Assessing the probability density function of urban wind-induced risk under typhoon conditions

Abstract

In the context of climate change, tropical cyclones (typhoons) of increasing frequency and intensity have become one of the primary challenges to cities located in the relevant geographic context. Enhancing these cities’ adaptability to climate risks and mitigating the catastrophic impacts of tropical cyclones require a thorough understanding of urban flow with extreme approaching wind conditions. This study employed Computational Fluid Dynamics (CFD) methods, in particular Embedded Large Eddy Simulation (ELES), to reconstruct the flow field in a real heterogeneous urban building complex featuring mixed low-rise buildings, and applied a statistical approach based on the 3-parameter Weibull distribution (3 W) to examine the probability density function (PDF) of wind-induced risks under typhoon conditions. Results showed that, with a fixed reference wind speed, the typhoon-condition pedestrian-level wind (PLW) environment strongly deteriorates due to larger upstream terrain roughness. Normalized time-averaged and extreme velocity at pedestrian level can increase locally up to 2.0 and 3.0 respectively with upstream terrain roughness length of 0.30 m, in contrast to 1.25 and 1.75 with that of 0.01 m. The building morphology however, does not significantly influence the PLW. There was a significant difference in the vertical variation of PDFs between different building morphologies. Deterioration of PLW brings higher wind-induced risk to pedestrians. Moreover, simple and steady criterion of identifying wind-induced risks proved to be insufficient. Apart from practical findings, this study demonstrated a feasible workflow of integrated statistical extraction of urban gust wind and wind-induced risk based on CFD results, while reducing the potential uncertainties brought by empirical methods. Such research workflow could be further applied to support city-scaled simulation and climate-disaster early-warning.