Riemann Zeta - Cones PDF
Riemann Zeta - Cones PDF
Riemann Zeta - Cones PDF
Let K be a local field, i.e. a nondiscrete locally compact field. The action
of K = GL1 (K) on the additive group K by multiplication,
(1)
(, x) x
K , x K ,
together with the uniqueness, up to scale, of the Haar measure of the additive
group K, yield a homomorphism,
(2)
a K |a| R+ ,
1
Mod(K) = {|| R+ ; K }
is a closed subgroup of R+ .
The fields R, C and H (of quaternions) are the only ones with Mod(K) =
R = {x K ; |x| 1} ,
Mod(K) = q Z R+ .
() = q ,
Mod(K) , 6= {1} .
K 0 = {x Kun ; (x) = x
} ,
with rather obvious notations so that q is the of (6). Let then WK be the
subgroup of Gal(Kab : K) whose elements induce on Kun an integral power
of the Frobenius automorphism. One endows W K with the locally compact
topology dictated by the exact sequence of groups,
(10)
and the main result of local class field theory asserts the existence of a
canonical isomorphism,
(11)
WK K ,
A = Mn (D) ,
Br(K) Q/Z ,
of the Brauer group of classes of central simple algebras over K (with tensor
product as the group law), with the group Q/Z of roots of 1 in C.
All the above discussion was under the assumption that K is non Archimedian. For Archimedian fields R and C the same questions have an idiotically simple answer. Since C is algebraically closed one has K = K and the
whole picture collapses. For K = R the only non trivial value of the Hasse
invariant is
(14)
(H) = 1 .
A=
kv ,
res
where the product is the restricted product of the local fields k v labelled by
the places of k.
When the characteristic of k is p > 1 so that k is a function field over
Fq , one has
k kun kab ksep k ,
(16)
() = q
The main theorem of global class field theory asserts the existence of a
canonical isomorphism,
(18)
Mod(M ) R+ ,
S(M ) =
Spec( ) R+
(x x) 0
x M , (1) = 1 ,
= S S ,
hx, yi = (y x) , x, y M .
In the case of local fields the module was a group homomorphism ((2)) from
K to R+ . The counterpart for factors is the group homomorphism, ([6])
(25)
from the additive group R viewed as the dual of R + for the pairing,
(26)
(, t) it
R+ , t R ,
Mod(M ) R+ ,
is the flow of weights ([25] [15] [8]) of M . It is obtained from the module
as the dual action of R+ on the abelian algebra,
(28)
C = Center of M >/ R ,
(30)
The condition (29) is obviously what one would expect for an approximately
finite dimensional algebra. Condition (30) is similar to amenability for discrete groups and the implication (30) (29) is a very powerful tool.
We refer to [5] [15] [12] for (C) and we just describe the actual construction of the central simple algebra M associated to a given virtual subgroup,
R+ .
(31)
M (C) ,
of all operators in Hilbert space. The second factor is the unique approximately finite dimensional factor of type II . It is
(33)
R0,1 = R M (C) ,
0 (xy) = 0 (yx)
x, y R .
( (a)) = (a)
a Domain , R+ .
Let first R+ be an ordinary closed subgroup of R+ . Then the corresponding factor R with modulo is given by the equality:
(36)
R = {x R0,1 ; (x) = x
} ,
an abelian von Neumann algebra A, and the formula (36) easily extends to,
(37)
R = {x R0,1 A ; ( ) x = x
R+ } .
(This reduces to (36) for the action of R + on the algebra A = L (X) where
X is the homogeneous space X = R+ /.)
7
The pair (R0,1 , ) arises very naturally in geometry from the geodesic
flow of a compact Riemann surface (of genus > 1). Let V = S be the
unit cosphere bundle of such a surface , and F be the stable foliation of
the geodesic flow. The latter defines a one parameter group of automorphisms of the foliated manifold (V, F ) and thus a one parameter group of
automorphisms of the von Neumann algebra L (V, F ).
This algebra is easy to describe, its elements are random operators T =
(Tf ), i.e. bounded measurable families of operators T f parametrized by the
leaves f of the foliation. For each leaf f the operator T f acts in the Hilbert
space L2 (f ) of square integrable densities on the manifold f . Two random
operators are identified if they are equal for almost all leaves f (i.e. a set of
leaves whose union in V is negligible). The algebraic operations of sum and
product are given by,
(38)
and that the geodesic flow satisfies (35). Indeed the foliation (V, F )
admits up to scale a unique transverse measure and the trace is given
(cf. [4]) by the formal expression,
(40)
(T ) =
Trace(Tf ) d(f ) ,
since the geodesic flow satisfies () = one obtains (35) from simple geometric considerations. The formula (37) shows that most approximately finite dimensional factors already arise from foliations, for instance the unique
approximately finite dimensional factor R such that,
(41)
Mod(R ) = R+ ,
one proves (cf. [4]) that for any codimension one foliation F of a compact
manifold V with non vanishing Godbillon-Vey class one has,
(43)
s .
