A method for extracting coupled patterns of predictable and chaotic components in pairs of climate variables

Abstract

Interannual variability in the seasonal mean of a climate variable can arise from a number of sources which can be categorized as (a) slowly varying (interannual/supra-annual) external forcing (for example, sea surface temperature forcing) and internal dynamics and (b) internal dynamics within the season. The former is generally assumed to be potentially predictable and the latter unpredictable. Here, a method is proposed for extracting coupled patterns of interannual variability that relate the predictable and unpredictable components in pairs of climate variables. The method is applied to observed Australian summer surface air temperature and the global 500 hPa geopotential height for the period 1951--1999.