Compact Riemannian Manifolds with Homogeneous Geodesics
A homogeneous Riemannian space (M = G/H,g) is called a geodesic orbit space (shortly, GO-space) if any geodesic is an orbit of one-parameter subgroup of the isometry group G. We study the structure of compact GO-spaces and give some sufficient conditions for existence and non-existence of an invaria...
Збережено в:
| Опубліковано в: : | Symmetry, Integrability and Geometry: Methods and Applications |
|---|---|
| Дата: | 2009 |
| Автори: | , |
| Формат: | Стаття |
| Мова: | Англійська |
| Опубліковано: |
Інститут математики НАН України
2009
|
| Онлайн доступ: | https://nasplib.isofts.kiev.ua/handle/123456789/149121 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Цитувати: | Compact Riemannian Manifolds with Homogeneous Geodesics / D.V. Alekseevsky, Y.G. Nikonorov // Symmetry, Integrability and Geometry: Methods and Applications. — 2009. — Т. 5. — Бібліогр.: 28 назв. — англ. |
Репозитарії
Digital Library of Periodicals of National Academy of Sciences of Ukraine| _version_ | 1859652593168416768 |
|---|---|
| author | Alekseevsky, D.V. Nikonorov, Y.G. |
| author_facet | Alekseevsky, D.V. Nikonorov, Y.G. |
| citation_txt | Compact Riemannian Manifolds with Homogeneous Geodesics / D.V. Alekseevsky, Y.G. Nikonorov // Symmetry, Integrability and Geometry: Methods and Applications. — 2009. — Т. 5. — Бібліогр.: 28 назв. — англ. |
| collection | DSpace DC |
| container_title | Symmetry, Integrability and Geometry: Methods and Applications |
| description | A homogeneous Riemannian space (M = G/H,g) is called a geodesic orbit space (shortly, GO-space) if any geodesic is an orbit of one-parameter subgroup of the isometry group G. We study the structure of compact GO-spaces and give some sufficient conditions for existence and non-existence of an invariant metric g with homogeneous geodesics on a homogeneous space of a compact Lie group G. We give a classification of compact simply connected GO-spaces (M = G/H,g) of positive Euler characteristic. If the group G is simple and the metric g does not come from a bi-invariant metric of G, then M is one of the flag manifolds M₁ = SO(2n+1)/U(n) or M₂ = Sp(n)/U(1)·Sp(n–1) and g is any invariant metric on M which depends on two real parameters. In both cases, there exists unique (up to a scaling) symmetric metric g₀ such that (M,g0) is the symmetric space M = SO(2n+2)/U(n+1) or, respectively, CP²n⁻¹. The manifolds M₁, M₂ are weakly symmetric spaces.
|
| first_indexed | 2025-12-07T13:35:21Z |
| format | Article |
| fulltext |
Symmetry, Integrability and Geometry: Methods and Applications SIGMA 5 (2009), 093, 16 pages
Compact Riemannian Manifolds
with Homogeneous Geodesics?
Dmitrĭı V. ALEKSEEVSKY † and Yurĭı G. NIKONOROV ‡
† School of Mathematics and Maxwell Institute for Mathematical Studies, Edinburgh University,
Edinburgh EH9 3JZ, United Kingdom
E-mail: D.Aleksee@ed.ac.uk
‡ Volgodonsk Institute of Service (branch) of South Russian State University of Economics
and Service, 16 Mira Ave., Volgodonsk, Rostov region, 347386, Russia
E-mail: nikonorov2006@mail.ru
Received April 22, 2009, in final form September 20, 2009; Published online September 30, 2009
doi:10.3842/SIGMA.2009.093
Abstract. A homogeneous Riemannian space (M = G/H, g) is called a geodesic orbit space
(shortly, GO-space) if any geodesic is an orbit of one-parameter subgroup of the isometry
group G. We study the structure of compact GO-spaces and give some sufficient conditions
for existence and non-existence of an invariant metric g with homogeneous geodesics on
a homogeneous space of a compact Lie group G. We give a classification of compact simply
connected GO-spaces (M = G/H, g) of positive Euler characteristic. If the group G is
simple and the metric g does not come from a bi-invariant metric of G, then M is one of
the flag manifolds M1 = SO(2n + 1)/U(n) or M2 = Sp(n)/U(1) · Sp(n − 1) and g is any
invariant metric on M which depends on two real parameters. In both cases, there exists
unique (up to a scaling) symmetric metric g0 such that (M, g0) is the symmetric space
M = SO(2n + 2)/U(n + 1) or, respectively, CP 2n−1. The manifolds M1, M2 are weakly
symmetric spaces.
Key words: homogeneous spaces, weakly symmetric spaces, homogeneous spaces of posi-
tive Euler characteristic, geodesic orbit spaces, normal homogeneous Riemannian manifolds,
geodesics
2000 Mathematics Subject Classification: 53C20; 53C25; 53C35
1 Introduction
A Riemannian manifold (M, g) is called a manifold with homogeneous geodesics or geodesic
orbit manifold (shortly, GO-manifold) if all its geodesic are orbits of one-parameter groups of
isometries of (M, g). Such manifold is a homogeneous manifold and can be identified with
a coset space M = G/H of a transitive Lie group G of isometries. A Riemannian homogeneous
space (M = G/H, gM ) of a group G is called a space with homogeneous geodesics (or geodesic
orbit space, shortly, GO-space) if any geodesic is an orbit of a one-parameter subgroup of the
group G. This terminology was introduced by O. Kowalski and L. Vanhecke in [20], who initiated
a systematic study of such spaces.
Recall that homogeneous geodesics correspond to “relative equilibria” of the geodesic flow,
considered as a hamiltonian system on the cotangent bundle. Due to this, GO-manifolds can
be characterized as Riemannian manifolds such that all integral curves of the geodesic flow are
relative equilibria.
?This paper is a contribution to the Special Issue “Élie Cartan and Differential Geometry”. The full collection
is available at http://www.emis.de/journals/SIGMA/Cartan.html
mailto:D.Aleksee@ed.ac.uk
mailto:nikonorov2006@mail.ru
http://dx.doi.org/10.3842/SIGMA.2009.093
http://www.emis.de/journals/SIGMA/Cartan.html
2 D.V. Alekseevsky and Yu.G. Nikonorov
GO-spaces may be considered as a natural generalization of symmetric spaces, classified by
É. Cartan [10]. Indeed, a simply connected symmetric space can be defined as a Rieman-
nian manifold (M, g) such that any geodesic γ ⊂ M is an orbit of one-parameter group gt of
transvections, that is one-parameter group of isometries which preserves γ and induces the pa-
rallel transport along γ. If we remove the assumption that gt induces the parallel transport, we
get the notion of a GO-space.
The class of GO-spaces is much larger then the class of symmetric spaces. Any homogeneous
space M = G/H of a compact Lie group G admits a metric gM such that (M, gM ) is a GO-space.
