algebraic and differential geometry. computational...

69
Symmetric-like manifolds Applications in Inverse Spectral Geometry Modifying the classical setting: The Steklov spectrum Algebraic and Differential Geometry. Computational Algebra. Teresa Arias-Marco Department of Mathematics University of Extremadura, Badajoz, Spain January 14th, 2016 First Joint Meeting ´ Evora-Extremadura on Mathematics Teresa Arias-Marco GADAC

Upload: others

Post on 24-Jul-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Algebraic and Differential Geometry.Computational Algebra.

Teresa Arias-Marco

Department of MathematicsUniversity of Extremadura, Badajoz, Spain

January 14th, 2016

First Joint Meeting Evora-Extremaduraon Mathematics

Teresa Arias-Marco GADAC

Page 2: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The research group GADACweb page http://www.unex.es/investigacion/grupos/gadacGroup Coordinator IGNACIO OJEDAResearch Lines of IGNACIO OJEDA

Semigroup algebras. Toric geometry.Classification of projective curves.

Group Members and Research lines

Amelia Alvarez – Schemes algebras and their representations.Teresa Arias-Marco–Homogeneous spaces with additionalgeometric structures.Adrian Gordillo–Natural tensors, differential invariants andRiemannian metrics.Jose Navarro–Natural tensors, differential invariants andRiemannian metrics.Juan A. Navarro GonzalezPedro Sancho de Salas – Schemes algebras and their rep.David Sevilla–Algorithms for algebraic varieties.

Teresa Arias-Marco GADAC

Page 3: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The research group GADACweb page http://www.unex.es/investigacion/grupos/gadacGroup Coordinator IGNACIO OJEDAResearch Lines of IGNACIO OJEDA

Semigroup algebras. Toric geometry.Classification of projective curves.

Group Members and Research lines

Amelia Alvarez – Schemes algebras and their representations.Teresa Arias-Marco–Homogeneous spaces with additionalgeometric structures.Adrian Gordillo–Natural tensors, differential invariants andRiemannian metrics.Jose Navarro–Natural tensors, differential invariants andRiemannian metrics.Juan A. Navarro GonzalezPedro Sancho de Salas – Schemes algebras and their rep.David Sevilla–Algorithms for algebraic varieties.

Teresa Arias-Marco GADAC

Page 4: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The research group GADACweb page http://www.unex.es/investigacion/grupos/gadacGroup Coordinator IGNACIO OJEDAResearch Lines of IGNACIO OJEDA

Semigroup algebras. Toric geometry.Classification of projective curves.

Group Members and Research lines

Amelia Alvarez – Schemes algebras and their representations.Teresa Arias-Marco–Homogeneous spaces with additionalgeometric structures.Adrian Gordillo–Natural tensors, differential invariants andRiemannian metrics.Jose Navarro–Natural tensors, differential invariants andRiemannian metrics.Juan A. Navarro GonzalezPedro Sancho de Salas – Schemes algebras and their rep.David Sevilla–Algorithms for algebraic varieties.

Teresa Arias-Marco GADAC

Page 5: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The research group GADACweb page http://www.unex.es/investigacion/grupos/gadacGroup Coordinator IGNACIO OJEDAResearch Lines of IGNACIO OJEDA

Semigroup algebras. Toric geometry.Classification of projective curves.

Group Members and Research lines

Amelia Alvarez – Schemes algebras and their representations.Teresa Arias-Marco–Homogeneous spaces with additionalgeometric structures.Adrian Gordillo–Natural tensors, differential invariants andRiemannian metrics.Jose Navarro–Natural tensors, differential invariants andRiemannian metrics.Juan A. Navarro GonzalezPedro Sancho de Salas – Schemes algebras and their rep.David Sevilla–Algorithms for algebraic varieties.

Teresa Arias-Marco GADAC

Page 6: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The research group GADACweb page http://www.unex.es/investigacion/grupos/gadacGroup Coordinator IGNACIO OJEDAResearch Lines of IGNACIO OJEDA

Semigroup algebras. Toric geometry.Classification of projective curves.

Group Members and Research lines

Amelia Alvarez – Schemes algebras and their representations.Teresa Arias-Marco–Homogeneous spaces with additionalgeometric structures.Adrian Gordillo–Natural tensors, differential invariants andRiemannian metrics.Jose Navarro–Natural tensors, differential invariants andRiemannian metrics.Juan A. Navarro GonzalezPedro Sancho de Salas – Schemes algebras and their rep.David Sevilla–Algorithms for algebraic varieties.

Teresa Arias-Marco GADAC

Page 7: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The research group GADACweb page http://www.unex.es/investigacion/grupos/gadacGroup Coordinator IGNACIO OJEDAResearch Lines of IGNACIO OJEDA

Semigroup algebras. Toric geometry.Classification of projective curves.

Group Members and Research lines

Amelia Alvarez – Schemes algebras and their representations.Teresa Arias-Marco–Homogeneous spaces with additionalgeometric structures.Adrian Gordillo–Natural tensors, differential invariants andRiemannian metrics.Jose Navarro–Natural tensors, differential invariants andRiemannian metrics.Juan A. Navarro GonzalezPedro Sancho de Salas – Schemes algebras and their rep.David Sevilla–Algorithms for algebraic varieties.

