Forecasting rare stratospheric transitions using short simulations

Nonlinear atmospheric dynamics produce rare events that are hard to predict and attribute due to many interacting degrees of freedom. A sudden stratospheric warming is a spectacular example in which the winter polar vortex in the stratosphere rapidly breaks down, inducing a shift in midlatitude tropospheric weather patterns that persist for up to 2-3 months.  In principle, lengthy numerical simulations can be used to predict and understand these rare transitions.  For complex models, however, the cost of the direct numerical simulation approach is often prohibitive.  We describe an alternative approach which in principle only requires relatively short duration computer simulations of the system.  Applying this methodology to a classical idealized stratospheric model with stochastic forcing, we compute optimal forecasts of sudden warming events and quantify the limits of predictability.  Statistical analysis relates these optimal forecasts to a small number of easy-to-interpret physical variables.Remarkably, we are able to estimate these quantities using a data set of simulations much shorter than the return time of the warming event.