WorksMathematics
Boundaries of groups with compact-by-amenable decomposition
Abstract
We look at the Liouville property (triviality of harmonic functions) of topological groups G relative to a compact subgroup K. We show that Gelfand pairs have this property. Furthermore, we show that this property holds if and only if G admits a product decomposition G = K S for an amenable subgroup S. We identify the Poisson and Furstenberg boundaries of these groups. This applies in particular to reductive algebraic groups over local fields.
Introduction
Let be a probability measure on a locally compact group , and let be a compact subgroup. We say has the Liouville property relative to if the following condition holds.
For any bounded left--invariant function on , the equality implies that is constant.
Similarly, we say that has the Liouville property if there is such an admissible probability measure . We show that this property holds whenever is a Gelfand pair, in the following sense.
Theorem A. Let be a locally compact second countable group, and let be a compact subgroup such that the associated Hecke algebra is commutative. Then all admissible, bi--invariant probability measures have the Liouville property relative to .
For the above recovers the Choquet–Deny theorem [5]. Our direct motivation comes from Furstenberg’s celebrated work [7] on the boundaries of semisimple Lie groups , in which a result of the above type for the maximal compact subgroup plays a crucial role in the identification of the Poisson boundary of through analysis of random walks on the symmetric space .
Indeed, the relative Liouville property allows us to get a strong structural property for locally compact groups and their Poisson boundaries closely analogous to semisimple Lie groups. More precisely, we have the following results.
Theorem B. Let be a locally compact second countable group, and let be a compact subgroup.
(1) If holds for some closed amenable subgroup , then there exists an admissible probability measure on that has the Liouville property relative to .
(2) If there exists an admissible probability measure on that has the Liouville property relative to , then for a closed, amenable subgroup , and the Poisson boundary of is for some -quasi-invariant probability measure .
For the above recovers the main result of [17], which states that amenability of is characterized by the triviality of some Poisson boundary of . More recently, Monod proved that Gelfand pairs admit an Iwasawa decomposition [13]. The combination of Theorems A and B gives a random-walk proof of his result.
Finally, we also have a description of the Furstenberg boundary for groups satisfying the conclusion of the above theorem, as follows.
Theorem C. Let be a locally compact second countable group such that some Poisson boundary of is isomorphic to where is a cocompact, amenable subgroup and a -quasi-invariant probability measure. Then is the Furstenberg boundary of .
A prominent class of locally compact groups satisfying the assumption of Theorem A comes from connected reductive algebraic groups over non-Archimedean local fields , by taking the group of -points . Using Theorems B and C, we identify the Furstenberg boundary of such groups , see § 7.1.
As a concrete example, we can consider the determinant-one matrices over the -adic numbers and identify its Furstenberg boundary as , where is the Borel subgroup of upper triangular matrices. This is a direct analogue of the well-known result in the context of semisimple Lie groups established in Furstenberg’s foundational work [Fur73].
While this paper was in the final stage of preparation, Monod announced [14]. His work contains closely overlapping results particularly on the Furstenberg boundary of groups as above, although the proof strategy seems somewhat different from ours.
Acknowledgements. Part of this work was carried out during the first-named author’s stay at Institut Henri Poincaré for the Operator algebras: approximation, rigidity and dynamics program in the fall of 2026. We thank Sven Raum for a stimulating discussion. M.Y.’s research was funded, in part, by The Research Council of Norway [project 324944].
Declaration on AI usage. After developing the main ideas for this paper, we used a large language model for part of the brainstorming process to refine formulations.
Preliminaries
Let be a locally compact, second countable group. Throughout the paper we fix a left Haar measure on , denoted .
Poisson boundary
Poisson boundary. Let be a probability measure on . We say that:
is spread out if for some , that is, some positive convolution power of is absolutely continuous relative to the Haar measure on .
is adapted if generates as a semigroup, that is, .
is admissible if is spread out and adapted.
Let be a standard Borel -space. For a probability measure on , our convention of the convolution is , which is a new probability measure on . We say is -stationary if . By [15 Lemma 1.1], if is admissible, then is -quasi-invariant, that is and have the same null sets on for all .
