Eisenstein series

Eisenstein series, named after German mathematician Gotthold Eisenstein, are particular modular forms with infinite series expansions that may be written down directly. Originally defined for the modular group, Eisenstein series can be generalized in the theory of automorphic forms.

Eisenstein series for the modular group

The real part of

G 6

as a function of

q

on the unit disk. Negative numbers are black.

The imaginary part of

G 6

as a function of

q

on the unit disk.

Let $τ$ be a complex number with strictly positive imaginary part. Define the holomorphic Eisenstein series $G 2 k (τ)$ of weight $2 k$ , where $k \geq 2$ is an integer, by the following series:

G_{2k}(\tau )=\sum _{(m,n)\in \mathbb {Z} ^{2}\setminus (0,0)}{\frac {1}{(m+n\tau )^{2k}}}.

This series absolutely converges to a holomorphic function of $τ$ in the upper half-plane and its Fourier expansion given below shows that it extends to a holomorphic function at $τ = i \infty$ . It is a remarkable fact that the Eisenstein series is a modular form. Indeed, the key property is its $SL(2, ℤ)$ -invariance. Explicitly if $a, b, c, d \in ℤ$ and $ad - bc = 1$ then

G_{2k}\left({\frac {a\tau +b}{c\tau +d}}\right)=(c\tau +d)^{2k}G_{2k}(\tau )

(Proof)

{\begin{aligned}G_{2k}\left({\frac {a\tau +b}{c\tau +d}}\right)&=\sum _{(m,n)\in \mathbb {Z} ^{2}\setminus (0,0)}{\frac {1}{\left(m+n{\frac {a\tau +b}{c\tau +d}}\right)^{2k}}}\\&=\sum _{(m,n)\in \mathbb {Z} ^{2}\setminus (0,0)}{\frac {(c\tau +d)^{2k}}{(md+nb+(mc+na)\tau )^{2k}}}\\&=\sum _{\left(m',n'\right)=(m,n){\begin{pmatrix}d\ \ c\\b\ \ a\end{pmatrix}} \atop (m,n)\in \mathbb {Z} ^{2}\setminus (0,0)}{\frac {(c\tau +d)^{2k}}{\left(m'+n'\tau \right)^{2k}}}\end{aligned}}

If $ad - bc = 1$ then

{\begin{pmatrix}d&c\\b&a\end{pmatrix}}^{-1}={\begin{pmatrix}\ a&-c\\-b&\ d\end{pmatrix}}

so that

(m,n)\mapsto (m,n){\begin{pmatrix}d&c\\b&a\end{pmatrix}}

is a bijection $ℤ 2 \to ℤ 2$ , i.e.:

\sum _{\left(m',n'\right)=(m,n){\begin{pmatrix}d\ \ c\\b\ \ a\end{pmatrix}} \atop (m,n)\in \mathbb {Z} ^{2}\setminus (0,0)}{\frac {1}{\left(m'+n'\tau \right)^{2k}}}=\sum _{\left(m',n'\right)\in \mathbb {Z} ^{2}\setminus (0,0)}{\frac {1}{(m'+n'\tau )^{2k}}}=G_{2k}(\tau )

Overall, if $ad - bc = 1$ then

G_{2k}\left({\frac {a\tau +b}{c\tau +d}}\right)=(c\tau +d)^{2k}G_{2k}(\tau )

and $G 2 k$ is therefore a modular form of weight $2 k$ . Note that it is important to assume that $k \geq 2$ , otherwise it would be illegitimate to change the order of summation, and the $SL(2, ℤ)$ -invariance would not hold. In fact, there are no nontrivial modular forms of weight 2. Nevertheless, an analogue of the holomorphic Eisenstein series can be defined even for $k = 1$ , although it would only be a quasimodular form.

Relation to modular invariants

The modular invariants $g 2$ and $g 3$ of an elliptic curve are given by the first two Eisenstein series:

{\begin{aligned}g_{2}&=60G_{4}\\g_{3}&=140G_{6}.\end{aligned}}

The article on modular invariants provides expressions for these two functions in terms of theta functions.

Recurrence relation

Any holomorphic modular form for the modular group can be written as a polynomial in $G 4$ and $G 6$ . Specifically, the higher order $G 2 k$ can be written in terms of $G 4$ and $G 6$ through a recurrence relation. Let $d k = (2 k + 3) k! G 2 k + 4$ , so for example, $d 0 = 3 G 4$ and $d 1 = 5 G 6$ . Then the $d k$ satisfy the relation

\sum _{k=0}^{n}{n \choose k}d_{k}d_{n-k}={\frac {2n+9}{3n+6}}d_{n+2}

for all $n \geq 0$ . Here, $(n k)$ is the binomial coefficient.

The $d k$ occur in the series expansion for the Weierstrass's elliptic functions:

{\begin{aligned}\wp (z)&={\frac {1}{z^{2}}}+z^{2}\sum _{k=0}^{\infty }{\frac {d_{k}z^{2k}}{k!}}\\&={\frac {1}{z^{2}}}+\sum _{k=1}^{\infty }(2k+1)G_{2k+2}z^{2k}.\end{aligned}}

Fourier series

G 4

G 6

G 8

G 10

G 12

G 14

Define $q = e 2π iτ$ . (Some older books define $q$ to be the nome $q = e π iτ$ , but $q = e 2 π iτ$ is now standard in number theory.) Then the Fourier series of the Eisenstein series is

G_{2k}(\tau )=2\zeta (2k)\left(1+c_{2k}\sum _{n=1}^{\infty }\sigma _{2k-1}(n)q^{n}\right)

where the coefficients $c 2 k$ are given by

{\begin{aligned}c_{2k}&={\frac {(2\pi i)^{2k}}{(2k-1)!\zeta (2k)}}\\[4pt]&={\frac {-4k}{B_{2k}}}={\frac {2}{\zeta (1-2k)}}.\end{aligned}}

Here, $B n$ are the Bernoulli numbers, $ζ (z)$ is Riemann's zeta function and $σ p (n)$ is the divisor sum function, the sum of the $p$ th powers of the divisors of $n$ . In particular, one has

{\begin{aligned}G_{4}(\tau )&={\frac {\pi ^{4}}{45}}\left(1+240\sum _{n=1}^{\infty }\sigma _{3}(n)q^{n}\right)\\[4pt]G_{6}(\tau )&={\frac {2\pi ^{6}}{945}}\left(1-504\sum _{n=1}^{\infty }\sigma _{5}(n)q^{n}\right).\end{aligned}}

The summation over $q$ can be resummed as a Lambert series; that is, one has

\sum _{n=1}^{\infty }q^{n}\sigma _{a}(n)=\sum _{n=1}^{\infty }{\frac {n^{a}q^{n}}{1-q^{n}}}

for arbitrary complex $| q | < 1$ and $a$ . When working with the $q$ -expansion of the Eisenstein series, this alternate notation is frequently introduced:

{\begin{aligned}E_{2k}(\tau )&={\frac {G_{2k}(\tau )}{2\zeta (2k)}}\\&=1+{\frac {2}{\zeta (1-2k)}}\sum _{n=1}^{\infty }{\frac {n^{2k-1}q^{n}}{1-q^{n}}}\\&=1-{\frac {4k}{B_{2k}}}\sum _{n=1}^{\infty }\sigma _{2k-1}(n)q^{n}\\&=1-{\frac {4k}{B_{2k}}}\sum _{d,n\geq 1}n^{2k-1}q^{nd}.\end{aligned}}

Identities involving Eisenstein series

As theta functions

Given $q = e 2 π iτ$ , let

{\begin{aligned}E_{4}(\tau )&=1+240\sum _{n=1}^{\infty }{\frac {n^{3}q^{n}}{1-q^{n}}}\\E_{6}(\tau )&=1-504\sum _{n=1}^{\infty }{\frac {n^{5}q^{n}}{1-q^{n}}}\\E_{8}(\tau )&=1+480\sum _{n=1}^{\infty }{\frac {n^{7}q^{n}}{1-q^{n}}}\end{aligned}}

and define

{\begin{aligned}a&=\theta _{2}\left(0;e^{\pi i\tau }\right)=\vartheta _{10}(0;\tau )\\b&=\theta _{3}\left(0;e^{\pi i\tau }\right)=\vartheta _{00}(0;\tau )\\c&=\theta _{4}\left(0;e^{\pi i\tau }\right)=\vartheta _{01}(0;\tau )\end{aligned}}

where $θ m$ and $ϑ ij$ are alternative notations for the Jacobi theta functions. Then,

