Watch this video to learn about the basics of climate modeling.
Constructing a climate model involves a number of steps:
Building a 3-D map of the Earth's climate system. The basic building blocks of climate models are 3-D "grid cells" that contain climate-related physical information about a particular location. Within each grid cell, key physical, chemical, geological, and biological climate processes are represented by mathematical equations, many of which are based on physical laws, including the conservation of momentum, mass, and energy. Processes that cannot be depicted by physical or chemical laws are represented by equations derived by observing how things behave in the real world (empirically derived).
Developing computer code. The mathematical equations representing Earth's climate system are then translated into computer code, which is typically divided up by climate component--the atmosphere, ocean, clouds, etc. Because these various parts of the climate system interact, additional computer code is needed to allow different pieces of the model to talk to each other--both between components and from grid cell to grid cell.
Making the model run through time. Models also incorporate the element of time, simulating what happens within each grid cell at one time, then stepping forward in time (hours, days, months, or years at a time). Shorter time steps require more computing power.
How are climate equations developed and tested? The accuracy of equations that describe the behavior of the climate system are evaluated by putting them into models and evaluating the model results validating them against past observations. When an equation is shown to accurately represent the climate system, it is used again and again.