Skip to content
Licensed Unlicensed Requires Authentication Published by De Gruyter October 14, 2015

Intrinsic flat stability of the positive mass theorem for graphical hypersurfaces of Euclidean space

  • Lan-Hsuan Huang EMAIL logo , Dan A. Lee and Christina Sormani

Abstract

The rigidity of the Positive Mass Theorem states that the only complete asymptotically flat manifold of nonnegative scalar curvature and zero mass is Euclidean space. We study the stability of this statement for spaces that can be realized as graphical hypersurfaces in 𝔼n+1. We prove (under certain technical hypotheses) that if a sequence of complete asymptotically flat graphs of nonnegative scalar curvature has mass approaching 0, then the sequence must converge to Euclidean space in the pointed intrinsic flat sense. The appendix includes a new Gromov–Hausdorff and intrinsic flat compactness theorem for sequences of metric spaces with uniform Lipschitz bounds on their metrics.

Award Identifier / Grant number: 0932078 000

Award Identifier / Grant number: DMS 1308837

Award Identifier / Grant number: DMS 1452477

Award Identifier / Grant number: DMS 1309360

Funding statement: This material is also based upon work supported by the NSF under Grant No. 0932078 000, while all three authors were in residence at the Mathematical Sciences Research Institute in Berkeley, California, during the Fall 2013 program in Mathematical General Relativity. Huang is partially supported by NSF DMS 1308837 and DMS 1452477. Lee is partially supported by a PSC CUNY Research Grant. Sormani is partially supported by a PSC CUNY Grant and NSF DMS 1309360

A Appendix

The following theorem concerning intrinsic flat limits of integral current spaces with varying metrics may be applicable in other settings as well. The Gromov–Hausdorff part of this theorem was already proven by Gromov in [8] but with a completely different proof in which the common metric space Z is the disjoint union. The fact that one also obtains an intrinsic flat limit which agrees with the Gromov–Hausdorff limit is new.

Theorem A.1.

Fix a precompact n-dimensional integral current space (X,d0,T) without boundary (e.g. T=0) and fix λ>0. Suppose that dj are metrics on X such that

(A.1)λdj(p,q)d0(p,q)1λ.

Then there exists a subsequence, also denoted dj, and a length metric d satisfying (A.1) such that dj converges uniformly to d

(A.2)ϵj=sup{|dj(p,q)-d(p,q)|:p,qX}0.

Furthermore,

(A.3)limjdGH((X,dj),(X,d))=0

and

(A.4)limjd((X,dj,T),(X,d,T))=0.

In particular, (X,d,T) is an integral current space and set(T)=X so there are no disappearing sequences of points xj(X,dj).

In fact, we have

(A.5)dGH((X,dj),(X,d))2ϵj

and

(A.6)d((X,dj,T),(X,d,T))2n+12λn+12ϵj𝑴(X,d0)(T).

To prove this theorem we need a series of lemmas:

Lemma A.2.

Under the hypothesis of Theorem A.1, there exists a subsequence, also denoted dj, and a length metric d satisfying (A.1) such that dj converges uniformly to d,

limjsup{|dj(p,q)-d(p,q)|:p,qX}0,

and (X,d,T) is an integral current space.

Proof.

Observe that the functions dj may be extended to the metric completion

dj:X¯×X¯[0,diamdj(X)][0,λdiamd0(X)].

By (A.1) they are equicontinuous and so by the Arzela–Ascoli Theorem they have a subsequence converging uniformly to a function

d:X¯×X¯[0,λdiamd0(X)].

Taking the limit of (A.1) we see that d satisfies (A.1) as well. In particular, d is a metric on X.

Furthermore, the Ambrosio–Kirchheim mass measure defined using d0 and defined using dj may be related as follows:

λnT0TjT0λ-n.

Recall that X=set0(T)X¯ because (X,d0,T) is an integral current space (by the definition of integral current space). In general the set of positive density also depends upon the metric just as the mass measure done. Here we have

X=set0(T)
={pX¯:lim infr0T0(Bp(r))rn>0}
={pX¯:lim infr0Tj(Bp(r))rn>0}
=setj(T)

and so (X,dj,T) is also an integral current space. This is true for all j=1,2,,. ∎

Lemma A.3.

Given two metric spaces (X,dj) and (X,d) there exists a common metric space

Zj=[-ϵj,ϵj]×X,

where

(A.7)ϵj=sup{|dj(p,q)-d(p,q)|:p,qX}

with a metric dj on Zj such that

(A.8)dj((-ϵj,p),(-ϵj,q))=dj(p,q)

and

(A.9)dj((ϵj,p),(ϵj,q))=d(p,q).