Therefore given an admissible probability measure on , for every standard Borel -space where is a -stationary measure, we have a well-defined map between von Neumann algebras
Automatically, is linear, unital, positive, normal, contractive and -equivariant. The image of can be as small as , and by evaluating indicator functions one finds that is isometric if and only if holds for all Borel set satisfying .
By [8 Theorem 3.1], there exists a standard Borel -space , called the Poisson boundary of , which has the property that is isometric and the image of is the set of the -harmonic functions . Here is an established but unfortunate notation, because reverses the order of convolution, where are admissible measures on .
There is a separable, metrizable topological -space that recovers by taking its Borel -algebra, see for example [22 Theorem 2.1.19].
Furstenberg boundary
Let be a compact space with a continuous -action. We say that:
is minimal if it has no closed -invariant subspaces, except and .
is strongly proximal if for every probability measure on there is a point mass contained in the weak-closure of the -orbit .
is a -boundary if it is minimal and strongly proximal.
By [Fur73, Proposition 4.6] there exists a Furstenberg boundary or universal boundary , that is an up to isomorphism unique -boundary with the universal property that for any -boundary there exists a -equivariant surjection .
Liouville measures from Gelfand pairs
In this section we prove Theorem A. We recall that is a Gelfand pair, that means the Hecke algebra is commutative, if and only if all bi--invariant measures on commute. For any the measure is an example of a bi--invariant measure, where denotes the Haar probability measure on the compact subgroup .
The only observation we need to make is that whenever is a bi--invariant probability measure and is a -harmonic function, also are -harmonic for all :
Also, for all bi--invariant on we may write
Proof of Theorem A. Since is spread out, there exists for some such that . So any -harmonic has a continuous representative
Also, we recall that -harmonic is automatically right--invariant because .
Consider
where
is convex and bounded. Because and are weak-continuous, the latter being the Banach-space adjoint of a contraction on , we find that is weak-closed. By the Banach–Alaoglu theorem is weak-compact. So by the Krein–Milman theorem, it suffices to show the extreme points of are constant.
Consider an extreme point . For and we have
Because , and limits of convex combinations remain in , extremality implies
Note that for this holds trivially because . By continuity, we may evaluate both sides at and use the bi--invariance of in the form to obtain that for all . This implies both and must be disjoint from . Since is adapted, this shows is constant on all of . □
Remark 3.1. In the case when is a compact, open subgroup of , the countable set measures over double-cosets forms a hypergroup under convolution. Combining results of [20] and [10], commutative hypergroups with dimension function can be shown to be weakly amenable in the sense of [11], which is witnessed by an ergodic probability measure. By an adaptation of this proof, on the group one can show that in fact all absolutely continuous, bi--invariant probability measures on with full support have the Liouville property relative to .
As we remarked in the introduction, connected reductive algebraic groups over totally disconnected local fields satisfy the assumption for Theorem A, see § 7.1.
Liouville measure from Iwasawa decomposition
In this section, we prove the part of Theorem B that constructs an admissible probability measure with the Liouville property from a compact-by-amenable decomposition. We closely follow the proof of [KV83, Theorem 4.3]. The only additional ingredient is the following:
Lemma 4.1. Assume with compact and closed and amenable. Then for every compact subset and there exists a probability measure on such that
Here is always the total variation norm for measures on .
Proof of Lemma 4.1. The set is compact. By Reiter’s property, using the amenability of , there exists a probability density on such that satisfies
for all . Here is the Haar measure on .
Now for all we have a decomposition where and , and by the -invariance of the Haar measure on . Hence
□
To construct a Liouville measure relative to we only have to construct such that
This is because for all left--invariant, -harmonic we have . So for all bounded, left--invariant, -harmonic functions , we get an estimate
for all , from (1).
There is a generic argument from the proof of [KV83, Theorem 4.3] which shows how to combine probability measures obtained from Lemma 4.1 into a single probability measure
for certain weights with such that (1) holds. Choosing with strictly positive density relative to gives full support, and all from Lemma 4.1 can be chosen absolutely continuous by convolving on the right with some fixed probability .
