(\square \)
Remark A.2
Let us highlight that the statement of the above theorem is slightly different than the well-known Theorem 3.1 in [9]. The difference relies in the fact that, in [9], it does not seem to be explicitly stated that, a priori, the decay rate \(r^{2-n}\) cannot be improved, regardless of how large \(\tau \) may be. Nevertheless, as can be seen in the above proof, the above mild correction comes about from results contained in the same reference, as is also clear from a careful examination of the proof of Theorem 3.1 in [9]. Similar comments apply to the following theorem, which can be found in [9] as Corollary 3.2, and presents the relation between two different structures of infinity.
Theorem A.3
Let (M, g) an asymptotically Euclidean manifold, \(g\in W^{k,q}_{loc}\), \(k\ge 1\) and \(q>n\), with two structures at infinity \(\phi ,\psi :M\setminus K \rightarrow {\mathbb {R}}^n\setminus \overline{B_1(0)}\) with decay rates \(\tau _\phi \) and \(\tau _\psi \), where of each of these weights satisfies that \(1-\tau \) is non-exceptional. There exists \((O,a)\in O(n)\times {\mathbb {R}}^n\) such that
which implies
where \(\tau \doteq \min \{\tau _\phi ,\tau _\psi ,n-2\}\), \(x=\phi ^{-1}\) and \(z=\psi ^{-1}\).
Proof
Let \(y^i\) and \(w^i\) be the harmonic coordinates constructed in the previous theorem associated to \(\phi \) and \(\psi \) respectively. Then, since \(H_1\) is intrinsic to M and \(\{1,y^i\}\) and \(\{1,w^i\}\) are bases for this space, we get that
where \(A\in \mathrm {GL}(n,{\mathbb {R}})\) a priori, but actually, from the construction of the previous theorem, we know that \(z^i=A^i_jx^j + a^i - A^i_jv^j + {\bar{v}}^i\), with \(v^i\in W^{k,q}_{1-\tau ^{*}_{\phi }}\) and \({\bar{v}}^i\in W^{k,q}_{1-\tau ^{*}_{\psi }}\), where \(\tau ^{*}\) was defined in the previous theorem. Since these last two systems are Cartesian, then \(A\in O(n)\) and we explicitly see that \(z^i-A^i_jx^j-a^i\in W^{k,q}_{1-\tau }\). \(\square \)
The above estimate is the best possible and it is determined by the Ricci curvature as shown by the next theorem which corresponds to Proposition 3.3 in [9], see also [44].
Theorem A.4
Let (M, g) an asymptotically Euclidean manifold with a structure at infinity \(\phi :M\setminus K \rightarrow {\mathbb {R}}^n\setminus B(0,1)\) with decay rate \(\tau \) such that \((\phi _{*} g-\delta )\in W^{2,q}_{-\tau }({\mathbb {R}}^n\setminus B(0,1))\) for \(q>n\) and such that
Then there exists a structure at infinity \(\Theta :M\setminus K' \rightarrow {\mathbb {R}}^n\setminus B_1\) such that \((\Theta _* g-\delta )\in W^{2,q}_{-\eta }({\mathbb {R}}^n\setminus B_1)\).
Finally, the following statement concerning the conformal Laplacian will be important in our analysis. It is a variant of theorem 9.2 of [11]
Corollary A.1
Let (M, g) an asymptotically Euclidean manifold of order \(\tau >0\) and assume \(R_g\ge 0\). If \(2-n<\delta <0\) and \(q>1\), then
is an isomorphism.