Thus we have metric isometric embeddings φj:(X,dj)(Zj,dj), φj:(X,d)(Zj,dj) such that

φj(p)=(-ϵj,p)𝑎𝑛𝑑φj(p)=(ϵj,p).

In addition, if d0,dj satisfy (A.1), then

(A.10)dj(z1,z2)d0((t1,p1),(t2,p2)):=|t1-t2|+λd0(p1,p2).

More precisely, we define dj by

dj(z1,z2):=min{d,d-,d+,d-+,d+-}

where

d=d(z1,z2)=|t1-t2|+max{dj(p1,p2),d(p1,p2)},
d-=d-(z1,z2)=|t1+ϵj|+|t2+ϵj|+dj(p1,p2),
d+=d+(z1,z2)=|t1-ϵj|+|t2-ϵj|+d(p1,p2),
d-+=d-+(z1,z2)=inf{d-(z1,z)+d+(z,z2):zZj},
d+-=d+-(z1,z2)=inf{d+(z1,z)+d-(z,z2):zZj}.

Note that Zj need not be a complete metric space, even if X is complete with respect to both metrics. See Example A.4. However, we may always take the metric completion of Zj if we need a complete metric space.

Before proving this lemma we apply it to prove Theorem A.1:

Proof of Theorem A.1.

First apply Lemmas A.2 and A.3 and take the metric completion of Zj if it is not yet complete. Observe that

dj((-ϵj,p),(ϵj,p))d-((-ϵj,p),(ϵj,p))
=0+2ϵj+dj(p,p)
=2ϵj.

Thus

dHZj(φj(X),φj(X))2ϵj

and we have (A.5) which implies (A.3).

To obtain (A.6), we take Bj=Iϵ×T to be the product integral current on Zj=Iϵ×X, where Iϵ=[-ϵj,ϵj] (see [21] for the precise definition of such products of intervals with currents). When T is just integration over a smooth manifold M, then Iϵ×T is just integration over Iϵ×M.

In [21] it is proven that

(Iϵ×T)=Iϵ×(T)+(Iϵ)×T.

Since T=0, we have

Bj=φj#T-φj#T.

Then by the definition of the intrinsic flat distance,

d((X,dj,T),(X,d,T))dFZj(φj#T,φj#T)
𝐌(Zj,dj)(B)+0.

So we need only estimate the mass of Bj.

In [21] it is shown that

𝐌(Zj,Dj)([-ϵj,ϵj]×T)=2ϵj𝐌(X,λd0)(T)

when the distance Dj is the isometric product metric on Zj defined with d0:

Dj((t1,p1),(t2,p2))=|t1-t2|2+(λd0(p1,p2))2.

Since

dj(z1,z2)d0((t1,p1),(t2,p2)):=|t1-t2|+λd0(p1,p2)
2Dj((t1,p1),(t2,p2)).

We have

𝐌(Zj,dj)(B)𝐌(Zj,2D)(B)
2n+12𝐌(Zj,Dj)(B)
2n+122ϵj𝐌(X,λd0)(T)
2n+12λn+12ϵj𝐌(X,d0)(T).

Thus we have (A.6) which implies (A.4).

This completes the proof of Theorem A.1. ∎

Finally, we prove Lemma A.3:

Proof.

First note that

d-+=|t1+ϵj|+|t2-ϵj|+2ϵj+inf{dj(p1,p)+d(p,p2):pX},
d+-=|t1-ϵj|+|t2+ϵj|+2ϵj+inf{d(p1,p)+dj(p,p2):pX}.

Observe that dj is immediately symmetric and nonnegative. It is positive definite because

min{d(z1,z2),d-(z1,z2),d+(z1,z2)}|t1-t2|+min{dj(p1,p2),d(p1,p2)}

and clearly d-+(z1,z2),d+-(z1,z2)>2ϵj for distinct z1,z2.

Before proving the triangle inequality, we apply (A.7) to prove (A.8):

d-((-ϵj,p1),(-ϵj,p2))=dj(p1,p2),
d((-ϵj,p1),(-ϵj,p2))dj(p1,p2),
d+((-ϵj,p1),(-ϵj,p2))=4ϵj+d(p1,p2)4ϵj+dj(p1,p2)-ϵjdj(p1,p2),
d-+((-ϵj,p1),(-ϵj,p2))=0+2ϵj+2ϵj+inf{dj(p1,p)+d(p,p2):pX}
4ϵj+dj(p1,p2)-ϵjdj(p1,p2),
d+-((-ϵj,p1),(-ϵj,p2))=2ϵj+0+2ϵj+inf{d(p1,p)+dj(p,p2):pX}
4ϵj+dj(p1,p2)-ϵjdj(p1,p2).