Teresa Arias-Marco GADAC

Page 8: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The research group GADACweb page http://www.unex.es/investigacion/grupos/gadacGroup Coordinator IGNACIO OJEDAResearch Lines of IGNACIO OJEDA

Semigroup algebras. Toric geometry.Classification of projective curves.

Group Members and Research lines

Amelia Alvarez – Schemes algebras and their representations.Teresa Arias-Marco–Homogeneous spaces withadditional geometric structures.Adrian Gordillo–Natural tensors, differential invariants andRiemannian metrics.Jose Navarro–Natural tensors, differential invariants andRiemannian metrics.Juan A. Navarro GonzalezPedro Sancho de Salas – Schemes algebras and their rep.David Sevilla–Algorithms for algebraic varieties.

Teresa Arias-Marco GADAC

Page 9: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Homogeneous spaces with additional geometric structures

T. Arias-Marco, D. Schueth.On inaudible curvature properties of closed Riemannianmanifolds.Ann. Glob. Anal. Geom., (2010), 37:339-349.

T. Arias-Marco, D. Schueth.Local symmetry of harmonic spaces as determined by thespectra of small geodesic spheres.Geom. Funct. Anal. (GAFA), (2012), 22:1-21.

1 Symmetric-like manifolds

2 Applications in Inverse Spectral GeometryThe settingOur Results and open problems

3 Modifying the classical setting: The Steklov spectrum

Teresa Arias-Marco GADAC

Page 10: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Homogeneous spaces with additional geometric structures

T. Arias-Marco, D. Schueth.On inaudible curvature properties of closed Riemannianmanifolds.Ann. Glob. Anal. Geom., (2010), 37:339-349.

T. Arias-Marco, D. Schueth.Local symmetry of harmonic spaces as determined by thespectra of small geodesic spheres.Geom. Funct. Anal. (GAFA), (2012), 22:1-21.

1 Symmetric-like manifolds

2 Applications in Inverse Spectral GeometryThe settingOur Results and open problems

3 Modifying the classical setting: The Steklov spectrum

Teresa Arias-Marco GADAC

Page 11: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Homogeneous spaces with additional geometric structures

T. Arias-Marco, D. Schueth.On inaudible curvature properties of closed Riemannianmanifolds.Ann. Glob. Anal. Geom., (2010), 37:339-349.

T. Arias-Marco, D. Schueth.Local symmetry of harmonic spaces as determined by thespectra of small geodesic spheres.Geom. Funct. Anal. (GAFA), (2012), 22:1-21.

1 Symmetric-like manifolds

2 Applications in Inverse Spectral GeometryThe settingOur Results and open problems

3 Modifying the classical setting: The Steklov spectrum

Teresa Arias-Marco GADAC

Page 12: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Einstein manifolds and Local homogeneity

Definition

(M, g) is an Einstein manifold ifric = C · g for a certain constant Cat each point of M.

Definition.(M, g) is an homogeneous

Riemannian manifold if I(M)acts transitively on M.

Definition

(M, g) is locally homogeneous if thepseudogroup of local isometries actstransitively on M.

Teresa Arias-Marco GADAC

Page 13: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Einstein manifolds and Local homogeneity

Definition

(M, g) is an Einstein manifold ifric = C · g for a certain constant Cat each point of M.

Definition.(M, g) is an homogeneous

Riemannian manifold if I(M)acts transitively on M.

Definition

(M, g) is locally homogeneous if thepseudogroup of local isometries actstransitively on M.

Teresa Arias-Marco GADAC

Page 14: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Spaces of type A

Definition (A. Gray, 1978)

(M, g) is of type A (TAm) ifric is cyclic-parallel; i.e.(∇X ric)(X ,X ) = 0 for all X ∈ TM.

Szabo 1993. TAm are analytic.

D’Atri, Nickerson 1969. TAmhave constant scalar curvature.

Kowalski 1983, Pedersen, Tod1999. TAm up to dimension 3are homogeneous and classified.

-, Kowalski 2008. Classifiedhomogeneous TAm of dim 4.

Pedersen, Tod 1999. Indimensions ≥ 5, there existnon-locally homogeneous TAm.

Tod 1999. In dim 4, TAmcondition doesn’t imply localhomogeneity.

Teresa Arias-Marco GADAC

Page 15: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Spaces of type A

Definition (A. Gray, 1978)

(M, g) is of type A (TAm) ifric is cyclic-parallel; i.e.(∇X ric)(X ,X ) = 0 for all X ∈ TM.

Szabo 1993. TAm are analytic.

D’Atri, Nickerson 1969. TAmhave constant scalar curvature.

Kowalski 1983, Pedersen, Tod1999. TAm up to dimension 3are homogeneous and classified.

-, Kowalski 2008. Classifiedhomogeneous TAm of dim 4.

Pedersen, Tod 1999. Indimensions ≥ 5, there existnon-locally homogeneous TAm.

Tod 1999. In dim 4, TAmcondition doesn’t imply localhomogeneity.

Teresa Arias-Marco GADAC

Page 16: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Spaces of type A

Definition (A. Gray, 1978)

(M, g) is of type A (TAm) ifric is cyclic-parallel; i.e.(∇X ric)(X ,X ) = 0 for all X ∈ TM.

Szabo 1993. TAm are analytic.

