The above examples demonstrate that the concept of a catastrophe is fundamental to atmospheric dynamics. While some of the specifics of these examples may later turn out to be incomplete or even incorrect, the general concept of a catastrophe will still remain useful. When we study (quasi-)equilibria such as Rossby waves, cloud clusters, and gravity waves, we need to ask not only how they are maintained, but also whether and how they can become unstable, and how they transition to other (quasi-)equilibria.

Why are there so many catastrophes in the atmosphere? The answer has two components. First, as seen from the examples in the preceding section, multiple (quasi-)equilibria exist because one or both of the forcings that maintain a quasi-equilibrium are highly nonlinear functions of a gross measure of the physical system.

The atmosphere is full of factors contributing to nonlinearity. Among them are the earth's rotation, resonance, the structure of the external heating, air-sea interaction, etc. Second, there is an abundance of mechanisms that can act as triggers in moving a system from one (quasi-) equilibrium to another. Among these mechanisms are the annual cycle, the diurnal cycle (in spontaneous catastrophes), and various perturbations (in triggered catastrophes.) (The annual cycle and the diurnal cycle can be viewed as triggers in a larger sense.)

The reasons for nonlinearity are different for different atmospheric catastrophes. Finding them is the core challenge of studying atmospheric catastrophes. The examples given in this review may not be sufficient to serve as a complete guide to studying other atmospheric catastrophes, but they do point in the general direction.

Finally, how can we turn the knowledge gained from studying atmospheric catastrophes into practical use, such as improving the performance of forecast models? One possibility is, rather than assessing the performance of global models by comparing the seasonal mean state of the model with observations — the most common practice — one should compare the catastrophes simulated by the models with those observed — such as stratospheric sudden warming and monsoon onset. Most models have difficulties in simulating these events with fidelity. Through studying the models' failures, one can strive for improvements.

[Received 27 October 2006; Revised 29 March 2007; Accepted 27 May 2007.]

0 0

Post a comment