Naturally, (A.9) follows in a similar way.

It suffices now to prove the triangle inequality. In (A.13)–(A.16) we prove the triangle inequality in the case where

(A.11)dj(z1,z2)=min{d(z1,z2),d-(z1,z2),d+(z1,z2)}

and

(A.12)dj(z2,z3)=min{d(z2,z3),d-(z2,z3),d+(z2,z3)}.

Observe that

(A.13)dj(z1,z3)d(z1,z3)
=|t1-t3|+max{dj(p1,p3),d(p1,p3)}
|t1-t2|+|t2-t3|
+max{dj(p1,p2)+dj(p2,p3),d(p1,p2)+d(p2,p3)}
|t1-t2|+max{dj(p1,p2),d(p1,p2)}
+|t2-t3|+max{dj(p2,p3),d(p2,p3)}
=d(z1,z2)+d(z2,z3),
dj(z1,z3)d-(z1,z3)
=|t1+ϵj|+|t3+ϵj|+dj(p1,p3)
|t1-t2|+|t2+ϵj|+|t3+ϵj|+dj(p1,p2)+dj(p2,p3)
|t1-t2|+max{dj(p1,p2),d(p1,p2)}
+|t2+ϵj|+|t3+ϵj|+dj(p2,p3)
d(z1,z2)+d-(z2,z3),

and similarly

(A.14)dj(z1,z3)d(z1,z2)+d+(z2,z3).

Clearly,

|t1+ϵj|+|t3+ϵj||t1+ϵj|+2|t2+ϵj|+|t3+ϵj|

so

(A.15)dj(z1,z3)d-(z1,z2)+d-(z2,z3),
dj(z1,z3)d+(z1,z2)+d+(z2,z3).

Immediately by the definition we have

(A.16)dj(z1,z3)d-+(z1,z3)d-(z1,z2)+d+(z2,z3),
dj(z1,z3)d+-(z1,z3)d+(z1,z2)+d-(z2,z3).

Thus we have shown the triangle inequality holds as long as (A.11)–(A.12) hold.

We need only prove the triangle inequality for all the five cases where

dj(z1,z2)=d-+(z1,z2).

The rest of the cases will follow by symmetry in the definitions of d-+ and d+- and in swapping of the points z1,z2 with z3,z2. We have

dj(z1,z3)d-+(z1,z3)
=|t1+ϵj|+|t3-ϵj|+2ϵj+inf{dj(p1,p)+d(p,p3):pX}
|t1+ϵj|+|t2-ϵj|+2ϵj+|t3-t2|
+inf{dj(p1,p)+d(p,p2)+d(p2,p3):pX}
d-+(z1,z2)+d(z2,z3),
dj(z1,z3)d-+(z1,z3)
=inf{d-(z1,z)+d+(z,z3):zZ}
inf{d-(z1,z)+d+(z,z2)+d+(z2,z3):zZj}
=d-+(z1,z2)+d+(z2,z3).

Below we will use the following inequality, which follows from (A.7),

dj(p1,p2)inf{dj(p1,p)+d(p,p2):pX}+ϵj.

So

dj(z1,z3)d-(z1,z3)
=|t1+ϵj|+|t3+ϵj|+dj(p1,p3)
|t1+ϵj|+|t3+ϵj|+dj(p1,p2)+dj(p2,p3)
|t1+ϵj|+|t3+ϵj|+inf{dj(p1,p)+d(p,p2):pX}
+ϵj+dj(p2,p3)
d-+(z1,z2)+d-(z2,z3)

and

dj(z1,z3)d-+(z1,z3)
=|t1+ϵj|+|t3-ϵj|+2ϵj+inf{dj(p1,p)+d(p,p3):pX}
|t1+ϵj|+|t3-ϵj|+2ϵj+inf{dj(p1,p)+dj(p,p2):pX}
+inf{dj(p2,p)+d(p,p3):pX}
|t1+ϵj|+|t3-ϵj|+3ϵj+inf{dj(p1,p)+d(p,p2):pX}
+inf{dj(p2,p)+d(p,p3):pX}
d-+(z1,z2)+d-+(z2,z3),
dj(z1,z3)d-(z1,z3)
=|t1+ϵj|+|t3+ϵj|+dj(p1,p3)
|t1+ϵj|+|t3+ϵj|+dj(p1,p2)+dj(p2,p3)
|t1+ϵj|+|t3+ϵj|+2ϵj+inf{dj(p1,p)+d(p,p2):pX}
+inf{dj(p2,p)+d(p,p3):pX}
d-+(z1,z2)+d+-(z2,z3).

