"The answer is both yes and no", says Jörg Schwinger
Schwinger, from the Bjerknes Centre and NORCE, was one of the panel members at a debate during the final meeting of the Norwegian climate model project EVA in Oslo this week. His "yes" concerns the climate models.
"Climate models are our most important scientific tool. But the results may come too late to lead us out of a crisis."
The models are absolutely necessary to allow us to say anything about the future climate. Jörg Schwinger's "no" concerns the consequences. That models show increasing temperatures, is not enough. For society to take action, social science is also needed. Despite the "no", he thinks the situation is improving.
"The Paris agreement was possible only because knowledge about negotiations already existed", he says.
More details still needed
Trond Iversen, from the Norwegian Meteorological Institute and the University of Oslo, introduced the debate by pointing to the degree of detail needed to ensure trust in model results. The models must provide information that is detailed enough to allow society to adjust to local effects of global climate change.
In climate models, the Earth and its atmosphere are divided into cells, both vertically through the atmosphere and the ocean, and horizontally. The models calculate temperature, precipitation, ocean currents, wind and other variables in each cell. If the cells are too large, errors may occur in phenomena like clouds, heavy rain and extreme drought. This is just like the resolution of a photograph. Many small pixels or dots provide more information and give a sharper picture, than fewer, larger dots.
As computers have become more powerful, it has become possible to increase both the number of cells and to include new processes in the models. Still, as long as there are limits to this power, an increase in one will come at the cost of the other.
Trond Iversen emphasized the importance of prioritizing model versions with small cells and high resolution. Such models are needed to reduce the deviation between the calculated climate and the climate people experience locally, especially in Europe and parts of the Arctic.
He also asked for changes in climate variables to be presented with probability estimates, as risks, not as absolute values.
Nadine Goris from the Bjerknes Centre and NORCE agrees.
"We should be careful not to give politicians the impressions that climate models are already perfect and do not need to be developed further", she says.
A better model biology
EVA is the last of three large research projects that have worked to develop, run and manage the Norwegian climate model, or Earth system model, NorESM. This has been a cooperation between seven institutions (see box).
Leader of the EVA project, Christoph Heinze from the Bjerknes Centre and the Geophysical Institute at the University of Bergen is satisfied with the results.
"The project has contributed to improving the projections of our future climate", he says.
He points to the biological and chemical processes in the model as something that has been improved by this project.
That EVA now comes to an end, does not mean that the development of the Norwegian Earth system model will stop. The meeting in Oslo continues as a startup meeting for the successor: the infrastructure project INES.