'Jumpiness' of the ECMWF and UK Met Office EPS control and ensemble-mean forecasts