Thus dj is a metric. ∎

The metric space Zj constructed in Lemma A.3 is not necessarily complete even if X is complete with respect to both dj and d:

Example A.4.

Let X={0,12,14,}{1}. Let

dj(p1,p2)=|p1-p2|.

Let F:XX be the identity map on X{0,1} and F(0)=1 and F(1)=0. Let

d(p1,p2)=|F(p1)-F(p2)|.

Both (X,dj) and (X,d) are complete but with different limits for the sequence {12,14,}:

dj(1i,0)0andd(1i,1)0as i.

Observe that ϵj=1 because

1ϵjlimi|dj(1i,1)-d(1i,1)|=1.

So

Zj=[-1,1]×X.

Take the sequence of points zi=(0,1i). This sequence is Cauchy in Zj because

dj(zi,zk)d(zi,zk)=0+|1i-1k|for all i,k>1.

Assume on the contrary that this sequence of points converges to a point

z=(t,p)Zj.

Observe that for any zZj,

d+(zi,z)|0-1|+|t-1|1,
d-(zi,z)|0+1|+|t+1|1,
d-+(zi,z)=inf{d-(zi,z)+d+(z,z):zZj}1,
d+-(zi,z)=inf{d+(zi,z)+d-(z,z):zZj}1.

Therefore, for i sufficiently large,

dj(zi,z)=d(zi,z)=|0-t|+max{dj(1i,p),d(1i,p)}0.

Thus p is the limit of the sequence {1i} with respect to both metrics dj,d, which is a contradiction. Thus Zj is not complete.

The metric completion of Zj is

Z¯j=[-1,1]×(X{p})|,

where (-1,p)(-1,0) and (1,p)(1,1). For ti[-1,1] and piX we have

dj((t1,p1),(t2,p2))= as in Lemma A.3,
dj((t1,p1),(t2,p))=limkdj((t1,p1),(t2,1k)),
dj((t1,p),(t2,p))=limkdj((t1,1k),(t2,1k)).

Note that with this distance

dj((-1,0),(-1,p))=limkdj((-1,0),(-1,1k))
=limkdj(0,1k)=dj(0,0)=0,
dj((1,1),(1,p))=limkdj((1,1),(1,1k))
=limkd(1,1k)=dj(0,1k)=0,

and that is why (-1,p)(-1,0) and (1,p)(1,1).

Acknowledgements

The three authors appreciate the Mathematical Sciences Research Institute for the wonderful research environment there and the opportunity to begin working together on this project. We are grateful to Jim Isenberg, Yvonne Choquet-Bruhat, Piotr Chrusciel, Greg Galloway, Gerhard Huisken, Sergiu Klainerman, Igor Rodnianski, and Richard Schoen for their organization of the program in General Relativity at MSRI in Fall 2013. We also thank the referee for careful reading and for helpful comments.

References

[1] L. Ambrosio and B. Kirchheim, Currents in metric spaces, Acta Math. 185 (2000), no. 1, 1–80. 10.1007/BF02392711Search in Google Scholar

[2] H. Bray and F. Finster, Curvature estimates and the positive mass theorem, Comm. Anal. Geom. 10 (2002), no. 2, 291–306. 10.4310/CAG.2002.v10.n2.a3Search in Google Scholar

[3] J. Corvino, A note on asymptotically flat metrics on 3 which are scalar-flat and admit minimal spheres, Proc. Amer. Math. Soc. 133 (2005), no. 12, 3669–3678, (electronic). 10.1090/S0002-9939-05-07926-8Search in Google Scholar

[4] H. Federer and W. H. Fleming, Normal and integral currents, Ann. of Math. (2) 72 (1960), 458–520. 10.2307/1970227Search in Google Scholar

[5] F. Finster, A level set analysis of the Witten spinor with applications to curvature estimates, Math. Res. Lett. 16 (2009), no. 1, 41–55. 10.4310/MRL.2009.v16.n1.a5Search in Google Scholar

