The Added Value Index: A new metric to quantify the added value of regional models

A metric to quantify the value added by high-resolution models is introduced. It is based on a characteristic spatial distribution of skill rather than the averages of skill values. Normal distribution functions are fit to the model skill distribution of coarse-and fine-resolution models, and a new metric (Added Value Index, AVI) is defined as the area enclosed by the two distribution functions, with information on the way the two curves cross each other. The AVI is computed for a case of downscaling seasonal forecasts and is shown to properly provide a different degree of added value by high-resolution models.