It is sufficient to take the metric gM which is induced with a bi-invariant Riemannian metric g
on the Lie group G such that (G, g) → (M = G/H, gM ) is a Riemannian submersion with totally
geodesic fibres. Such GO-space (M = G/H, gM ) is called a normal homogeneous space.
More generally, any naturally reductive manifold is a geodesic orbit manifold. Recall that
a Riemannian manifold (M, gM ) is called naturally reductive if it admits a transitive Lie
group G of isometries with a bi-invariant pseudo-Riemannian metric g, which induces the met-
ric gM on M = G/H, see [18, 8]. The first example of non naturally reductive GO-manifold had
been constructed by A. Kaplan [16]. An important class of GO-spaces consists of weakly sym-
metric spaces, introduced by A. Selberg [22]. A homogeneous Riemannian space (M = G/H, gM )
is a weakly symmetric space if any two points p, q ∈ M can be interchanged by an isom-
etry a ∈ G. This property does not depend on the particular invariant metric gM . Weakly
symmetric spaces M = G/H have many interesting properties (for example, the algebra of
G-invariant differential operators on M is commutative, the representation of G in the space
L2(M) of function is multiplicity free, the algebra of G-invariant Hamiltonians on T ∗M with
respect to Poisson bracket is commutative) and are closely related with spherical spaces, com-
mutative spaces and Gelfand pairs etc., see the book by J.A. Wolf [26]. The classification of
weakly symmetric reductive homogeneous spaces was given by O.S. Yakimova [28], see also [26].
In [20], O. Kowalski and L. Vanhecke classified all GO-spaces of dimension ≤ 6. C. Gor-
don [14] reduced the classification of GO-spaces to the classification of GO-metrics on nilmani-
folds, compact GO-spaces and non-compact GO-spaces of non-compact semisimple Lie group.
She described GO-metrics on nilmanifolds. They exist only on two-step nilponent nilmanifolds.
She also presented some constructions of GO-metrics on homogeneous compact manifolds and
non compact manifolds of a semisimple group.
Many interesting results about GO-spaces one can find in [7, 12, 27, 23, 24], where there are
also extensive references.
Natural generalizations of normal homogeneous Riemannian manifolds are δ-homogeneous
Riemannian manifolds, studied in [3, 4, 5]. Note that the class of δ-homogeneous Riemannian
manifolds is a proper subclass of the class of geodesic orbit spaces with non-negative sectional
curvature (see the quoted papers for further properties of δ-homogeneous Riemannian mani-
folds).
In [1], a classification of non-normal invariant GO-metrics on flag manifolds M = G/H was
given. The problem reduces to the case when the (compact) group G is simple. There exist
only two series of flag manifolds of a simple group which admit such metric, namely weakly
symmetric spaces M1 = SO(2n + 1)/U(n) and M2 = Sp(n)/U(1) · Sp(n − 1), equipped with
any (non-normal) invariant metric (which depends on two real parameters). Moreover, there
exists unique (up to a scaling) invariant metric g0, such that the Riemannian manifolds (Mi, g0)
are isometric to the symmetric spaces SO(2n + 2)/U(n + 1) and CP 2n−1 = SU(2n)/U(2n− 1),
respectively.
The main goal of this paper is a generalization of this result to the case of compact ho-
mogeneous manifolds of positive Euler characteristic. We prove that the weakly symmetric
manifolds M1, M2 exhaust all simply connected compact irreducible Riemannian non-normal
GO-manifolds of positive Euler characteristic.
Compact Riemannian Manifolds with Homogeneous Geodesics 3
We indicate now the idea of the proof. Let (M = G/H, gM ) be a compact irreducible non-
normal GO-space of positive Euler characteristic. Then the stability subgroup H has maximal
rank, which implies that G is simple. We prove that there is rank 2 regular simple subgroup G′
of G (associated with a rank 2 subsystem R′ of the root system R of the Lie algebra g = Lie(G))
such that the orbit M ′ = G′o = G′/H ′ of the point o = eH ∈ M (with the induced metric) is
a non-normal GO-manifold. Using [1, 3], we prove that the only such manifold M ′ is SU(5)/U(2).
This implies that the root system R is not simply-laced and admits a “special” decomposition
R = R0 ∪ R1 ∪ R2 into a disjoint union of three subsets, which satisfies some properties. We
determine all such special decompositions of irreducible root systems and show that only root
systems of type Bn and Cn admit special decomposition and associated homogeneous manifolds
are M1 and M2.
The structure of the paper is the following. We fix notations and recall basic definitions
in Section 2. Some standard facts about totally geodesic submanifolds of a homogeneous Rie-
mannian spaces are collected in Section 3. We discuss some properties of compact GO-spaces
in Section 4. These results are used in Section 5 to derive sufficient conditions for existence
and non-existence of a non-normal GO-metric on a homogeneous manifold of a compact group.
Section 6 is devoted to classification of compact GO-spaces with positive Euler characteristic.
2 Preliminaries and notations
Let M = G/H be a homogeneous space of a compact connected Lie group G. We will denote
by b = 〈·, ·〉 a fixed AdG-invariant Euclidean metric on the Lie algebra g of G (for example, the
minus Killing form if G is semisimple) and by
g = h + m (1)
the associated b-orthogonal reductive decomposition, where h = Lie(H). An invariant Rieman-
nian metric gM on M is determined by an AdH -invariant Euclidean metric g = (·, ·) on the
space m which is identified with the tangent space ToM at the initial point o = eH.
If p is a subspace of m, we will denote by Xp the b-orthogonal projection of a vector X ∈ g
onto p, by bp the restriction of the symmetric bilinear form to p and by Ap = prp ◦A ◦ prp the
projection of an endomorphism A to p. If g is a AdH -invariant metric, the quotient
A = b−1
m ◦ g
is an AdH -equivariant symmetric positively defined endomorphism on m, which we call the
metric endomorphism. Conversely, any such equivariant positively defined endomorphism A
of m defines an invariant metric g = b ◦ A = b(A·, ·) on m, hence an invariant Riemannian
metric gM on M .
Lemma 1. Let (M = G/H, gM ) be a compact homogeneous Riemannian space with metric
endomorphism A and
m = m1 ⊕m2 ⊕ · · · ⊕mk, (2)
the A-eigenspace decomposition such that A|mi = λi · 1mi. Then
(mi,mj) = 〈mi,mj〉 = 0 (3)
and AdH-modules mi satisfy [mi,mj ] ⊂ m for i 6= j.
4 D.V. Alekseevsky and Yu.G. Nikonorov
Proof. Since A commute with AdH , eigenspaces mi are AdH -invariants and for X ∈ mi, Y ∈ mj ,
i 6= j, we get
λi〈X, Y 〉 = 〈AX, Y 〉 = (X, Y ) = 〈X, AY 〉 = λj〈X, Y 〉.
This implies (3). The inclusion [mi,mj ] ⊂ m follows from the fact that mj is AdH -invariant and
〈[mi,mj ], h〉 = 〈mi, [mj , h]〉 = 0. �
For any subspace p ⊂ m we will denote by p⊥ its orthogonal complement with respect to the
metric g and by 1p the identity operator on p.
Recall that AdH -submodules p, q are called disjoint if they have no non-zero equivalent
submodules. If AdH -module m is decomposed into a direct sum
m = m1 + · · ·+ mk
of disjoint submodules, then any AdH -invariant metric g and associated metric endomorphism A
have the form
g = gm1 ⊕ · · · ⊕ gmk
, A = Am1 ⊕ · · · ⊕Amk .
Let (M = G/H, gM ) be a compact homogeneous Riemannian space with the reductive de-
composition (1) and metric endomorphism A ∈ End(m).
We identify elements X, Y ∈ g with Killing vector fields on M . Then the covariant deriva-
tive ∇XY at the point o = eH is given by
∇XY (o) = −1
2 [X, Y ]m + U(Xm, Ym), (4)
where the bilinear symmetric map U : m×m → m is given by
2(U(X, Y ), Z) = (adm
Z X, Y ) + (X, adm
Z Y ) (5)
for any X, Y, Z ∈ m and Xm is the m-part of a vector X ∈ g [8].
Definition 1. A homogeneous Riemannian space (M = G/H, gM ) is called a space with ho-
mogeneous geodesics shortly, GO-space if any geodesic γ of M is an orbit of 1-parameter
subgroup of G. The invariant metric gM is called GO-metric.
If G is the full isometry group, then GO-space is called a manifold with homogeneous
geodesics or GO-manifold.
Definition 2. A GO-space (M = G/H, gM ) of a simple compact Lie group G is called a proper
GO-space if the metric gM is not G-normal, i.e. the metric endomorphism A is not a scalar
operator.
Lemma 2 ([1]). A compact homogeneous Riemannian space (M = G/H, gM ) with the reductive
decomposition (1) and metric endomorphism A is GO-space if and only if for any X ∈ m there
is HX ∈ h such that one of the following equivalent conditions holds:
i) [HX + X, A(X)] ∈ h;
ii) ([HX + X, Y ]m, X) = 0 for all Y ∈ m.
This lemma shows that the property to be GO-space depends only on the reductive decom-
position (1) and the Euclidean metric g on m. In other words, if (M = G/H, gM ) is a GO-space,
then any locally isomorphic homogeneous Riemannian space (M ′ = G′/H ′, gM ′
) is a GO-space.
Also a direct product of Riemannian manifolds is a manifold with homogeneous geodesics if and
only if each factor is a manifold with homogeneous geodesics.
Compact Riemannian Manifolds with Homogeneous Geodesics 5
3 Totally geodesic orbits in a homogeneous Riemannian space
In this section we deal with totally geodesic submanifolds of compact homogeneous Riemannian
spaces. This is a useful tool for study of GO-spaces due to the following
Proposition 1 ([3, Theorem 11]). Every closed totally geodesic submanifold of a Riemannian
manifold with homogeneous geodesics is a manifold with homogeneous geodesics.
Let (M = G/H, gM ) be a compact Riemannian homogeneous space with the reductive de-
composition (1).
Definition 3. A subspace p ⊂ m is called totally geodesic if it is the tangent space at o of
a totally geodesic orbit Ko ⊂ G/H = M of a subgroup K ⊂ G.
Proposition 2. A subspace p ⊂ m is totally geodesic if and only if the following two conditions
hold:
a) p generates a subalgebra of the form k = h′ + p, where h′ is a subalgebra of h;
b) the endomorphism adp
Z ∈ End(p) for Z ∈ p⊥ is g-skew-symmetric or, equivalently,
U(p, p) ⊂ p.
Proof. If p is the tangent space of the orbit Ko = K/H ′, then Lie(K) = k = h′ + p, where
h′ = Lie(H ′) is a subalgebra of h. Moreover, the formulas (4) and (5) imply U(p, p) ⊂ p.
Conversely, the conditions a) and b) imply that p is the tangent space of the totally geodesic
orbit Ko of the subgroup K generated by the subalgebra k. �
Corollary 1.
i) A subspace p ⊂ m is totally geodesic if a) holds and Ap = p.
ii) If a totally geodesic subspace p is adh-invariant and A-invariant, then
[h + p, p⊥] ⊂ p⊥.
Proof. i) Assume that Ap = p. Then Ap⊥ = p⊥ and 〈p, p⊥〉 = 0. From i) and Ap = p we get
〈Z, [X, AX]〉 = 0 for any X ∈ p and Z ∈ p⊥. This implies
0 = 〈[Z,X], AX〉 = 〈[Z,X]m, AX〉 = ([Z,X]m, X)
= ([Z,X]p, X) = (U(X, X), Z).
ii) follows from the fact that the endomorphisms adh+p are b-skew-symmetric and preserves
the subspace p. Hence, they preserve its b-orthogonal complement p⊥. �
Corollary 2. Let (M = G/H, g) be a compact Riemannian homogeneous space and K a con-
nected subgroup of G. The orbit P = Ko = K/H ′ is a totally geodesic submanifold if and only if
the Lie algebra k is consistent with the reductive decomposition (1) (that is k = k∩h+k∩m = h′+p)
and
U(p, p) ⊂ p
or, equivalently, the endomorphisms adp
Z ∈ End(p), Z ∈ p⊥ are g-skew-symmetric.
6 D.V. Alekseevsky and Yu.G. Nikonorov
4 Properties of GO-spaces
Lemma 3. Let (M = G/H, gM ) be a GO-space with the reductive decomposition (1) and m =
p + q a g-orthogonal AdH-invariant decomposition. Then
U(p, p) ⊂ p, U(q, q) ⊂ q
and the endomorphisms adq
p, adp
q are skew-symmetric.
Proof. For X ∈ p, Y ∈ q we have
0 = ([Y + HY , X]m, Y ) = − (adX Y, Y ) = − (U(Y, Y ), X) ,
where HY is as in Lemma 2. This shows that adq
X is skew-symmetric and U(q, q) ⊂ q. �
Lemma 3 together with Proposition 2 implies
Proposition 3. Let (M = G/H, gM ) be a GO-space with the reductive decomposition (1). Then
any connected subgroup K ⊂ G which contains H has the totally geodesic orbit P = Ko = K/H
which is GO-space (with respect to the induced metric). Moreover, if the space p := k ∩ m is
A-invariant, then
[k,m⊥] ⊂ m⊥
and the metric ḡ := g|p⊥ is AdK-invariant and defines an invariant GO-metric gN on the ho-
mogeneous manifolds N = G/K. The projection π : G/H → G/K is a Riemannian submersion
with totally geodesic fibers such that the fibers and the base are GO-spaces.
Proof. The first claim follows from Lemma 3, Lemma 2 and Proposition 2. If Ap = p, then m =
p + p⊥ is a b-orthogonal decomposition and since the metric b is AdG-invariant, AdK p⊥ = p⊥.
