The synthetic seismograms put us a few steps closer toward being able to accurately anticipate the ground shaking and other effects of likely earthquakes. The sounds presented here are predictions of ground shaking for seismic recording stations in southern Nevada, many in the Las Vegas urban area. On the map at right showing the 3-d computation area including Las Vegas Valley and the location of the 1992 magnitude-5.5 Little Skull Mountain earthquake, the locations of some of these stations are shown as blue squares and yellow and green triangles. Yellow lines are potential earthquake fault traces. Brown diamonds mark recent earthquakes. (You can map recent Nevada earthquakes yourself at http://mapserver.library.unr.edu/ website/seismoweb/RealtimeEQ/viewer.htm.)
The synthetic recordings have been speeded up by a factor of 200, so you hear each of the 23 stations' 200-second recordings in just one second. Since the time sampling of the synthetic is at just 40 Hz, the waves have a maximum frequency of 20 Hz. In practice, the mechanics of the finite-difference computation yield an upper frequency limit to the waves of only 0.5 Hz, or 2 seconds period. (Mathematical details of this limitation can be found in a UNR Geophysics course exercise for seniors.) The objectives of the modeling process here are similar to those of a music synthesizer, or the rendering of audio from MIDI instruction codes. We seek to re-create the physical representation of a wave-generation process in the computer (an earthquake instead of a musical instrument), and propagate those waves through the medium between the source and the receiver (a seismometer instead of an ear). Our work emphasizes the accurate representation of the propagation, so we develop detailed geological and geophysical model volumes, where an acoustic engineer strives to accurately represent the concert-hall environment.
Sped up by a factor of 200, you will hear the waves at about 100 Hz. In the left channel you will hear the NW component of ground vibration, and the NE component in the right channel. The vertical, up-down component has been added to both channels. The recordings proceed approximately in the order of station distance from the earthquake, starting at the epicenter. Every second you listen to the same absolute time interval, and continually step away from the epicenter.
At the beginning of the recordings you will hear the sharp thumps of the body waves arriving first. You may hear muted echoes from the sides of the computation box. But the energy soon converts to drawn-out horizontal vibrations of energy trapped in the soft sedimentary basins that are sprinkled through this region like the holes in Swiss cheese. This trapped energy has the highest amplitude and presents the greatest shaking hazard to Las Vegas. Though the trapped energy sounds like noise, these synthetics have clean linear wave propagation with no noise or stochastic effects added.
The movies showing the wave propagation illustrate the trapping and amplification well. As in the graphic on the right, each frame of the movies present a map of 3-component ground motions for the tilted region on the map. The movie frames are 300 km wide from NW to SE and 200 km high from SW to NE. The movies proceed at two and a half times the real modeled time. The three primary computer display colors of red, green, and blue (RGB) are used to represent the three directional components of ground vibration X, Y, and Z, respectively. Each color is given an intensity related to the intensity of shaking motion in the respective direction. A black color indicates very little ground motion; red is motion in the X direction (horizontal on the screen); green is motion in Y (vertical on the screen); and blue is motion in Z (in and out of the screen). Where shaking directions combine, the colors combine according to the rules of colored light- yellow indicates combined horizontal motion (relative to the ground) of X and Y, adding red and green light, so could be north-south or east-west. White color, adding red, green, and blue all together, indicates high-intensity shaking on all components, including up and down. With these colors, P waves will be mostly blue, S waves red, green, or yellow; and the Rayleigh surface wave is identifiable by having blue up-down motion between the red, green, or yellow radial motions (elliptical particle motions).
We present two different models here, to explore the role of all the various basins in the region. The first model has all the basins known from geological and geophysical studies. The second model only has the Las Vegas basin. Certainly you can hear a substantial difference between the models for the stations not in Las Vegas. But our research question is whether there are important differences between the predictions of these models within the Las Vegas basin.
Synthetic shaking audio, all regional basins (768 kb)
Synthetic shaking video, all regional basins (2.1 Mb)
Synthetic shaking audio, no regional basins (532 kb)
Synthetic shaking video, no regional basins (3.7 Mb)
These recordings and videos are also available as Podcast episodes for Apple iTunes and iPod listeners. Subscribe to http://crack.seismo.unr.edu/sounds/sound-of-seismic.xml. A new episode will be posted each month.Each of the plots below is in the approximate order of the stations' distance from the epicenter, increasing from top to bottom. Time increases toward the right in these plots. The top plot is the model with all regional basins; the model with the Las Vegas basin only is below: