Low-order models (LOM's), which are systems of ordinary differential equations which have been simplified by extreme reduction of the number of dependent variables, are often capable of representing atmospheric processes in a qualitatively correct manner. With a LOM it is generally possible to obtain a much more extended time-dependent solution, or a much larger ensemble of solutions, than would be economically feasible with a larger model.
A general procedure for constructing LOM's is described. A selection of LOM's is presented, to illustrate the many forms which these models may take and the many uses to which they may be put. The step-by-step construction of a LOM is illustrated with a model of the large-scale circulation of a moist atmosphere.