D’Atri, Nickerson 1969. TAmhave constant scalar curvature.

Kowalski 1983, Pedersen, Tod1999. TAm up to dimension 3are homogeneous and classified.

-, Kowalski 2008. Classifiedhomogeneous TAm of dim 4.

Pedersen, Tod 1999. Indimensions ≥ 5, there existnon-locally homogeneous TAm.

Tod 1999. In dim 4, TAmcondition doesn’t imply localhomogeneity.

Teresa Arias-Marco GADAC

Page 17: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Symmetric spaces

Definition. sp at p ∈ M is a local geodesic symmetry if for allX ∈ N(p) in TpM, sp(expp(X )) = expp(−X ).

Definition.

(M, g) is locally symmetric if for allp ∈ M, sp is an isometry.

≡ ∇R = 0.

Definition

(M, g) is a Riemannian symmetricspace if for all p ∈ M, sp is globallydefined and it is an isometry.

⇒ Riemannian symmetric spaces are

spaces with metric - preserving

geodesic symmetries.

Teresa Arias-Marco GADAC

Page 18: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Symmetric spaces

Definition. sp at p ∈ M is a local geodesic symmetry if for allX ∈ N(p) in TpM, sp(expp(X )) = expp(−X ).

Definition.

(M, g) is locally symmetric if for allp ∈ M, sp is an isometry.

≡ ∇R = 0.

Definition

(M, g) is a Riemannian symmetricspace if for all p ∈ M, sp is globallydefined and it is an isometry.

⇒ Riemannian symmetric spaces are

spaces with metric - preserving

geodesic symmetries.

Teresa Arias-Marco GADAC

Page 19: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Harmonic Spaces

Definition

(M, g) is an harmonic space iff

the volume density function ofthe geodesic exponential map isradial around each point.

l

at each m ∈ M there exists anormal neighborhood of M onwhich ∆u = 0 admits a realsolution depending only uponthe distance r to m and beinganalytic for r 6= 0.

Besse, 1978: Every Harmonicmanifold is Einstein.

Open problem: Are harmonicspaces always locallyhomogeneous?

Teresa Arias-Marco GADAC

Page 20: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Harmonic Spaces

Definition

(M, g) is an harmonic space iff

the volume density function ofthe geodesic exponential map isradial around each point.

l

at each m ∈ M there exists anormal neighborhood of M onwhich ∆u = 0 admits a realsolution depending only uponthe distance r to m and beinganalytic for r 6= 0.

Besse, 1978: Every Harmonicmanifold is Einstein.

Open problem: Are harmonicspaces always locallyhomogeneous?

Teresa Arias-Marco GADAC

Page 21: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Weakly symmetric spaces

Definition (Selberg, 56, Szabo, 93)

(M, g) is called weakly symmetric ifeach p ∈ M and each nontrivial γstarting in p there exists an isometryof M which fixes p and reverses γ.

Berndt,Vanhecke,96. Weaklysym.examples non-symmetric.

Classifications3, 4-dim by Berndt,Vanhecke,96.5-dim by Kowalski,Marinosci,97.

Weakly symm. spaces are:Homogeneous by Selberg,56.Type A by Kowalski, Prufer,89.

Remark.

If a Riemannian covering of(M, g) is weakly symmetric

⇓(M, g) is Type A and

locally weakly symmetric.

Teresa Arias-Marco GADAC

Page 22: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Weakly symmetric spaces

Definition (Selberg, 56, Szabo, 93)

(M, g) is called weakly symmetric ifeach p ∈ M and each nontrivial γstarting in p there exists an isometryof M which fixes p and reverses γ.

Berndt,Vanhecke,96. Weaklysym.examples non-symmetric.

Classifications3, 4-dim by Berndt,Vanhecke,96.5-dim by Kowalski,Marinosci,97.

Weakly symm. spaces are:Homogeneous by Selberg,56.Type A by Kowalski, Prufer,89.

Remark.

If a Riemannian covering of(M, g) is weakly symmetric

⇓(M, g) is Type A and

locally weakly symmetric.

Teresa Arias-Marco GADAC

Page 23: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Weakly symmetric spaces

Definition (Selberg, 56, Szabo, 93)

(M, g) is called weakly symmetric ifeach p ∈ M and each nontrivial γstarting in p there exists an isometryof M which fixes p and reverses γ.

Berndt,Vanhecke,96. Weaklysym.examples non-symmetric.

Classifications3, 4-dim by Berndt,Vanhecke,96.5-dim by Kowalski,Marinosci,97.

Weakly symm. spaces are:Homogeneous by Selberg,56.Type A by Kowalski, Prufer,89.

Remark.

If a Riemannian covering of(M, g) is weakly symmetric

⇓(M, g) is Type A and

locally weakly symmetric.

Teresa Arias-Marco GADAC

Page 24: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Weakly locally symmetric spaces

Definition

(M, g) is weakly locally symmetric if∀p ∈ M, ∃ε > 0 such that ∀γ in M,γ(0) = p, ∃ an isometry of Bε(p)which fixes p and reverses γ|(−ε,ε).

(M, g) is locally symmetric

⇓(M, g) is Weakly locally sym.

Weakly symmetric examples byBerndt,Vanhecke,96. are alsonon locally symmetric.

