When the world has a question about dinosaurs, a bunch of paleontologists get out their little shovels and picks and go to town in the dirt. We can see the work they put it. This work in the field is called in situ. But they are not required to make predictions about whether or not dinosaurs will return. What about when the world has a question about bacteria growth in the walls of our children’s schools? Biologists pull out a stack of Petri dishes and do lots of experiments in controlled environments to learn exactly how the bacteria behaves. Laboratory work is called in vitro. Then the biologist is expected to make a prediction about how the bacteria will act in the walls of the school. So, what happens when the world wants to know more about climate?
Climate scientists have a bit of trouble with their charge. The world wants climate scientists to predict the climate. But how much evidence is really available? The paleontologists have very limited evidence (only what they can dig up that is still in good condition) but they are not expected to predict the future. However, the biologist can run many many experiments and must make an educated guess about the future. Climate scientists are also restricted to gathering evidence based on what we can dig up and what we can measure today. But a measurement today can’t tell us about the climate. This is because climate is an average. So a scientist must make as many careful observations about the past and present as possible and then extrapolate into the future. Because in order to do a real experiment on the climate, we would actually have to add a ton of CO2 to the atmosphere and then sit around for hundreds of years to see what happens. This is not an acceptable solution!
Fundamentally, climate scientists have a real problem. We are living in our only Petri-dish.
The only solution is to build models! Climate scientists can use computer models to simulate the climate as they believe it to works. Then the computer models serve as petri-dishes. They can attempt any number of experiments on the models to determine exactly how the planet will behave. Well, not quite. They can determine exactly how they think the planet will behave. The petri-dishes are still constructed out of human minds- unlike the biologists who have actual bacteria to observe.
So climate scientists have a hard problem. How do they predict the future with very little evidence about the past? Through the use of models, climate scientists can make better predictions which will improve as the models improve. This has begun to be referred to as in silico (presumably for the silicon in computers!). Some mathematicians are quite good at improving models. These mathematicians are involved with climate change because we need to believe these models are representing truth. The quality and structure of the model is extremely important!