28.04.2014 Views

Mock-modular forms of weight one - UCLA Department of Mathematics

Mock-modular forms of weight one - UCLA Department of Mathematics

Mock-modular forms of weight one - UCLA Department of Mathematics

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

MOCK-MODULAR FORMS OF WEIGHT ONE 25<br />

Using the Fourier expansion <strong>of</strong> Φ B (ĝ ψ , z) and Rankin-Selberg unfolding, we can rewrite equation<br />

(61) as<br />

( )<br />

k<br />

(62) a B (ĝ ψ , m) + 24σ 1 (m)a B (ĝ ψ , 0) = lim δ(k)r B (pm + k)r ψ (k)φ s ,<br />

where<br />

φ s (µ) :=<br />

∫ ∞<br />

1<br />

s→0<br />

∑<br />

k≥1<br />

du<br />

(1 + µu) 1+s u .<br />

Comparing Q s−1 (t) and φ s−1 (µ), we see that φ 0 (µ) = 2Q 0 (1 + 2µ) and<br />

φ s−1 (µ) − 2Γ(2s) Q<br />

sΓ(s) 2 s−1 (1 + 2µ) = O(µ −1−s ).<br />

( ) ( )<br />

2<br />

That means we can substitute Q<br />

sΓ(s) 2 s−1 1 + 2k<br />

k<br />

for φ<br />

pm s−1 and in the limit as s approaches<br />

1. Together with (8) and (60), equation (61)<br />

pm<br />

becomes<br />

(63)<br />

〈j lift,B<br />

m , g ψ 〉 reg +24σ 1 (m)〈ϑ B , g ψ 〉 =<br />

− lim<br />

s→1<br />

∑<br />

A<br />

ψ(A) ∑ k≥1<br />

pm<br />

( ))<br />

δ(k)r B (pm + k)r A (k)<br />

(−2Q s−1 1 + 2k .<br />

pm<br />

Now the counting argument in Proposition (3.11) in [23] tells us that<br />

δ(k)r B (pm + k)r A (k) = ρ m (A, B, k),<br />

where ρ m (A, B, k) counts the number <strong>of</strong> γ ∈ M 2 (Z)/{±1} such that det(γ) = m and<br />

cosh d(τ √ AB −1 , γτ √ AB ) = 1 + 2k<br />

pm .<br />

In particular when k = 0, the number <strong>of</strong> γ ∈ M 2 (Z)/{±1} such that det(γ) = m and<br />

γτ √ AB = τ√ AB<br />

is exactly r −1 B (pm) = r B (m). Since k ≥ 1 in the summation, equation (63)<br />

becomes<br />

(64)<br />

〈jm<br />

lift,B ,g ψ 〉 reg + 24σ 1 (m)〈ϑ B , g ψ 〉 = − lim<br />

= − lim<br />

ɛ→0<br />

lim<br />

s→1<br />

∑<br />

A<br />

s→1<br />

∑<br />

A<br />

ψ(A)<br />

∑<br />

(<br />

g s τ<br />

√<br />

AB −1, γτ √ AB<br />

γ∈M 2 (Z)/{±1}<br />

det(γ)=m,<br />

γτ √ AB≠τ√ AB −1<br />

ψ(A) ( G m s (τ √ AB −1 + ɛ, τ √ AB ) − r B(m)g s (τ √ AB −1 + ɛ, τ √ AB −1 ) )<br />

where g s is defined in (57). From this expression, it is clear that the choice <strong>of</strong> these CM<br />

points do not matter.<br />

)<br />

Since g 1 (τ + ɛ, τ) = − log<br />

(1 + 4Im(τ)2 and ψ is non-trivial, we see that<br />

ɛ 2<br />

lim lim<br />

ɛ→0<br />

s→1<br />

∑<br />

A<br />

ψ(A)g s (τ √ AB −1 + ɛ, τ √ AB −1 ) = − ∑ A<br />

Also by equation (59), we have<br />

∑<br />

lim ψ(A)G m s (τ √<br />

s→1<br />

AB<br />

+ ɛ,τ √ −1 AB ) = ∑<br />

A<br />

A<br />

4πσ 1 (m) ∑ A<br />

ψ(A) log(y 2 √<br />

AB −1).<br />

ψ(A) log |Ψ m (τ √ AB −1 + ɛ, τ √ AB )|2 −<br />

ψ(A) lim<br />

s→1<br />

(E(τ √ AB −1 + ɛ, s) + E(τ √ AB , s)).<br />

)

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!