Lemma.Let (M, g) be complete, simplyconnected and, weakly locallysymmetric V M is weakly sym.

Complete Weakly locallysymmetric spaces are spacesof type A.

Teresa Arias-Marco GADAC

Page 25: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Weakly locally symmetric spaces

Definition

(M, g) is weakly locally symmetric if∀p ∈ M, ∃ε > 0 such that ∀γ in M,γ(0) = p, ∃ an isometry of Bε(p)which fixes p and reverses γ|(−ε,ε).

(M, g) is locally symmetric

⇓(M, g) is Weakly locally sym.

Weakly symmetric examples byBerndt,Vanhecke,96. are alsonon locally symmetric.

Lemma.Let (M, g) be complete, simplyconnected and, weakly locallysymmetric V M is weakly sym.

Complete Weakly locallysymmetric spaces are spacesof type A.

Teresa Arias-Marco GADAC

Page 26: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Weakly locally symmetric spaces

Definition

(M, g) is weakly locally symmetric if∀p ∈ M, ∃ε > 0 such that ∀γ in M,γ(0) = p, ∃ an isometry of Bε(p)which fixes p and reverses γ|(−ε,ε).

(M, g) is locally symmetric

⇓(M, g) is Weakly locally sym.

Weakly symmetric examples byBerndt,Vanhecke,96. are alsonon locally symmetric.

Lemma.Let (M, g) be complete, simplyconnected and, weakly locallysymmetric V M is weakly sym.

Complete Weakly locallysymmetric spaces are spacesof type A.

Teresa Arias-Marco GADAC

Page 27: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Others symmetric-like manifolds

Definition (D’Atri,Nickerson,69,74)

(M, g) is called D’Atri space if thelocal geodesic symmetries areRiemannian volume-preserving.

On complete Riemannianmanifolds

Weak local symmetry⇓

D’Atri property,

the C property,

probabilistic commutativity,

the TC property,

the GC property,

⇓The type A property.

AM

Teresa Arias-Marco GADAC

Page 28: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Others symmetric-like manifolds

Definition (Berndt,Vanhecke,92)

(M, g) is called C-space if for each γthe eigenvalues of the associatedJacobi operator are constant along γ.

On complete Riemannianmanifolds

Weak local symmetry⇓

D’Atri property,

the C property,

probabilistic commutativity,

the TC property,

the GC property,

⇓The type A property.

AM

Teresa Arias-Marco GADAC

Page 29: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Others symmetric-like manifolds

Definition (Kowalski, Prufer, 82, 89)

(M, g) is probabilistic commutativeif all Euclidean Laplacian ∆(k),k ∈ N, commute.

On complete Riemannianmanifolds

Weak local symmetry⇓

D’Atri property,

the C property,

probabilistic commutativity,

the TC property,

the GC property,

⇓The type A property.

AM

Teresa Arias-Marco GADAC

Page 30: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Others symmetric-like manifolds

Definition (Berndt, Vanhecke, 1993)

(M, g) is a TC-space if ∀m ∈ M and∀p ∈ M suff. close to m, Tp(m) andTsm(p)(m) have the same eigenvalues

On complete Riemannianmanifolds

Weak local symmetry⇓

D’Atri property,

the C property,

probabilistic commutativity,

the TC property,

the GC property,

⇓The type A property.

AM Teresa Arias-Marco GADAC

Page 31: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Others symmetric-like manifolds

Definition

(M, g) is a GC-space if ∀m ∈ M and∀p ∈ M suff. close to m, Tm(p) ands−1m∗ ◦Tm(sm(p)) ◦ sm∗ have the sameeigenvalues

On complete Riemannianmanifolds

Weak local symmetry⇓

D’Atri property,

the C property,

probabilistic commutativity,

the TC property,

the GC property,

⇓The type A property.

AM Teresa Arias-Marco GADAC

Page 32: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Open problems on symmetric-like manifolds

Teresa Arias-Marco GADAC

Page 33: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

1 Symmetric-like manifolds

2 Applications in Inverse Spectral GeometryThe settingOur Results and open problems

3 Modifying the classical setting: The Steklov spectrum

Teresa Arias-Marco GADAC

Page 34: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Eigenvalues problems for the Laplace-Beltrami operator

Let (M, g) be a compact, connected, n-dimensional smoothRiemannian manifold, possibly with boundary.

Let ∆ = −div grad be the Laplace operator associated with g ,acting on functions.

Eigenvalues Problems: Find all real numbers λ for which thereexists a nontrivial solution f ∈ C2(M) to ∆f + λf = 0 when

∂M = ∅. Namely, closed eigenvalue problem.

f = 0 on ∂M. Namely, Dirichlet eigenvalue problem.

∂ν f = 0 on ∂M where ν is the inward unit normal vector fieldon ∂M. Namely, Neumann eigenvalue problem.

Throughout our work, (M, g) will be a closed manifold.

Teresa Arias-Marco GADAC

Page 35: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Eigenvalues problems for the Laplace-Beltrami operator

Let (M, g) be a compact, connected, n-dimensional smoothRiemannian manifold, possibly with boundary.

Let ∆ = −div grad be the Laplace operator associated with g ,acting on functions.