[6] F. Finster and I. Kath, Curvature estimates in asymptotically flat manifolds of positive scalar curvature, Comm. Anal. Geom. 10 (2002), no. 5, 1017–1031. 10.4310/CAG.2002.v10.n5.a6Search in Google Scholar

[7] M. Gromov, Groups of polynomial growth and expanding maps, Publ. Math. Inst. Hautes Études Sci. 53 (1981), 53–73. 10.1007/BF02698687Search in Google Scholar

[8] M. Gromov, Metric structures for Riemannian and non-Riemannian spaces, Progr. Math. 152, Birkhäuser, Boston 1981. Search in Google Scholar

[9] K. Grove and P. Petersen, Manifolds near the boundary of existence, J. Differential Geom. 33 (1991), no. 2, 379–394. 10.4310/jdg/1214446323Search in Google Scholar

[10] L.-H. Huang and D. A. Lee, Stability of the positive mass theorem for graphical hypersurfaces of Euclidean space, Comm. Math. Phys. 337 (2015), no. 1, 151–169. 10.1007/s00220-014-2265-9Search in Google Scholar

[11] L.-H. Huang and D. Wu, Hypersurfaces with nonnegative scalar curvature, J. Differential Geom. 95 (2013), no. 2, 249–278. 10.4310/jdg/1376053447Search in Google Scholar

[12] L.-H. Huang and D. Wu, The equality case of the penrose inequality for asymptotically flat graphs, Trans. Amer. Math. Soc. 367 (2015), 31–47. 10.1090/S0002-9947-2014-06090-XSearch in Google Scholar

[13] M.-K. G. Lam, The graph cases of the Riemannian positive mass and Penrose inequalities in all dimensions, Ph.D. thesis, Duke University; ProQuest LLC, Ann Arbor 2011, http://arxiv.org/abs/1010.4256. Search in Google Scholar

[14] D. A. Lee, On the near-equality case of the positive mass theorem, Duke Math. J. 148 (2009), no. 1, 63–80. 10.1215/00127094-2009-021Search in Google Scholar

[15] D. A. Lee and C. Sormani, Stability of the positive mass theorem for rotationally symmetric Riemannian manifolds, J. reine angew. Math. 686 (2014), 187–220. 10.1515/crelle-2012-0094Search in Google Scholar

[16] P. LeFloch and C. Sormani, Nonlinear stability of rotationally symmetric spaces with low regularity, J. Funct. Anal. 268 (2015), no. 7, 2005–2065. 10.1016/j.jfa.2014.12.012Search in Google Scholar

[17] R. C. Reilly, Variational properties of functions of the mean curvatures for hypersurfaces in space forms, J. Differential Geometry 8 (1973), 465–477. 10.4310/jdg/1214431802Search in Google Scholar

[18] R. Schoen and S. T. Yau, On the proof of the positive mass conjecture in general relativity, Comm. Math. Phys. 65 (1979), no. 1, 45–76. 10.1007/BF01940959Search in Google Scholar

[19] C. Sormani, Friedmann cosmology and almost isotropy, Geom. Funct. Anal. 14 (2004), no. 4, 853–912. 10.1007/s00039-004-0477-4Search in Google Scholar

[20] C. Sormani, Intrinsic flat Arzela–Ascoli theorems, preprint (2014), http://arxiv.org/abs/1402.6066. 10.4310/CAG.2018.v26.n6.a3Search in Google Scholar

[21] C. Sormani and J. Portegies, Properties of the intrinsic flat distance, preprint (2015), http://arxiv.org/abs/1210.3895v4. Search in Google Scholar

[22] C. Sormani and S. Wenger, The intrinsic flat distance between Riemannian manifolds and other integral current spaces, J. Differential Geom. 87 (2011), no. 1, 117–199. 10.4310/jdg/1303219774Search in Google Scholar

[23] S. Wenger, Compactness for manifolds and integral currents with bounded diameter and volume, Calc. Var. Partial Differential Equations 40 (2011), no. 3–4, 423–448. 10.1007/s00526-010-0346-ySearch in Google Scholar

[24] E. Witten, A new proof of the positive energy theorem, Comm. Math. Phys. 80 (1981), no. 3, 381–402. 10.1007/BF01208277Search in Google Scholar

Received: 2015-1-12
Revised: 2015-5-23
Published Online: 2015-10-14
Published in Print: 2017-6-1

© 2017 Walter de Gruyter GmbH, Berlin/Boston

Downloaded on 2.5.2024 from https://www.degruyter.com/document/doi/10.1515/crelle-2015-0051/html
Scroll to top button