Then Lemma 3 shows that the metric g|p⊥ is AdK-invariant and defines an invariant metric gN
on N = G/K such that N becomes GO-space. �
Note that a subgroup K ⊃ H is compatible with any invariant metric on G/H if AdH -
modules p and m/p are strictly disjoint. This remark implies
Proposition 4. Let (M = G/H, g) be a compact homogeneous Riemannian space. Then the
connected normalizer N0(Z) of a central subgroup Z of H and the connected normalizer N0(H)
are subgroups consistent with any invariant metric on M .
Proposition 5. Let (M = G/H, gM ) be a compact GO-space with metric endomorphism A.
i) Let X, Y ∈ m be eigenvectors of the metric endomorphism A with different eigenvalues λ, µ.
Then
[X, Y ] =
λ
λ− µ
[H,X] +
µ
λ− µ
[H,Y ]
for some H ∈ h.
ii) Assume that the vectors X, Y belong to the λ-eigenspace mλ of A and X is g-orthogonal
to the subspace [h, Y ]. Then
[X, Y ] ∈ h + mλ.
Compact Riemannian Manifolds with Homogeneous Geodesics 7
Proof. i) Let X, Y ∈ m be eigenvectors of A with different eigenvalues λ, µ and H = HX+Y ∈ h
the element defined in Lemma 2. Then
[H + X + Y, A(X + Y )] = [H + X + Y, λX + µY ]
= λ[H,X] + µ[H,Y ] + (µ− λ)[X, Y ] ∈ h.
By Lemma 1, [H,X], [H,Y ], [X, Y ] ∈ m and the right hand side is zero.
ii) Assume now that X, Y ∈ mλ satisfy conditions ii) and Z is an eigenvector of A with an
eigenvalue µ 6= λ. Then we have
([X, Y ]m, Z) = µ〈[X, Y ], Z〉 = µ〈X, [Y, Z]〉 =
µ
λ
(X, [Y, Z]m)
=
µ
λ
(
X,
λ
λ− µ
[H,Y ] +
µ
λ− µ
[H,Z]
)
= 0.
This shows that [X, Y ] ∈ h + mλ. �
Corollary 3. Let (M = G/H, gM ) be a compact GO-space with the reductive decomposition (1)
and metric endomorphism A and
m = m1 + · · ·+ mk (6)
the A-eigenspace decomposition such that A|mi = λi1mi. Then for any AdH-submodules pi ⊂ mi,
pj ⊂ mj, i 6= j, we have
[pi, pj ] ⊂ pi + pj .
Moreover, if p, p′ are g-orthogonal AdH-submodules of mi then
[p, p′] ⊂ h + mi.
5 Some applications
5.1 A sufficient condition for non-existence of GO-metric
Here we consider some applications of results of the previous section.
Definition 4. Let (M = G/H, gM ) be a compact homogeneous Riemannian space. A connected
closed Lie subgroup K ⊂ G which contains H is called compatible with the metric gM if
the subspace p = k ∩m of m is invariant under the metric endomorphism A.
Let K, K ′ be two subgroups of G which are compatible with the metric of a homogeneous
Riemannian space (M = G/H, gM ). Then we can decompose the space m into a g-orthogonal
sum of A-invariant AdH -modules
m = q + p1 + p2 + n (7)
where
q = p ∩ p′, p = k ∩m = q + p1, p′ = k′ ∩m = q + p2
and n is the orthogonal complement to
p + p′ = q + p1 + p2
in m.
8 D.V. Alekseevsky and Yu.G. Nikonorov
Proposition 6. Let (M = G/H, gM ) be a homogeneous Riemannian space, K, K ′ two subgroups
of G which are compatible with gM and (7) the associated decomposition as above. Then p1, p2, n
are Ad
H̃
-modules, where H̃ = K ∩K ′ is the Lie group with the Lie algebra h̃ = h + q, and
[p1, p2] ⊂ n.
Moreover, if (M = G/H, gM ) is a GO-space, then the restriction Am̃ of the metric endomorphism
to m̃ = p1 + p2 + n commutes with Ad
H̃
|m̃ and for any H̃-irreducible submodules p′1 ⊂ p1, and
p′2 ⊂ p2 such that
[p′1, p
′
2] 6= 0,
the metric endomorphism A is a scalar on the space
p′1 + p′2 + [p′1, p
′
2].
Proof. Since the decomposition (7) is b-orthogonal, we conclude that it is Ad
H̃
-invariant and
[p, p⊥] = [p, p2 + n] ⊂ p2 + n,
[p′, (p′)⊥] = [p′, p1 + n] ⊂ p1 + n,
by Proposition 2. This implies
[p1, p2] ⊂ n.
If (M, gM ) is a GO-space, then by Proposition 3 the metric endomorphism Am̃ is H̃-invariant.
If modules p′1, p′2 belong to A-eigenspaces with different eigenvalues, then by Corollary 3,
[p′1, p
′
2] ⊂ p1 + p2.
Together with the previous inclusion, it implies [p′1, p
′
2] = 0. If these modules belong to the same
eigenspace mλ, then by Corollary 3,
[p′1, p
′
2] ⊂ mλ. �
As a corollary, we get the following sufficient condition that a homogeneous manifold M =
G/H does not admit a proper GO-metric.
Proposition 7. Let M = G/H be a homogeneous space of a compact group G with the reductive
decomposition g = h+m. Assume that the Lie algebra g has two subalgebras k = h+p, k′ = h+p′
which contain h and generate g. Let
m = q + p1 + p2 + n, q = p ∩ p′ (8)
be the associated b-orthogonal decomposition. Assume that there is no commuting adh+q sub-
modules of p1 and p2. Then for any GO-metric, defined by an operator A which preserves this
decomposition, A is a scalar operator on p1+p2+n. In particular, if q is trivial and AdH-modules
p1, p2, n are strictly non-equivalent, then the only GO-metric on M is the normal metric.
Proof. Let A be an operator on m which preserves the decomposition (8) and defines a GO-
metric. Then by Proposition 6,
A|p1+p2+[p1,p2] = λ · 1
for some λ. Now p1 and [p1, p2] ⊂ n are two g-orthogonal submodules of the A-eigenspace mλ.
Applying Corollary 3, we conclude that
[p1, [p1, p2]] ⊂ mλ.
Iterating this process, we prove that
n = [p1, p2] + [p1, [p1, p2]]n + [p2, [p1, p2]]n + · · · ⊂ mλ
and A = λ · 1 on p1 + p2 + n. �
Compact Riemannian Manifolds with Homogeneous Geodesics 9
5.2 A sufficient condition for existence of GO-metric
Lemma 4. Let M = G/H be a homogeneous space of a compact Lie group with a reductive
decomposition g = h + m. Assume that AdH-module m has a decomposition
m = m1 + · · ·+ mk
into invariant submodules, such that for any i < j
[mi,mj ] = 0
or this condition valid with one exception (i, j) = (1, 2) and in this case
[m1,m2] ⊂ m2
and for any X ∈ m1, Y ∈ m2 there is H ∈ h such that adH Y = adX Y and
adH(m1 + m3 + · · ·+ mk) = 0.
Then any metric endomorphism of the form A =
∑
xi · 1mi defines a GO-metric on M .
