Overview

Much of our current understanding of the Arctic climate system comes from numerical models. The beauty of models, which range from the simple to the very complex, is that they help us to understand interactions and feedbacks between variables and system components that can be difficult, and often impossible, to address from observations alone. A powerful tool in this regard is sensitivity experiments, whereby results from a control simulation, with “standard” model inputs and physical treatments, are compared to results in which particular inputs or physical treatments are perturbed, thereby capturing the importance of the perturbation on the modeled system. Models can also provide information on variables which are only sparsely or infrequently observed, such as evapo-transpiration. From previous chapters, the reader should already be familiar with some of the uses of models. In Chapter 5, we discussed aspects of cloud radiative forcing based on output from radiative transfer models. Chapter 7 briefly examined the application of thermodynamic models to understand sea ice growth. Studies of the atmospheric circulation in Chapter 4 made wide use of the NCEP/NCAR reanalysis system, which represents an optimal blending of a global atmospheric model and observations. Other applications will be reviewed in following chapters.

Modeling is probably the most rapidly evolving field in climate science and any attempt to define the “state of the art” is in danger of quickly becoming dated.