WunderBlog Archive » Dr. Ricky Rood's Climate Change Blog

Category 6 has moved! See the latest from Dr. Jeff Masters and Bob Henson here.


By: Dr. Ricky Rood, 7:07 AM GMT on March 20, 2007


Models are used in just about every field of human endeavor--think architecture. The basic dictionary definition of a model has the following ideas in it; it's a description of a system or phenomena that accounts for its known or inferred properties and may be used for further studies of its characteristics. Models are used to both diagnose cause and effect in observations and to make predictions. Since it is difficult to set up experiments in the atmosphere, the ability to make predictions with models takes on (part of) the role of experimentation in the classical definition of science. If a model makes a bad prediction, then we often get information about how to improve models. If a model makes a good prediction, then... then, we might know something about cause and effect. But, we might have the right answer for the wrong reason. Therefore the ability to make predictions again and again coupled with the ability of multiple independent groups to make high quality predictions add confidence that we do know something about cause and effect. This idea of confidence contributes to the specification of certainty estimates in IPCC 2007.

Since models are at the heart and soul of the climate change problem, I will take some time to work through the concepts of a model. First, there are different types of models, some simple, some complex. There are models based on physical principles like Newton's Laws of Motion; there are models based on statistical descriptions of observed variability(for example, average and standard deviations). All of them have their role. It is, however, the physical model that we care most about. This is because if the model can use the laws of physics to both explain cause and effect and to make a good prediction, then we have a solid foundation for determining our confidence in the predictions. Physics: I remember telling people I was a physics major in college, and it being a definitive conversation ender. The physics of the Earth is both simple and complex. It is simple because it is based on well established, tested ideas that have been around hundreds of years. Complex because there are so many things that are important to the climate, and they exist on large and small spatial scales, on fast and slow time scales. The last 3 or 4 blogs are just beginning to hint at this complexity.

I usually start to explain models with a household budget. If you have income and expenses, then you have the basic idea. The amount of money that you have tomorrow is equal to the amount that you have today, plus any income, minus any expenses. You could write an equation for this ... but I'll try to use this figure.

Figure 1: Schematic of the Earth Sun System to illustrate the conservation of energy during a stable climate.

A budget is the idea behind the conservation principle; for instance, the conservation of energy. Consider the Earth sitting in space; draw a circle around the Earth. Assume that the Earth is at an approximately steady temperature over, say, a year. If this balance is true then the energy that comes into the Earth (income) must equal the energy that leaves the Earth (expense). It balances, like a budget. The figure represents this straightforward and meaningful model.

The energy that comes to the Earth comes from the Sun. Without the Sun we'd freeze. Yes, there is some geothermal energy, but it would not keep us warm. So if our energy comes from the Sun, then there are two basic ideas that are important for the energy balance. How much of the Sun's energy is reflected and how much of it is absorbed. If absorption and reflection are changed then the temperature of the Earth would change.

Models: So with a model we have to identify and quantify all of things that change absorption and reflection. Then we have determine how they relate to each other--if one thing changes, does the other change? This should begin to hint at how model experiments are set up to disentangle, for instance, how we decide whether or not we are seeing a trend or variability due to the North Atlantic Oscillation. We'll do more next time.


The views of the author are his/her own and do not necessarily represent the position of The Weather Company or its parent, IBM.