{\begin{aligned}E_{4}(\tau )&={\tfrac {1}{2}}\left(a^{8}+b^{8}+c^{8}\right),\\[4pt]E_{6}(\tau )&={\tfrac {1}{2}}\left(-3a^{8}\left(b^{4}+c^{4}\right)+b^{12}+c^{12}\right)\\[4pt]&={\tfrac {1}{2}}{\sqrt {\frac {\left(a^{8}+b^{8}+c^{8}\right)^{3}-54(abc)^{8}}{2}}},\end{aligned}}

thus,

E_{4}^{3}-E_{6}^{2}={\tfrac {27}{4}}(abc)^{8}

an expression related to the modular discriminant,

\Delta =g_{2}^{3}-27g_{3}^{2}=(2\pi )^{12}\left({\tfrac {1}{2}}abc\right)^{8}

Also, since $E 8 = E 24$ and $a 4 - b 4 + c 4 = 0$ , this implies

E_{8}(\tau )={\tfrac {1}{2}}\left(a^{16}+b^{16}+c^{16}\right).

Products of Eisenstein series

Eisenstein series form the most explicit examples of modular forms for the full modular group $SL(2, ℤ)$ . Since the space of modular forms of weight $2 k$ has dimension 1 for $2 k = 4, 6, 8, 10, 14$ , different products of Eisenstein series having those weights have to be equal up to a scalar multiple. In fact, we obtain the identities:

E_{4}^{2}=E_{8},\quad E_{4}E_{6}=E_{10},\quad E_{4}E_{10}=E_{14},\quad E_{6}E_{8}=E_{14}.

Using the $q$ -expansions of the Eisenstein series given above, they may be restated as identities involving the sums of powers of divisors:

\left(1+240\sum _{n=1}^{\infty }\sigma _{3}(n)q^{n}\right)^{2}=1+480\sum _{n=1}^{\infty }\sigma _{7}(n)q^{n},

hence

\sigma _{7}(n)=\sigma _{3}(n)+120\sum _{m=1}^{n-1}\sigma _{3}(m)\sigma _{3}(n-m),

and similarly for the others. The theta function of an eight-dimensional even unimodular lattice $Γ$ is a modular form of weight 4 for the full modular group, which gives the following identities:

\theta _{\Gamma }(\tau )=1+\sum _{n=1}^{\infty }r_{\Gamma }(2n)q^{n}=E_{4}(\tau ),\qquad r_{\Gamma }(n)=240\sigma _{3}(n)

for the number $r Γ (n)$ of vectors of the squared length $2 n$ in the root lattice of the type $E 8$ .

Similar techniques involving holomorphic Eisenstein series twisted by a Dirichlet character produce formulas for the number of representations of a positive integer $n$ ' as a sum of two, four, or eight squares in terms of the divisors of $n$ .

Using the above recurrence relation, all higher $E 2 k$ can be expressed as polynomials in $E 4$ and $E 6$ . For example:

{\begin{aligned}E_{8}&=E_{4}^{2}\\E_{10}&=E_{4}\cdot E_{6}\\691\cdot E_{12}&=441\cdot E_{4}^{3}+250\cdot E_{6}^{2}\\E_{14}&=E_{4}^{2}\cdot E_{6}\\3617\cdot E_{16}&=1617\cdot E_{4}^{4}+2000\cdot E_{4}\cdot E_{6}^{2}\\43867\cdot E_{18}&=38367\cdot E_{4}^{3}\cdot E_{6}+5500\cdot E_{6}^{3}\\174611\cdot E_{20}&=53361\cdot E_{4}^{5}+121250\cdot E_{4}^{2}\cdot E_{6}^{2}\\77683\cdot E_{22}&=57183\cdot E_{4}^{4}\cdot E_{6}+20500\cdot E_{4}\cdot E_{6}^{3}\\236364091\cdot E_{24}&=49679091\cdot E_{4}^{6}+176400000\cdot E_{4}^{3}\cdot E_{6}^{2}+10285000\cdot E_{6}^{4}\end{aligned}}

