Statistical models were developed to emulate ensembles of process-based crop models.
They describe the between-crop model variability of the simulated yield data.
They can be used to compute mean yield loss and probabilities of yield loss.
Their interests were illustrated for maize, wheat, and rice.