Proof. Under the assumptions of lemma, for H ∈ h and Xi ∈ mi we have[
H +
∑
Xi,
∑
xiXi
]
=
∑
i<j
(xj − xi)[Xi, Xj ] +
∑
i
xi adH Xi
= (x2 − x1) adX1 X2 + x2 adH X2 + adH
(
x1X1 +
∑
k≥3
xkXk
)
.
The right-hand side is zero if H is chosen as in the lemma (where Y = x2X2 and X = (x1 −
x2)X1). Now, it suffices to apply Lemma 2. �
Example 1. The homogeneous space M = SUp+q/SUp × SUq is a GO-space with respect to
any invariant metric.
We have the reductive decomposition
sup+q = h + m = (sup + suq) + (Ra + p),
where p ' Cp ⊗ Cq and ada |p = i · 1p. Any metric endomorphism A = λ · 1Ra + µ · 1p defines
a GO-metric by above lemma since for any X ∈ p there is H ∈ h such that adH X = iX. Note
that for p 6= q these manifolds are weakly symmetric spaces [26].
5.3 GO-metrics on a compact group G
Proposition 8. A compact Lie group G with a left-invariant metric g is a GO-space if and only
if the corresponding Euclidean metric (·, ·) on the Lie algebra g is bi-invariant.
Proof. The condition that (G, g) is a GO-space can be written as
0 = (X, [X, Y ]) = −(adY X, X) = 0.
This shows that the metric (·, ·) is bi-invariant. �
Note that a compact Lie group G can admit a non-bi-invariant left-invariant metrics g with
homogeneous geodesics. But the corresponding GO-space will have the form L/H where the
group L will contain G as a proper subgroup. See [11] for details.
10 D.V. Alekseevsky and Yu.G. Nikonorov
6 Homogeneous GO-spaces with positive Euler characteristic
6.1 Basic facts about homogeneous manifolds of positive Euler characteristic
Here we recall some properties of homogeneous spaces with positive Euler characteristic (see,
for example, [21] or [4]). A homogeneous space M = G/H of a compact connected Lie group G
has positive Euler characteristic χ(M) > 0 if and only if the stabilizer H has maximal rank
(rk(H) = rk(G)).
If the group G acts on M almost effectively, then it is semisimple and the universal covering
M̃ = G̃/H̃ is a direct product
M̃ = G1/H1 × · · · ×Gk/Hk,
where G̃ = G1 ×G2 × · · · ×Gk is the decomposition of the group G̃ (which is a covering of G)
into a direct product of simple factors and Hi = H̃ ∩Gi.
Any invariant metric gM on M defines an invariant metric gM̃ on M̃ and the homogeneous
Riemannian space (M̃ = G̃/H̃, gM̃ ) is a direct product of homogeneous Riemannian spaces
(Mi = Gi/Hi, g
Mi), i = 1, . . . , k, of simple compact Lie groups Gi, see [19]. We have
Proposition 9 ([19]). A compact almost effective homogeneous Riemannian space (M =
G/H, gM ) of positive Euler characteristic is irreducible if and only if the group G is simple.
If the group G acts effectively on M , it has trivial center.
This proposition shows that a simply connected compact GO-space (M = G/H, gM ) of posi-
tive Euler characteristic is a direct product of simply connected GO-spaces (Mi = Gi/Hi, g
Mi)
of simple Lie groups with positive Euler characteristic. So it is sufficient to classify simply
connected GO-spaces of a simple compact Lie group with positive Euler characteristic.
A description of homogeneous spaces G/H of positive Euler characteristic reduces to descrip-
tion of connected subgroups H of maximal rank of G or equivalently, subalgebras of maximal
rank of a simple compact Lie algebra g, see [9] and also Section 8.10 in [25]. An important
subclass of compact homogeneous spaces of positive Euler characteristic consists of f lag mani-
folds. They are described as adjoint orbits M = AdG x of a compact connected semisimple
Lie group G or, in other terms as quotients M = G/H of G by the centrelizer H = ZG(T ) of
a non-trivial torus T ⊂ G.
Note that every compact naturally reductive homogeneous Riemannian space of positive Euler
characteristic is necessarily normal homogeneous with respect to some transitive semisimple
isometry group [4].
6.2 The main theorem
Let G be a simple compact connected Lie group, H ⊂ K ⊂ G its closed connected subgroups.
We denote by b = 〈·, ·〉 the minus Killing form on the Lie algebra g and consider the following
b-orthogonal decomposition
g = h⊕m = h⊕m1 ⊕m2,
where
k = h⊕m2
is the Lie algebra of the group K. Obviously, [m2,m1] ⊂ m1. Let gM = gx1,x2 be a G-invariant
Riemannian metric on M = G/H, generated by the Euclidean metric g = (·, ·) on m of the form
g = x1 · bm1 + x2 · bm2 , (9)
Compact Riemannian Manifolds with Homogeneous Geodesics 11
where x1 and x2 are positive numbers, or, equivalently, by the metric endomorphism
A = x1 · 1m1 + x2 · 1m2 . (10)
We consider two examples of such homogeneous Riemannian spaces (M = G/H, gx1,x2):
a) (G, K,H) = (SO(2n + 1), U(n), SO(2n)), n ≥ 2. The group G = SO(2n + 1) acts tran-
sitively on the symmetric space Com(R2n+2) = SO(2n + 2)/U(n) of complex structures
in R2n+2 with stabilizer H = U(n), see [15]. So we can identify M = G/H with this
symmetric space, but the metric gx1,x2 is not SO(2n + 2)-invariant if x2 6= 2x1 [17].
b) (G, K,H) = (Sp(n), Sp(1) · Sp(n − 1), U(1) · Sp(n − 1)), n ≥ 2. The group G = Sp(n)
acts transitively on the projective space CP 2n−1 = SU(2n + 2)/U(2n + 1) with stabilizer
H = U(1) · Sp(n − 1). So we can identify M = G/H with CP 2n−1, but the metric gx1,x2
is not SU(2n + 2)-invariant if x2 6= 2x1, see [15, 17].
Now we can state the main theorem about compact GO-spaces of positive Euler characteristic.
Theorem 1. Let (M = G/H, gM ) is a simply connected proper GO-space with positive Euler
characteristic and simple compact Lie group G. Then M = G/H = SO(2n + 1)/U(n), n ≥
2, or G/H = Sp(n)/U(1) × Sp(n − 1), n ≥ 2, and gM = gx1,x2 is any G-invariant metric
which is not G-normal homogeneous. The metric gM is G-normal homogeneous (respectively,
symmetric) when x2 = x1 (respectively, x2 = 2x1). Moreover, these homogeneous spaces are
weakly symmetric flag manifolds.
The non-symmetric metrics gx1,x2 have G as the full connected isometry group of the con-
sidered GO-spaces (M = G/H, gx1,x2), see discussion in [17, 21, 1]. The claim that all these
homogeneous Riemannian spaces are weakly symmetric spaces was proved in [27]. Note also
that Theorem 1 allows to simplify some arguments in the paper [3].
