TESLA, A Tool for Automatic Earthquake Low-Frequency Spectral Level Estimation: The Study of 2013 St. Gallen Earthquake Fault-Plane Solutions

One of the challenges of seismicity monitoring is to achieve multiparametric catalogs complete down to small magnitude using automatic procedures. This can be obtained using seismic networks with high performance and robust, automatic algorithms able to process large data sets, limiting the manual operations of the analysts. The charac-terization of microseismicity is fundamental to study its spatial and temporal evolution and to define the seismic activity of fault systems. Among the source parameters of microseismic events, focal mechanisms are not generally calculated and, when available in the seismic catalog, their reliability may be dubious. We propose a new tool, named Tool for automatic Earthquake low-frequency Spectral Level estimAtion (TESLA), to automatically calculate the P- and S-wave low-frequency spectral levels. Indeed, it has been shown that these levels can be inverted together with P-phase polarities to better constrain the focal mechanism or to estimate the seismic moment. TESLA is designed to invert the P-and S-displacement spectra searching the optimal signal window to use for the spectral analysis. Using a signal window of fixed duration, although variable according to the earthquake magnitude, is not always the appropri-ate choice, especially when microseismicity is analyzed. TESLA performs three main tasks for both P and S phases: (1) a systematic exploration of several signal windows to use for the computation of displacement spectra, (2) the spectral analysis for all the selected signal windows, and (3) the evaluation of the best-displacement spectra through quantitative criteria and the estimation of the low-frequency spectral levels. The tool is first validated and then applied to the 2013 St. Gallen, Switzerland, induced seismic sequence to calculate the P and S low-frequency spectral level ratios, which are inverted to estimate focal mechanisms. Our results show the robustness of the tool to process microseismicity and the benefit of using it to automatically analyze large wave-form data sets.