Eigenvalues Problems: Find all real numbers λ for which thereexists a nontrivial solution f ∈ C2(M) to ∆f + λf = 0 when

∂M = ∅. Namely, closed eigenvalue problem.

f = 0 on ∂M. Namely, Dirichlet eigenvalue problem.

∂ν f = 0 on ∂M where ν is the inward unit normal vector fieldon ∂M. Namely, Neumann eigenvalue problem.

Throughout our work, (M, g) will be a closed manifold.

Teresa Arias-Marco GADAC

Page 36: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Eigenvalues problems for the Laplace-Beltrami operator

Let (M, g) be a compact, connected, n-dimensional smoothRiemannian manifold, possibly with boundary.

Let ∆ = −div grad be the Laplace operator associated with g ,acting on functions.

Eigenvalues Problems: Find all real numbers λ for which thereexists a nontrivial solution f ∈ C2(M) to ∆f + λf = 0 when

∂M = ∅. Namely, closed eigenvalue problem.

f = 0 on ∂M. Namely, Dirichlet eigenvalue problem.

∂ν f = 0 on ∂M where ν is the inward unit normal vector fieldon ∂M. Namely, Neumann eigenvalue problem.

Throughout our work, (M, g) will be a closed manifold.

Teresa Arias-Marco GADAC

Page 37: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The spectrum of ∆ and the geometry of M

Definition. The heat kernel K (t, x , y) is the fundamentalsolution of ( ∂∂t + ∆)f = 0. K (t, x , y) is analytic in t > 0 andC∞ in (x , y) ∈ M ×M.

The Sturm-Liouville decomposition:∃ a complete orthonormal basis {ϕ0, ϕ1, ....} of C∞(M), with ϕj

having eigenvalue λj of ∆ satisfying 0 ≤ λ0 ≤ λ1 ≤ λ2 ≤ · · · → ∞.Each eigenvalue has finite multiplicity. Finally,

K (t, x , y) =∞∑j=0

e−λj tϕj(x)ϕj(y)

with convergence absolute, and uniform, for each t > 0. Inparticular, ∫

MK (t, x , x)dV (x) =

∞∑j=0

e−λj t .

Teresa Arias-Marco GADAC

Page 38: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The spectrum of ∆ and the geometry of M

Definition. The heat kernel K (t, x , y) is the fundamentalsolution of ( ∂∂t + ∆)f = 0. K (t, x , y) is analytic in t > 0 andC∞ in (x , y) ∈ M ×M.

The Sturm-Liouville decomposition:∃ a complete orthonormal basis {ϕ0, ϕ1, ....} of C∞(M), with ϕj

having eigenvalue λj of ∆ satisfying 0 ≤ λ0 ≤ λ1 ≤ λ2 ≤ · · · → ∞.Each eigenvalue has finite multiplicity. Finally,

K (t, x , y) =∞∑j=0

e−λj tϕj(x)ϕj(y)

with convergence absolute, and uniform, for each t > 0. Inparticular, ∫

MK (t, x , x)dV (x) =

∞∑j=0

e−λj t .

Teresa Arias-Marco GADAC

Page 39: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The spectrum of ∆ and the geometry of M

Minakshisundaram-Pleijel Asymptotic Expansion, 1949:

K (t, x , x) v (4πt)−n2

∞∑k=0

uk(x)tk

where uk(x) are C∞(M) and polynomials in the components of thecurvature tensor R and its covariant derivatives. Moreover,

∞∑j=0

e−λj t v (4πt)−n2

∞∑k=0

aktk ,

where the coefficients ak are called classical heat invariants. Inparticular,

a0 =

∫Mω = Vol(M, g), a1 =

1

6

∫Mτω,

a2 =1

360

∫M

(2|R|2 − 2|ric|2 + 5τ2)ω, ...

where τ denote the scalar curvature.Teresa Arias-Marco GADAC

Page 40: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The spectrum of ∆ and the geometry of M

Minakshisundaram-Pleijel Asymptotic Expansion, 1949:

K (t, x , x) v (4πt)−n2

∞∑k=0

uk(x)tk

where uk(x) are C∞(M) and polynomials in the components of thecurvature tensor R and its covariant derivatives. Moreover,

∞∑j=0

e−λj t v (4πt)−n2

∞∑k=0

aktk ,

where the coefficients ak are called classical heat invariants. Inparticular,

a0 =

∫Mω = Vol(M, g), a1 =

1

6

∫Mτω,

a2 =1

360

∫M

(2|R|2 − 2|ric|2 + 5τ2)ω, ...

where τ denote the scalar curvature.Teresa Arias-Marco GADAC

Page 41: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The spectrum of ∆ and the geometry of M

V The spectrum of ∆ depends only on the Riemannianstructure of (M, g).

Direct and inverse problems on spectral theory.

Direct: To use geometrical properties of M to determineinformation about the eigenvalues and eigenfunctions of ∆.

Theorem (Lichnerowicz)

(M, g) closed and ric ≥ (n − 1)k > 0, for k ∈ N.Then, the first positive eigenvalue λ satisfies λ ≥ nk.

Inverse: one assumes knowledge about Spec(∆,M) and attemptsto determine information about the geometry of (M, g).

Teresa Arias-Marco GADAC

Page 42: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The spectrum of ∆ and the geometry of M

V The spectrum of ∆ depends only on the Riemannianstructure of (M, g).

