Probabilistic forecasting from ensemble prediction systems: Improving upon the best‐member method by using a different weight and dressing kernel for each member

Ensembles of meteorological forecasts can both provide more accurate long‐term forecasts and help assess the uncertainty of these forecasts. No single method has however emerged to obtain large numbers of equiprobable scenarios from such ensembles. A simple resampling scheme, the ‘best member’ method, has recently been proposed to this effect: individual members of an ensemble are ‘dressed’ with error patterns drawn from a database of past errors made by the ‘best’ member of the ensemble at each time step. It has been shown that the best‐member method can lead to both underdispersive and overdispersive ensembles. The error patterns can be rescaled so as to obtain ensembles which display the desired variance. However, this approach fails in cases where the undressed ensemble members are already overdispersive. Furthermore, we show in this paper that it can also lead to an overestimation of the probability of extreme events. We propose to overcome both difficulties by dressing and weighting each member differently, using a different error distribution for each order statistic of the ensemble. We show on a synthetic example and using an operational ensemble prediction system that this new method leads to improved probabilistic forecasts, when the undressed ensemble members are both underdispersive and overdispersive.