Full moment tensor of North American (depth 0-20km; M3-6; 1995-2018) with a focus on induced seismicity

The excel spreadsheet (provided separately “MT_all.xlsx”) includes the earthquake data used in this study. Each sheet contains a certain type of earthquake (I-induced, N-natural/tectonic) in a region. Due to the complexity of identifying the nature of a certain earthquake, the “induced” earthquakes are only used when one or more peer-reviewed publication has persuasively suggested its induced nature (see the comments in each sheet). The following explains the data source in each sheet. • WCanada_N: The focal mechanisms for the natural events in the WCSB are contributed by John Ristau (Ristau et al., 2007) through personal communication. Time range: 1995.6-2006.3. N=17. • WCanada_I: All the focal mechanisms for the induced events in the WCSB are resolved by University of Alberta broadband seismology group and have been revised/confirmed by Alberta Geological Survey. The lowest moment magnitude is 2.99 and the lowest Variance Reduction is 65%. Such criterions enhance our confidence in the resolved moment tensors. The focal mechanism of the event on November 10, 2016, is constrained as a pure double-couple and is not used for the nonDC analysis. Time range: 2013.12-2017.8. N=16. • CentralUS_I: For induced events in the central US, we limited the magnitude > 4 to ensure enough stations has been used during GCMT solution, therefore, maximize the accuracy of the non-DC events. Induced events are selected based on the HiQuake catalog (Wilson et al., 2017). We also use the focal mechanism from USGS as well as Saint Louise University Earthquake center for reference. Time range: 2011.2-2016.12. N=20 • CentralUS_N: Natural events only used when depth < 10 km, with the hypocenter location reported outside the popular induced seismicity regions (see https://www.usgs.gov/news/induced-earthquakes-raise-chances-damaging-shaking-2016 ) and has one or more peer-reviewed publication to prove its tectonic nature (e.g., Huang et al., 2017). With this over-restricted criterion, we minimize the possibility of including the potential induced earthquakes in the central US (time range: 2010.1-2017.8, 20< latitude <40, -112< longitude