Local fluctuations of critical Mandelbrot cascades Konrad Kolesko joint with D. Buraczewski and P. Dyszewski Warwick, 18-22 May, 2015 Random measures µ µ1 µ2 For given random variables X1 , X2 s.t. E e−X1 + e−X2 = 1 we are interested in random measures µ on [0, 1) satisfying self similar property: µ(B) = e−X1 µ1 2(B ∩ [0, 1/2)) + e−X2 µ2 2(B ∩ [1/2, 1) − 1) , where µ1 ⊥⊥ µ2 ⊥⊥ X1 , X2 and Lµ = Lµ1 = Lµ2 . Goal: Understand local properties of µ Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 2 / 19 Random measures µ µ1 µ2 For given random variables X1 , X2 s.t. E e−X1 + e−X2 = 1 we are interested in random measures µ on [0, 1) satisfying self similar property: µ(B) = e−X1 µ1 2(B ∩ [0, 1/2)) + e−X2 µ2 2(B ∩ [1/2, 1) − 1) , where µ1 ⊥⊥ µ2 ⊥⊥ X1 , X2 and Lµ = Lµ1 = Lµ2 . Goal: Understand local properties of µ Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 2 / 19 Random measures µ µ1 µ2 For given random variables X1 , X2 s.t. E e−X1 + e−X2 = 1 we are interested in random measures µ on [0, 1) satisfying self similar property: µ(B) = e−X1 µ1 2(B ∩ [0, 1/2)) + e−X2 µ2 2(B ∩ [1/2, 1) − 1) , where µ1 ⊥⊥ µ2 ⊥⊥ X1 , X2 and Lµ = Lµ1 = Lµ2 . Goal: Understand local properties of µ Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 2 / 19 Phase transition Define ψ(t) := log2 E e−tX1 + e−tX2 . By assumption ψ(0) = 1, ψ(1) = 0, ψ % ∞. Supercritical Critical Subcritical Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 3 / 19 Subcritical case ψ 0 (1) = −m < 0 Theorem (Liu) Suppose that ψ 0 (1) = −m < 0, then for any δ > 0, almost all realization of µ, µ-almost all x and sufficiently large n √ 2 µ(B(x, 2−n )) ≥ 2−nm−(1+δ) 2σ n log log n √ 2 µ(B(x, 2−n )) ≤ 2−nm+(1+δ) 2σ n log log n , where σ 2 = ψ 00 (1) − ψ(1). Moreover µ(B(x, 2−n )) ≤ 2−nm−(1−δ) µ(B(x, 2−n )) ≥ 2−nm+(1−δ) Konrad Kolesko √ 2σ 2 n log log n √ 2σ 2 n log log n Local fluctuations of critical Mandelbrot cascades i.o. i.o. 18-22 May, 2015 4 / 19 Subcritical case ψ 0 (1) = −m < 0 Theorem (Liu) Suppose that ψ 0 (1) = −m < 0, then for any δ > 0, almost all realization of µ, µ-almost all x and sufficiently large n √ 2 µ(B(x, 2−n )) ≥ 2−nm−(1+δ) 2σ n log log n √ 2 µ(B(x, 2−n )) ≤ 2−nm+(1+δ) 2σ n log log n , where σ 2 = ψ 00 (1) − ψ(1). Moreover µ(B(x, 2−n )) ≤ 2−nm−(1−δ) µ(B(x, 2−n )) ≥ 2−nm+(1−δ) Konrad Kolesko √ 2σ 2 n log log n √ 2σ 2 n log log n Local fluctuations of critical Mandelbrot cascades i.o. i.o. 18-22 May, 2015 4 / 19 Supercritical & critical case Barral, Rhodes, Vargas When ψ 0 (1) > 0 then µ is purely atomic Barral, Kupiainen, Nikula, Saksman, Webb If ψ 0 (1) = 0, then the random measure µ almost surely has no atoms. Moreover for any k and δ > 0, µ-a.e. x √ √ − 6 log 2 n(log n+(1/3+δ) log log n) −n for µ(B(x, 2 )) ≥ e sufficiently large n √ √ µ(B(x, 2−n )) ≥ e−( 2 log 2+δ) n log n i.o. µ(B(x, 2−n )) ≤ n−k for sufficiently large n. Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 5 / 19 Supercritical & critical case Barral, Rhodes, Vargas When ψ 0 (1) > 0 then µ is purely atomic Barral, Kupiainen, Nikula, Saksman, Webb If ψ 0 (1) = 0, then the random measure µ almost surely has no atoms. Moreover for any k and δ > 0, µ-a.e. x √ √ − 6 log 2 n(log n+(1/3+δ) log log n) −n for µ(B(x, 2 )) ≥ e sufficiently large n √ √ µ(B(x, 2−n )) ≥ e−( 2 log 2+δ) n log n i.o. µ(B(x, 2−n )) ≤ n−k for sufficiently large n. Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 5 / 19 Supercritical & critical case Barral, Rhodes, Vargas When ψ 0 (1) > 0 then µ is purely atomic Barral, Kupiainen, Nikula, Saksman, Webb If ψ 0 (1) = 0, then the random measure µ almost surely has no atoms. Moreover for any k and δ > 0, µ-a.e. x √ √ − 6 log 2 n(log n+(1/3+δ) log log n) −n for µ(B(x, 2 )) ≥ e sufficiently large n √ √ µ(B(x, 2−n )) ≥ e−( 2 log 2+δ) n log n i.o. µ(B(x, 2−n )) ≤ n−k for sufficiently large n. Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 5 / 19 Supercritical & critical case Barral, Rhodes, Vargas When ψ 0 (1) > 0 then µ is purely atomic Barral, Kupiainen, Nikula, Saksman, Webb If ψ 0 (1) = 0, then the random measure µ almost surely has no atoms. Moreover for any k and δ > 0, µ-a.e. x √ √ − 6 log 2 n(log n+(1/3+δ) log log n) −n for µ(B(x, 2 )) ≥ e sufficiently large n √ √ µ(B(x, 2−n )) ≥ e−( 2 log 2+δ) n log n i.o. µ(B(x, 2−n )) ≤ n−k for sufficiently large n. Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 5 / 19 Main theorem Theorem (D. Buraczewski, P. Dyszewski, K.K.) Let ψ 0 (1) = 0, k ∈ N and δ > 0. Then for almost all realizations of µ, µ-almost all x ∈ [0, 1) and sufficiently large n we have p µ(B(x, 2−n )) ≥ exp −(1 + δ) 2σ 2 n log log n ! √ − n µ(B(x, 2−n )) ≤ exp Qk 2 i=1 log(i) n (log(k+1) n) Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 6 / 19 Main theorem Theorem (D. Buraczewski, P. Dyszewski, K.K.) Let ψ 0 (1) = 0, k ∈ N and δ > 0. Then for almost all realizations of µ, µ-almost all x ∈ [0, 1) and sufficiently large n we have p µ(B(x, 2−n )) ≥ exp −(1 + δ) 2σ 2 n log log n ! √ n − µ(B(x, 2−n )) ≤ exp Qk 2 i=1 log(i) n (log(k+1) n) e−K1 √ n(log n+(1/3+δ) log log n) √ e−(K2 +δ) Konrad Kolesko n−k n log n Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 6 / 19 Notations Dyadic intervals on [0,1) ! vertices of a binary tree T x ∈ [0, 1) ! θ ∈ ∂T . B(v) = {θ ∈ ∂T : v ∈ oθ}. For a random measure µω we are interested in a pointwise estimates of µω (B(θn )) on the enlarged measure space e P(dω, dθ) := P(dω)µω (dθ) i.e. we are looking for deterministic φ1 , φ2 s.t. φ1 (n) ≤ µω (B(θn )) ≤ φ2 (n), e for P-almost all (ω, θ) and large n. e can be replaced by P(dω, b P dθ) = P(dω)µω (dθ), where µω is the normalized measure µω . Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 7 / 19 Notations Dyadic intervals on [0,1) ! vertices of a binary tree T x ∈ [0, 1) ! θ ∈ ∂T . B(v) = {θ ∈ ∂T : v ∈ oθ}. For a random measure µω we are interested in a pointwise estimates of µω (B(θn )) on the enlarged measure space e P(dω, dθ) := P(dω)µω (dθ) i.e. we are looking for deterministic φ1 , φ2 s.t. φ1 (n) ≤ µω (B(θn )) ≤ φ2 (n), e for P-almost all (ω, θ) and large n. e can be replaced by P(dω, b P dθ) = P(dω)µω (dθ), where µω is the normalized measure µω . Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 7 / 19 Notations Dyadic intervals on [0,1) ! vertices of a binary tree T x ∈ [0, 1) ! θ ∈ ∂T . B(v) = {θ ∈ ∂T : v ∈ oθ}. For a random measure µω we are interested in a pointwise estimates of µω (B(θn )) on the enlarged measure space e P(dω, dθ) := P(dω)µω (dθ) i.e. we are looking for deterministic φ1 , φ2 s.t. φ1 (n) ≤ µω (B(θn )) ≤ φ2 (n), e for P-almost all (ω, θ) and large n. e can be replaced by P(dω, b P dθ) = P(dω)µω (dθ), where µω is the normalized measure µω . Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 7 / 19 Branching random walk o P – probability measure on the set of labeled trees X(v) – the sum of random variables along the path between o and v (X(u) = X2 + X21 + X212 ) {X(v)}v∈T – Branching random walk (BRW) Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 8 / 19 Branching random walk o X1 X2 P – probability measure on the set of labeled trees X(v) – the sum of random variables along the path between o and v (X(u) = X2 + X21 + X212 ) {X(v)}v∈T – Branching random walk (BRW) Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 8 / 19 Branching random walk o X1 X11 X2 X12 P – probability measure on the set of labeled trees X(v) – the sum of random variables along the path between o and v (X(u) = X2 + X21 + X212 ) {X(v)}v∈T – Branching random walk (BRW) Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 8 / 19 Branching random walk o X1 X11 X2 X12 X21 X22 P – probability measure on the set of labeled trees X(v) – the sum of random variables along the path between o and v (X(u) = X2 + X21 + X212 ) {X(v)}v∈T – Branching random walk (BRW) Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 8 / 19 Branching random walk o X1 X11 X2 X12 X21 X22 X212 u P – probability measure on the set of labeled trees X(v) – the sum of random variables along the path between o and v (X(u) = X2 + X21 + X212 ) {X(v)}v∈T – Branching random walk (BRW) Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 8 / 19 Branching random walk o X1 X11 X2 X12 X21 X22 X212 u P – probability measure on the set of labeled trees X(v) – the sum of random variables along the path between o and v (X(u) = X2 + X21 + X212 ) {X(v)}v∈T – Branching random walk (BRW) Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 8 / 19 Stopping line P P Since E |v|=1 e−X(v) = 1 and E |v|=1 X(v)e−X(v) = 0 the equation X Ef (Y ) := E f (X(v))e−X(v) |v|=1 defines distribution of a driftless r.v. Y Let h be a harmonic function on some set A (a solution of a Dirichlet problem), Vn = Y1 + · · · + Yn and σ = min{k : s + Vk ∈ / A}. Then the process h(s + Vmin(n,σ) ) is a martingale. Let τ = {w : s + X(w) ∈ / A for the first time}, vτ = min(v, τ ). Then X Wns = h(s + X(vτ ))e−X(vτ ) is a martingale. Konrad Kolesko |v|=n Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 9 / 19 Stopping line P P Since E |v|=1 e−X(v) = 1 and E |v|=1 X(v)e−X(v) = 0 the equation X Ef (Y ) := E f (X(v))e−X(v) |v|=1 defines distribution of a driftless r.v. Y Let h be a harmonic function on some set A (a solution of a Dirichlet problem), Vn = Y1 + · · · + Yn and σ = min{k : s + Vk ∈ / A}. Then the process h(s + Vmin(n,σ) ) is a martingale. Let τ = {w : s + X(w) ∈ / A for the first time}, vτ = min(v, τ ). Then X Wns = h(s + X(vτ ))e−X(vτ ) is a martingale. Konrad Kolesko |v|=n Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 9 / 19 Stopping line P P Since E |v|=1 