Equivalently the orbits of the left action of 0 on /0 are all finite. One
defines the Hecke algebra,
(2)
H(, 0 ) ,
Hk (, 0 ) = H(, 0 ) Z k ,
1
0
b
; a Q+ , b Q
a
0 =
1
0
n
; nZ
1
`2 (0 \) ,
kk2 =
|()|2 .
0 \
0 \/0
t R
t R.
For fixed the KMS states form a Choquet simplex and thus decompose
uniquely as a statistical superposition from the pure phases given by the
extreme points. For interesting systems with nontrivial interaction, one
expects in general that for large temperature T , (i.e. small since = T1
up to a conversion factor) the disorder will be predominant so that there will
exist only one KMS state. For low enough temperatures some order should
set in and allow for the coexistence of distinct thermodynamical phases so
that the simplex K of KMS states should be non trivial. A given symmetry
group G of the system will necessarily act trivially on K for large T since
K is a point, but acts in general non trivially on K for small T so that
it is no longer a symmetry of a given pure phase. This phenomenon of
spontaneous symmetry breaking as well as the very particular properties of
the critical temperature Tc at the boundary of the two regions are corner
stones of statistical mechanics.
In our case we just let A be the C algebra which is the norm closure of
HC (, 0 ) in the algebra of operators in `2 (0 \). We let t Aut(A) be
the unique extension of the automorphisms t of (8).
For = 1 it is tautological that is a KMS state since we obtained
t precisely this way ([24]). One proves ([3]) that for any 1 (i.e. for
T = 1) there exists one and only one KMS state.
The compact group G,
(11)
G = CQ /DQ ,
HQ (, 0 ) A .
(13)
i.e. the von Neumann algebra crossed product of the L functions on Adeles
of Q by the action of Q by multiplication.
The one parameter group of automorphisms, Aut(R0,1 ), is obtained
as the restriction to,
DQ = R+ ,
(15)
(g, x) g x
g CQ , x AQ /Q ,
12
Global fields k provide a natural context for the Riemann Hypothesis on the
zeros of the zeta function and its generalization to Hecke L-functions. When
the characteristic of k is non zero this conjecture was proved by A. Weil.
His proof relies on the following dictionary (put in modern language) which
provides a geometric meaning, in terms of algebraic geometry over finite
fields, to the function theoretic properties of the zeta functions. Recall that
k is a function field over a curve defined over F q ,
Algebraic Geometry
Function Theory
Eigenvalues of action of
1 (, Q )
Frobenius on Het
`
Zeros of
Poincare duality in
`-adic cohomology
Functional equation
Explicit formulas
Castelnuovo positivity
Riemann Hypothesis
We shall describe a third column in this dictionary, which will make sense
for any global field. It is based on the geometry of the Adele class space,
(1)
X = A/k , A = Adeles of k .
This space is of the same nature as the space of leaves of the horocycle
foliation (section I) and the same geometry will be used to analyse it.
Our spectral interpretation of the zeros of zeta involves Hilbert space.
The reasons why Hilbert space (apparently invented by Hilbert for this purpose) should be involved are manifold, let us mention three,
13
(A) Let N (E) be the number of zeros of the Riemann zeta function satisfying
0 < Im < E, then ([22])
(2)
hN (E)i =
E
2
log
E
7
1 + + o(1) ,
2
8
1
Nosc (E) = Im log
1
+ iE
2
The numbers xj = hN (j )i where n is the imaginary part of the nth zero are
of average density one and behave like the eigenvalues of a random Hermitian
matrix. This was discovered by H. Montgomery [18] who conjectured (and
proved for suitable test functions) that when M , with , > 0,
(5) # {(i, j) {1, . . . , M }2 ; xi xj [, ]} M
sin u
u
2
du
which is exactly what happens in the Gaussian Unitary Ensemble. Numerical tests by A. Odlyzko [20] and further theoretical work by Katz-Sarnak
[17] and J. Keating give overwhelming evidence that zeros of zeta should be
the eigenvalues of a hermitian matrix.
