Seismic signature of rain and wind inferred from seismic data

Seismic stations are increasingly used to monitor river activity and to quantify sediment transport during flood events. In tropical regions, cyclone-induced floods are often associated with heavy rain and strong wind episodes, generating complex seismic records involving the simultaneous signature of water, sediment, rainfall and wind. Hence, seismic characterization of rain and wind is then required to better decipher each process and improve our understanding of the river seismic signature. In this study, we investigate experimentally the seismic response of rain and wind using data recorded by geophones deployed in various soil types and at different burial depth (BD), co-located with various meteorological instruments. Our results show that the power spectral density (PSD) of the seismic noise intensifies at a frequency between 60 and 500 Hz for rain and 5 and 500 Hz for wind, in the presence of rain precipitation as low as 0.025 mm/min and/or wind speed >= 3 m/s. PSD analysis indicates that the seismic signal associated with rain decreases with the BD with a value of similar to 2-5 dB in a depth difference of 10 cm. We also observe that each soil type has its own seismic signature. The 4-min root mean square correlation between the seismic signal amplitude and the rain precipitation suggests that they best correlate with Pearson coefficient >0.90 at BD of 30 cm. The transfer function between the precipitation rate (or kinetic energy) and the seismic signal amplitude shows that the signal recorded by the geophone can be used as a robust proxy for these parameters. Plain Language Summary In addition to recording deep Earth activity such as earthquakes or volcanic eruptions, seismic sensors can measure numbers of naturally-induced ground vibrations such as those generated by rivers activities. In tropical areas during cyclonic events, heavy rain (and strong wind) leads to the presence of severe flooding, during which several activities in and on the river (e.g., sediment transport, wind, water, and rain) generate ground vibration, making the overall signals recorded by seismic sensors complex to interpret. To better decipher the river seismic signature, investigating the wind and rain signatures are then essential. In this study, we investigate experimentally the seismic response of rain and wind from data recorded by geophones and complemented by different meteorological sensors. The obtained results show that (a) the strength of the seismic noise intensifies with frequency for rain and wind; (b) the seismic signal associated with rain decreases with the burying depth of the sensor; (c) the amplitude of the seismic signal correlates with rain precipitation when the sensor is well buried and (d) the signal recorded by the geophone can be used as a robust proxy for rain precipitation.