e−X(v) = 1 and E |v|=1 X(v)e−X(v) = 0 the equation X Ef (Y ) := E f (X(v))e−X(v) |v|=1 defines distribution of a driftless r.v. Y Let h be a harmonic function on some set A (a solution of a Dirichlet problem), Vn = Y1 + · · · + Yn and σ = min{k : s + Vk ∈ / A}. Then the process h(s + Vmin(n,σ) ) is a martingale. Let τ = {w : s + X(w) ∈ / A for the first time}, vτ = min(v, τ ). Then X Wns = h(s + X(vτ ))e−X(vτ ) is a martingale. Konrad Kolesko |v|=n Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 9 / 19 Martingales Some natural martingales h ≡ 1: Wn = P h(x) = x: Dn = |v|=n e−X(v) P |v|=n X(v)e−X(v) A = [0,P ∞), h(x) ≈ x ∨ 0: Wns = |v|=n h(s + X(v))1[s+X(v0 )>0, for v0 ≤v] e−X(v) µ(B(v)) := lim n Konrad Kolesko X X(w)e−X(w) |w|=n, w<v Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 10 / 19 Martingales Some natural martingales h ≡ 1: Wn = P h(x) = x: Dn = |v|=n e−X(v) → 0 P |v|=n X(v)e−X(v) → D > 0 A = [0,P ∞), h(x) ≈ x ∨ 0: Wns = |v|=n h(s + X(v))1[s+X(v0 )>0, for v0 ≤v] e−X(v) µ(B(v)) := lim n Konrad Kolesko X X(w)e−X(w) |w|=n, w<v Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 10 / 19 Martingales Some natural martingales h ≡ 1: Wn = P h(x) = x: Dn = |v|=n e−X(v) → 0 P |v|=n X(v)e−X(v) → D > 0 A = [0,P ∞), h(x) ≈ x ∨ 0: Wns = |v|=n h(s + X(v))1[s+X(v0 )>0, for v0 ≤v] e−X(v) µ(B(v)) := lim n Konrad Kolesko X X(w)e−X(w) |w|=n, w<v Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 10 / 19 Martingales Some natural martingales h ≡ 1: Wn = P h(x) = x: Dn = |v|=n e−X(v) → 0 P |v|=n X(v)e−X(v) → D > 0 A = [0,P ∞), h(x) ≈ x ∨ 0: Wns = |v|=n h(s + X(v))1[s+X(v0 )>0, for v0 ≤v] e−X(v) µ(B(v)) := lim n Konrad Kolesko X X(w)e−X(w) |w|=n, w<v Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 10 / 19 Wns Wns = X h(s + X(v))1[s+X(v0 )>0, for v0 ≤v] e−X(v) |v|=n Wns → W s P-a.s. and L1 For any s ∈ D define Ps := Ws ·P h(s) P[supp W s ] & 1 − 1/s Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 11 / 19 Ps =⇒ BRW ∞ 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 12 / 19 Ps BRW ∞ =⇒ 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 12 / 19 Ps BRW ∞ =⇒ 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 12 / 19 Spinal decomposition of Ps BRW ∞ s 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ s 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ X11 X21 s 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ X11 1 e−X1 h(s + X11 ) X21 1 e−X2 h(s + X21 ) s 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S1 = s+X21 S0 = s 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ X12 S1 X22 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ X12 2 e−X1 h(S1 + X12 ) S1 X22 2 e−X2 h(S1 + X22 ) S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S1 S2 = S1 +X21 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S1 S2 X13 X23 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S1 S2 X13 e −X13 h(S2 + X13 ) X23 −X23 e S0 h(S2 + X23 ) 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S1 S3 = S2 +X31 S2 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S1 S3 S2 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S4 S1 S3 S2 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S4 S1 S3 S2 