What we would like to know is how warming will proceed and affect future climate.
There are two ways to make predictions about warming and its effects:
Make assumptions about the future release of greenhouse gases and calculate, using computer models, how warming will proceed and climate will respond.
Use the geologic record to reconstruct periods of unusually warm climates and find out what the Earth’s environment looked like and how plants and animals responded.
These techniques for dealing with the question of climate change each have their strengths and weaknesses. They work best when combined.
The first approach, a simulated climate machine in the form of a computer program, contains the best ingredients of our knowledge, incorporating everything from the principles of radiation balance to the equations of motion. Such a machine consists of thousands of instructions of the kind “IF (A) THEN (B),” converting a momentary condition (A) into another condition (B) in one computing step. For example, (A) might describe the temperature of a small area on the ground, while (B) might refer to the radiation given off to the atmosphere, the evaporation experienced at this place, or the transfer of sensible heat to a neighboring area. The "forcing" of the model is ultimately caused by the known seasonal changes in solar radiation. The machine "responds" by producing seasonal cycles of surface temperature, rainfall, snowfall, cloud cover and winds. The more this "output" resembles the real world of changing seasons, the more credence is given to the program as a valid simulation of what is happening in the real world.
As mentioned, different computer models give somewhat different predictions when experiments are made to test their response to changing conditions. The reason is not that the physics of climate change is different from one scientist to another. The reason is that many of the physical processes are not sufficiently understood, or resolved in sufficient detail, to be amenable to exact computation. To fill in such gaps in knowledge, climate physicists make assumptions that differ somewhat from one experiment to the next.
An example of climate model design, illustrating the various factors that influence results of a “run” of the climate model. Different models may contain different factors depending on their complexity and the computing power that is available.
The chief benefit of the second approach - reconstructing past climate history - is the expansion of our knowledge of how the Earth behaves over time. Without the historical perspective we are trapped within the limited wisdom provided by physics and the experience of individual observers. Physics can tell us a lot about how the climate machine works. But only climate history, that is, long-term experience, can tell us whether we are witnessing highly unusual conditions or not.
When the results from the computer models and the knowledge about past climate are combined they show that the Earth is warming, and that it is doing so at a highly unusual rate and toward highly unusual conditions.