Improved understanding of wind gust climates may be of great value to the wind energy industry, and is currently hampered by a lack of high-quality in situ data in wind resource rich environments. Thus, we are examining the potential to supplement anemometry with data from seismometers, including those deployed as part of the USArray Transportable Array (TA). Two models of the relationship between gust magnitude and ground motion are evaluated based on their skill at describing the distribution of gust wind speeds over 1 year using seismic data.
The approach is illustrated, using observed gust magnitudes obtained from sonic anemometers located at or near the 15 TA seismic stations. One deterministic and one probabilistic wind-seismic model are conditioned using one year of 5-minute resolution data and tested on a second year of independent data. Both models relate the variance of ground acceleration in the frequency range of 0.01 to 0.1 Hz (P) to gust speed (Ug) but differ in their functional form. The probabilistic model is found to perform well in predicting the gust distribution in independent data, and at the 15 sites considered herein has an integrated error across the entire cumulative probability distribution that is only 5% of the mean gust magnitude.