Many relationships between products of Eisenstein series can be written in an elegant way using Hankel determinants, e.g. Garvan's identity

\Delta ^{2}=-{\frac {691}{1728^{2}\cdot 250}}\det {\begin{vmatrix}E_{4}&E_{6}&E_{8}\\E_{6}&E_{8}&E_{10}\\E_{8}&E_{10}&E_{12}\end{vmatrix}}

where

\Delta ={\frac {E_{4}^{3}-E_{6}^{2}}{1728}}

is the modular discriminant.[1]

Ramanujan identities

Srinivasa Ramanujan gave several interesting identities between the first few Eisenstein series involving differentiation. Let

{\begin{aligned}L(q)&=1-24\sum _{n=1}^{\infty }{\frac {nq^{n}}{1-q^{n}}}&&=E_{2}(\tau )\\M(q)&=1+240\sum _{n=1}^{\infty }{\frac {n^{3}q^{n}}{1-q^{n}}}&&=E_{4}(\tau )\\N(q)&=1-504\sum _{n=1}^{\infty }{\frac {n^{5}q^{n}}{1-q^{n}}}&&=E_{6}(\tau ),\end{aligned}}

then

{\begin{aligned}q{\frac {dL}{dq}}&={\frac {L^{2}-M}{12}}\\q{\frac {dM}{dq}}&={\frac {LM-N}{3}}\\q{\frac {dN}{dq}}&={\frac {LN-M^{2}}{2}}.\end{aligned}}

These identities, like the identities between the series, yield arithmetical convolution identities involving the sum-of-divisor function. Following Ramanujan, to put these identities in the simplest form it is necessary to extend the domain of $σ p (n)$ to include zero, by setting

{\begin{aligned}\sigma _{p}(0)={\tfrac {1}{2}}\zeta (-p)\quad \Longrightarrow \quad \sigma (0)&=-{\tfrac {1}{24}}\\\sigma _{3}(0)&={\tfrac {1}{240}}\\\sigma _{5}(0)&=-{\tfrac {1}{504}}.\end{aligned}}

Then, for example

\sum _{k=0}^{n}\sigma (k)\sigma (n-k)={\tfrac {5}{12}}\sigma _{3}(n)-{\tfrac {1}{2}}n\sigma (n).

Other identities of this type, but not directly related to the preceding relations between $L$ , $M$ and $N$ functions, have been proved by Ramanujan and Giuseppe Melfi,[2][3] as for example

{\begin{aligned}\sum _{k=0}^{n}\sigma _{3}(k)\sigma _{3}(n-k)&={\tfrac {1}{120}}\sigma _{7}(n)\\\sum _{k=0}^{n}\sigma (2k+1)\sigma _{3}(n-k)&={\tfrac {1}{240}}\sigma _{5}(2n+1)\\\sum _{k=0}^{n}\sigma (3k+1)\sigma (3n-3k+1)&={\tfrac {1}{9}}\sigma _{3}(3n+2).\end{aligned}}

Generalizations

Automorphic forms generalize the idea of modular forms for general Lie groups; and Eisenstein series generalize in a similar fashion.

Defining $O K$ to be the ring of integers of a totally real algebraic number field $K$ , one then defines the Hilbert–Blumenthal modular group as $PSL(2, O K)$ . One can then associate an Eisenstein series to every cusp of the Hilbert–Blumenthal modular group.

References

Milne, Steven C. (2000). "Hankel Determinants of Eisenstein Series". arXiv:math/0009130v3.
Ramanujan, Srinivasa (1962). "On certain arithmetical functions". Collected Papers. New York, NY: Chelsea. pp. 136–162.
Melfi, Giuseppe (1998). "On some modular identities". Number Theory, Diophantine, Computational and Algebraic Aspects: Proceedings of the International Conference held in Eger, Hungary. Walter de Grutyer & Co. pp. 371–382.

Eisenstein series