Monitoring subsurface velocity variation using seismic ambient noise generated by wind turbines: A feasibility study

The ambient-noise tomography has been widely used to characterize subsurface models across scales. The wind farm generates a large amount of seismic energy, which provides a powerful cultural seismic source. The wind turbines (WTs) are typically distributed inhomogeneously in space, which makes it difficult to use the conventional ambient noise tomography that assumes a uniform distribution of noise sources in space. However, because the seismic noise generated by WTs is consistent in time, the wind farm can be used as a stable source although inhomogeneous in space, to monitor the variation of the subsurface model. In this paper, we used a seismic survey line to monitor the subsurface velocity variation of a wind farm area in Hubei Province of China in three months by using ambient noise generated by WTs. Based on the analysis of the dominant backazimuth of the noise source, we verified the relatively high spatial and temporal stability of ambient-noise data in the frequency range between 5 and 10 Hz, which is confirmed to be mainly generated by WTs. On this basis, we obtained the cross-correlation functions and surface-wave dispersion spectra from the ambient-noise data. We analyzed the temporal change of surface-wave phase velocity over three months, which shows the velocity variation of the subsurface model. This study proved the feasibility of using wind turbines as a noise source for the monitoring of subsurface properties.