6.3 Proof of the main theorem
Using results from [1] and [3], we reduce the proof to a description of some special decompositions
of the root system of the Lie algebra g of the isometry group G.
Let M = G/H be a homogeneous space of a compact simple Lie group of positive characte-
ristic and
g = h + m
associated reductive decomposition. The subgroup H contains a maximal torus T of G. We
consider the root space decomposition
gC = tC +
∑
α∈R
gα
of the complexification gC of the Lie algebra g, where tC is the Cartan subalgebra associated
with T and R is the root system.
For any subset P ⊂ R we denote by
g(P ) =
∑
α∈P
gα
the subspace spanned by corresponding root space gα. Then H-module mC is decomposed into
a direct sum
mC = g(R1) + · · ·+ g(Rk)
12 D.V. Alekseevsky and Yu.G. Nikonorov
of disjoint submodules, where R = R1 ∪ · · · ∪Rk is a disjoint decomposition of R and subsets Ri
are symmetric, i.e. −Ri = Ri. Moreover, real H-modules g∩g(Ri) are irreducible. Any invariant
metric on M is defined by the metric endomorphism A on m whose extension to mC has the
form
A = diag(x1 · 1p1 , . . . , x` · 1p`
),
where xi are arbitrary positive numbers, xi 6= xj and pi is a direct sum of modules g(Rm).
We will assume that A is not a scalar operator (i.e. ` > 1) and it defines an invariant metric
with homogeneous geodesics. We say that a root α corresponds to eigenvalue xi of A if gα ⊂ pi.
Lemma 5. There are two roots α, β which correspond to different eigenvalues of A such that
α + β is a root.
Proof. If it is not the case, [p1, pi] = 0 for i 6= 1 and g1 = p1 + [p1, p1] would be a proper ideal
of a simple Lie algebra g. �
Now, consider the roots α and β as in the previous lemma. Since R(α, β) := R ∩ span{α, β}
is a rank 2 root system, we can always choose roots α, β ∈ R which form a basis of the root
system R(α, β). Then the subalgebra
gα,β := tC +
∑
γ∈R(α,β)
gγ
of gC is the centralizer of the subalgebra t′ = kerα ∩ ker β ⊂ tC.
Then the orbit Gα,βo ⊂ M of the corresponding subgroup Gα,β = T ′ ·G′
α,β ⊂ G is a totally
geodesic submanifold (see Corollary 2), hence a proper GO-space with the effective action of
the rank two simple group G′
α,β associated with the root system R(α, β) (see Proposition 1).
Note that it has positive Euler characteristic since the stabilizer of the point o contains the
two-dimensional torus generated by vectors Hα,Hβ ∈ tC associated with roots α, β. Recall that
Hα = 2
〈α,α〉 b−1 · α.
Proposition 10. Every proper GO-space (M = G/H, gM ) with positive Euler characteristic of
a simple group G of rank 2 is locally isometric to the manifold M = SO(5)/U(2) with the metric
defined by the metric endomorphism
A = x1 · 1g(Rs) + x2 · 1g(R`), x1 6= x2 > 0
where
Rs = {±ε1,±ε2}, R` = {±ε1 ± ε2},
are the sets of short and, respectively, long roots of the Lie algebra so(5). We may assume also
that
mC = g(Rs ∪ {ε1 + ε2}) and hC = tC + gε1−ε2 .
Proof. Proof of this proposition follows from results of the papers [1] and [3]. Indeed, the
group G has the Lie algebra g isomorphic to su(3) = A2, so(5) = sp(2) = B2 = C2 or g2. Since
the universal Riemannian covering of a GO-space is a GO-space (Lemma 2), we may assume
without loss of generality that G/H is simply connected.
If g = su(3), then G/H = SU(3)/S(U(2)×U(1)) (a symmetric space) or G/H = SU(3)/T 2,
where T 2 is a maximal torus in SU(3). Both these spaces are flag manifolds, and results of [1]
show that any GO-metric on these spaces is SU(3)-normal homogeneous.
Compact Riemannian Manifolds with Homogeneous Geodesics 13
If g = so(5) = sp(2), then (g, h) = (so(5), R2), (g, h) = (so(5), R⊕su(2)l), (g, h) = (so(5), R⊕
su(2)s), or (g, h) = (so(5), su(2)l⊕su(2)l), where su(2)l, (respectively, su(2)s) stands for a three-
dimensional subalgebras generated by all long (respectively, short) roots of g. The last pair
corresponds to the irreducible symmetric space SO(5)/SO(4), which admits no non-normal
invariant metric. All other spaces are flag manifolds. Results of [1] implies that the only
possible pair is (g, h) = (so(5), R ⊕ su(2)l), which corresponds to the space SO(5)/U(2) =
Sp(2)/U(1) · Sp(1).
For g = g2 the statement of proposition is proved in [3, Proposition 23]. �
Corollary 4. Let G be a simple compact Lie group and M = G/H a proper GO-space with
positive Euler characteristic. Then the root system R of the complex Lie algebra gC admits
a disjoint decomposition
R = R0 ∪R1 ∪R2,
where R0 is the root system of the complexified stability subalgebra hC,with the following proper-
ties:
i) If α ∈ R1, β ∈ R2 and α + β ∈ R then α− β ∈ R and the rank 2 root system R(α, β) has
type B2 = C2.
ii) Moreover, if α, β is a basis of R(α, β) (that is 〈α, β〉 < 0), then one of the roots α, β is
short and the other is long and one of the long roots α± β belongs to R0 and second one
belongs to R1 ∪R2.
iii) If both roots α, β are short, then one of the long roots α ± β belongs to R0 and the other
belongs to R1 ∪R2.
iv) If α ∈ R1 and β ∈ R2 are long roots, then α± β /∈ R.
We will call a decomposition with the above properties a special decomposition. Corol-
lary 4 implies
Corollary 5. There is no proper GO-spaces of positive Euler characteristic with simple isometry
group G = SU(n), SO(2n), E6, E7, E8 (these are all simple Lie algebras with all roots of the
same length (simply-laced root system)).
Corollary 6 ([3, Proposition 23]). Any GO-space (G/H, µ) of positive Euler characteristic
with G = G2 is normal homogeneous.
Now, we describe all special decompositions of the root systems of types Bn, Cn, F4. We
will use notation from [13] for root systems and simple roots.
Lemma 6. The root system
R(F4) = {±εi, 1/2(±ε1 +∓ε2 ± ε3 +±ε4,±εi ± εj), i, j = 1, 2, 3, 4, i 6= j}
does not admit a special decomposition.