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S5 S4 S1 S3 S2 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S5 S4 S1 S3 S2 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S5 S4 S1 S3 S2 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S5 S4 S1 S3 S2 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S5 S4 S1 S3 S2 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps BRW ∞ S5 S4 S1 S3 S2 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps Random tree Ts with a distinguished ray Θ ∈ ∂Ts BRW ∞ S5 S4 S1 S3 S2 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps s s P(T ∈ dω) = P (dω) BRW ∞ S5 S4 S1 S3 S2 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 Spinal decomposition of Ps bs (dω, dθ) := P P(Ts ∈ dω, Θ ∈ dθ) BRW ∞ S5 S4 S1 S3 S2 S0 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 13 / 19 bs vs. P b P We have bs P. b P The converse is not true, but for any set A bs (A) = 1 ⇒ P(A) b P ≥ 1 − (s), for some (s) → 0 as s → ∞. In particular, if for any s > 0 bs (A) = 1 P then b P(A) = 1. Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 14 / 19 Probabilistic interpretation of local fluctuations e−(S0 −s) ∞ C0 e−(S1 −s) S1 S3 S2 C1 R0 S5 S4 e−(S2 −s) C2 S0 R1 R2 0 Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 15 / 19 Probabilistic interpretation of local fluctuations bs the sequence µω (B(θn )) has the same law as Under P X e−(Sk −s) Ck Rk , k≥n where Sk is a random walk starting form s conditioned to stay positive, Rk independent random variables. Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 15 / 19 Useful information We are looking for LIL for P k≥n e−Sk +s Ck Rk , Hambly, Kersting, Kyprianou √ ψ(t) dt < ∞ iff Sn > nψ(n) t lim supn √ 2Sn =1 2nσ log log n R∞ √ eventually p nψ(n) < Sn < (1 + δ) 2nσ 2 log log n Ps -a.s. for n sufficiently large Kyprianou 1 − Ee−tR ∼ tL(t) near 0 =⇒ ERγ < ∞ for γ < 1 In particular, by Borel-Cantelli lemma, Rn < n2 for all but finitely many n. Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 16 / 19 Useful information We are looking for LIL for P k≥n e−Sk Rk , Hambly, Kersting, Kyprianou √ ψ(t) dt < ∞ iff Sn > nψ(n) t lim supn √ 2Sn =1 2nσ log log n R∞ √ eventually p nψ(n) < Sn < (1 + δ) 2nσ 2 log log n Ps -a.s. for n sufficiently large Kyprianou 1 − Ee−tR ∼ tL(t) near 0 =⇒ ERγ < ∞ for γ < 1 In particular, by Borel-Cantelli lemma, Rn < n2 for all but finitely many n. Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 16 / 19 Useful information We are looking for LIL for P k≥n e−Sk Rk , Hambly, Kersting, Kyprianou √ ψ(t) dt < ∞ iff Sn > nψ(n) t lim supn √ 2Sn =1 2nσ log log n R∞ √ eventually p nψ(n) < Sn < (1 + δ) 2nσ 2 log log n Ps -a.s. for n sufficiently large Kyprianou 1 − Ee−tR ∼ tL(t) near 0 =⇒ ERγ < ∞ for γ < 1 In particular, by Borel-Cantelli lemma, Rn < n2 for all but finitely many n. Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 16 / 19 Useful information We are looking for LIL for P k≥n e−Sk Rk , Hambly, Kersting, Kyprianou √ ψ(t) dt < ∞ iff Sn > nψ(n) t lim supn √ 2Sn =1 2nσ log log n R∞ √ eventually p nψ(n) < Sn < (1 + δ) 2nσ 2 log log n Ps -a.s. for n sufficiently large Kyprianou 1 − Ee−tR ∼ tL(t) near 0 =⇒ ERγ < ∞ for γ < 1 In particular, by Borel-Cantelli lemma, Rn < n2 for all but finitely many n. Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 16 / 19 Upper bound Take any ψ such that R∞ ψ(t)dt t X < ∞. We have that e−Sk Rk k≥n is eventually bounded by X e− √ kψ(k) 2 k . k≥n On the other hand, if X R∞ ψ(t)dt t = ∞ then i.o. √ e−Sk Rk ≥ e−Sn Rn ≥ e− nψ(n) i.o. Rn ≥ δe− √ nψ(n) k≥n Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 17 / 19 Upper bound Take any ψ such that R∞ ψ(t)dt t X < ∞. We have that e−Sk Rk k≥n is eventually bounded by X √ e− kψ(k) . k≥n On the other hand, if X R∞ ψ(t)dt t = ∞ then i.o. √ e−Sk Rk ≥ e−Sn Rn ≥ e− nψ(n) i.o. Rn ≥ δe− √ nψ(n) k≥n Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 17 / 19 Upper bound Take any ψ such that R∞ ψ(t)dt t X < ∞. We have that e−Sk Rk k≥n is eventually bounded by X e−k · polynomial(k). √ k≥ nψ(n) On the other hand, if X R∞ ψ(t)dt t = ∞ then i.o. √ e−Sk Rk ≥ e−Sn Rn ≥ e− nψ(n) i.o. Rn ≥ δe− √ nψ(n) k≥n Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 17 / 19 Upper bound Take any ψ such that R∞ ψ(t)dt t X < ∞. We have that e−Sk Rk k≥n is eventually bounded by X e−k . √ k≥ nψ(n) On the other hand, if X R∞ ψ(t)dt t = ∞ then i.o. √ e−Sk Rk ≥ e−Sn Rn ≥ e− nψ(n) i.o. Rn ≥ δe− √ nψ(n) k≥n Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 17 / 19 Upper bound Take any ψ such that R∞ ψ(t)dt t X < ∞. We have that e−Sk Rk k≥n is eventually bounded by √ e− On the other hand, if X R∞ ψ(t)dt t nψ(n) . = ∞ then i.o. √ e−Sk Rk ≥ e−Sn Rn ≥ e− nψ(n) i.o. Rn ≥ δe− √ nψ(n) k≥n Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 17 / 19 Upper bound Take any ψ such that R∞ ψ(t)dt t X < ∞. We have that e−Sk Rk k≥n is eventually bounded by √ e− On the other hand, if X R∞ ψ(t)dt t nψ(n) . = ∞ then i.o. √ e−Sk Rk ≥ e−Sn Rn ≥ e− nψ(n) i.o. Rn ≥ δe− √ nψ(n) k≥n Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 17 / 19 Lower bound Take q > 1 and by N (n) := q dlogq ne . Borel-Cantelli’s lemma implies that for sufficiently large n, sup Rk ≥ δ0 . q n <k≤q n+1 X e−Sk Rk ≥ k≥n X e−(1+δ) √ 2kσ 2 log log k Rk k≥n qN (n) ≥ X √ 2 e−(1+δ) 2kσ log log k Rk k=N (n) √ 2 ≥ δ0 e−(1+δ) 2qN (n)σ log log(qN (n)) √ 2 2 2 ≥ δ0 e−(1+δ) 2q nσ log log(q n) √ 2 ≥ e−(1+2δ) 2nσ log log n , for some q. Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 18 / 19 Lower bound Take q > 1 and by N (n) := q dlogq ne . Borel-Cantelli’s lemma implies that for sufficiently large n, sup Rk ≥ δ0 . q n <k≤q n+1 X e−Sk Rk ≥ k≥n X e−(1+δ) √ 2kσ 2 log log k Rk k≥n qN (n) ≥ X √ 2 e−(1+δ) 2kσ log log k Rk k=N (n) √ 2 ≥ δ0 e−(1+δ) 2qN (n)σ log log(qN (n)) √ 2 2 2 ≥ δ0 e−(1+δ) 2q nσ log log(q n) √ 2 ≥ e−(1+2δ) 2nσ log log n , for some q. Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 18 / 19 Lower bound Take q > 1 and by N (n) := q dlogq ne . Borel-Cantelli’s lemma implies that for sufficiently large n, sup Rk ≥ δ0 . q n <k≤q n+1 X e−Sk Rk ≥ k≥n X e−(1+δ) √ 2kσ 2 log log k Rk k≥n qN (n) ≥ X √ 2 e−(1+δ) 2kσ log log k Rk k=N (n) √ 2 ≥ δ0 e−(1+δ) 2qN (n)σ log log(qN (n)) √ 2 2 2 ≥ δ0 e−(1+δ) 2q nσ log log(q n) √ 2 ≥ e−(1+2δ) 2nσ log log n , for some q. Konrad Kolesko Local fluctuations of critical Mandelbrot cascades 18-22 May, 2015 18 / 19