Proof of part (1) in Theorem B. We pick compact sets with . Pick with and decreasing to and such that for all . Now, for any , given consider the finite set of probability measures
We may select (e.g. among the ) a large enough compact set such that for all it holds . Now finally, we pick using Lemma 4.1 a probability measure such that
Now for write and for words . Then
Here the first summand contains only words where all the letters are less than . The second summand contains the rest of all words, where there is some letter that is at least and we choose the first such occurrence. So there where and is some probability measure made of the remaining factors . For all it holds
Here is the only place we use the choice of , to ensure the first sum over the words is small, where we have used the crude bound for the difference of probability measures.
For the integral expression, we have
since the case allows us to replace by with an -error, while the case has small -measure bounded by again. So we have shown for every . □
Poisson boundary from a Liouville measure
We are now ready to give applications to boundary theory. In this section, we prove the part of Theorem B that constructs the Iwasawa decomposition and Poisson boundary.
Proof of part (2) in Theorem B. Let . It is closed, and -invariant because was -quasi-invariant. Because is a null set, we have a -equivariant isomorphism of measure spaces
Similarly as in [7 Lemma 5.1], for every bounded continuous , we consider the function
It is bounded, and -invariant by right invariance of (compact groups are unimodular). By continuity of the action and dominated convergence, is continuous. Then is a bounded, left--invariant -harmonic function. By the Liouville property of it is constant, say . Because is linear, unital and injective, holds -a.e. So the set is an open null set. By definition, it must be disjoint from . This shows
Fix . By -invariance of we have . Suppose there is with . Then the -orbits and are disjoint compact sets, and we find with and . Because the topology on is metrizable, one does not even need Urysohn’s lemma but can construct concretely. In any case and , contradicts (2), so . Since is -invariant, we have , so we have shown
Define
For all we know that , so for some , that is , i.e. . That means . Importantly, the compact is Hausdorff because is closed. Now the orbit map
is a well-defined -equivariant bijection. It is continuous by the universal property of the quotient topology, and it is actually a homeomorphism because continuous bijections from compact spaces onto Hausdorff spaces are closed. So we get a -equivariant isomorphism of measure spaces
By equivariance, again is -quasi-invariant.
The action is Zimmer-amenable [21 Theorem 5.2]. The Zimmer-amenability does not feel null sets or equivariant isomorphisms of measure spaces, so is Zimmer-amenable. By [22 Proposition 4.3.2] this is equivalent to being amenable. □
The following is essentially the computation leading up to [7 Theorem 5.4].
Corollary 5.1. In the setting of Theorem B part (2), if in addition is left--invariant, then is the unique -invariant probability measure on . Namely, .
Proof. If is left--invariant, then is -invariant, so also is -invariant. Let be any -invariant probability measure on and . By -invariance and Fubini
Every is of the form with , because . By right--invariance of (compact groups are unimodular) we have . Hence .
Furstenberg boundary from homogeneous Poisson boundary
Let us next prove Theorem C, which gives an application to the Furstenberg boundary. From [6 Proposition 6.4] we know that has a cocompact amenable subgroup if and only if acts transitively on the Furstenberg boundary. By [16 Proposition 10 and the preceding discussion], in the setting of Theorem C we know already that the Furstenberg boundary of must be of the form for some maximal cocompact, relatively amenable subgroup that is unique up to conjugation.
In order to prove Theorem C, namely that can be taken to be , we have to show:
Proposition 6.1. Suppose some Poisson boundary of is isomorphic to with some closed, cocompact subgroup and -quasi-invariant . Then is a -boundary.
Theorem C will follow from this proposition: One may use that is amenable, to invoke the fixed-point property and find some -invariant probability measure on . Since is compact, the -orbit of already contains a point mass, which forces for some . Then the map , is continuous and -equivariant. By the universal property we have a -equivariant surjection in the other direction. But on every -boundary , the identity map is the unique continuous, -equivariant self map. So and are inverses of each other, and .
Lemma 6.2. Let be -quasi-invariant. If then : .
Proof. Let and Borel. If , then by Fubini and -quasi-invariance
This shows the claim for some . By -transitivity and we get it for all .
In particular, Lemma 6.2 reminds that quasi-invariant have full support: One just notes that for all nonempty, open subsets again is nonempty, open. And nonempty, open subsets of have positive Haar measure, so would be a contradiction. So because its complement is .