Proof. Assume that such a decomposition exists. Then we can choose roots α ∈ R1, β ∈ R2
such that α±β is a root. Then α, β has different length and we may assume that |α| < |β| and
〈α, β〉 < 0. Then we can include α, β into a system of simple roots δ, α, β, γ, see [13]. Since all
such systems are conjugated, we may assume that α = ε4, β = −ε4 + ε3, see [13]. Then we get
contradiction, since α− β is not a root. �
14 D.V. Alekseevsky and Yu.G. Nikonorov
Now we describe two special decompositions for the root systems
R(Bn) = {±εi, ±εi ± εj , i, j = 1, . . . , n}
and
R(Cn) = {±2εi, ±εi ± εj , i, j = 1, . . . , n}
of types Bn and Cn. Note that in both cases RA = {±(εi − εj)} is a closed subsystem. We set
R+
A = {±(εi + εj)}.
We denote by R+ the standard subsystem of positive roots of a root system R and by Rs
and R` the subset of short and, respectively, long roots of R. Then there is a special decomposi-
tion R = R0 ∪R1 ∪R2 of the systems R(Bn), R(Cn) which we call the standard decomposition:
R(Bn) = RA ∪Rs ∪R+
A,
R(Cn) = RA ∪R` ∪R+
A.
These decompositions define the following reductive decompositions of the homogeneous spaces
SO(2n + 1)/U(n) and Sp(n)/U(n):
so(2n + 1) = h + (m1 + m2) = g(RA) + (g(Rs) + g(R+
A)),
sp(n) = h + (m1 + m2) = g(RA) + (g(R`) + g(R+
A)),
where m1, m2 are irreducible submodules of m. It is known [1] that any metric endomorphism
A = diag(x1 · 1m1 , x2 · 1m2) defines a metric with homogeneous geodesics on the corresponding
manifold M = G/H (see a discussion before the statement of Theorem 1). Now, the proof of
Theorem 1 follows from the following proposition.
Proposition 11. Any special decomposition of the root systems RB, RC is conjugated to the
standard one.
Proof. We give a proof of this proposition for R(Bn). The proof for R(Cn) is similar.
Let
R(Bn) = R0 ∪R1 ∪R2
be a special decomposition of R(Bn). We may assume that there are roots α ∈ R1 and β ∈ R2
with 〈α, β〉 < 0 and |α| < |β|. Then we can include α, β into a system of simple roots, which,
without loss of generality, can be written as
α1 = ε1 − ε2, . . . , εn−2 − εn−1, εn−1 + εn = β, −εn = α.
Then (εn−1 − εn) ∈ R0. We need the following lemma.
Lemma 7. Let R(Bn) = R0 ∪ R1 ∪ R2 be a special decomposition as above, V ′ = ε⊥n the
orthogonal complement of the vector εn and R(Bn−1) = R′ := R∩ V ′ the root system induced in
the hyperspace V ′. Then the induced decomposition R′ = R′
0 ∪R′
1 ∪R′
2, where R′
i := Ri ∩ V ′, is
a special decomposition.
Proof. It is sufficient to check that subsets R′
1, R′
2 are not empty.
We say that two roots γ, δ are R0-equivalent (γ ∼ δ) if their difference belongs to R0.
The equivalent roots belong to the same component Ri. The root εn−1 = εn − (εn−1 − εn) is
R0-equivalent to α = εn. Hence it belongs to R1.
Compact Riemannian Manifolds with Homogeneous Geodesics 15
We say that a pair of roots γ, δ with 〈γ, δ〉 < 0 is special if one of the roots belongs to R1
and another to R2. Then they have different length (say, |γ| < |δ|). Moreover, the root γ + δ is
short and it belongs to the same part Ri, i = 1, 2 as the short root δ and the root 2γ + δ is long
and it belongs to R0.
Consider the roots σ± = ±εn−2 + εn−1. They have negative scalar product with εn−1 ∈ R1
and β = εn−1 + εn ∈ R2. They can not belong to R1 since then we get a special pair δ±, β which
consists of long roots. They both can not belong to R0 since otherwise the root εn−2 ∼ εn−1 ∈ R1
and ±εn−2 + εn ∼ εn−1 + εn ∈ R2 and we get a special pair
γ = εn−2 ∈ R1, δ = −εn−2 + εn ∈ R2,
such that 2γ + δ ∈ R0, which is impossible. We conclude that one of the roots σ± = ±εn−2 +
εn−1 ∈ R′ must belongs to R2. Since the root εn−1 ∈ R′ belongs to R1, the lemma is proved. �
Now we prove the proposition by induction on n. The claim is true for n = 2 by Proposi-
tion 10. Assume that it is true for R(B(n − 1)) and let R(Bn) = R0 ∪ R1 ∪ R2 be a special
decomposition as above. By lemma, the decomposition R′ = R′
0 ∪ R′
1 ∪ R′
2, indiced in the hy-
perplane V ′ = en ⊥, is a special decomposition. By inductive hypothesis we may assume that
it has the standard form:
R0 = {±(εi − εj)}, R1 = {±εi}, R2 = {±(εi + εj), i, j = 1, . . . , n− 1}.
This implies that the initial decomposition is also standard. �
Acknowledgements
The first author was partially supported by the Royal Society (Travel Grant 2007/R3). The
second author was partially supported by the State Maintenance Program for the Leading Scien-
tific Schools of the Russian Federation (grant NSH-5682.2008.1). We are grateful to all referees,
whose comments and suggestions permit us to improve the presentation of this article.
References
[1] Alekseevsky D.V., Arvanitoyeorgos A., Riemannian flag manifolds with homogeneous geodesics, Trans.
Amer. Math. Soc. 359 (2007), 3769–3789.
[2] Akhiezer D.N., Vinberg E.B., Weakly symmetric spaces and spherical varieties, Transform. Groups 4 (1999),
3–24.
[3] Berestovskii V.N., Nikonorov Yu.G., On δ-homogeneous Riemannian manifolds, Differential Geom. Appl.
26 (2008), 514–535, math.DG/0611557.
[4] Berestovskii V.N., Nikonorov Yu.G., On δ-homogeneous Riemannian manifolds. II, Siber. Math. J. 50 (2009),
214–222.
[5] Berestovskii V.N., Nikitenko E.V., Nikonorov Yu.G., The classification of δ-homogeneous Riemannian man-
ifolds with positive Euler characteristic, arXiv:0903:0457.
[6] Berger M., Les variétés riemanniennes homogènes normales à courbure strictement positive, Ann. Scuola
Norm. Sup. Pisa (3) 15 (1961), 179–246.
[7] Berndt J., Kowalski O., Vanhecke L., Geodesics in weakly symmetric spaces, Ann. Global Anal. Geom. 15
(1997), 153–156.
[8] Besse A.L., Einstein manifolds, Springer-Verlag, Berlin, 1987.
[9] Borel A., de Siebenthal J., Les sous-groups fermés de rang maximum des groups de Lie clos, Comment.
Math. Helv. 23 (1949), 200–221.
[10] Cartan É., Sur une classe remarquable d’espaces de Riemann, Bull. Soc. Math. France 54 (1926), 214–264.
Cartan É., Sur une classe remarquable d’espaces de Riemann. II, Bull. Soc. Math. France 55 (1927), 114–134.
http://arxiv.org/abs/math.DG/0611557
http://arxiv.org/abs/0903:0457
16 D.V. Alekseevsky and Yu.G. Nikonorov
[11] D’Atri J.E., Ziller W., Naturally reductive metrics and Einstein metrics on compact Lie groups, Mem. Amer.