Direct and inverse problems on spectral theory.

Direct: To use geometrical properties of M to determineinformation about the eigenvalues and eigenfunctions of ∆.

Theorem (Lichnerowicz)

(M, g) closed and ric ≥ (n − 1)k > 0, for k ∈ N.Then, the first positive eigenvalue λ satisfies λ ≥ nk.

Inverse: one assumes knowledge about Spec(∆,M) and attemptsto determine information about the geometry of (M, g).

Teresa Arias-Marco GADAC

Page 43: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The spectrum of ∆ and the geometry of M

V The spectrum of ∆ depends only on the Riemannianstructure of (M, g).

Direct and inverse problems on spectral theory.

Direct: To use geometrical properties of M to determineinformation about the eigenvalues and eigenfunctions of ∆.

Theorem (Lichnerowicz)

(M, g) closed and ric ≥ (n − 1)k > 0, for k ∈ N.Then, the first positive eigenvalue λ satisfies λ ≥ nk.

Inverse: one assumes knowledge about Spec(∆,M) and attemptsto determine information about the geometry of (M, g).

Teresa Arias-Marco GADAC

Page 44: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Inverse problems in spectral geometry

To which extent does the spectrum of ∆ on (M, g)determine its geometry?

Important fact:

If (M, g) is isometric to (M ′, g ′) V Spec(∆,M) = Spec(∆′,M ′).

Natural question:

Does Spec(∆,M) = Spec(∆′,M ′) imply M isometric to M ′?

Negative answer, 1964: Milnor’s example(a pair of 16-dim flat tori)

Famous question by M. Kac, 1966:

Can one hear the shape of a drum?

Partial positive result: Spec(∆,M) = Spec(∆′, a ball) implies M isa ball of the same radius so isometric to that ball.

Teresa Arias-Marco GADAC

Page 45: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Inverse problems in spectral geometry

To which extent does the spectrum of ∆ on (M, g)determine its geometry?

Important fact:

If (M, g) is isometric to (M ′, g ′) V Spec(∆,M) = Spec(∆′,M ′).

Natural question:

Does Spec(∆,M) = Spec(∆′,M ′) imply M isometric to M ′?

Negative answer, 1964: Milnor’s example(a pair of 16-dim flat tori)

Famous question by M. Kac, 1966:

Can one hear the shape of a drum?

Partial positive result: Spec(∆,M) = Spec(∆′, a ball) implies M isa ball of the same radius so isometric to that ball.

Teresa Arias-Marco GADAC

Page 46: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Inverse problems in spectral geometry

To which extent does the spectrum of ∆ on (M, g)determine its geometry?

Important fact:

If (M, g) is isometric to (M ′, g ′) V Spec(∆,M) = Spec(∆′,M ′).

Natural question:

Does Spec(∆,M) = Spec(∆′,M ′) imply M isometric to M ′?

Negative answer, 1964: Milnor’s example(a pair of 16-dim flat tori)

Famous question by M. Kac, 1966:

Can one hear the shape of a drum?

Partial positive result: Spec(∆,M) = Spec(∆′, a ball) implies M isa ball of the same radius so isometric to that ball.

Teresa Arias-Marco GADAC

Page 47: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Inverse problems in spectral geometry

To which extent does the spectrum of ∆ on (M, g)determine its geometry?

Important fact:

If (M, g) is isometric to (M ′, g ′) V Spec(∆,M) = Spec(∆′,M ′).

Natural question:

Does Spec(∆,M) = Spec(∆′,M ′) imply M isometric to M ′?

Negative answer, 1964: Milnor’s example(a pair of 16-dim flat tori)

Famous question by M. Kac, 1966:

Can one hear the shape of a drum?

Partial positive result: Spec(∆,M) = Spec(∆′, a ball) implies M isa ball of the same radius so isometric to that ball.

Teresa Arias-Marco GADAC

Page 48: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Basic definitions

Definition

Two compact Riemannian manifolds (M, g) and (M ′, g ′) are saidto be isospectral if Spec(∆,M) = Spec(∆′,M ′).

Definition

A geometric property of a compact Riemannian manifold M can beheard if it can be determined from Spec(∆,M).

Definition

A geometric property is inaudible, i.e. not determined by thespectrum, if there exist pairs of isospectral manifolds which differwith respect to this property.

Teresa Arias-Marco GADAC

Page 49: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Basic definitions

Definition

Two compact Riemannian manifolds (M, g) and (M ′, g ′) are saidto be isospectral if Spec(∆,M) = Spec(∆′,M ′).

Definition

A geometric property of a compact Riemannian manifold M can beheard if it can be determined from Spec(∆,M).

Definition

A geometric property is inaudible, i.e. not determined by thespectrum, if there exist pairs of isospectral manifolds which differwith respect to this property.

Teresa Arias-Marco GADAC

Page 50: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Some Positive and Negative Results

Some positive results:

Heat invariants can be heard:

a0 =

∫Mω = Vol(M, g), a1 =

1

6

∫Mτω,

a2 =1

360

∫M

(2|R|2 − 2|ric|2 + 5τ2)ω, ...

V Therefore, in dim 2, the total curvature and the Eulercharacteristic of the manifold can be heard.

