A climate model is a computer program where you can test the world. What happens if you turn up the greenhouse effect in the model? Does it rain more or less? Does the temperature increase more in the Arctic than at the equator? How much ice melts in Antarctica?

Climate research institutes around the world have their own climate models. Why? Any Earth system model is a simplification of the real world, and the researchers have to choose how to prioritize and perform these simplifications. The structure of the models is similar, and they are all built on the same physics, but each model is a little bit different from the others. This means that their results may also vary. Together, the models cover more possible outcomes than any single model can do.

In the Climate Model Intercomparison Project (CMIP), the institutions share the data from the models, allowing anyone to analyze and compare the results. They also agree on specific tests to run. This may be for historic and future periods in time or for different conditions such as only natural or only man-made changes.

CMIP5, the fifth study in the series, was the foundation for the fifth assessment report from the Intergovernmental Panel on Climate Change (IPCC AR5), which was released in 2013.

By running all of these models, the collective results provide a better overview of the world than any single model could to. It would be impossible for any of the climate research centers to run all the models on its own. Thus, CMIP is an international team effort to produce as much knowledge about the Earth’s climate as possible.

Results from the Norwegian Earth System Model (NorESM) were part of CMIP5, and our scientists are now developing results for CMIP6.

The CMIP models can be used to run various kinds of experiments. Some are described in other articles on this web page, but the most well-known are the future scenarios – or projections – run for the IPCC.