资源环境科技发展态势分析平台

Seismic and wind data from 27 stations are used to evaluate the generalizability of gust-seismic response functions and determine whether gust occurrence and magnitudes can be retrieved from seismic time series. Using wavelet analyses it is shown that there is a well-defined gust signature in a seismic frequency band of 0.01Hz-0.1Hz. The magnitude of this seismic response scales with the wind gust magnitude, so a gust-seismic relationship is developed using a least-squares polynomial fit wherein the variance of the acceleration of the horizontal seismic components is related to the gust intensity. A new gust detection method is then developed using the threshold obtained from the polynomial fit. Contingency tables indicate that the average proportion correct and hit rates for gust detection are 84% and 80%, respectively. This approach could be applied to high-resolution seismic arrays to detect, characterize, and quantify wind gusts with unprecedented spatial and temporal details.