You are here: README - 2002 Interactive Deaggregations

README - 2002 Interactive Deaggregations

Printer Friendly Version

Contents

Introduction

Thank you for coming to visit us on the interactive seismic hazards deaggregation web page.

The current National Seismic Hazards Mapping Project models (2002 edition) are being used in these interactive deaggregations.

At this web page, seismic hazard deaggregations can be run for sites in the conterminous (48) states. If the user uses this web page to perform deaggregations for non-US sites, it is likely that half or more of the hazard sources will be omitted, even near US borders. Please contact the Geological Survey of Canada for seismic hazard information for Canadian sites, and similar agencies in Mexico, Cuba, the Bahamas, and Bermuda for seismic hazard inforation in those countries. You can deaggregate hazard at sites in Alaska and Hawaii at another web address (http://eqint.cr.usgs.gov/deaggint/1996/index.php ).

Site conditions are assumed to be rock, with average shear wave velocity of 760 m/s in the uppermost 30 m. It is known that differing site conditions can affect response. In general, site conditions are not expected to affect the relative contributions from sources of seismic hazard very strongly, but they can affect probabilistic ground-motion levels, SA 0, strongly.

What is New in 2002

This web page is intended to allow the user to determine the principal sources that contribute to seismic hazard at a specified site, according to the USGS seismic hazard model documented in Frankel et al (2002). The previous USGS seismic hazard model, often called the 1996 model, may be deaggregated at another web address (http://eqint.cr.usgs.gov/deaggint/1996/index.php ).

What Is New in the Layout and Design?

We have attempted to maintain a similar style of analysis and appearance as is presented the interactive deaggregation of the 1996 USGS seismic hazard model. As before, two types of deaggregation are possible, one in distance - magnitude -and epsilon, or ground-motion uncertainty (R,M,e) , and the other in geographic location - average magnitude - and relative contribution.

For the (R,M, e) option, the magnitude bin width has been reduced from 0.5 Mw units to 0.2 units. The distance bin width is either 10 km or 25 km, 10 km where seismic activity is relatively high, such as in parts of the Western U.S., the New Madrid Seismic Zone, and the Charleston, S.C. seismic zone; and 25 km elsewhere. The plot color scheme has changed. Now, the hazard columns at each (R,M) bin location are color coded to the average e0 for sources in that bin. e0 is the number of standard deviations that a given ground motion is from the median motion for a given source. Binned contributions with a negative e0, which means that the motion is less than the median motion that might be expected from a source, are assigned warm colors, and binned contributions with a positive e0 are assigned cool colors, with yellow and green for bins whose e0 are just below or just above the median, respectively. The front face of each column provides information on e binning, but is no longer presented as a set of colors. Instead, gray regions show the relative contributions with lines between the 1-sigma demarcation points. The top of each column is left in the original column color, and is the contribution from exceedances of mean + 2 sigma motions. The first gray rectangle down from the colored top is the 1 to 2 sigma contribution, below which is the 0 to 1 sigma contribution, and so on. Thus all of the information available in the 1996 deaggregation plots has been preserved, but a new emphasis has been placed on e0 in the updated version.

For the geographic deaggregations, the gridded seismicity contributions are now binned into 45-degree sections. The region around the site is divided into concentric cylinders (circles) with radii 25 km, 50 km, 100 km, 200 km, etc. Each of these regions is divided into eights and the gridded contribution is lumped in each of these subregions. The reason for presenting the hazard contributions this way is to show how gridded hazard compares to specific fault hazard. Often there is comparable hazard from faults as from gridded sources when the latter are lumped in this manner. However, if a finer spatial binning is used, the relative contribution from random or gridded seismicity sources tends to be difficult to compare to fault hazard. Also, the current way of spatial binning allows the user to determine which if any azimuths exhibit higher gridded hazard than the others. Light blue circles and rays outline the boundaries of the new gridded-hazard geographic bins.

The fault traces that appear on the geographic hazard plots are or should be the 2002 update map faults. They include several newly recognized faults and omit some faults that are no longer considered a seismic hazard.

The contribution from Cascadia subduction sources is now based on four models of the easternmost location of seismogenic rupture. In the geographic deaggregations for sites near these sources, the down dip limits of Cascadia sources are plotted as orange lines. The location of the nearest point on each of these surfaces to the site is used to calculate the distance associated with that hazard. We are currently working on the azimuth for these sources. The interim solution is to plot the hazard due west of the site, except for sites south of Cape Mendocino, for which the source azimuth is assigned a northwest direction. This approximation will be corrected shortly in our software.

The text files contain information for PGA and the two spectral periods for which PSHA analysis was initially performed for the 2002 updates, namely 0.2 sec SA and 1.0 sec SA. The text files are similar to those for the 1996 deaggregations. A few additional comment lines are now included. One new line shows the probability that the specified PGA or SA will be less than the median motion of an earthquake source that might be recorded at the given site in a random 50-year period. Sometimes this probability is so near zero that it is reported as zero. Other new lines at the bottom of each SA period's analysis summarize the source types that principally contribute to the hazard. There are 24 source types. Deaggregation by these source types helps to show the new features in the 2002 PSHA model as well as features that were in the earlier PSHA. For example, you can see the contributions associated with the narrow Charleston S.C. source zone (new) and the broad Charleston S.C. source zone (like that of the 1996 maps) by looking at the bottom of the analysis, for sites reasonably near Charleston, S.C. New categories (compared to 1996) include extensional gridded sources and extensional faults, among others. Following the source type lines are lines for specific faults, fault zones, or fault segments. These can get quite complex, especially in western California. Fault descriptions or segmentation models are sometimes in a terse code that can be hard to decipher, for example SG is San Gregorio, and SH is south segment of Hayward fault. These codes may be improved on later, or a glossary may be provided.

