Simulation of non-stationary winds in typhoon field based on turbulent coherent structure and AR-GARCH model

A key challenge in simulating nonstationary wind speeds lies in accurately modeling the time-varying mean wind speeds and/or the standard deviation of the fluctuating wind components. This difficulty is primarily due to limited understanding of the mechanisms and evolutionary characteristics of wind speed nonstationary, which can currently only be described statistically in the time-frequency domain. In this study, ten sets of field-measured wind speed data from five typhoon events were analyzed to extract turbulent coherent structures, which are generally regarded as intrinsic drivers of wind nonstationarity. Their spatial and temporal distributions, as well as their evolution within typhoon wind fields, were systematically examined. Based on these insights, a simulation method for nonstationary typhoon wind speeds was proposed. Specifically, all datasets were decomposed into nonstationary coherent structure components and stationary fluctuating wind components. The parameters of an autoregressive generalized autoregressive conditional heteroskedasticity (AR-GARCH) model and a normalized spectral model were then fitted separately to these decomposed components. The simulated results were evaluated using the cumulative distribution function, cumulative normalized arias intensity, and power spectral density. The results show that the proposed method can effectively reproduce the time-frequency characteristics of measured nonstationary wind speeds associated with typhoon events.