Abstract

One of the key aspects of our climate is the seasonal cycle; often the largest signal in most of climate time series. Climate time-series reveal the combination of the seasonal cycle, short-time processes related to weather, and slowly-varying signals caused by ocean circulation or global warming. Despite the myriad of processes involved in shaping climate time-series, the most parsimonious physical framework to describe them is that of Brownian particles flowing through a slowly-varying seasonal cycle, which is described with a periodic non-autonomous stochastic dynamical system. This mathematical model consists of the deterministic contribution, generating a reliable seasonal cycle and a stochastic forcing, capturing the impact of short-time scale processes. The generality of the method affords applicability to a wide range of systems/subsystems. First, I will show how this model can be used to understand the seasonal variability of Arctic sea ice. Analytic solutions constructed from a stochastic perturbation method reveal the basic physical processes controlling the seasonal variability. Second, I will use this formalism to construct a stochastic model to regenerate the statistics of monthly-average surface temperature data, which spans around 133 years from 1880 to present. I will use this framework to discuss issues such as climate sensitivity and predictability.