Electronic Supplement to
SCEC Broadband Platform: System Architecture and Software Implementation

by Philip J. Maechling, Fabio Silva, Scott Callaghan, and Thomas H. Jordan

The file that may be downloaded below contains text-based Broadband Platform (BBP) input files that were used to perform a validation ground-motion simulation of the 1989 Mw 6.94 Loma Prieta earthquake. The figures below show the output data products produced by the BBP using the SDSU simulation method. The input files are provided as an example of the earthquake parameters used to define a BBP ground-motion simulation. The output data product images are provided to show the variety of results that can be produced by the platform, including output files that compare simulated ground-motion results against observed ground-motion data for the Loma Prieta earthquake.


Broadband Platform Input and Output Files for Ground-Motion Simulation

Download: bbp_input_files.zip [Zipped Broadband Platform (BBP) input configuration files; 5 KB]. This file contains the following ASCII-text based BBP input files that were used to configure and run a BBP ground-motion validation simulation for the Loma Prieta earthquake:

SRC Source Description (lomap-sdsu-0021.src): This file contains the definition for one source realization for the 1989 Loma Prieta earthquake. It consists of several key-value pairs describing the rupture, including the following:

  1.   FAULT_WIDTH (km)
  2.   FAULT_LENGTH (km)
  3.   DEPTH_TO_TOP (km)
  4.   LAT_TOP_CENTER (degrees)
  5.   LON_TOP_CENTER (degrees)
  6.   MAGNITUDE (Mw)
  7.   STRIKE (degrees)
  8.   DIP (degrees)
  9.   RAKE (degrees)
  10.   HYPO_ALONG_STK (km)
  11.   HYPO_DOWN_DIP (km)
  12.   DWID (km)
  13.   DLEN (km)
  14.     CORNER_FREQ (Hz)
  15.   SEED (numerical seed value)

STL Site List (lomap_v13_1.stl): Site list (in ASCII BBP Site List format) used in the Loma Prieta validation simulation. The columns in this file correspond to longitude, latitude, station name, Rrup (km), VS30 (m/s), low-pass frequency (Hz), high-pass frequency (Hz), and comments.

Velocity Model (genslip_lp_generic1d-gp01.vmod): Velocity model file for the northern California region used in the Loma Prieta simulation. The velocity model consists of 18 layers, and the six columns of the file are depth (km), VP (km/s), VS (km/s), rho (g/cm3), QP, QS.

The BBP produces simulation results as text-based data files and as PNG images. Examples of BBP output data files, including ground-motion time series (seismograms), are provided in the zipped file. The figures below plot contents of the BBP data files. User can use the text-based data files to reprocess the BBP output data or to reformat the images for their own purposes.

Download: bbp_output_files.zip [Zipped Broadband Platform (BBP) output data files; 17.2 MB].. This file contains text-based output data products produced by the BBP Loma Prieta validation simulation.

XML Workflow Specification (10000021.xml): The BBP uses this XML-based file to define which input files, and which computational modules, are used in a BBP simulation. These XML files can be created for the user by using the BBP text-based interface. Once created, the BBP XML-based workflow specifications can be used to run and re-run broadband simulations.

SRF Rupture File (loma-sdsu-0021.srf): This is a Standard Rupture Format (SRF) rupture file produced by the GP rupture generator for the Loma Prieta earthquake. This SRF format file defines the slip time history of the rupture, defined in the input SRC file.

Velocity seismogram (10000021.8001-CLS.vel.bbp): Loma Prieta three-component velocity seismogram calculated by the Broadband Platform for station 8001-CLS using the SDSU method. The four columns in the file are time (s), north–south (cm/s/s), east–west (cm/s/s), and up–down (cm/s/s).

Acceleration seismogram (10000021.8001-CLS.acc.bpp): Loma Prieta three-component acceleration seismogram calculated by the Broadband Platform for station 8001-CLS using the SDSU method. The four columns in the file are time (s), north–south (cm/s/s), east–west (cm/s/s), and up–down (cm/s/s).

RotD50 amplitudes (10000021.8001-CLS.rd50): RotD50 amplitudes calculated by the Broadband Platform for station 8001-CLS. The four columns on the file are period (s), PSA 5% north (g), PSA 5% East (g), and RotD50 (g).


Figures

Figure S1. Station map showing the rupture (black), the stations (red), and the hypocenter projection on the surface (yellow star).

Figure S2. SRF slip plot generated by the GP method for the 1989 Loma Prieta earthquake. The slip contours correspond to slip propagation away from the hypocenter across the fault plane at 1 s intervals within the rupture.

Figure S3. Three-component velocity seismogram calculated by the Broadband Platform for station 8001-CLS.

Figure S4. Three-component acceleration seismogram calculated by the Broadband Platform for station 8001-CLS.

Figure S5. Comparison plot of velocity seismograms between simulated data (red) and recorded observations (black). The two horizontal components are shown. The bottom plots feature a comparison of the integral of Arias intensity as a function of time (red, simulated data; black, observed data).

Figure S6. RotD50 PSA comparison plot showing the observed amplitudes (blue) against the calculated amplitudes (green). The dashed red line on the right marks the period after which the comparison should not be considered (due to limitations of the recorded data).

Figure S7. The standard goodness-of-fit bias plot produced by the Broadband Platform for a validation simulation. It compares calculated data against observed data for periods from 0.01 to 10 s. In a perfectly matching simulation, the red line would be a flat line over zero. Where the red line stands above zero, the simulation is underpredicting the ground motion; and, where the red line goes below the zero line, the simulation is overpredicting the recorded data. These bias plots are based on ground-motion amplitudes calculated for multiple station sites that detected the earthquake. The yellow band represents the 95% confidence interval of the mean for all sites, and the cyan band shows the standard deviation for all sites. The top plot shows RotD50, while the bottom two plots show the two horizontal components.

Figure S8. Linear distance goodness-of-fit plot over distance using a linear scale on the x axis. Each of the eight plots shows a comparison of recorded data against the calculated data. Each dot represents a station.

Figure S9. Log distance goodness-of-fit plot over distance using a log scale on the x axis. Each of the eight plots shows a comparison of recorded data against the calculated data. Each dot represents a station.

Figure S10. Map goodness-of-fit plot showing how calculated data compares against recorded observations. Each dot represents a station. The rupture is shown as a black line, and the yellow star points to the projection of the hypocenter location. In the color scale on the right, green indicates a match between calculated and observed amplitudes. Dots colored toward the blue end of the spectrum represent stations where the simulation underpredicts the observed ground motions, and stations on the red side indicate locations where the simulation overpredicts the recordings. Eight different periods are shown, one on each subplot.

[ Back ]