Predictability of Tropical Cyclones: A Comparison of Typhoon Haiyan (2013) and Typhoon Hagupit (2014) Using Convection-Permitting Ensemble Forecasts

Although forecasting the intensity of tropical cyclones (TCs) remains challenging, track forecasts have improved considerably over the past few decades. However, there remain cases where numerical weather prediction models are unable to accurately forecast the motion of a TC. Typhoon Haiyan (2013) and Typhoon Hagupit (2014) are examples of two storms where, despite similarities in the track and intensity, the predictability of the storms differed greatly. Both storms made landfall over the Philippines having followed a similar track across the Pacific and both reached intensities in excess of 60 m s^-1. Operational ensemble forecasts showed large uncertainty in the track of Hagupit, whereas the ensemble spread for Haiyan was considerably less.

Using the Met Office's Unified Model, cutting-edge 4.4-km, limited area, convection-permitting, 5-day ensemble forecasts were produced for both storms along with 5-day global ensemble forecasts. Each ensemble (both convection-permitting and global) includes 12 members and forecasts are initialised 12 hours apart over a 6 day period. Consistent with the operational forecasts produced at the time of the storms, the spread of tracks is greater in the forecasts produced for Hagupit than Haiyan. The reasons for these differences are investigated using several techniques: (i) Comparisons of the three-dimensional structures of the storms, their intensities and their environments in different ensemble members; (ii) Steering flow calculations in the simulations both with the storm and following storm removal; (iii) Three-dimensional Lagrangian trajectories, to investigate the dynamical interactions between the storm and other large-scale weather systems, such as the outflow of convective systems over the Maritime Continent during Typhoon Hagupit. In particular, it is shown that prior to making landfall Hagupit slowed down as it entered an area between two high pressure systems, where the environmental flow advecting the storm was weak. It is at this point that the track forecasts begin to diverge.