Toward a refined estimation of typhoon wind hazards: Parametric modeling and upstream terrain effects

Abstract Parametric and stochastic typhoon model enable a rapid estimation of wind hazards in typhoon-prone regions. It always needs a large amount of historical track information to develop the track, intensity as well as wind field parameters. This study describes a technique for estimating two commonly used typhoon wind field parameters, i.e. R m a x , s and B s using the observed wind information from the best track dataset coupled with a semi-analytical wind field model. At each timestep of every typhoon event, the radial wind speed profile is well reproduced with an optimal pair of R m a x , s and B s . The correlation analyses of R m a x , s and B s with other parameters are conducted. The R m a x , s and B s at different timesteps allows the development of recursive models accounting for their autocorrelations between adjacent timesteps. Linearly weighted progressive formulas of R m a x , s and B s using all data extracted in the Western Pacific domain are developed. This idea is similar to the track and intensity models during stochastic typhoon simulations, which provides a forward step towards the more rational estimation of typhoon hazards. Moreover, the typhoon-event-specific R m a x , s and B s enables the reconstruction of historical wind hazards. By introducing the underlying terrain effects on wind speeds in terms of a directional roughness length and a topographic speed-up factor, 184 observed landed or offshore typhoon-scale storms along the China coastline from 1977 to 2015 are investigated. A dataset regarding the wind hazard footprints for over-water, roughness only and roughness and topography combined conditions of these 184 storms is developed to facilitate risk assessment and disaster mitigation during typhoon events.