Some negative results:

Gordon, Gornet, Schueth, Webb, Wilson, 1998:

The maximum of the scalar curvature τ is inaudible.

Schueth, 1999: ∫M |ric|

2 is inaudible.

Teresa Arias-Marco GADAC

Page 51: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Some Positive and Negative Results

Some positive results:

Heat invariants can be heard:

a0 =

∫Mω = Vol(M, g), a1 =

1

6

∫Mτω,

a2 =1

360

∫M

(2|R|2 − 2|ric|2 + 5τ2)ω, ...

V Therefore, in dim 2, the total curvature and the Eulercharacteristic of the manifold can be heard.

Some negative results:

Gordon, Gornet, Schueth, Webb, Wilson, 1998:

The maximum of the scalar curvature τ is inaudible.

Schueth, 1999: ∫M |ric|

2 is inaudible.

Teresa Arias-Marco GADAC

Page 52: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Some Positive and Negative Results

Some positive results:

Heat invariants can be heard:

a0 =

∫Mω = Vol(M, g), a1 =

1

6

∫Mτω,

a2 =1

360

∫M

(2|R|2 − 2|ric|2 + 5τ2)ω, ...

V Therefore, in dim 2, the total curvature and the Eulercharacteristic of the manifold can be heard.

Some negative results:

Gordon, Gornet, Schueth, Webb, Wilson, 1998:

The maximum of the scalar curvature τ is inaudible.

Schueth, 1999: ∫M |ric|

2 is inaudible.

Teresa Arias-Marco GADAC

Page 53: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Our Results

Main Theorem (Ann. Glob. Anal. Geom. 37(2010), 339–349.)

Each of the following properties is an inaudible property ofRiemannian manifolds:

Weak local symmetry,

D’Atri property,

the C property,

probabilistic commutativity,

the TC property,

the GC property,

the type A property.

Teresa Arias-Marco GADAC

Page 54: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Our Results

Main Theorem (Ann. Glob. Anal. Geom. 37(2010), 339–349.)

Each of the following properties is an inaudible property ofRiemannian manifolds:

Weak local symmetry,

D’Atri property,

the C property,

probabilistic commutativity,

the TC property,

the GC property,

the type A property.

Teresa Arias-Marco GADAC

Page 55: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Our Results

Main Theorem (Ann. Glob. Anal. Geom. 37(2010), 339–349.)

Each of the following properties is an inaudible property ofRiemannian manifolds:

Weak local symmetry,

D’Atri property,

the C property,

probabilistic commutativity,

the TC property,

the GC property,

the type A property.

Teresa Arias-Marco GADAC

Page 56: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

An interesting question

Problem by B.Y. Chen, L. Vanhecke in 1981:

To what extent do the properties of sufficiently small geodesicspheres determine the Riemannian geometry of the ambient space?

Tp(m):Shape op.(at m)of the geodesic sph. with center p, radius r

Characterization (Vanhecke, Willmore, 1983)

M is locally symmetric space iff ∀m ∈ M and ∀p ∈ Msufficiently close to m, Tp(m) = Tsm(p)(m).

Characterization (Ledger, Vanhecke, 1987)

M is locally symmetric space iff ∀m ∈ M and ∀p ∈ Msufficiently close to m, sm∗ ◦ Tm(p) = Tm(sm(p)) ◦ sm.

TC

↗↘TP

GC

↗↘GP

Teresa Arias-Marco GADAC

Page 57: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

An interesting question

Problem by B.Y. Chen, L. Vanhecke in 1981:

To what extent do the properties of sufficiently small geodesicspheres determine the Riemannian geometry of the ambient space?

Tp(m):Shape op.(at m)of the geodesic sph. with center p, radius r

Characterization (Vanhecke, Willmore, 1983)

M is locally symmetric space iff ∀m ∈ M and ∀p ∈ Msufficiently close to m, Tp(m) = Tsm(p)(m).

Characterization (Ledger, Vanhecke, 1987)

M is locally symmetric space iff ∀m ∈ M and ∀p ∈ Msufficiently close to m, sm∗ ◦ Tm(p) = Tm(sm(p)) ◦ sm.

TC

↗↘TP

GC

↗↘GP

Teresa Arias-Marco GADAC

Page 58: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Open problems

Question 1

Are the properties of being locally symmetric, Einstein or harmonicspectrally determined on closed manifold?

Question 2

To what extent do the spectra of small geodesic spheres in a(possibly noncompact) Riemannian manifold M determine thegeometry of M ?

Question 3

Is the property of being locally symmetric determined by thespectrum of sufficiently small geodesic spheres?

Teresa Arias-Marco GADAC

Page 59: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Open problems

Question 1

Are the properties of being locally symmetric, Einstein or harmonicspectrally determined on closed manifold?

Question 2

To what extent do the spectra of small geodesic spheres in a(possibly noncompact) Riemannian manifold M determine thegeometry of M ?

Question 3

Is the property of being locally symmetric determined by thespectrum of sufficiently small geodesic spheres?

Teresa Arias-Marco GADAC

Page 60: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Open problems

Question 1

Are the properties of being locally symmetric, Einstein or harmonicspectrally determined on closed manifold?

Question 2

To what extent do the spectra of small geodesic spheres in a(possibly noncompact) Riemannian manifold M determine thegeometry of M ?

