Supermodelling climate

Seminar by Dr. Noel Keenlyside from Geophysical Institute, University of Bergen and Bjerknes Centre Nansen Environmental and Remote Sensing Centre

01 July 2024
KST 14:00 – 15:00

The Seminar is being held in Room 1010 (Jasmin) – Integrated mechanical engineering building. Click here for the campus map.

Climate models are plagued by long-standing biases. Increasing resolution to km-scales is not the solution, as recent massive ongoing efforts have shown. In contrast, supermodelling—an advanced physics-informed machine learning approach—can radically reduce biases and improve predictions. In supermodelling, different models are combined interactively so that errors can be corrected, and large-scale biases do not develop. In this talk, I will introduce the concept along with the progress we have made in supermodelling using state-of-the-art Earth System Models. I will show marked improvements in the simulation of tropical climate, and in seasonal predictions. In addition, I will demonstrate the potential of supermodels to improve the simulation of global climate and to reduce uncertainties in climate change projections.