Model independence in multi-model ensemble prediction
Gab Abramowitz · Australian Meteorological and Oceanographic Journal · 2010
This paper outlines the emerging research area of model independence in multi-model ensemble prediction, explains why it is important, and contextualises several published attempts to characterise it. In particular, we highlight model independence and model performance as two unrelated properties of model prediction, and by doing so address the common misconception that striving for independent model ensemble members promotes divergence from observed behaviour.