Ensemble calibration of ECMWF's medium-range forecasts

A post-processing method is presented, that is aimed at enhancing the value of ensemble forecasts through ensemble calibration. This approach can reduce model errors. In March 2008, ECMWF introduced the VarEPS/Monthly system. Part of this service provides a weekly reforecast dataset for the last 18 years. This dataset has been successfully applied for ensemble calibration at the Hungarian Meteorological Service. The calibration was performed for surface meteorological variables such as 2m temperature, minimum and maximum temperature, 10m wind speed, and daily precipitation. An objective verification was performed on the basis of several deterministic and ensemble verification scores. The calibration method was introduced operationally in mid-2009.