A Complexity Perspective on Climate Change

Abstract

This talk introduces a new approach to understanding and predicting climate change by using tools from information theory and the new field of computational complexity. As an orientation to the approach, the tools are applied to simpler complex dynamical systems including the Lorenz model, and some ETL field data. An interesting climate result has already been discovered by applying one of the tools (mutual information) to NCEP reanalysis data over the northern hemisphere: over the last fifty years, the most noticeable increase in the complexity of winter, geopotential heights occurred in the arctic.