3. How accurate or reliable are the results of climate models?

A central aspect of climate research is the question of how realistic the models actually are. How reliable are, for example, statements on the consequences or the speed of climate change? To find out how well the different climate models reflect the reality, the various climate modeling teams from all over the world regularly compare their models.

For this purpose, they feed their climate models with the same data, such as the development of greenhouse gas concentrations over a certain period, and then compare the results of the simulations - for example average temperatures for Northwest Europe in 2100. The differences in these simulations are a valuable indication of the uncertainty of the models.

Since no one knows what the climate will actually be 100 years from now, the scientists also take into account how well the models simulate the climate of the past years and decades for which accurate measurements exist. How well the models resemble reality depends above all on how well the physical, biological and chemical processes in the atmosphere and in the oceans are represented by the program codes.

It is particularly difficult, for example, to simulate how the clouds change with climate warming. Such changes, however, could significantly intensify or mitigate global warming. Clouds interact differently with the sunlight and with the thermal radiation emitted from the earth surface. On the one hand, clouds strongly reflect the sunlight and thereby cool the planet. On the other hand, they can also warm the earth by acting in a similar way as greenhouse gases. How well the climate models depict such complex interactions depends on how the physical processes are simulated. Since the different climate modelers program their models in slightly different ways, the results are automatically different.