Climate models attempt to simulate the behaviour of the climate system. Through understanding the climate system, it is possible to obtain

• a clearer picture of past climates by comparison with empirical observation
• indications about future climate change scenarios

Models can be used to simulate climate on a variety of spatial and temporal scales. Sometimes one may wish to study regional climates; at other times global-scale climate models, which simulate the climate of the entire planet, will be desired. The ultimate purpose of a model is to identify the likely response of the climate system to a change in any of the parameters and processes which control the system. For example, the climate system may be perturbed by the radiative forcing associated with an increase in carbon dioxide (a greenhouse gas) in the atmosphere. The aim of the model is then to assess how the climate system will respond to this perturbation, in an attempt to restore equilibrium.There are three major sets of processes which must be considered when constructing a climate model:

• radiative - the transfer of radiation through the climate system (e.g. absorption, reflection)
• dynamic - the horizontal and vertical transfer of energy (e.g. advection, convection, diffusion)
• surface process - processes involving land/ocean/ice, and the effects of albedo, emissivity and surface-atmosphere energy exchange

All models must simplify what is a very complex climate system. This is in part due to the limited understanding that exists of the climate system, and partly the result of computational restraints. Simplification may be achieved in terms of spatial dimensionality, space and time resolution, or through parameterisation of the processes that are simulated.

General circulation models (GCMs) represent the most sophisticated attempt to simulate the climate system. The 3-D model formulation is based on the fundamental laws of physics (conservation of energy, conservation of momentum, conservation of mass, and the Ideal Gas Law). To compute the basic atmospheric variables at each gridpoint requires the storage, retrieval, recalculation and re-storage of 10⁵ figures at every time-step. Since the models contain thousands of grid points, GCMs are computationally expensive. However, being 3-D, they can provide a reasonably accurate representation of the planetary climate, and unlike simpler models, can simulate global and continental scale processes (e.g. the effects of mountain ranges on atmospheric circulation) in detail. The spatial resolution of GCMs, however, is limited in the vertical dimension. Figure 1 shows that different GCMs do not exactly show the same results e.g concerning humidity. Therefore, it is crucial to compare distinct models and their results in order to improve them. (Joussaume and Taylor, 1998)
 

1- Comparison of GCM models
 
 

Through much of the history of climate modelling, division between modelling and observational studies has hampered the development of both sides. The coupling of these theoretical and empirical disciplines, to test both model accuracy and understanding gained from analysis of observational data, has only recently been addressed. Such a divide must be bridged if accurate forecasts of future climates are to be produced.