Proof
Since \(L_g\) is a compact perturbation of the Laplacian, it is also Fredholm, hence thanks to its self-adjointness it suffices to proof injectivity to get the result. Consider \( u\in W_\delta ^{2,q}\) such that \(\Delta _g u=c_n R_g u\), and first notice that by elliptic regularity we can assume that \(u\in W^{2,q}_{\delta }\) with \(q>n\). Also, notice that \(R_g u \in W^{2,q}_{\delta '-2}\) for any \(\delta '>\delta -\tau \). Since \(\tau >0\), we can pick \(\delta '\) satisfying \(\max \{\delta -\tau ,2-n\}<\delta '<\delta \) and appeal to Theorem A.1 to obtain \(u\in W^{2,q}_{\delta '}\). Since we can repeat the argument as much as necessary, we can assume that \(u\in W^{2,q}_{\delta '}\) with \(\delta '> 2-n\) taken arbitrary close to \(2-n\). Then, we can multiply \(L_g(u)=0\) by u and integrate by parts to obtain
Since \(u=O_2(r^{\delta '})\) near infinity, we can estimate \(\vert \nabla u \vert _g^2=O(r^{2\delta '-2})\) and \(R_gu^2=O(r^{2\delta '-\tau -2})\) which implies these quantities are in \(L^1(M)\) as long as chose \(2-n<\delta '<\frac{2-n}{2}\). Furthermore, under these condition one finds \(u\partial _\nu u=O(r^{2\delta '-1})\) and therefore
so that we can pass to the limit as \(R\rightarrow +\infty \), which gives that \(u\equiv 0\) since \(R_g{\ge }0\) and \(u\rightarrow 0\) as \(|x|\rightarrow \infty \), which concludes the proof. \(\square \)
Appendix: Conventions on Q-curvature
Let us adopt the following general definition of Q-curvature for an arbitrary Riemannian manifold \((M^n,g)\) with \(n\ge 3\):
where \(S_g\doteq \frac{1}{n-2}\left( \mathrm {Ric}_g - \frac{1}{2(n-1)}R_gg \right) \) stands for the Schouten tensor and \(\sigma _k(S_g)\) stands for the k-th elementary symmetric function of the eigenvalues of \(S_g\). In this context, the Paneitz operator is defined by
for all \(u\in C^{\infty }(M)\). In this context, the following relations hold:
which implies that
In particular, for \(n\ne 4\), if \({\bar{g}}=u^{\frac{4}{n-4}}g\), then
In the case of \(n=4\) we can apply the above definitions to get
and in this case, if \({\bar{g}}=e^{2u}g\), we have that [42]
Now, let us notice that it is also quite standard to redefine the Q-curvature in dimension 4 via (see [29, 41])
and in this case, it follows that
We will not adopt this redefinitions and keep a unified notation via (B.1) along this paper.
Appendix: Conformal normal coordinates
In order to deliver a presentation as self-cointained as possible, this section is meant to summarise some of the results concerning conformal normal coordinates presented in [11], which are used in the core of this paper. The basic construction is given on a smooth Riemannian manifold \((M^n,g)\) where we intend to expand g around a fixed point \(p\in M\). In particular, we are interested in finding an element \({\tilde{g}}\) within the conformal class [g] for which, in \({\tilde{g}}\)-normal coordinates \(\{x^i\}\), the following expansion holds around p:
for any chosen \(N\ge 2\), where \(r=|x|\) (see Theorem 5.1 in [11]). This type of expansion is achieved by first noticing that, for any Riemannian metric g, in g-normal coordinates \(\{x^i \}_{i=1}^n\) around \(p\in M\), the following holds
where all the coefficients are evaluated at p. From this expression it is possible to compute \(\det (g)\) around p for any such metric, so as to get
where again all the coefficients are evaluated at p. Assuming an expanssion of the form \(\det (g)=1+O(r^N)\) with \(N\ge 2\) (the case \(N=2\) is valid for any metric in its own normal coordinates) and appelaing to Theorem 5.2 in [11], the authors can find a homogeneous polynomial \(f\in {\mathcal {P}}_N\) so that the expansion of \(\det ({\tilde{g}})=1+O(r^{N+1})\) for \({\tilde{g}}=e^{2f}g\), which establishes an inductive proof (see the proof of Theorem 5.1). But notice that this implies that the symmetrization of the coefficients in (C.2) of order up to N must vanish for \({\tilde{g}}\). Thus, in the case we do this construction for \(N\ge 4\), we see that this implies
Putting together the second condition above with the contracted Bianchi identities, we also see that