Characterizing Wind Noise and Spatial Variability on Near-Surface Broadband Seismometers

By coupling with the ground, wind causes ground motion that appears on seismic records as noise at a variety of frequencies, which can drown out features such as small earthquakes and prevent observation of normal modes. Understanding the processes by which wind-induced noise is generated in detail will help improve our ability to mitigate locally-generated seismic noise. Since wind speed and direction are such locally variable phenomena due to obstacles, diurnal heating, and changes in topography, it is likely that there will be differences in the wind speed and direction at locations only a few meters apart. This locally-heterogeneous wind field could cause substantial differences in the signals induced on seismometers even in relatively close proximity to one another, making it difficult to use wind speed recordings to mitigate wind-induced seismic noise. We investigated the spatial variability of wind using two weather sensors approximately 100 m apart collocated with four near-surface broadband seismometers. Consistent with other studies, we found that at long periods (>5 s), increasing wind speed increases noise on the horizontal components of these seismometers. We then quantitatively estimated the linear relationship between root mean square (RMS) wind speed and RMS seismic velocity for our sensors. While this increase in horizontal noise at long periods has been previously observed, we also found low coherence between the wind speed, wind direction, and pressure recorded by our two weather stations. Furthermore, we identified a previously-undocumented loss of coherence between nearby seismometers on the vertical component between 8 and 20 s periods, the intensity of which appears to depend on the substrate surrounding the seismometer. These data support an existing theoretical model of vertical ground displacement due to wind, where this displacement depends on the properties of the medium in which the instrument is emplaced. We also investigated the frequency dependence of wind-induced noise, and found that the dominant source of high-frequency seismic noise at some sites could be cultural activity rather than wind.