Question 3

Is the property of being locally symmetric determined by thespectrum of sufficiently small geodesic spheres?

Teresa Arias-Marco GADAC

Page 61: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Our question and result on geodesic spheres

Our Question

Do the spectra of geodesic spheres distinguish symmetric(∇R = 0) harmonic spaces from nonsymmetric (∇R 6= 0)harmonic spaces?

Main Theorem (Geom. Funct. Anal. (GAFA) 22(2012), 1–21.)

Let M1 and M2 be harmonic spaces, and let p1 ∈ M1, p2 ∈ M2.If there exists ε > 0 such that for each r ∈ (0, ε) the geodesicspheres Sr (p1) and Sr (p2) are isospectral, then

|∇R|2p1 = |∇R|2p2

Consequently,

M1 is locally symmetric if and only if M2 is locally symmetric.

Teresa Arias-Marco GADAC

Page 62: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Our question and result on geodesic spheres

Our Question

Do the spectra of geodesic spheres distinguish symmetric(∇R = 0) harmonic spaces from nonsymmetric (∇R 6= 0)harmonic spaces?

Main Theorem (Geom. Funct. Anal. (GAFA) 22(2012), 1–21.)

Let M1 and M2 be harmonic spaces, and let p1 ∈ M1, p2 ∈ M2.If there exists ε > 0 such that for each r ∈ (0, ε) the geodesicspheres Sr (p1) and Sr (p2) are isospectral, then

|∇R|2p1 = |∇R|2p2

Consequently,

M1 is locally symmetric if and only if M2 is locally symmetric.

Teresa Arias-Marco GADAC

Page 63: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Our question and result on geodesic spheres

Our Question

Do the spectra of geodesic spheres distinguish symmetric(∇R = 0) harmonic spaces from nonsymmetric (∇R 6= 0)harmonic spaces?

Main Theorem (Geom. Funct. Anal. (GAFA) 22(2012), 1–21.)

Let M1 and M2 be harmonic spaces, and let p1 ∈ M1, p2 ∈ M2.If there exists ε > 0 such that for each r ∈ (0, ε) the geodesicspheres Sr (p1) and Sr (p2) are isospectral, then

|∇R|2p1 = |∇R|2p2

Consequently,

M1 is locally symmetric if and only if M2 is locally symmetric.

Teresa Arias-Marco GADAC

Page 64: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Our question and result on geodesic balls

Our Question

Do the spectra of geodesic balls distinguish symmetric (∇R = 0)harmonic spaces from nonsymmetric (∇R 6= 0) harmonic spaces?

Main Theorem (Geom. Funct. Anal. (GAFA) 22(2012), 1–21.)

Let M1 and M2 be harmonic spaces, and let p1 ∈ M1, p2 ∈ M2.If there exists ε > 0 such that for each r ∈ (0, ε) the geodesic ballsBr (p1) and Br (p2) are Dirichlet or Neumann isospectral, then

|∇R|2p1 = |∇R|2p2

Consequently,

M1 is locally symmetric if and only if M2 is locally symmetric.

Teresa Arias-Marco GADAC

Page 65: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The Steklov problem

In 1895 V. A. Steklov posed the following problem:

Let M be a smooth manifold of dimension ≥ 2 with smoothboundary bd(M). Find functions u on M and scalars σ ∈ R thatsatisfy,

∆u = 0 in M

∂νu = σu on bd(M)

The Steklov spectrum of M is set of all such σ.

Teresa Arias-Marco GADAC

Page 66: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The Steklov problem

In 1895 V. A. Steklov posed the following problem:

Let M be a smooth manifold of dimension ≥ 2 with smoothboundary bd(M). Find functions u on M and scalars σ ∈ R thatsatisfy,

∆u = 0 in M

∂νu = σu on bd(M)

The Steklov spectrum of M is set of all such σ.

Teresa Arias-Marco GADAC

Page 67: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

The Steklov problem, rephrased

The Steklov spectrum is the spectrum of the“Dirichlet-to-Neumann” operator,

D : C∞(bd(M))→ C∞(bd(M)),

defined as follows.

Take u ∈ C∞(bd(M)).

Let u be the harmonic extension of u to M.

D(u) = (∂ν u)|bd(M).

Operator D is an elliptic pseudodifferential operator of order onewith the same principal symbol as

√∆bd(M).

Teresa Arias-Marco GADAC

Page 68: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Symmetric-like manifoldsApplications in Inverse Spectral Geometry

Modifying the classical setting: The Steklov spectrum

Applied context: Electrical impedance tomography

In electrical impedance tomography electrical stimuli are placedaround a body. Resulting surface voltages are recorded and used toinfer the structure of objects inside the body.

This applied (numerical) technique relies on use of theDirichlet-to-Neumann operator.

Teresa Arias-Marco GADAC

Page 69: Algebraic and Differential Geometry. Computational …matematicas.unex.es/wp-content/uploads/2016/01/1EvoraExt...2016/01/01  · Algebraic and Di erential Geometry. Computational Algebra

Algebraic and Differential Geometry.Computational Algebra.

Teresa Arias-Marco

Department of MathematicsUniversity of Extremadura, Badajoz, Spain

January 14th, 2016

First Joint Meeting Evora-Extremaduraon Mathematics

Teresa Arias-Marco GADAC