This data set lists earthquakes that produce large ground motions recorded at seven small aperture arrays. For each station and each channel within the networks, we catalog events that produce the largest peak ground velocity (PGV) and largest peak ground acceleration (PGA) out of a suite of 4028 earthquakes (magnitude 2.5+; September 2010 - June 2023) that occurred within southern California.  All stations in the network must record data on all three channels. Data from seven networks are evaluated: Jackass Flat (JF), Ramona (RA), Transect (TR), Pinyon Flat (YN), Blackburn (BB), Sage Brush Flat (SG), and Dry Wash (DW).  Two of these seven networks (PY and TR) are located away from the primary fault system, and five span the San Jacinto fault (BB, DW, JF, RA, and SG).  Because these networks’ spatial footprints are small (< 3 km in diameter), the expectation is that stations within a given array would report the same earthquake as producing the largest ground motions across the entire array. That is not what we found. These results indicate that extrapolation from one location to regions within only a few kilometers away could be in error.

Peak ground motions derived from seismic waveforms (acceleration HN* channels and velocity HH* channels) that are outliers with respect to theoretical expectations (Abrahamson et al., 2014). Quantitative measurements are derived from the ratio of the theoretically predicted ground motions to the observed values. Ratios below (above) zero indicate the observed ground motions were greater than (less than) expected. These ratios are evaluated using MATLAB’s isoutlier routine, using the ‘mean’ method that flags elements that are three or more standard deviations from the mean.  The CSV file includes Network, Sta.Chan, Time, Mag, Depth_km, Lon, Lat, Epidist_km, Az_degree, MaxAmp, Rat_val, ID, high/low, keep/toss.

Peak ground motions derived from seismic waveforms (acceleration HN* channels and velocity HH* channels) that are outliers with respect to theoretical expectations (Abrahamson et al., 2014). Quantitative measurements are derived from the ratio of the theoretically predicted ground motions to the observed values. Ratios below (above) zero indicate the observed ground motions were greater than (less than) expected. These ratios are evaluated using MATLAB’s isoutlier routine, using the ‘mean’ method that flags elements that are three or more standard deviations from the mean.  The CSV file includes Network, Sta.Chan, Time, Mag, Depth_km, Lon, Lat, Epidist_km, Az_degree, MaxAmp, Rat_val, ID, high/low, keep/toss.

This data set lists earthquakes that produce large ground motions recorded at seven small aperture arrays. For each station and each channel within the networks, we catalog events that produce the largest peak ground velocity (PGV) and largest peak ground acceleration (PGA) out of a suite of 4028 earthquakes (magnitude 2.5+; September 2010 - June 2023) that occurred within southern California.  All stations in the network must record data on all three channels. Data from seven networks are evaluated: Jackass Flat (JF), Ramona (RA), Transect (TR), Pinyon Flat (YN), Blackburn (BB), Sage Brush Flat (SG), and Dry Wash (DW).  Two of these seven networks (PY and TR) are located away from the primary fault system, and five span the San Jacinto fault (BB, DW, JF, RA, and SG).  Because these networks’ spatial footprints are small (< 3 km in diameter), the expectation is that stations within a given array would report the same earthquake as producing the largest ground motions across the entire array. That is not what we found. These results indicate that extrapolation from one location to regions within only a few kilometers away could be in error.

This dataset contains an earthquake catalog for Southern California compiled from the USGS ComCat database, including events with magnitude ≥ 2.5 recorded between September 2010 and June 2023. The CSV file includes eleven columns: year, month, day, hour, minute, second, latitude (degrees), longitude (degrees), depth (km), magnitude, and event ID.Column DescriptionYearMonthDayHourMinuteSecondLatitude (degrees)Longitude (degrees)Depth (km)MagnitudeEvent ID (ComCat) Sample data:2010 09 20 00 30 02.400  32.0261667 -115.1083333 9.939 4.75 ci148442442010 09 20 00 50 53.990  32.0893333 -115.0796667 9.97 2.67 ci148442682010 09 20 00 54 18.960  32.0875 -115.0841667 9.976 2.69 ci148442842010 09 20 15 12 03.880  32.2143333 -115.2776667 8.712 2.89 ci148446602010 09 20 19 28 33.450  32.0831667 -115.1258333 9.936 2.67 ci14844796

This dataset contains an earthquake catalog for Southern California compiled from the USGS ComCat database, including events with magnitude ≥ 2.5 recorded between September 2010 and June 2023. The CSV file includes eleven columns: year, month, day, hour, minute, second, latitude (degrees), longitude (degrees), depth (km), magnitude, and event ID.Column DescriptionYearMonthDayHourMinuteSecondLatitude (degrees)Longitude (degrees)Depth (km)MagnitudeEvent ID (ComCat) Sample data:2010 09 20 00 30 02.400  32.0261667 -115.1083333 9.939 4.75 ci148442442010 09 20 00 50 53.990  32.0893333 -115.0796667 9.97 2.67 ci148442682010 09 20 00 54 18.960  32.0875 -115.0841667 9.976 2.69 ci148442842010 09 20 15 12 03.880  32.2143333 -115.2776667 8.712 2.89 ci148446602010 09 20 19 28 33.450  32.0831667 -115.1258333 9.936 2.67 ci14844796

The LHPOST, located about 15 km east of the UC San Diego main campus, is a unique shake table facility. It is the largest capacity 6-DOF shake table in the world. The LHPOST is designed to permit the accurate simulation of far- and near-field ground motions, and when used in combination with the two existing large laminar soil boxes allows for the testing of soil-foundation-structure interaction systems.

Deployment of 300 nodes on Piñon Flat in southern California

Calibration of reaction mass for LHPST seismic simulator funded by the National Science Foundation. The shake table will be able to reproduce multi-dimensional earthquake motions with unprecedented accuracy.

Nodal array deployed near Ridgecrest California. Funded by the ERC project FaultScan (PI Florent Brenguier)