A Simple Model for Simulating Regional Short-Term Climatic Changes

Abstract

A method is proposed for simulating the fluctuations on time scales from a month to a few years in climatic properties such as circulation strengths and temperatures in specific regions. The model represents explicitly the variables of interest, together with any variables that influence them and exhibit persistence. All variables not represented explicitly or specified are parameterized in terms of short-term random forcing. The validity of the model depends on the adequacy of this parameterization. All processes are parameterized in terms of the influence of one variable on another; this draws attention to the role of feedback in determining the statistics of the variables. Development of the model is illustrated by a pilot study. Physical guidance is used to propose a two-variable model with constant parameters to simulate the Southern Oscillation. Observed statistics require that the model be modified to include an annual variation in one parameter. The model fails to simulate the observed negative lag correlations. Suggestions for further development are outlined.