(B) The equivalence between RH and the positivity of the Weil distribution
on the Idele class group Ck shows that Hilbert space is implicitly present.
(C) The deep arithmetic significance of the work of A. Selberg on the spectral
analysis of the Laplacian on L2 (G/) where is an arithmetic subgroup of
a semi simple Lie group G.
Direct atempts (cf. [2]) to construct the Polya-Hilbert space giving a
spectral realization of the zeros of using quantum mechanics, meet the
following sign problem: Let H be the Hamiltonian of the quantum mechanical system obtained by quantizing the classical system,
(6)
(X, Ft )
Nosc (E)
1
1
1 X X
sin(T# m E)
m = 1 m 2sh m
2
where the are the periodic orbits of the flow F , the T # are their periods
and the the unstability exponents of these orbits.
One can compare ([2]) (8) with the equally heuristic asymptotic expansion of (4) using the Euler product of which gives, using log(1 x) =
X
xm
,
m
m=1
(9)
1 1
1 XX
sin((log p) m E) .
p m=1 m pm/2
Comparing (8) and (9) one gets precious information on the hypothetical
Riemann flow of M. Berry. The periodic orbits should be labelled by the
primes p, the periods should be the log p as well as the unstability exponents
p . Also, in order to avoid duplication of orbits, the flow should not be time
reversal symmetric, i.e. non isomorphic to the time reversed:
(10)
(X, Ft ) .
L(,) = 0
b
h(,
) b
h(0) b
h(1) =
XZ
v
0
kv
h(u1 )
d u,
|1 u|
b is its Fourier
where h is a test function on the Idele class group C k , h
transform,
(12)
b
h(x,
z) =
R0
(14)
0 H 0 H1 H 0 .
The example we have in mind for (14) is the assembled Euler complex for
a Riemann surface, where H0 is the codimension 2 subspace of differential
forms of even degree orthogonal to harmonic forms, where H 1 is the space
of 1-forms and where T = d + d is the sum of the de Rham coboundary
with its adjoint d .
Since we want to obtain the spectral interpretation not only for zeta
functions but for all L-functions with Grossencharakter we do not expect to
have only an action of Z for char(k) > 0 corresponding to the Frobenius, or
of the group R+ if char(k) = 0, but to have the equivariance of (14) with
respect to a natural action of the Idele class group C k = GL1 (A)/k .
Let X = A/k be the Adele class space. Our basic idea is to take for H 0
a suitable completion of the codimension 2 subspace of functions on X such
that,
(15)
f (0) = 0 ,
f dx = 0 ,
while H1 = L2 (Ck ) and T is the restriction map coming from the inclusion
Ck X, multiplied by |a|1/2 ,
(16)
(U (g)f )(x) = f (g 1 x)
g Ck
f fq , fq (x) = f (xq) ,
16
q k .
Here the natural function space is the Bruhat-Schwarz space S(A) and by
(15) the codimension 2 subspace,
(19)
f dx = 0
T (f )(a) = |a|1/2
f (aq)
a Ck .
q k
kk2 =
Trace(W (h)) =
(
L , 1 + = 0
2
i R/N
b
h(, )
R
b is defined in
where N = Mod(k), W (h) = W (g) h(g) d g, h S(Ck ), h
(12) and the multiplicity is counted as in the theorem.
17
kf k2 =
(24)
(up to normalization) between the additive Haar measure and the multiplicative one. In the global case one has,
dx = lim |x|1+ d x ,
(25)
0
and (23) ignores the divergent normalization constant which plays no role in
the computation of traces or of adjoint operators. The exponent 12 in (16)
turns T into an isometry,
T : L2 (X)0 L2 (Ck ) .
(26)
(F f )(x) =
f (y) (xy) dy
A
f S(A)0
( )(a) = (a1 )
a Ck .
The Poisson formula means exactly that T commutes with the operation.
This is just a reformulation of the work of Tate and Iwasawa on the proof
of the functional equation, but we shall now see that if we follow the proof
by Atiyah-Bott ([1]) of the Lefschetz formula we do obtain a clear geometric
meaning for the Weil distribution. One can of course as in [10] define inner
18
(U )(x) = ((x))
x .
One has Trace (U ) = k(x, x) dx, where k(x, y) dy is the Schwarz kernel
associated to U , i.e. the distribution on such that,
(30)
(U )(x) =
k(x, y) (y) dy .
k(x, y) = (y (x)) ,
Trace (U ) =
(x) = x
1
,
|1 0 (x)|
where 0 is the Jacobian of and | | stands for the absolute value of the
determinant.
