Earth’s climate is a complex system with many interacting parts, including the atmosphere, oceans, land surfaces, ice, and the sun. To better understand how it works and to make projections for its future behavior, scientists use computer-based simulations known as climate models.
These models, according to the World Meteorological Organization (WMO), incorporate the physics and chemistry of the climate system’s various components. Using mathematical representations, they aim to answer questions such as when the next El Niño or La Niña event might develop and how the climate will respond to increasing amounts of greenhouse gas concentrations.
Developed from weather forecasting models, climate models look further out in time and involve an extremely large number of calculations. To run, they require very powerful computers. In fact, the WMO says it can take several months to complete a 50-year projection.
Dividing the planet into a three-dimensional grid, a climate model simulates the movement of air, water, and heat energy within each of its grid cells and evaluates interactions between them. All of these processes are based on the laws of physics. When all the processes from all the grids are linked together over time and space, the model simulates Earth’s climate.
As with weather, there are a number of different climate models in use around the globe. Some, according to the IPCC, perform better than others for particular aspects of the climate system. The variations in predictions that they produce are generally the result of differences in initial conditions, different parameters for interactions between parts of the system, and different estimates of future greenhouse gas emissions. That said, all the climate models included in the IPCC report agree that Earth’s average temperature is rising and expect it to continue to rise in the future.
To see if a climate model will perform well, it is tested against the past. This is a process called Hindcasting. If a model accurately predicts climate trends that have already taken place, it is expected to predict what might happen in the future with a reasonable amount of certainty. Models can also be refined as new sources of data become available.
Below is a short video by The National Academy of Sciences on the basics of climate modeling. Credit: NAS and YouTube.