Proof of Proposition 6.1. Write . Being -transitive, is minimal, so we only have to show strong proximality. In fact, showing some point-measure is in the -orbit of a given probability measure on will imply that all point masses are contained. So we can assume and are given, and we have to construct a net such that in the weak* sense.
Having that is a Poisson boundary, the Poisson transform is isometric, so for every open neighborhood of we have . Since is metrizable, we can choose a sequence of neighborhoods with radius for some fixed metric, and elements such that . Then by the Borel-Cantelli lemma the following is -null:
So for -a.e. , namely for all , we know that . This already means we have weak*ly. Using the quasi-invariance, this computation transfers to all measures on the homogeneous space, after possibly applying a shift:
Lemma 6.2 gives for all . By Fubini
So there exists with . That means for -almost every . This shows in the weak* sense using dominated convergence theorem. □
In practice, one may read the combination of Theorems B and C as a way to compute Poisson and Furstenberg boundaries for compact-by-amenable groups. However, starting from a factorization with closed, amenable one obtains a potentially larger amenable subgroup with where identifies the boundaries. The following makes precise how much larger the subgroup becomes:
Corollary 6.3. Let be locally compact, second countable, with compact subgroup . If with closed, amenable subgroup , then the Furstenberg boundary of can be computed:
Here means “the smallest closed subgroup of containing ”.
Proof. We can find a cocompact amenable subgroup such that where , and such that is maximal in the sense that if any other amenable contains , then contains .
We have . The inclusion “” holds because . The inclusion “” holds because . Namely, if with , , then already .
We have . The inclusion “” holds because and because closed subgroups of amenable groups are amenable. The inclusion “” holds because , contains . So if is amenable then , so . □
Example 6.4. One may consider . It contains copies of the compact circle group
and the upper half-plane
which is a semi-direct product of abelian groups, so solvable and hence amenable. One can write down a unique factorization with .
Now one computes the Furstenberg boundary as where : In fact, for one checks that is closed and still solvable, while for all other it is not difficult to write down a formula that expresses a generic element of in terms of only , and elements of , but is famously non-amenable.
Examples
Reductive algebraic groups over local fields
7.1. Reductive algebraic groups over local fields. Let be a connected reductive algebraic group over an infinite non-Archimedean local field . We consider the locally compact group . Then, by the Bruhat–Tits theory, we obtain a compact open subgroup as the stabilizer of an appropriately chosen point in the enlarged Bruhat–Tits building for [3, 19]. In the case of split semisimple groups, we can take for the ring of integers .
By a result of Satake [18] (see also [4 Theorem 4.1]), the spherical Hecke algebra is commutative.
Proposition 7.1. Let be the minimal parabolic subgroup defined by a choice of root data for . Then the subgroup is maximally amenable in .
Proof. The groups and become part of a Tits system, see for example [1 paragraph IV.14.15 and Theorem V.21.15]. The intermediate subgroups are classified by subsets of the set of positive simple roots [2 Théorème IV.2.5.3]. If is not empty, the corresponding parabolic subgroup is indeed nonamenable, since it contains a closed subgroup locally isomorphic to in its Levi factor.
From any initial cocompact amenable subgroup , [Fur73, Proposition 4.4] determines that all -boundaries are of the form for some intermediate amenable subgroup .
Corollary 7.2. In the above setting, the Furstenberg boundary of is .
Type A groups over p-adic numbers
For the type A groups , over the p-adic numbers there are more elementary ways to compute the Furstenberg boundary.
It will be enough to show that is strongly proximal, with the subgroup of upper triangular matrices of . Because when is a -boundary, from the universal property of , we obtain an equivariant surjection . When for then from the equivariance we get also for with . But the upper triangular subgroup does not contain any conjugate except itself, so implies , and thus , so .
To show the strong proximality of , one may consider the “large” open subset
One can show that for every probability measure on one may apply a shift , such that is supported on . Furthermore, one can show the sequence converges weak to the point mass , where can be any diagonal matrix of the form
For example, when the group acts on the projective line via the Moebius transform
Here , and since is uncountable, every probability measure must vanish on some , so one can translate its support into . Then positive powers of
contract to .
