Comparison of ensemble-MOS methods in the Lorenz '96 setting

A suite of methods that have been proposed for statistical postprocessing of ensemble forecasts based on historical verification data (i.e., ensemble-MOS methods) are compared with each other, and to direct probability estimates using ensemble relative frequencies, in the idealized Lorenz '96 setting. The three most promising methods are logistic regressions predicting probabilities associated with selected quantiles, ensemble dressing (a kernel density estimation approach), and linear regressions with nonconstant prediction errors that depend on the ensemble variance.