With more work ([11]) one obtains a similar formula for the distributional
trace of the action of a flow,
(33)
x , t R.
Trace (U (h)) =
XZ
h(u)
d u ,
|1 (Fu ) |
where U (h) = h(t) U (t) dt, h is a test function on R, the labels the
periodic orbits of the flow, including the fixed points, I is the corresponding
isotropy subgroup, and (Fu ) is the tangent map to Fu on the transverse
space to the orbits, and finally d u is the unique Haar measure on I which
is of covolume 1 in (R, dt).
19
Now it is truly remarkable that when one analyzes the periodic orbits of
the action of Ck on X one finds that not only it qualifies as a Riemann flow
in the above sense, but that (34) becomes,
(35)
Trace (U (h)) =
XZ
v
kv
h(u1 )
d u.
|1 u|
d u log
for ,
1|u|
(37)
Trace (U (h)) 0
h , h(1) = 0 , h(u) 0
u C
simplest case of the action of K on L2 (K) for a local field K. In this local
case one lets,
(39)
R = Pb P , R+ ,
{ L2 (K) ; (x) = 0
x, |x| > } ,
h(u1 )
d u + o(1)
|1 u|
R
CS =
kv /OS ,
vS
XS =
kv /OS .
vS
The Hilbert space L2 (XS ), its Fourier transform F and the orthogonal projection P , Pb = F P F 1 continue to make sense, with
(44)
Im P = { L2 (XS ) ; (x) = 0
x , |x| > } .
X Z
vS
h(u1 )
d u + o(1)
|1 u|
kv
where the notations are as above and the finite values 0 depend on the additive character of kv defining the Fourier transform F . When Char (k) = 0
the projectors P , Pb commute on L2 for large enough so that one can
replace R by the orthogonal projection Q on Im P Im Pb . The situation for Char (k) = 0 is more delicate since P and Pb do not commute
(for large) even in the local Archimedian case. But fortunately [21] these
operators commute with a specific second order differential operator, whose
eigenfunctions, the Prolate Spheroidal Wave functions provide the right filtration Q . This allows to replace R by Q and to state the global trace
formula
(45)
XZ
v
kv
h(u1 )
d u + o(1) .
|1 u|
Our final result is that the validity of this trace formula implies (in fact
is equivalent to) the positivity of the Weil distribution, i.e. RH for all Lfunctions with Grossencharakter. Moreover the filtration by Q allows to
define the Adelic cohomology and to complete the dictionary between the
function theory and the geometry of the Adele class space.
Function Theory
Geometry
Eigenvalues of action of C k
on Adelic cohomology
Functional Equation
operation
Explicit formula
Lefschetz formula
RH
Trace formula
References
[1] M.F. Atiyah and R. Bott, A Lefschetz fixed point formula for elliptic
complexes: I, Annals of Math, 86 (1967), 374-407.
22
[17] N. Katz and P. Sarnak, Zeros of zeta functions, their spacings and
spectral nature, (1997), to appear.
[18] H. Montgomery, The pair correlation of zeros of the zeta function, Analytic Number Theory, AMS (1973).
[19] M.L. Mehta, Random matrices, Academic Press,(1991).
[20] A. Odlyzko, On the distribution of spacings between zeros of zeta functions, Math. Comp. 48 (1987), 273-308.
[21] D. Slepian and H. Pollak, Prolate spheroidal wave functions, Fourier
analysis and uncertainty I, Bell Syst. Tech. J. 40 (1961).
[22] B. Riemann, Mathematical Werke, Dover, New York (1953).
[23] A.Selberg, Collected papers, Springer (1989).
[24] M. Takesaki, Tomitas theory of modular Hilbert algebras and its applications, Lecture Notes in Math. bf 128, Springer (1989).
[25] M. Takesaki, Duality for crossed products and the structure of von
Neumann algebras of type III, Acta Math. 131 (1973), 249-310.
[26] A. Weil, Basic Number Theory, Springer, New York (1974).
[27] A. Weil, Sur les formules explicites de la theorie des nombres, Izv. Mat.
Nauk., (Ser. Mat.) 36, 3-18.
[28] A. Weil, Sur la theorie du corps de classes, J. Math. Soc. Japan, 3,
(1951).
[29] D. Zagier, Eisenstein series and the Riemannian zeta function, Automorphic Forms, Representation Theory and Arithmetic, Tata, Bombay
(1979), 275-301.
24