Typhoon Fanapi (2010) and its Interaction with Taiwan Terrain – Evaluation of the Uncertainty in Track and Rainfall Simulations

Typhoon Fanapi (2010) produced heavy rainfall over the mountain area of Taiwan, with the maximum 2-day accumulated rainfall of 1127 mm over the ridge of southern Central Mountain Range (CMR). A revised version of the WRF-based EnKF data assimilation system and all available ITOP data are integrated to investigate Fanapi's track, in particular the southward deflection before landfall, and to evaluate the impact of CMR on the uncertainty of typhoon structure and the associated rainfall.

The simulation results show that the characteristics of typhoon track, intensity and structure are generally well captured and the corresponding ensemble spread remain small throughout the integration, implying a case of high predictability. It is shown that Fanapi's distinct track deflection, which occurs just before the Taiwan landfall, can be attributed to the terrain-induced TC asymmetric mean flow in the mid-troposphere. The track deflection is smaller in a simulation where Taiwan terrain is replaced by the flat ocean. The ensemble simulations also suggest that the terrain-induced track deflection is more sensitive to the TC size and translation speed than to TC intensity.

The sensitivity experiments show that the rainfall distribution is markedly changed when the terrain of Taiwan is removed. The effects of Taiwan's topography increase the track deflection and enhance rainfall and intensity variability, but not the spread of TC positions. A quantitative comparison based on the Equitable Threat Score (ETS) and pattern correlations calculated in two specific mountainous areas (southern Taiwan and northeastern Taiwan) with heavy rainfall signature is conducted to evaluate the rainfall forecast skill. Though the control experiment shows better skill in rainfall forecast in the two selected area than two other simulations, i.e., simulations with half of Taiwan terrain and no Taiwan terrain, the rainfall skill is still limited. The result indicates that the model fails to demonstrate enough skill in producing the extreme rainfall amounts in the mountainous areas. In contrast to the low skill regarding ETS and pattern correlation analyses, the small spreads among ensemble members demonstrate the predictability of Fanapi's track and the associated rainfall over Taiwan.