Multi-scale Modeling of Extreme Urban Wind Conditions under Climate Change

Many coastal cities, such as Hong Kong, have high population densities and numerousbuildings and are located in tropical cyclone (TC)-prone areas. Current standards andcodes generally provide design wind speeds based on typical terrain categories for anentire city. However, urban wind conditions during TC events are significantlyinfluenced by complex urban landscapes so that high-resolution estimation of the urbanwind environment is desired. To date, fine-scale modeling of extreme urban wind fieldsremains challenging due to high computational costs and uncertainties associated withatmospheric boundary layer (ABL) conditions. Additionally, the impact of climate changeon extreme urban wind conditions is not well understood and related research is limited.To this end, this proposed project will be aimed at developing a novel multi-scalemodeling framework to assess the urban wind fields during TCs under various climatechange scenarios.The proposed study will integrate the advanced Weather Research and Forecasting(WRF) model with Computational Fluid Dynamics (CFD) simulations to capture complexinteractions between large-scale atmospheric processes and fine-scale urban features.The WRF module will provide real ABL conditions for the CFD module. This frameworkwill allow for detailed estimates of urban wind environments, such as the wind velocitydistribution on a specific street or region and the wind effects on large-scale buildingsor infrastructures.To investigate the influence of climate change on extreme urban wind conditions, a newmachine learning-based TC detection and tracking model will be developed based onhistorical reanalysis datasets. Subsequently, the model will be applied to find out all TCevents in a multi-model ensemble consisting of multiple Global Climate Models (GCMs)with various Shared Socioeconomic Pathways (SSPs) provided by the Coupled ModelIntercomparison Project Phase 6 (CMIP6). And all TC events that pose potential threatsto Hong Kong will be identified and simulated using the proposed WRF-CFDframework. In addition, uncertainties throughout this modeling pipeline will bequantified and analyzed.Finally, this research will establish a high-resolution dataset characterizing urban windconditions for different return periods. The dataset will provide invaluable informationfor urban planners, engineers, and policymakers to improve the design of resilient citiesand hazard managements, which are particularly relevant for Hong Kong. Additionally,the research findings will contribute to broader scientific understanding of urbanmeteorology and climate adaptation strategies, offering new perspectives on how citiescan better address the threats of TCs and challenges posed by climate changes.