The PGA and SA values that you get at this web site are the same as those from the 2002 maps at grid locations where they were calculated, and for the 2% in 50 year and the 10% in 50 year probabilities of exceedance. For other locations, and for lower and higher probabilities, this web site's analysis actually solves for the ground motion associated with that probability. The solution is iterative and may take a few minutes, so please be patient.

Other Remarks:

The generation of synthetic seismograms in 2002 is for a point source having either the modal magnitude and distance or the mean magnitude and distance. In fact, this part of the web page has not yet been changed from the 1996 version. Synthetic seismograms are no longer scaled to the probabilisitc ground motion, and this is a change from the 1996 web site. For more details, read the 2002 stochastic seismograms discussion.

Site Name Format

The site name is used to label the plot(s) only. Please specify a valid site name. Only underscore (_), comma (,) and alphanumeric characters are allowed, use of any other characters may cause the computation to fail. If you use blanks, we'll replace them with underscores. Site name is limited to 16 characters (including spaces). Here are some examples of valid site names:

  • Boulder_City,NV
  • New_York,New_York
  • NYC
  • San_Francisco
  • SanFranciscoCA
  • SF,CA
  • San Francisco CA (converted to San_Francisco_CA)
  • San Francisco, CA (converted to San_Francisco,_C truncated to 16 chars)

Latitude Format

Please try to be accurate in your specification of site coordinates. Here are some valid latitudes:

  • 36.123
  • 36.1
  • 36.
  • 36

It is best to use floating point numbers, in the range 25. to 49. degrees for the conterminous US or 17. to 19.9 for Puerto Rico.

Longitude Format

Please try to be accurate in your specification of site coordinates. Here are some valid longitudes:

  • -75.25
  • -122.987
  • -120

Valid longitudes are in the range -125. to -65 for the conterminous US or -64. to -68.5 for Puerto Rico. As with latitude, we are looking for a floating point number without embellishments.

Return Time Format

The most common PSHA probabilities are 10% Probability of Exceedance in 50 years (475 year return time) and 2% Probability of Exceedance in 50 years (2475 year return time). We have added four more, to let you look longer and shorter return times.

  • 1% PE in 50 years (4975 year return time)
  • 5% PE in 50 years (975 year return time)
  • 20% PE in 50 years (224 year return time)
  • 50% PE in 75 years (108 year return time)

With a little experience, you should be able to predict how the deaggregation plots would change as you increase/decrease the probability of exceedance.

Spectral Acceleration or Peak Ground Acceleration

Do you want to deaggregate a spectral acceleration (SA) frequency, or peak ground acceleration (PGA)? If your project requires 5% damped pseudo-spectral acceleration, we can provide you with detailed deaggregation results for the following frequencies (periods):

  • 1 Hz (1.0-s)
  • 5 Hz (0.2-s)

We currently have no other "frequencies" on line. Peak ground acceleration deaggregations are also available.

Hazard Matrices

Our software computes several SA-frequency data simultaneously. We have decided to give you an ascii file of everything we calculate, even though you may have only specified one frequency. Each frequency's (or period's) dataset is separated by header info and asterisks. Your plot or plots, however, will only be for the specified SA frequency (or PGA, if that is what you ask for). The "decimal exponent" number format is used in some header fields. This format is base 10 (not e). For example .404e-03 equals .000404.

Geographic Deaggregation Maps

These maps are included to help you visualize the relative hazard from different faults, source zones, and plate subduction events. Your site is at the center of the map, and is shown as a yellow disk. The basemap is a fairly good approximation to an equal-area projection, but it is not exact. Please do not rely on it if you require very precise map locations. The hazard columns are displayed on a true equal-area projection. For perspective, the camera is oriented 35 degrees above the horizon. This causes north-south foreshortening.

The region plotted is square. Its scale depends on one quantity: What is the greatest distance to your site of a hazard bin with at least a 0.005% contribution to seismic hazard?

The color scale adjusts automatically to your data. The smallest magnitude (M) in the color scale is (approximately) the smallest average magnitude (average M is what is plotted at each location) in the data, and the largest M in the color scale is the largest average M in the data. Often, the largest M corresponds to a characteristic magnitude.

The map also includes schematic locations of active Quaternary faults (red lines) that contribute to hazard, and includes epicenters of historical earthquakes, M≥6 in the WUS, and M≥5 in the CEUS. The location of the edge of the Mississippi embayment is included on some maps.

Magnitude, Distance Plots

The magnitude, distance, epsilon deaggregation plots require a larger contribution for including a datum, 0.05%. Smaller contributions may appear in the ascii files but not in the plots. The "header" information on the plots give you some often-used statistical descriptors of the hazard data. These are mean (i.e., average) and modal values for magnitude, distance, and ground-motion uncertainty (epsilon). Modal epsilon is the average epsilon for sources in the modal (maximum) bin. Epsilon is the number of logarithmic standard deviations SA0 is from the logarithmic mean ground motion for a hazard source. Positive epsilon implies SA0 is above the median value, negative epsilon implies SA0 is below the median value, given the source magnitude and distance.

Binning Details

A note on binning: We use a distance increment of 10 km for deaggregations at sites having relatively high seismic hazard, and a distance increment of 25 km for other sites. Additionally, the geographic deaggregations have an azimuthal bin size that is relatively fine (1°) for faults and relatively coarse (45°) for gridded sources. For the distance, M, epsilon deaggregations, the M-increment is always 0.2(Mw) units, and the epsilon-increment is 1 (sigma). For the geographic deaggregations, we compute the weighted average M in each bin, and we integrate over ground-motion uncertainty (from SA0 to m+3s).

Conclusion

We hope you find this web page helpful for your project. We would of course appreciate your comments about this page.