Math. Soc. 18 (1979), no. 215, 1–72.
[12] Dus̆ek Z., Kowalski O., Nikc̆ević S., New examples of Riemannian g.o. manifolds in dimension 7, Differential
Geom. Appl. 21 (2004), 65–78.
[13] Gorbatzevich V.V., Onishchik A.L., Vinberg E.B., Lie groups and Lie algebras. III. Structure of Lie groups
and Lie algebras, Encyclopaedia of Mathematical Sciences, Vol. 41, Springer-Verlag, Berlin, 1994.
[14] Gordon C.S., Homogeneous Riemannian manifolds whose geodesics are orbits, in Topics in Geometry:
in Memory of Joseph D’Atri, Progr. Nonlinear Differential Equations Appl., Vol. 20, Birkhäuser Boston,
Boston, MA, 1996, 155–174.
[15] Helgason S., Differential geometry and symmetric spaces, Pure and Applied Mathematics, Vol. 12, Academic
Press, New York – London, 1962.
[16] Kaplan A., On the geometry of groups of Heisenberg type, Bull. London Math. Soc. 15 (1990), 32–42.
[17] Kerr M.M., Some new homogeneous Einstein metrics on symmetric spaces, Trans. Amer. Math. Soc. 348
(1996), 153–171.
[18] Kobayashi S., Nomizu K., Foundations of differential geometry, John Wiley & Sons, New York, Vol. I, 1963,
Vol. II, 1969.
[19] Kostant B., On holonomy and homogeneous spaces, Nagoya Math. J. 12 (1957), 31–54.
[20] Kowalski O., Vanhecke L., Riemannian manifolds with homogeneous geodesics, Boll. Un. Mat. Ital. B (7)
5 (1991), 189–246.
[21] Onishchik A.L., Topology of transitive transformation groups, Johann Ambrosius Barth Verlag GmbH,
Leipzig, 1994.
[22] Selberg A., Harmonic analysis and discontinuous groups in weakly symmetric Riemannian spaces, with
applications to Dirichlet series, J. Indian Math. Soc. (N.S.) 20 (1956), 47–87.
[23] Tamaru H., Riemannin G.O. spaces fibered over irreducible symmetric spaces, Osaka J. Math. 15 (1998),
835–851.
[24] Tamaru H., Riemannin geodesic orbit space metrics on fiber bundles, Algebra Groups Geom. 36 (1999)
835–851.
[25] Wolf J.A., Spaces of constant curvature, Publish or Perish, Inc., Houston, TX, 1984.
[26] Wolf J.A., Harmonic analysis on commutative spaces, Mathematical Surveys and Monographs, Vol. 142,
American Mathematical Society, Providence, RI, 2007.
[27] Ziller W., Weakly symmetric spaces, in Topics in Geometry: in Memory of Joseph D’Atri, Progr. Nonlinear
Differential Equations Appl., Vol. 20, Birkhäuser Boston, Boston, MA, 1996, 355–368.
[28] Yakimova O.S., Weakly symmetric Riemannian manifolds with a reductive isometry group, Sb. Math. 195
(2004), 599–614.
1 Introduction
2 Preliminaries and notations
3 Totally geodesic orbits in a homogeneous Riemannian space
4 Properties of GO-spaces
5 Some applications
5.1 A sufficient condition for non-existence of GO-metric
5.2 A sufficient condition for existence of GO-metric
5.3 GO-metrics on a compact group G
6 Homogeneous GO-spaces with positive Euler characteristic
6.1 Basic facts about homogeneous manifolds of positive Euler characteristic
6.2 The main theorem
6.3 Proof of the main theorem
References
|
| id | nasplib_isofts_kiev_ua-123456789-149121 |
| institution | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| issn | 1815-0659 |
| language | English |
| last_indexed | 2025-12-07T13:35:21Z |
| publishDate | 2009 |
| publisher | Інститут математики НАН України |
| record_format | dspace |
| spelling | Alekseevsky, D.V. Nikonorov, Y.G. 2019-02-19T17:31:47Z 2019-02-19T17:31:47Z 2009 Compact Riemannian Manifolds with Homogeneous Geodesics / D.V. Alekseevsky, Y.G. Nikonorov // Symmetry, Integrability and Geometry: Methods and Applications. — 2009. — Т. 5. — Бібліогр.: 28 назв. — англ. 1815-0659 2000 Mathematics Subject Classification: 53C20; 53C25; 53C35 https://nasplib.isofts.kiev.ua/handle/123456789/149121 A homogeneous Riemannian space (M = G/H,g) is called a geodesic orbit space (shortly, GO-space) if any geodesic is an orbit of one-parameter subgroup of the isometry group G. We study the structure of compact GO-spaces and give some sufficient conditions for existence and non-existence of an invariant metric g with homogeneous geodesics on a homogeneous space of a compact Lie group G. We give a classification of compact simply connected GO-spaces (M = G/H,g) of positive Euler characteristic. If the group G is simple and the metric g does not come from a bi-invariant metric of G, then M is one of the flag manifolds M₁ = SO(2n+1)/U(n) or M₂ = Sp(n)/U(1)·Sp(n–1) and g is any invariant metric on M which depends on two real parameters. In both cases, there exists unique (up to a scaling) symmetric metric g₀ such that (M,g0) is the symmetric space M = SO(2n+2)/U(n+1) or, respectively, CP²n⁻¹. The manifolds M₁, M₂ are weakly symmetric spaces. This paper is a contribution to the Special Issue “Elie Cartan and Differential Geometry”. The first author was partially supported by the Royal Society (Travel Grant 2007/R3). The second author was partially supported by the State Maintenance Program for the Leading Scientific Schools of the Russian Federation (grant NSH-5682.2008.1). We are grateful to all referees, whose comments and suggestions permit us to improve the presentation of this article. en Інститут математики НАН України Symmetry, Integrability and Geometry: Methods and Applications Compact Riemannian Manifolds with Homogeneous Geodesics Article published earlier |
| spellingShingle | Compact Riemannian Manifolds with Homogeneous Geodesics Alekseevsky, D.V. Nikonorov, Y.G. |
| title | Compact Riemannian Manifolds with Homogeneous Geodesics |
| title_full | Compact Riemannian Manifolds with Homogeneous Geodesics |
| title_fullStr | Compact Riemannian Manifolds with Homogeneous Geodesics |
| title_full_unstemmed | Compact Riemannian Manifolds with Homogeneous Geodesics |
| title_short | Compact Riemannian Manifolds with Homogeneous Geodesics |
| title_sort | compact riemannian manifolds with homogeneous geodesics |
| url | https://nasplib.isofts.kiev.ua/handle/123456789/149121 |
| work_keys_str_mv | AT alekseevskydv compactriemannianmanifoldswithhomogeneousgeodesics AT nikonorovyg compactriemannianmanifoldswithhomogeneousgeodesics |