References
References
- [1]Borel, A. Linear algebraic groups. Second. Vol. 126. Graduate Texts in Mathematics. Springer-Verlag, New York, 1991, pp. xii+288.
- [2]Bourbaki, N. Éléments de mathématique. Groupes et algèbres de Lie. Chapitre 4 à 6. Reprint of the 1981 (Masson) edition. Springer, Berlin-Heidelberg, 2007.
- [3]Bruhat, F. and Tits, J. “Groupes réductifs sur un corps local”. In: Inst. Hautes Études Sci. Publ. Math. 41 (1972), pp. 5–251.
- [4]Cartier, P. “Representations of p-adic groups: a survey”. In: Automorphic forms, representations and L-functions (Proc. Sympos. Pure Math., Oregon State Univ., Corvallis, Ore., 1977), Part 1. Vol. XXXIII. Proc. Sympos. Pure Math. Amer. Math. Soc., Providence, RI, 1979, pp. 111–155.
- [5]Choquet, G. and Deny, J. “Sur l’équation de convolution μ = μ ∗ σ”. In: C. R. Acad. Sci. Paris 250 (1960), pp. 799–801.
- [6]Caprace, P.-E., Kalantar, M., and Monod, N. “A type I conjecture and boundary representations of hyperbolic groups”. In: Proc. Lond. Math. Soc. (3) 127.2 (2023), pp. 447–486.arxiv.org/abs/2110.00190
- [7]Furstenberg, H. “A Poisson formula for semi-simple Lie groups”. In: Ann. of Math. (2) 77 (1963), pp. 335–386.DOI
- [8]Furstenberg, H. “Random walks and discrete subgroups of Lie groups”. In: Advances in Probability and Related Topics, Vol. 1. Dekker, New York, 1971, pp. 1–63.
- [10]Hayashi, T. “Harmonic function spaces of probability measures on fusion algebras”. In: Publ. Res. Inst. Math. Sci. 36.2 (2000), pp. 231–252.DOI
- [11]Hiai, F. and Izumi, M. “Amenability and strong amenability for fusion algebras with applications to subfactor theory”. In: Internat. J. Math. 9.6 (1998), pp. 669–722.DOI
- [13]Monod, N. “Gelfand pairs admit an Iwasawa decomposition”. In: Math. Ann. 378.1-2 (2020), pp. 605–611.arxiv.org/abs/1902.09497
- [14]Monod, N. Invariant measures, Gelfand pairs and integral representations. preprint. 2026.arxiv.org/abs/2611.14085
- [15]Nevo, A. and Zimmer, R. J. “Homogenous projective factors for actions of semi-simple Lie groups”. In: Invent. Math. 138.2 (1999), pp. 229–252.DOI
- [16]Ozawa, N. Lecture on the Furstenberg boundary and C*-simplicity. lecture note. 2014.
- [17]Rosenblatt, J. “Ergodic and Mixing Random Walks on Locally Compact Groups.” In: Mathematische Annalen 257 (1981), pp. 31–42.DOI
- [18]Satake, I. “Theory of spherical functions on reductive algebraic groups over p-adic fields”. In: Inst. Hautes Études Sci. Publ. Math. 18 (1963), pp. 5–69.DOI
- [19]Tits, J. “Reductive groups over local fields”. In: Automorphic forms, representations and L-functions (Proc. Sympos. Pure Math., Oregon State Univ., Corvallis, Ore., 1977), Part 1. Vol. XXXIII. Proc. Sympos. Pure Math. Amer. Math. Soc., Providence, RI, 1979, pp. 29–69.DOI
- [20]Yamagami, S. “Notes on amenability of commutative fusion algebras”. In: Positivity 3.4 (1999), pp. 377–388.DOI
- [21]Zimmer, R. J. “Amenable ergodic group actions and an application to Poisson boundaries of random walks”. In: J. Functional Analysis 27.3 (1978), pp. 350–372.DOI
- [22]Zimmer, R. J. Ergodic theory and semisimple groups. Vol. 81. Monographs in Mathematics. Birkhäuser Verlag, Basel, 1984, pp. x+209.DOI