The L2L^2-supercritical stochastic NLS blow-up conjecture

Let u0H1(R)u_0\in H^1(\mathbb R) and let u(t)u(t) be an evolution of the stochastic nonlinear Schrödinger equation with σ>2\sigma>2 and the specified space-time white-noise coefficient. Let Q=Q1,0Q=Q_{1,0} be the blow-up profile solving the referenced profile equation, and let aa be the corresponding constant. The L2L^2-supercritical blow-up conjecture. In the multiplicative (Stratonovich) noise case, sufficiently localized initial data blows up in finite positive random time with positive probability. In the additive noise case, any initial data leads to blow-up in finite random time almost surely. If a solution blows up at a random positive time T(ω)>0T(\omega)>0, then, for tt close to T(ω)T(\omega),

uc(t,x)L(t)1/σQ(xx(t)L(t))exp(iθ(t)+i2a(t)logTTt),u_c(t,x)\sim L(t)^{-1/\sigma}Q\left(\frac{x-x(t)}{L(t)}\right)\exp\left(i\theta(t)+\frac{i}{2a(t)}\log\frac{T}{T-t}\right),

where a(t)aa(t)\to a, θ(t)θ0\theta(t)\to\theta_0, x(t)xcx(t)\to x_c, and L(t)=(2a(Tt))1/2L(t)=(2a(T-t))^{1/2}. Consequently,

u(t,)Lx2L(t)(1s)(2a(Tt))12(12+1σ).\\|\nabla u(t,\cdot)\\|_{L_x^2}\sim L(t)^{-(1-s)}\equiv (2a(T-t))^{-\frac12(\frac12+\frac1\sigma)}.

Conditionally on finite-time blow-up, xcx_c is Gaussian; no conditioning is necessary in the additive case. Thus the blow-up has a polynomial rate without correction and the deterministic supercritical blow-up profile. The source presents this as a conjecture and reports numerical confirmation, but does not supply a general rigorous resolution.

Sources & referencesView supporting material

Primary source

Annie Millet, Svetlana Roudenko and Kai Yang, “Behavior of solutions to the 1D focusing stochastic L^2-critical and supercritical nonlinear Schrödinger equation with space-time white noise”, arXiv:2005.14266 (2020).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.