A strategy for high‐resolution ensemble prediction. I: Definition of representative members and global‐model experiments

Abstract

AbstractIn the last few years, tens of alternative weather forecasts have been made available to forecasters by operational ensemble prediction systems. In many forecasting applications, it is useful to identify (possibly in an objective way) a few representative ensemble members, deemed to represent the most interesting weather scenarios. In this paper, a strategy to select representative members (RMs hereafter) from an ensemble prediction is developed, and applied to four cases of medium‐range ensemble forecasts performed with the European Centre for Medium‐Range Weather Forecasts (ECMWF) Ensemble Prediction System (EPS). The four case‐studies correspond to events of very intense rainfall (leading to localized floods) in the Alpine region, selected as benchmarks for numerical simulations in the Mesoscale Alpine Programme. The RM selection procedure uses a cluster analysis of the ensemble forecasts as its first step. For each cluster, an RM is defined to be the member with the smallest ratio between its average distance from the members of its own cluster and its average distance from the members of the other clusters. Distances are computed either using an L2‐norm applied to 700 hPa geopotential height fields or an L1‐norm to precipitation fields.RMs are compared with cluster centroids in the four case‐studies of extreme rainfall. By definition, RMs are characterized by a synoptic‐scale atmospheric flow similar to the flow of the corresponding cluster centroid, but they contain more small‐scale features, especially in the prediction of weather parameters such as precipitation.RM initial conditions can be used to initiate higher‐resolution global forecasts; alternatively, RMs may be used to define initial and boundary conditions for nested high‐resolution forecasts with limited‐area models. Integrations of RMs with the ECMWF global model at T1 319 horizontal resolution (compared with the T1 159 resolution used in the EPS) were performed. Results indicate that each higher‐resolution forecast, started from RM initial conditions, remains closer to the low‐resolution RM than to other ensemble members, but provides a more detailed forecast of weather parameters, especially in regions of complex topography. Experiments with a nested limited‐area model, started from the same set of RMs, are described in a companion paper.