Sensitivity experiments for ensemble forecasts of the extratropical transition of typhoon Tokage (2004)

Abstract The extratropical transition (ET) of tropical cyclones often has a detrimental impact on predictability in the vicinity of the event and downstream. Ensemble forecasts provide an appropriate means by which to investigate both the uncertainty and the dynamical development leading to the different ET scenarios. Sensitivity experiments are presented using the European Centre for Medium‐Range Weather Forecasts (ECMWF) ensemble prediction system (EPS) to investigate different methods of perturbing the ensemble forecast of the ET of Typhoon Tokage (2004). During ET these perturbations have a notable impact on the ensemble spread representing the uncertainty. Three experiments were performed: one of them without singular vectors (SVs) targeted on the tropical cyclone, the second without stochastic physics and the third excluding both perturbation methods. The targeted perturbations are most important for sufficient spread in track and intensity. Without the targeted perturbations, the analysis is not contained within the ensemble spread. Stochastic physics leads to stronger reintensification of the ensemble members after ET. The higher track spread leads to higher variability in processes such as lower tropospheric latent heat release. This can be related to a higher spread in the upper‐level midlatitude flow for both perturbation methods. A connection is drawn between the strength of ET and the modification of the downstream midlatitude flow pattern. The uncertainty due to the targeted perturbations propagates downstream with a Rossby wave train excited during Tokage 's ET. For the case of stochastic physics, the uncertainty spreads to the ridge directly downstream of the ET system but is not evident further downstream. Copyright © 2010 Royal Meteorological Society